Psychology and Logic

J.R.KANTOR
Pnonsson or PSYCHOLOGY, INDIANA UNIVERSITY

VOLUME II

1950

\
THE PRINCIPIA PRESS
BLOOMINGTON, INDIANA
 ∙
∙ ≤ °

THE PRINCIPfAfi’zRESS, INC.

 CONTENTS
FOREWORD ......................................
CHAPTER XIII
\

ABSTRACTING AND GENERALIZING OPERATIONS ..........

A. and G. as Interbehavioral Operations, 1. Logical Forms
as Products of A. and G. Operations, 2. Constructional Print
ciples in A. and G. Operations, 4. A. and G. Levels, 7. Indi-
vidual Operatidns, 7. Cultural A. and G., 8. Interrelation of '
A. and G. Operations, 9. Specific Characteristics of A. Opera-
tions, 10. Typical A. Interbehavior, I 1. Specific Characteristics
of G. Operations, 3. Typical G. Interbehavior, I5. Analo-
gizing, 15. Magnification, I6. Reiteration, 16. Interpolation,
17. Extrapolation, I7. Combined G. Operations, 18. Linguistic
and Symbolic Techniques in A. and G. Operations, 18. Inter-
relation of Linguistic and Nonlinguistic Behavior in A. and G.
Situations, 19. Role of Language in A. and G. Situations, 2o.
Criteria and Limits for A. and G. Operations, 20. Illicit A. and
G. Products, 22. A. and G. Operations in System Building, 26.
CHAPTER XIV
DEFINING CLASSIFYING OPERATIONS ...............
AND

Definition and Classification in Logic, 28. The Definitional

One and Many, 29.'I. Definition, 29. A. Definition as Pursuit
of Essences, 3o. Nominal Definition, 31. Real Definition, 32.
Concept Definition, 32. B. Definition as Pursuit of Form, 33.
Syntactical Definition, 33. Semantic Definition,'34. Definition
as Interbehavioral Operation, 34. Range of Defining Opera-
tions, 35. Definitional Criteria, 35. Analysis of Definitional
Situations, 36. A. Stimulus Objects, 36. B. Defining Acts, 36.
C. Definitional Products, 37. D. Linguistic Aspects of Defini-
tion, 38. Defining Operations and System Building, 40. Defi-
nition and Description, ‘40. Definition and Meaning, 42. The
Indefinable, 44. Definition, Vagueness, and Ambiguity, 46.
Samples of Defining Procedures, 49. Defining by Naming, 49.
Defining by Delimitation, 49. Defining. by Classification, 49.
Defining by Postulation, 5o. Analyzing, and Synthesizing Defi-
nitiOn, 50. Defining by Relating an‘d Analogizing, 51. Defin-
ing by Measuring, 52. Defining by Prescription, 52. Defining
by Specific Operations, 53. Rules for Definition, 57. II. Classi-
fication, 58.
vi PSYCHOLOGY AND LOGIC

CHAPTER XV
SYSTEMS As
CONSTRUCTION PRODUCTS ...............∙ 60
Selection and Analysis of System Products, 60. Investigative
Schema for System Products, 60. Four Types of System Prod-
ucts, 61. A. Material or Content Systems, 61. (1) Things, 61.
(2) Event Systems, 71. (3) Relations as System Materials, 72.
(4) Classes as System Materials, 72. (5) Acts as System Ma—
terials, 73. (6) Linguistic Things as System Materials, 7 3.
B. Linguistic Systems, 7 3. (I) General Reference Systems, 74.
(2) Method Systems, 75. (3) Metalogical Systems, 76. (4)
Autological Cosmic Systems, 76. C. Formal Systems, 76. (I)
Implicatory Systems, 77. (2) Truth-function (Matrix) Sys-
tems, 78. (3) Complete Induction, 79. (4) Mathematical
Systems, 79. D. Operational Systems, 84. (1) Chemical Analy-
sis, 84. (2) Solution of a Mathematical Problem, 85. Scien-
tific Structure and Deductive Systems, 86. Interrelation of
Systemizing Operations and Materials, 86. Characteristic Vari-
ations of System Products, 88. Tight and Loose Systems, 88.
Complete and Expanding Systems, 89. Discoveredyand Con-
structed Systems, 89. Witting and Unwitting Systems, 89.

CHAPTER XVI
INSTRUMENTS FOR SYSTEM CONSTRUCTION 91
Models, Schemata, and Formulae as Logical Instruments,
91. Logical and Scientific Instruments, 91. Logical Instruments
as Products, 92.1nterbehavioral Sources of Logical Instruments,
92. Specificity of Logical Instruments, 94. Logical Instruments
in Operation, 94. Logical and Nonlogical Instruments Com—
pared, 95. Logical Instruments as Miniature Systems, 96. Logi—
cal Instruments as System Items, 97. Materials Of Systemological
Instruments, 97. Orientational Function of Logical Instruments,
98. Representational Tools, 98 . Analogically Descriptive Tools,
99. EXplanatory Tools, 99. Speculative Fictional Tools, 99.
Instruments in Various Orientational Situations, 100. Logic, 100.
Mathematics, 101. Physics, 102. Biology, 103. Psychology,
104. Social Sciences, 1 06. Techniques in Instrument Production,
106. Iconography, 106. Abstraction, 106. De Novo Construc-
tion, 107.
CHAPTER XVII
CATEGORIES AND SYSTEM BUILDINC\ 108
Categories and Logic, 108. Nature and Operation of Cater
gories, 108. Evolution of Categorial Systems, 109. Pro-Aris-
totelian Categorization, 109. Aristotelian Categorization, 109.
 CONTENTS vii
Post-Aristotelian Categorization, I 10. Kantian Categorization,
1 10. Categorial System of von 'Hartmann, 11 1. Current Atti-
tudes toward Categories, 112. Categorial Sources, 1 12. Cate-
gories and Concepts, 114. Categories, Classes, and Universals,
114. Categories, Language, and Symbols, 115. Categories as
Interbehavioral Products, 1 I 5. Categorial Continuity, 1 1 6. Cate-
gorial Specificity, 1 1 7. Special Categorial Functions, 1 1 7. Difl’er-a
entiating Functions, 1 I7. Classifying Functions, 118. Descrip-
tive Functions, 1 I8, Explanatory Functions, 1 18. Range of
Categorial Systems, 118. Scientific Departments, 118. Specific
Researches, 1 19. Property and Dimension Systems, 1 I9. Illus-
trative Categorial Operations, 1 19. 1. Mechanics, 1 19. 2. Elec-
tricity, 120. 3. Geometry, 120. 4. Everyday Categories, 121. 5.
Religious Categories, 121. 6. Dimensional Analysis, 122. 7.
Axiological Categories, 124. 8. Thermodynamics, I25. 9.
Biology, 126. 10. Ethical Categories, 127. 11. Logical Cate-
gories, 128. 12. Grammatical Categories, 128. 13. Philosophical
Categories, 129. 14. Aesthetic Categories, 129. 15. Mathemati-
cal Analysis, 130.
CHAPTER XVIII
UNIVERSALS: SYSTEMOLOGICAL COMPONENTS AND PRODUCTS 1 3 I
The Logic of Universals, 131. Historical Treatment of Uni-
versals, 131. Universals among the Greeks, I31. Medieval .

Concern with Universals, I32. (1) The Realists, 132. (2)

The Nominalists, I32. ( 3) The Conceptualists, 132. Current
Treatment of Universals, 133. (I) .Spatiotemporality, I33.
(2)‘ Character Value, 134. ( 3) Participation, 134. Logical
Forms as Things, Terms, or Thoughts, 136. Interbehavioral
Character of Universals, 137. Specificity of Forms, 138. Forms
as Products, 139. (a) Thing or Event Sources, 139. (b) Action
Sources, 141. (c) Product Sources, 141 . Forms as System Com-
ponents, 142. PrOpositions as Form Compounds. 143.Propo-
sitional Complexes, 143. Abstract Forms as Total Systems, 143
Dialectical Logics,144, Analytical Logics, 144. Symbolic Log-
ics, I44. Methodological Logics, 144.

CHAPTER XIX
INTERBEHAVIORAL LOGIC AND CAUSATION .............. 145
Causation as Applied Logic, 145. Is Causation a Logical Prob-
lem? 145. Emergence of Causality in Methodistic Logic, I46.'
Cultural Background of Causality, I47. Classical Causal
Theory, 149. (a) Formal Cause Theory,\149. (b) Epistemo-
lOgical Cause Theory, 150. (c) Ontological Cause Theory,
viii PSYCHOLOGY AND LOGIC

15o. Causality: Universal and Specific, 152. Logical Necessity

and Factual Cause, 154. Causality, Functionality, and CorreJ
lation, 156. Operational Causal. Theory, 158. Classical Psy-
chology and Causal Theory, 1 5 9. Pattern of Causal Investiga-
tion, .161. A. Data, 161. (I) Causal Constructs, 161. (2)
Causal Events, 163. B. Investigative Operations, 165. C. Con-
structions, 166. Cause and Effect, 166.’Causes and Conditions,
166. Causes and Causal Interpretation, 166. Cause and System
Building, 167. A. General Causal Structurization, 167. (a)
Event Level, 168. (b) Linguistic Level, 168. (c) Formal
Level, 170. B. Systems as Specific Causal Instruments, I71.
Causation and Indeterminism, 172.

CHAPTER XX
THE LAWS OF THOUGHT AND THINGS ................ 175
Logic and Law, 175. Varying Formulabnons of Logical
Laws, 177. Logical Laws as Products of Cultural Development,
179. Universality and Necessity of Logical Laws, 183.(1)
Nature and Number of Logical Principles, 184. (2) Source,
’of Logical Laws, 184. ( 3) Universality and Necessity of
Application, 185. Specificity and Contingency of Logical Laws,
186. Logical Laws in Reasoning and Thinking, I86. Logical
Laws and Existence, 188. Logical Laws and Data, 190. Logi-
cal Laws and Language, 194. Logical Laws and Mathematics,
196. Logical Laws as Factors in System Building, 199; Logical
Laws and Systemological Behavior, 202.

CHAPTER XXI
I
PROBABILITY AND
INTERBEHAVIORAL LOGIC ............ 204
Probability: Guide to Logic, 204. ‘The Domain of Proba-
bility, 205. Philosophical Presuppositions, 206. Calculative
Presuppositions, 207. Probability Events and Situations, 207.
Probability Range and Continuity, 208. Probability: Univocal
or Multivocal? 210. Generality and Specificity of Probability
Systems, 212. Relative Value of Probability Systems, 214. His-
torical Evolution of Probability systems, 215. Pragmatic Be-
ginnings, 216. ReSponse-Centered Systems, 217. Event-Cen-
tered Systems, 218. Propositional Probability, 219. Probability
as Statistical Operations, 221. Basic Orientation of Probability
Systems, 221. Probability, Certainty, and Necessity, 222. Prob-
ability, Truth, and Error, 223. Probability, Possibility, and
Impossibility, 224. Probability, Potentiality, and Actuality, 225.
Probability in Science, 225. Probability and Induction, 228.
Logic and Events, 228, Relation of Induction‘and Deduction,
 CONTENTS ix
228. Probability and Inductive Principles, 229. Probability and
Prediction, 230.
CHAPTER XXII
ANALYSIS OF
PROBABILITY SYSTEMS ............... ∙ ∙ 232
Design for Probability System Analysis, 232. (I) The Sys-
tem Builder, 232. (2) Methods and Procedures, 232. (3)
Materials, 233. (4) Working Auspices, 233. (5) Work Prod-
ucts, 233.Types of Probability Systems, 234(1) Probability
as Rational Responses, 234(2) Probability as Knowledge Re-
sponses, 237. (3) Probability as Laws of Things, 241. (4)
Probability as Propositional Relations, 246. (5) The Truth-
Frequency Theory, 248. (6) Probability as Factorial Coordina-
tion, 253. Probability Systems and Logical Multiplexity, 255.
Probability, Epistemology, and Ontology, 256. General Philo-
sophical Problems, 257. Special PhilosophicalProblems, 257.
Interbehavioral Probability versus Epistemology, 2 57.. Interbe-
havioral Probability versus Ontology, 258. Probability Propo-
sitions versus Probability Events, 2 58. Probability Calculation:
Goal or Operation? 261. Statistics and Probability, 261. The
Nature of Calculation, 262.
CHAPTER XXIII
LOGICAL ASPECTS OF MEASUREMENT .................. 26 5
The Place of Measurement in Logic, 265. What is Meas-
urement? 265. Ideological Trends in Measurement Theory,
266. Measurement as Applied Mathematics, 267. The Power
of Mathematical Reasoning, 269. Numbers as Supreme Uni-
versals, 269. Numbers as Philosophical Abstractions, 27o. Num-
bers and External Reality, 270. Numbers and Mental States,
270. Mensurational Operations and Numeral Assignment, 270.
Specificity of Measurement, 271. (1) Specificity of Measured
Objects, 272. (2) Specificity of Operations and Techniques,
273. (3) Specificity of Mensurative Situations, 274. Levels of
Measurement, 275. Practical Measurement, 27 5. Technologi-
cal Measurement, 276. Formal Metrology, 277. Interrelation
of Levels, 278. Quality and Quantity, 278. (1) Quantity as
Property of Things, 280. (2) Quantity as Measure, 281. (3)
Quantity as Constructions, 28 1 . Variety and Specificity of Mag—
nitudes, 281. Measuring Systems, 282. A. Assumptions and
Postulates of Measurement, 283. B. Choosing and Refining
Units, 284. C. Instruments and Operations, 285. D. Record-
ing, 292. E. Statistical Operations,
∙ How Arbitrary are
Units or Standards? 293. The Nature OfQuantity, Number,
and Magnitude, 295. Measuring and Counting, 296.
X PSYCHOLOGY AND LOGIC

CHAPTER XXIV
METROLOGY AND METASCIENCE ...................... 299
Problems of Mensurational Logic, 299. System Type and
Metrological Evolution, 300. (1) Measuring Systems, 300.
(2) Weighing Systems, 301. (3) Counting Systems, 301. (4)
Calculating Systems, 302. ( 5) Ordering Systems, 302. (6)
Estimating Systems, 302. (7) Evaluating Systems, 303. (8)
Equating Systems, 304. (9) Probability Systems, 304. Measure-
ment: Abstract and Concrete, 306. Measurement: Construction
and Observation, 307. Measurement and Cause, 31 1. Measure-
ment and Objectivity, 312. Measurement and Reality, 313.
Metrology and Logic, 318. −
CHAPTER XXV
SYSTEM BUILDING AND METALOGIC .................. 321
The Continuity of Human Behavior, 32‘1. Metalogic and
Logical Practice, 322. A. Logic and Living, 323. B. Logic
and Science, 324. Constructing an External World, 327. C.
Logic and Philosophy, 328. Towers and Systems, 329. Tech-
niques of Unrestricted System Building, 330. (I) Obeisance
to Utterance, 33o. ( 2) Projection and Fixation of Utterance,
330. Abnegation Motive in Philosophy, 330. Problem of the
Inductive System, 331. Statistical Procedures and .Absolute
Systems, 332. Logic and the Plenum, 333.
BIBLIOGRAPHY 334
SUBJECT INDEX ...................................... 353
NAME INDEX ........................................ 3 56
 FOREWORD
This volume emphasizes the applied aspects of the relation be-
tween psychology and logic. Whereas in the first volume we were
chiefly occupied with logical theory, we now investigate the ori-
going processes and finished products of system building. Because
theory and practice are continuous, the materials of the two
volumes are, of course, intimately integrated. The topics of
Volume II are treated on the basis of the principles worked out
in Volume I, the results, therefore, in a significant way consti-
tute illustrations and tests of interbehavioral theory.
Applying logical principles means for us in no sense exploit-
ing them. Interbehavioral logic is not a unique logical system
ready for use in annexing truth and reality. Logics of that type
we reject, along with the various historical philosophies serv—
ing as their foundation. Specificity logic,'In both its theoretical and
applied aspects, constitutes a scientific study of systemizing situa-
tions. .Its application,:accordingly, consists in observing how well
descriptiOns of system making fit when they are projected back
into the fields from which they are drawn. Thus exploitation
gives way to verification.
The chapters of the present volume demonstrate the essential
"place in various logical Situations of (I) concrete Specific Opera-
tions and (2) the things, events, and relations which constitute
the raw materials for (3) systemic products. In many instances,
too, we clarify (4) the types of instruments employed in sys—
tem construction. The indispensability of these four factors is
amply exemplified1n the investigation of system products (chap.
I 5), as well as in many other chapters.

It13 only to be expected that the\ emphasis of the four factors

varies with the specific topics treated. For example, in Chapters 1 3
and in which we study abstracting, generalizing, defining,
and classifying operations, the focal point of observation'Is the
logiciap engaged'1n system building. By cognparison, Chapter 16
 xii PSYCHOLOGY AND LOGIC I

describes the instruments necessary for constructing systems—

namely, models, schemata, and formulae. In the chapters on
categories (17) and universals (18) we Show that these intel-
lectual products constitute, on the one hand, system components
which may be analyzed out of, and separated from, given sys-
tems and, on the other, tools useful in logical construction.
The seven remaining chapters '( 19-25) comprise the more
strictly applied materials of the volume. These chapters serve
their verification function by diSplaying the contrast between the
conventional logical and interbehavioral approaches in the treat-
ment of causation (19), laws of thought (20), probability (21-
22), measurement (23-24), and the perspectives of logic (25).
Outstanding are the difierentiations revealed in logical and
scientific results when classical mentalistic psychology, which has
paralleled the transcendent philosophies of historical absolute and
universal logic, is set aside in favor of interbehavior with things
._under Specific circumstances.

As in Volume I, the frequent use of illustrative quotations is in

all cases exclusively intended to exemplify doctrines. It ‘has not
been necessary, therefore, to inquire whether other writers share
some particular view, condemn it, or whether the quoted author
has retreated from the position specified.

I want to call attention to a change in style of foot-note

references. A number of book titles, which in the first volume are
cited by two or more words, have been reduced to a single word,
usually the first.

Circumstances long delaying the completion of the present

volume have not been without compensations, The interval has
enabled me to consider several criticisms made of the first
volume. While I have no intention of defending the materials
presented beyon’d the arguments in the text, I prize the op-
portunity to sharpen a few points.
First, consider the criticism that thO‘interbehavioral hypothesis
is contradictory. On the one hand, it- is charged that I deny
generality to logic, but, on the other, claim generality for the inter~
behavioral hypothesis. My answer is simply to invoke the specificity
 FOREWORD xiii
rule. The rejection of all absolute and general logical systems is in
no way incompatible With the construction of a generalizing hy-
pothesis. The assertion that there is no universal logic is not con-
tradicted by demonstrating that all logical systems are specific
creations of particular individuals under unique auspices. The
charge here is reminiscent of Bradley’s absolutistic argument
about metaphySics. Deny metaphysics and you are a metaphysician,
though a bad one, for deny that reality is attainable by knowl-
edge is to claim knowledge of an unknowable reality. Absolutists
lack nothing except the appreciation that their universalism is but
a specific kind of local creation. ∙
The same critic objects to my attempt to separate both logic
and psychology from traditional philosophy. On what ground?
On nothing less than that presuppositions are inevitable and that
presuppositions mean philosophy. Granted that philosophy is
synonymous With presuppositions, one must still difierentiate
between types of presuppositions. Current phiIOSOphy concerned
with the critical methods of science is vastly difierent from tradi—
tional metaphysics. I am thoroughly sympathetic .with philosophy
taken as a generalized science of science. But it is clear beyond a
doubt that the development of an objective psychology makes
possible the complete and final extrusion of all idealisms, nominal-
isms, and realisms from science and logic. As an independent
branch of study logic is in no sense connected with the classic
metaphysical systems called philosophy.
My suggestion that logic might profit from cooperation with
an objective psychology has prompted another writer to raise the
perennial cry: psychologism. Two comments are in order. First,
the present epithet psychologirm must be regarded as altogether
different from the classic one used by Husserl and others. From
the standpoint of interbehavioral psychology there are no psychic
elements or essences such as are involved in classic psychologism.
Specificity logicians, therefore, are farther removed from psycholo-
gism than are both the original antipsychologists and those they at-
tacked. Interbehavioral system building is completely free from
internal Spiritistic entities. Supported by an objective psychology
we are able to demonstrate how cosmic logic, whether psycho-
logistic or nonpsychologistic, is primarily verbal creation and pro-
JCCtlon.‘
Xiv PSYCHOLOGY AND LOGIC

Moreover, it is precisely an operationally rooted psychology

which can differentiate (a) the various types of human behavior,
(b) the materials interbehaved with, and (C) the products of
such behavior. On this basis the work of psychologists can be dis—
tinguished from that of logicians, scientists, and other workers,
While full justice can be done to the interrelations of the diverse
enterprises.
Considering that, unlike most logical theorists, I have formally
set down my postulates, it is anomalous to find a critic assert-
ing that I regard specificity logic as Without presuppositions. The
association with objective psychology of a logical theory lack-
ing preSUppositions is impossible. The point is precisely that inter-
behavioral presuppositions are difierent. What can be a more
definite presupposition than the rejection of classical logical pre-
suppositions? Because a disdain for autistic world systems is not
.pleasing to cosmologicians, they find such a presupposition un-
palatable.

∙
In another instance a writer Who does associate postulates with
presuppositions denies that most of the basic interbehavioral
suppbsitions are postulates. The assumption is that only classical—
logical postulates are postulates. This criticism is made _on the ex—
pressed ground that logic is an a priori discipline; and, Since speci-
ficity logic is not such a discipline, it simply is beyond the pale.
Strangely enough, the argument made for a priori logic concerns the
existence of scientific contents—namely, events—which are inde-
pendent'of scientific work. But this is exactly the interbehavioral po-
sition. All authentic scientific work consists of interbehavior with
events described in Jim, manipulated in the laboratory, and counted
and computed when possible. Descriptions and explanations can only
be structured on such events as an indispensable foundation.
Furthermore, all this activity goes on Within the bounds of human
interbehavior. If thereIs anything that one must call a priori, it
belongs inside this system. N0 contact with scientific work or with
scientific subject matter allows for any absolutes and transcendents,
whether empiricistic or rationalistic.
Several writers criticize my proposed cooperatiOn betweenslogic
and psychology on the ground that, if this cooperation is necessary
for logic, it. is likeWise necessary for physics and other sciences.
 FOREWORD XV

To such critics it seems axiomatic that cooperation of physics and

psychology is not necessary. But writings on physical theory Show
conclusively how an objective psychology can obviate the mysticism
and metaphysics filling the pages written by physicists. Think only
of the bifurcated universe of the classical physicists, of the in—
determinism and reality puzzles of relativists and of quantum
mechanists! This matter is amply treated in our chapters on proba-
bility and measurement.
Curious are the vagaries of reviewers. One writer indulged in
considerable dispraise of objective psychology because, obviously,
formal mathematical analysis cannot be replaced by psychological
investigation. Missed altogether was the fundamental point that
interbehavioral psychology not only marks the boundaries of
human behavior, but also differentiates the various activities Within
these limits. For example, because of the variations in materials,
I have stressed the specificities of system building in mathematics,
in science and in everyday life. I have also pointed out the contrast
between system building and other behavior——a contrast which
should separate sharply, say, the mathematician’s calculative and
analytic behavior from the logician’s structuring of operations and
products. In addition, behavior analysis distinguishes clearly the
work of the creative mathematician from that of the logician who
builds symbol or word systems describing the products of such
mathematical work.
Quite aside from the above, it is reprehensible to overlook the
fact that, because of the continuity running through all intellectual
activity, new developments in one domain are reflected in all the
others. What science can escape the revolutionary consequences
of the physicist’s relativity and discontinuity? Thus the emergence
of objective psychology cannot but leave its mark on other disci-
plines. Only at the risk of intellectual peril can one cavalierly
wave it aside.

When Paul Valéry (Hadamard, an Essay on the Psychology of

Invention in the Mathematical Field, p. 60) was asked about his
reaction to his completed works, heireplied, “It always turns out
badly; Je divorce.” I admit that I. have been unable to grasp
as high as I have reached. Still, the unfettered reader will be able
xvi PSYCHOLOGY AND- LOGIC

to see that an objective psychological treatment of language,

thought, and reasoning is of considerable consequence for logic.

Once more it is my pleasure to acknowledge the day-by—day

collaboration of H. R. Kantor._ I Wish also to thank Mr. Ernest
Lilienstein for his efiective services while this book was being
written and printed. Likewise I am indebted to Deans Payne and
Thompson of the Graduate School of Indiana University for
grants required for library and secretarial assistance. AS in the\
case of Volume I, I want to acknowledge the help received from
Dr. A. E. Kanter, and the Directors of the Principia Press who
have generously contributed to publication costs.
J. R. K.
April .1950
 CHAPTER XIII
ABSTRACTING AND GENERALIZING OPERATIONS

ABSTRACTING AND GENERALIZING As INTERBEHAVIORAL

\
OPERATIONS

ALL logicians recognize that, in some form or other, abstracting

and generalizing are indispensable operations in complex
system making. Especially writers on mathematical theory have
pointed out how vital for the evolution of mathematics and logic
was the original achievement of abstracting numbers from thing‘s.1
What is now required is to treat these operations as definite inter-
behavior in concrete fields. Such an Objective treatment runs
counter to the tradition that logical abstracting and generalizing
comprise acts of “reason” or “mind,” whether or not dependent
on autonomous ontological entities.
Naturally, in conventional logic, which is universalistic and ne-
cessitarian, we can not expect abstracting and generalizing to be
described as specific interbehavior of persons with stimulus objects.
Logicians for whom the materials of system making are formal or
verbal things automatically accept the interpretation that forms
and propositions constitute autonomous entities. Thus they over—
look the competing View that conventional logical materials con-
sist of products of concrete operations (vol. I, p. 2f.). In this work
we treat abstracting behavior as analytic operations upon a con-
tinuum, and generalizing as synthetic interbehavior with presented
materials.
Failing to appreciate the field character of abstracting and
generalizing operations, logicians tend to pay more attention to the
products achieved than to the constructive processes by which
products are obtained. Once obtained, products appear not only
independent but established, and may easily conceal the specific
operations through which they. were produced. For instance,
though it is quite proper to regard mathematical equations as
purely formal and autonomous struct‘iIres, they are nevertheless

1Russell, Introduction, p. 3; Whitehead, Science, p. 29f.; Bell, Development, p.

23f. −
2 PSYCHOLOGY AND LOGIC
derived from interbehavioral field operations upon definite ma-
terials. That all equations must be accompanied by a text2 points
to two interbehavioral situations: (I) In applied mathematics n0
equation is valid unless it can be carried back to the original set of
measuring operations from which it springs; ( 2) In pure or formal
mathematics, equations go back to contacts with particular systems
of relations.
The confusion of formal objects and systems with unproduced—
entities is an acute instance of the antigenetic fallacy. Because some—
thing exists in the fullness Of its structure and use there is a
tendency to ignore the elaborate processes of its evolution. Granted
that the narrated history of a thing seldOm suffices for its com-
plete explication, there is still no warrant for assuming ingenetic
essences among its properties,
Abstracting and generalizing operations are frequently over-
looked because insufiicient account is taken of their specific char-
acteristics. Frege supports his belief that Platonic numbers inhabit
a timeless realm of Pure Being on the ground that numbers,
un-
like color, are not abstracted from things.3 But to grant this point
is certainly not to sanction the View that numbers are not, in some
way, constructed. The ontological argument that “nicht jeder
Gegenstand ist raumlich’” proves nothing contrary t6 the theory'
that such objects and their nonspace are created by abstracting and
generalizing operations upon concrete things.
The interbehavioral standpoint represented in this treatise im-
plies that constructional operations may be either refined or crude
depending upon the kind Of materials ope'rated upon. Whether
the materials are concrete or abstract, and whatever the motiva-
tion, the abstracting analysis is definitely selective. By the same
token, generalizationlS constantly subject to the materials and
problems at hand.

LOGICAL FORMS As PRODUCTS OF A. AND G. OPERATIONS

With the development in recent decades of mathematical or
symbolic logic we might have hoped for an increased eluc1dat10n
of abstracting and generalizing operations. Mathematicalfilogic,
2Bridgman, Nature, chap 6.

Ibid.
p.
Grundlagen, vol. I, 27.
 ABSTRACTING AND GENERALIZING OPERATIONS 3
like mathematics itself, has steadily become pervaded by the twin
characters of abstractness and generality.° Has not logic itself
been claimed as the science of pure form, and mathematics as re-
ducible to logic becauseIt can be exhibited as a completely logical
structure?“ That logical proof or demonstration'18 independent of
subject matter has become a greatly lauded principle.7
No appreciable clarification of abstracting and generalizing
processes, however, has materialized. And why not? Probably be-
cause the early introduction of mathematical logic coincided with
the resurgence of mystic doctrines and with the notion that logic
is concerned with Platonic essences. But do not forms really con-
stitute systems and organizations? Are not. forms structures, and
therefore products of ( I) investigation, in the case of objects and
events, and (2) constructed relations between things, when the
latter are submerged in their relations?
However, the conventional doctrine of logical forms has its
value for our' present study. For one thing, the most extreme ab—
straction is replete with material suggesting the paradox that
nothing is so full as the emptiness of a construct. The maxim
maker may define mathematics “as the subject in which we never
know what we are talking about, nor whether what we are saying
is true,”3 but obviously he cannot conceal his cap and bells.9 Why
Should an eminent mathematician10 Confound visible and palpable
things with things, or overlook the concreteness and actuality of
relations?
Formalism in logic and mathematics openly flouts the fact that
such creations as “zero,” “number,” and “successor” have been
used to demonstrate the reductiOn of mathematics to arithmetic
and arithmetic to logic. It seems highly anomalous to deal with
structures, especially such Obviously made and remade structures,
only to forget about the eEort and the plan responsible for their
production. To assert that mathematics is structural is entirely dif-
ferent from saying it reaches out toward empty and eternal forms.
a Bell, Development, p. 153.
\
Stebbing, Modern, chap. Io.
Nicod, Foundations, Part I.
aRussell, Mysticism,p. 75.
’This definition appears in an article designed to be “as romantic as possible,”
ibid., Preface.
1°
Eddington, Space, p. 14 n.
4 PSYCHOLOGY AND LOGIC

In the former instance it is tacitly acknowledged that mathematical

logic is system building, with setS'of relations as the building ma-
terial. Is it preferable to interpret mathematics as the science of
forms instead of the science of quantities? Yes, and here again
the work of formalization and system construction is evident. ‘To
describe or define mathematics as structure or system building,
however, does not eradicate the difference between mathematical
work and mathematical logic. The former involves many processes
besides system building.
The proposition that mathematics is logic, or reduces to logic,
is in itself an excellent illustration of abstracting and generalizing
operations. Note that the constructor creates the proposition by
abstracting all but a certain residue from the elaborate work and
materials of mathematics, and then legislatively generalizing that
the residue is the whole. The procedure in all respects is equivalent
to the creation of frictionless surfaces, perfectly iigid bodies, genic
determiners, sensations, and other such abstracted products.

CONSTRUCTIONAL PRINCIPLES IN
. A.
G. OPERATIONS
AND

In this work we are primarily concerned with abstracting and

generalizing designed to produce system-building materials. These
operations, of course, are only members of a general class. In all
cases the constructive feature, the development of a product, is
the outstanding characteristic. For objective psychology, abstract—
ing and generalizing Operations constitute instances of creative
imagination.11
Basic abstracting and generalizing operations can be very efiec-
tively studiedIn considering the Euclidean dimensionless point,
breadthless line, straight line, etc. Let us compare three interpreta-
tions of how abstracting and generalizing operated in the develop—
ment of these materials for Euclidean geometry.
Mac/z. As a confirmed empiricist within traditional philosophical
dualism, Mach stressed sensations as the basic factors of reality.

∙
Accordingly he attempted to derive all constructs (concepts) by a
process of emptying out sensuous Euclidean pointsvand
1’
Cf. Kantor, Principles, vol. II, chap. I8.
 ABSTRACTING AND GENERALIZING OPERATIONS ‘
5
lines, he argued,12 were end results of eliminating all other con-
tent from actual things. Thus, geometry as the science of space
must be concerned with bodies, but bodies from which all the
properties and characteristics are removed, until only a point or
a breadthless line is left. Geometry, therefore, is concerned with
ideal objects produced by the schematization of experientnl ob—
jects."
∙
i

Despite the fallacies inherent in Mach’s philosophy of sensa-

tions, consciousness and mind, he hit, in a genuine way, upon the
proper description of the evolution of the abstractive and general-
ized constructs Of geometry. Undoubtedly he clearly understood
the enormous cultural developments included in so elaborate a
humanistic system as Euclidean or other sorts of geometry.
Poincare. Armed with equipment obtained from a different
type of philosophy—namely, Kantian conceptualism—Poincaré
argued that a geometrical point iS in no sense derived from evacu-
ating concrete objects. Points are elementary and original ideal
elements, derived from rationally constructed systems 5 the process
of abstraction-is the isolation of the point element from the re—
mainder of the organization. The basic operation for the develop-
ment of points consists of intuition.
What Poincaré especially stands for is freedom of invention and
construction. For him, mathematical constructions are conventions
and conveniences of systematic organization. What Poincaré es-
pecially inveighs against is the opposite extreme—namely, the
purely formal and empty“ structures made up of logical proposi—
tions and symbols.
Strong. Both of the above statements concerning the basic sources
of geometric things, Strong believes, are derived from a theory
concerning geometry, rather than from a study of actual geometric
operations.” He suggests, that from the standpoint of actual mathe-'
matical operations, geometric points are relations (positions). In-
Book I of Euclid’s Elements, he argues, points are consistently
treated as positions, either indeterminate or determinate ones.
'3
Mach, Space, p. 48f.
Ibid., p. 68.
1‘
Foundations, pp. 41 3-485.
1‘
Strong, Procedures, Appendix.
6 PSYCHOLOGY AND LOGIC

Most appealing in Strong’s discussion is the emphasis upon 10-

cating Euclid’s abstractional processes in his actual geometric work.
This is tantamount to studying abstracting and generalizing as the
behavior of persons in specific situations. Certainly it is important
to study Euclid’s points, lines, and other constructions as derived
from his subject matter and procedures. Especially is this true
when the system elements, 'as products of work, do not match those
verbally and formally stated. But possibly Strong does not carry
this valuable point far enough. When we study an individual at
work, particularly when the work is of such giant proportions as
the systemization of a geometry, the interbehavioral field
for analysis.
out
cries
HowIncongruous is the system—building work of persons—their
actual operations upon materials—with glib verbal or symbolic
formulations. That there must be somethingin common between
the structure of sentences and facts6 is an eXtremely defective
proposition. Notice Euclid’s vacillation in his definitions, in the
first, for example, he declares. “A point is that which has no part,‘::
whilein the third he says: “The extremities of a line are points.”
∆ Heath17 indicates, Euclid was influenced by the-necessity both‘
of Organizing elements and refraining from violating the estab—
lished principles of logic and Science.
N ow the difiiculty with Mach’s and Poincaré’s discussions of the
creation of points and lines is probably not that they establish
meanings for these elements on the basis of their particular epis—
temologies, but that they attempt to make the latter valid for
Euclid. What they accomplish, then, is the explanation of wrong
epistemology. As creative workers, both Mach and Poincaré are
operating upon the basis of their characteristic intellectual traits,
and these traits inevitably Show in their work and products. Simi—
larly, Euclid’s epistemology influenced his work. At the same
time, however, he operated on a commonsense and objective basis.
Nonspatial mental states (sensations) or organized faculties did
not guide his constructions. Rather, he made his abstractions ac—
cording to the geometric knowledge and 'traditions at hand, and
in connection with the need to formalize and systematize them.
1'
Wittgenstein, Tractatus, 4.01, 5.05, 5.06, 5.12, ct passim.
1'
Thirteen, vol. I, pp. 155, 165, et passim.
 ABSTRACTING AND GENERALIZING OPERATIONS
7
Envisaged in this way, the inaccuracies and inconsistencies of his
∙ and generalizations
abstractions ∙ ∙ are easy to understand.
ABSTRACTING AND GENERALIZING LEVELS
Because all systems are specific it follows that particular ab-
stracting and generalizing operations are influenced by the type
of system being. evolved. More important, we need to emphasize
the work of particular individuals (vol. I), not only in a personal
capacity, but also as agents in the cumulative historical evolution
of cultural elements and systems.
Individual Operations. Individual abstracting and generalizing
operations may be schematized as linking the crude data with
which the worker begins and the constructional products repre—
senting the results of his action. This is indicated in the following
schema.
Construction Constructional Material worked upon
(Product) Operations (Crude Data)
The operations always may, and sometimes must, be organized
in a hierarchical scale of stages or levels.
The first or most elementary stage is a simple form of selec—
tive reaction with respect to things, the assumption of an attitude,
say, concerning some sort of material. Next in the series is the
descriptiOn 'of the selected event as a phase of the worker’s orienta-
tion toward it. Then follows the fixation of this description in
some linguistic or symbolic form. Obviously, the interbehavior
becomes increasingly abstractive, in the sense that the selection and
description may be regarded as definitely constructional.
As this abstractional process progresses it is possible to manipu—
late the descriptions, whether elementary protocols or highly
formulated laws, with an increased independence of the original
materials. To be sure, there are distinct limits to this increasing
remoteness from original contacts with things. In system building,
illimitable abstraction is a cardinal error leading to utter Sim—
plicity, mystic absolutes, ~tautology, and all the paradoxes of
mathematical infinites. Again, the individual may construct gen—
eralizations by analogy or other techniques,.and in diverse styles.
At the same time, the worker, in proteeding, should check his
generalizations by referring to the original data. These are propo—
8 PSYCHOLOGY AND LOGIC

sitions few will attempt to dispute, but the various attitudes tO—
ward the checking criterion are another matter. Within recent
years, since the advent of relativity and quantum mechanics, even
physical scientists question the existence of an objective world—
that is, events independent of the scientist’s descriptive and inter.—
pretative constructions (p. 3I2f). In paper after paper, in book
after book, physicists discuss whether the events with which they
deal are anything but sensations.18 Such discussions always skirt the
abyss of meaningless problems.
It is hoped that the interbehavioral position adhered to in this
work may serve to illuminate the character of the criterion by
which we check our constructions. Recall that the physicist, as well
as every other intellectual worker, operates in a definite reference
field. It is very easy, then, to determine whether or not he is in—
fluenced by presuppositions describable as cultural influences or
simply by local school traditions. As pOinted out in Volume I
(chaps. 5, 6), any notion that things and events are reducible to
sensations (psychic constructs), or that scientific propositions are
based upon sensations, can be definitely traced to the dualism
hampering scientific thinking. In this sense we can'identify the
fluences leading such a physicist as Mach to build upon Hume’s in}.
formulations, or the mathematician Poincaré to retain the conven—‘
tional conceptualism derived from Kant.
Cultural Abstracting and Generalizing. The evolution of the
number system provides the most efiective sample of the pyramid-
ing of individual abstracting and generalizing operations. The base
of the pyramid, obviously, is the series of positive integers. Per-
haps the next few levels should include the negative integers,
ratios, fractions, irrational numbers, and zero. Disregarding exact
chronology, we may point out another series of levels comprising
literate numbers, complex numbers, and ideal numbers. At the
apex, as of our time, we might place the numbers belonging to the
transfinite domains (p. 23). Clearly, these levels represent not
only possible temporal successions but also degrees of abstraction.
Only the most elementary levels of numbers are abstracted from
things. For therest, abstractionS‘and generalizations are produced

1'Houston, Philosophy; Eddington, Philosophy; Dingle, Through, Lenzen, Na-

ture, Bridgman, Nature.
 ABSTRACTING AND GENERALIZING OPERATIONS 9
by means of constructive procedures, occupying varying degrees
of remoteness from original Objects.9
Scientific and intellectual histories constitute, for the most part,
narrations concerning the evolution of interrelated items forged
through longer or shorter time periods, these items in their com-
pleted stages comprise cultural institutions. The history of logic
consists primarily of instances of cumulative abstractions and gen-
eralization growths. A noteWorthy example is the development
and elaboration of mathematical and symbolic items in modern
Formal Logic.

THE INTERRELATION OF A. G. OPERATIONS

AND

Our exposition so far has indicated that abstracting invOlves the

manipulation of concrete or abstract stimulus objectsIn order to
analyze out certain factors. Generalizing, on the other hand, is
primarily the creation of a product based upon concrete or abstract
materials either previously analyzed out of a continuum or ap-
proached in their integral character.
Despite variations in detail these operations are often very,
closely interrelated, a fact well exemplified in the building of
philosophic systems. For instance, by a series of abstractive pro-
cedures persons are separated on the ground of their varying in-
terbehavior with visually presented objects, say, variously colored
or shaped things. Then, by similar techniques, variations are em-
phasized which are based upon differing relations between persons
and obj ectsf Other abstracting procedures yield sharp differentia—
tions between the perceiving individuals and the Objects per-
ceived.
Next, a number of generalizing procedures are brought into
play. First, the relativity of perceiving is created. There are color-
weak and color—strong perceivings. Thus an inevitable diversity
between perceiving and perceived is set up, along with discrepan-
cies betWeen the relations of perceiving acts and perceived objects.
By devious creative generalizations propositions are formulated
concerning a possible real world different frOm anything Observed.
In brief, such conjoint operations are typical Of‘all speculative
theologies and metaphysical philosophies. The term metaphysics
1"The science of pure geometry is generally regarded as having been evolved
from) the humble source of land-measuring techniques. See Nagel, Formation.
IO PSYCHOLOGY AND LOGIC

itself suggests the historical development of the coordinate Opera;

tions of abstracting and generalizing. In Aristotelian times meta-
physics was simply the Study of the generalized aspects of events
included in physics and other distinct sciences. Later, unfortu-
nately, metaphysics became the exercise field of those who created
systems without much, if any, regard to actual events.
But not only the illegitimate metaphysical domain is rich in
illustrations of the conjoint operations of abstracting and gen:
eralizing behavior. The same situation prevails in every complex
system—building situation. It is therefore to our advantage to con-
sider the Specific characteristic of each operation separately;

SPECIFIC CHARACTERISTICS OF ABSTRACTING OPERATIONS

Abstracting means formalizing. In calculating we must abstract -
from the specific characteristics of things. In mensuration the units
must be taken as qualitatively indifferent, numbers are viewed
as ordered points without regard to what they specifically repre-
se.nt However, abstracting itself is a concrete operation, and‘the
abstractions are products of such operations. While at any moment
we may say it is indifferent what particu1ar things we are calculat-
ing or measuring, there could be nooperations without materials
to work upon. That these materials are abstruse and relational
enforces rather than weakens the point. In all such abstract situa-
tions we merely interbehave with relations between events, rather
than with events, or we might say we operate upon relational
events instead of upon Obj ects.
The fact that most behavior is really abstractional proves that
abstractions are concrete. In everyday speech when we. refer to
things we already are selecting certain features out of a compre—
hensive continuum. Similarly, every act of perceptual discrimina-
tion involves not only one but several such abstractional proced-
ures. First, we select a certain object from its background and
related obj ects, then we interbehave with the Object in a specific
manner determined by the immediate reference frame or field of
action.‘
Because we are unduly impressedby the product of our abstract-
ing behavior (p. I), once we develop a symbolism to represent
the results of our operations we are apt to consider oUr symbols
 ABSTRACTING AND GENERALIZING OPERATIONS II

simply as records of actions, irrespective of the objects manipu—

lated. Words or symbols, however, are products of interbehavior
with things, not the results of a creatiOnal process independent of
those things.
To a great extent, abstractional behavior involves isolating
operations. In many cases, to observe what is going on we separate
things or eve‘nts from the matrix or field in which they are inevit—
ably imbedded. Frequently the isolated datum is itself a system
interrelated with Other systems. How difierent is the isolating
procedure when we are interested in a particle moving in a straight
line without acceleration and when we study a particle in acceler—
ated motion in a circular path.
Physicists20 raise the question whether, in View of the impossi—
bility of isolating a system in nature, we should not refer to the
process as focusing attention upon a body or a System. Despite the
utterly metaphorical character of this alternative suggestion the
interbehavioral feature of the situation is stressed.

_ TYPICAL ABSTRACTING INTERBEHAVIOR

Abstracting operations vary as the performances of individuals
and according to the specific character of the things with which
they interbehave. Illustrative classes called Divisive and Multipli—
cative- may be described as follows.
Divisive Aestractions. Because of our general cultural circum-
stances the divisive or analytic type of abstraction is probably the
more familiar. With our tendency toward simplification as a
means of carrying out various activities we look upon abstractions
as factors analyzed out of larger complexes. In physics, for ex-
ample, it seems natural to reduce mechanical events to space, time,
and mass. As a matter of fact, the notion is exceedingly common
that space and time are inevitable features of all happenings. While
we know that this attitude is merely a sign of our cultural cir-
cumstances, and, that with other propensities developed, we could
just as well make other factors fundamental, this is still an excel-
lent illustration of the divisive or reductive type of abstracting be-
havior. As we proceed from one situation and one scientific field
to‘another the instances multiply.
” Dadourian, Force.
12 PSYCHOLOGY AND LOGIC

The hierarchical arrangement in the conventional classification

of the sciences is a good example. It is Often asserted that psycho—
logical science reduces to biology, biology in turn to chemistry and '

physics. But these two sciences are not always regarded as end
points in this reductive procedure. Physics is made basic to chem—
istry, this discipline, in turn, is shaken down to mathematical rela—
tions. Nor does this terminate the process, Since mathematics is
often cut down to logic. Current conventions limit the fecundity.
necessary to reduce logic to something else.
Multiplicative Austractions. The operations we call multiplica—
tive abstraction occur when we not merely analyze given material
but‘immediately add interpretative components. The operations,
in other words, are constructive. We are reminded here of those
continuity physicists (Mach, Ostwald) who fifty years ago denied
that physical events could properly be analyzed into atoms. For
them the basic and fundamental data of physics° consisted of energy,
and energy was regarded as continuous. The atomists, of course,
constructed their analyzed factors, and, though physicists could not
have foreseen recent developments in quantum theory, we might
trade an unbroken line between atoms and energy quahta as ab—
stracted units endowed with more elaborate characteristics.
Every scientific field offers numerous illustrations of construc-
tive abstractions. Consider the multiplicative character of the ab—
stracting operations that produced “sensations” in traditional psy—
chology. Sensations were not only the atoms of the mind, but dif—
fered from other kinds of atoms depending upon how many types
there were in the system. For Wundt sensations contrasted only
with feelings. Other psychologists multiplied sensation .proper—
ties to contrast with images and with will elements also. These
imaginary sensations were even endowed with spatial attributes!
The entire process of abstractive behavior is neatly revealed in the
story of how elements of “consciousness”
can be provided with
spatial properties. But even these extravagantly multiplicative
creations count as nothing beside the creativity of physicists who
make sensations into the fons et origo of the material cosmos.
The multiplicative character of abst‘raction products is excellent-
ly shown by observing how they evolve in the first place. Take
the abstraction stimulus as it developed in psychology. Originally,
 .ABSTRACTING AND GENERALIZING OPERATIONS I3
the term was taken over from biology to refer to some condition—
for example, light rays—which incited the action of a receptor
(retina), and finally resulted in the appearance of the sensation in
consciousness. Then the property of inciting an organism to act, as
in performing a motor response, was added to the sensation. When
psychologists became interested in the behavior and adjustments of
organisms to‘ their surroundings, the properties of stimuli were
again increased to include Obj ects. Still later, the stimulus abstrac—
tion became modified in order to take on properties of field factors.
Another good example of the multiplication of abstractive'prod—
ucts centers around the changes in such products as matter and
form. Matter, as abstracted from an art object, is sometimes pre—
Sumed to have no properties until impressed with form by the
artist. But it is asked: “Had the block'of marble no form at all
when it came out Of the quarry?”21 Mill has pointed out that the
word form should refer only to bodily figure. Beyond that every
usage became vague and arbitrary.

SPECIFIC CHARACTERISTICS OF GENERALIZING OPERATIONS

Whether or not generalizing operates as a correlated process
with abstracting, it is always a specific interbehavioral procedure
within a particular reference frame or coordinate system. The
frame of reference marks ofi a field of activity conditioned by
certain purposes and problems. A typical illustration is that in
which objects, relations, or events are constructively magnified into
something more important or pervasive. Let us choose examples
from everyday life, as well as from science and mathematics.
One takes a particular act performed by a person and general-
izes it into a typical trait. This procedure is consonant with the
process of creating abstract matter from specific qualities of things
observed, or generalizing observed microsc0pic events as typical of
microscopichappenings. 'Similar illustrations abound in every de-
partment of intellectual life. Through the controlled observations
of repeated events scientific laws are created by generalizing proc-
esses. Consider that interesting type of generalization which be-
gins with the specification that a thifig is either A or not A, that is,
it 'haskor has not certain qualities, properties, dimensions or rela-
\

21 M\ill, Examination, vol. 2, p. 139.

_ 14 PSYCHOLOGY AND LOGIC

tions. This fact is generalized into a so-called law, that of ex-

cluded middle (chap. 20). As a statement of the factual condi-
tion that a certain description or classification is satisfactory this law
is unobjectionable. Next, the law is generalized, as in Aristotle’s
sea-fight example.22 The necessity that something should or Should
not take place is, of course, a verbally constructed necessity.
From interbehavior with spatial events we abstract Shapes or
line structures, or we regard space as merely an order of triplets
or point representatives. In these cases the emphasis is upon cer—
tain results, but these results 'are always derived from very
definite interbehavioral operations. Generalized algebra mayibe
said to consist of the manipulation of a set of derived symbols in
order ‘to achieve a greater degree of abstractness for certain pur-
poses. But the history of all generalizing operations 'proves that
the most abstruse results can only be reached through step-by—step
operations. Indeed it would be fatal to overlook the small inter-
behavioral steps.
Generalization is an inevitable and indispensable feature of all.
human operations. Without it there could be no prediction, no
commerce with the future, no expectation. While interbehavioral
psychology cannot tolerate faculties or powers, in a definitely
practical sense there is a generalizing propensity. In other words,
generalization is the basis for all organization of objects and
events.
In science a generalizing form of interbehavior results in the
production of laws which not only yield consistency and certainty,
but allow for the interrelationships of interbehavior over ranges
of time and place. Yet unless we regard generalization as a definite
interbehavioral process, dealing with specific stimulus objects, we
run into all sorts of metaphysical and other insalubrious predica-
ments. Our formalizations and various generalizing processes,
such as induction, deduction, etc., must be performed with strict
regard to available stimulus objects.
But how far can we go in our generalization? What are the
limits? Undoubtedly, a strict line can be drawn between autistic-
constructions and serviceable principles\which imply respectgfor
observable facts. The question is: How far can we extrapolate from
.actual contacts with events? To be sure, we\ have a completely
Cf. vol. I, p. 160.
 ABSTRACTING ANDGENERALIZING OPERATIONS 15
valid test of our extrapolations when we employ the’ criterion of
Specific interbehavior. It is no great task to differentiate between
extrapolation which generates and furthers our contacts with spe-
cific existential things, and extrapolation consisting of autistic con-
structions which implement the creation of ultimates and absolutes.
In connection with the latter type, logicians have built up a formal-
istic cult. By abstracting from subject matter they come to the
point where they emphasize form. Form, then, becomes material
for erecting all sorts of structures. In-itself this sort of extrapola-
tion is a legitimate kind of system building. When regarded as
significant for events, however, it results in futile attempts at
application.

TYPICAL GENERALIZING- INTERBEHAVIOR

Since generalizing interbehavior is continuous with the abstrac-
tional processes of all human enterprise it is impossible to indicate
all its different types. We can only illustrate, therefore, some
prominent instances.
Analogizing. Numerous generalizations are constructed by
pointing Ont or establishing similarities, correspondences or analo—
gies between One type of Object, relation or event, and another.
Note that in the constructional procedure the analogizing reactions
share equal prominencewith the characteristics of the stimulus
object. If COmparisons are stressed, it is done on the basis of the
leads provided by the things compared. The essential creative per—
,formance consists of carrying over or generalizing similarities or
congruences.
The typical analogizing processes of mathematics illustrate such
operations. Temporal succession, taken to be undimensional and
continuous, is analogized by the interminable succession of integer
series. Thus algebra becomes the science of time. Three dimen—
sional space, with its symmetries, is constructed into geometrical
science. Similarities and correspondences in groups of things be—
come the raw materials for the creation of classes.
Scientific models (chap. 16) of various sorts constitute similar
illustrations. It is said that a Scientific\model constitutes a substi—
a
tute or replacement of part of the universe'by. Simpler and simi—
lar thing:23 At any rate, the Rutherford—Bohr model of the atom

Rosenlilueth and Wiener, Role, p. 3 I 6.

2'
16 PSYCHOLOGY AND LOGIC

is merely the adaptation of the solar-system analogy to the rela-

tionship between protons and electrons in an atom. Other promi-
nent analogical generalizations are the development of electro—
dynamics on the basis of classical mechanics, and Ehrenfest’s
Adiabatic Hypothesis for eXplaining the alteration in a State of
motion in an isolated mechanical system. This last hypothesis is
directly borrowed from thermodynamics.“
The following schema efl'ectively suggests the analogical basis
for generalizing constructions:
Corresqondence Noncorrespomlence
I
I l
Identity Similarity Difference Contranety
(complete (partial (partial (complete
analogy) analogy) polarity) polarity)
Magnification. When the generatidhs of classical theologians
built up their theistic conception of omnipresence and omnipotence
they were simply enlarging and modifying what they regarded
as the most desirable personal qualities and properties.
Islamic construct of the universally merciful and compassi‘onatet
The‘
Being represents an expansion and glorification of man’s Superior
virtues. Generalizing by magnification is to stress stimulus ob—
jects, though the response factors can not be minimized or slighted.
From scientific domains we may gather Such examples as “mathe-
matics is the queen of the sciences and arithmetic the queen of
mathematics,” also that physics is the basis and foundation of the
other sciences. _
A more specific instance is the glorification of mechanics in phys-
ics, or of the conductive and coordinative processes of the nervous
System in neurological and psychological circles, to such an extent,
in fact, that these processes become generalized as the
points in the biological and psychological economy.
central
Reiteration. Such abstractive constructs as the mathematical
infinite and infinitesimal (p. 23) display a prominence of respon-
siveness'rather than stimulus objects. The infinite is developed by
recursive acts of addition, whereas the infinitesimallS abstracted by
subtractive action. ∙

Reiterative or recursive responses are illustrated by\ starting

Sommerfeld, Atomic, p. 341f.
 ABSTRACTING AND GENERALIZING OPERATIONS I7

with zero and indefinitely repeating an additive act to reach the

mathematical infinite. Going in the opposite direction by repetitive
subtractional steps leads from minus one to the infinitesimal. Prob—
ably a still better illustration is the Archimedian exhaustion pro—
cedure. The steps taken to reduce a polygon to a circle indicate
the process of repeating an action to reach a definite limit.
Repetitive behavior is a decided source of abstractional and gen—
eralizing products. For example, instants of time are reached by
dividing a continuum, or an infinite duration, and continuously
enumerating the items to a named but nonexistent stopping place.
Interpolation. The grand—scale form of interpolative operation
is found in the generalizing creation of a mathematical continuum
by interpolating a point between any two given points. Interpolat-
ting construction constitutes an elaborate manipulation of stimulus
obj ects in two directions. In the first place, the selection and specifi-
cation of interpolated items suggest a separative instead of a gen-
eralizing procedure. On the other hand, the continuing process'of
placing items between given items enlarges the interpolative series
to form an expanding system. The progressive filling in of parts
to expand the whole surely works toward extensive abstractive
generalization.
Smaller-scale interpolative operations take the form of work—
ing out all sorts of proportions and fitting in logarithmic values.
Keeping actual procedures before us, we will not betoo greatly
influenced by the fixity and prior determination of end points,
and shall thus avoid minimizing the creating and CXpanding op—
erations.
Extrapolation. The astronomical domain prOvides great scope
for the exercise of extrapolative processes. Begin‘ with the earth
or other planet of the solar system, and continue to~enlarge the
system to include all the nine planets and intervening spaces. The
whole solar system, however, is continuous with the galactic sys-
tem in which there are many other members. Galactic systems, as
members of larger systems of star groups in space, afford the means
of generalizing a definite and someWhatJimited astronomical uni—
verse. Those who do not fill out their systems beyond fairly def—
inite data nefrain from transcending scientifically. legitimate bonds.
Others, \mo‘re speculatively inclined, continue to \extrapolate be—
18 PSYCHOLOGY AND LOGIC

yond any possibly observable or calculable system, and thereby

attain systems of universes which can only be verbally proposed
and managed. .
Extrapolative generalization, therefore,18 not limited to data
or things, but involves a preponderant emphasis of acts and op-
erations. As such it is a more versatile and facile type of generaliza-
tion behavior. Statistical tables offer a telling example. Consider,
for instance, the responsible and irresponsible creation of things,
references, and propositions way beyond original data.“5
_
Comeined Generalizing Operations. Convenient though it be to
treat generalizing operations as separate and independent pro-
cedures, in all complex situations a number of them are combined
‘to bring about the final result. Consider such a procedure as
Whitehead’s23 method of extensive abbstraction. To employ the
principle of convergence to Simplicity requires more than abstrac-
tion and creative generalization; the transformation of observable
areas into unobservable points involves magnification, interpola-
tion, extrapolation, and other generalizing procedures. .
Outside the field of mathematics this combination of operations
'is well illustrated by the causal theory of perception.27 The‘ fact»
of seeing the sun is analyzed into an object, the sun, which is then“
considered as the first link in the causal chain. It is assumed that
the sun exists prior to seeing it; then radiation requires eight min-
utes to reach the eye. This radiation is set up as an independent
item. Also, as parts of the causal chain, there are physiological
effects when the radiation impinges on the eye. Finally, the efiect
is described as “seeing the sun.”

LINGUISTIC AND SYMBOLIC TECHNIQUES"

IN A. AND G. OPERATIONS

a
It is commonplace that all complicated intellectual enterprises
necessarily involve linguistic and symbolic techniques. But, as we
have seen,28 the precise manner in which linguistic and‘ symbolic
things and acts operate in logical situations is not so clearly. recog-
nized. This condition undoubtedly may be traced back to a lack of
adequate views concerning the nature of language and symbolism.
2'
Cf. Ginsburg, Finite.
Concept.
:1 Russell, Reply, p. 703.
'3
Vol. I, chap. 8.
 ABSTRACTING AND GENERALIZING OPERATIONS 19
These linguistic events are only very Slowly being appreciated as
essentially interbehavioral in character. Even those writers alive
to the behavioral and interoperational nature of language fail to
differentiate between language operations and the products of
those operations. Thus, to work out the place of language and
symbols in abstracting and generalizing we must separate referring
from denoting action, and these in turn from selecting and en—
larging behavior, as well as from any products of the conjoint
operations of all.

INTERRELATION OF LINGUISTIC AND NONLINGUISTIC

BEHAVIOR IN A. AND G. SITUATIONS
Simple abstracting and generalizing behavior, and the early
stages of its complex forms, may be accurately described as non-
linguistic or, certainly, nonsymbolic. It is nonlinguistic in the sense
of forming or building up intellectual attitudes or bringing to bear
such previously developed attitudes upon present problems.”
These attitudes may be simple cognitive orientations based upon
observed locations and comparisons, as between qualities, proper-
ties, shapes, and relations. Hume furnishes an example. Despite
his 18th century ideas of mind, he describes effectively the ab-
stracting of color and figure from successively observed white and
black globes and cubes.”
Among the simplest linguistic responses in abstracting and gen-
eralizing situations are those acts of referring to the fact that white
and black, and sphere and cube, are variously distributed and
interchangeable things, shapes, and qualities. Following the ab-
stractive separation of factors there may be referential responses
concerning the independence of qualities and things and of their
generalized existence.
The acts of stabilizing and establishing names, and treating
shapes and signs, are productive of symbolic types of language
things. Various processes and stages may be noted here. In one
case vocal referential acts, because they result in sound patterns,
constitute potential creative acts When references become word
names or symbols standing for things and qualities. In other in-
stances substitute things, in the form ofmarks, signs or sentences,
\

2°
3°
Cf. Xantor, Principles, vol. II, chap. 20.
Treatise, Bk. I, sec. 7.
20 PSYCHOLOGY AND LOGIC

are deliberately created for abstracting and generalizing opera-

tions. At this level Recorde’s creation of the equality sign “bicause
noe. 2. thynges can be moare equalle” is instructive."’1

THE ROLE OF LANGUAGE

A. IN G. ‘SITUATIONS
AND

Without all sorts of linguistic acts and products it is impossible

to deal effectively with attenuated and distant things. Hence all
complex abstracting and generalizing situations involve many
linguistic factors. The role of language is primarily to substitute
for absent and created things. Such substitutive roles difier accord—
ing to whether they: (I) represent other things, as happens when
we separate ourselves more and more from the objects and rela-
tions with which we have been in contact, or (2) constitute newly
created things. Extrapolative operations, forexample, may simply
extend a bit some definite manipulative process or carry it out
beyond any possible manipulation.

CRITERIA AND LIMITS FOR _

' A. AND G. OPERATIONS
The principle that abstracting and generalizing operations in-.
volve the manipulation of things in specific Situations suggests
both criteria and limits for such activity.
Criteria. First and foremost is the criterion of specificity. What
is desirable or permissible in one Situation is not so in another.
Furthermore, within any given situation the amount and char—
acter of abstracting and generalizing depend upon the material
and the significance attached to the procedure. In other words, the
specificity of criteria for all operations curbs the freedom to ab—
stract and generalize.
Consider Russell’s assertion that (x—l—y)2 = x2+2xy+y2, even
if for x and y we substitute Socrates and Plato.32 Invoking a cri—
terion of Significance we ask: what intellectual value does Such an
analogy have? In what way does this illustration illuminate the
problems of mathematical abstraction, generalization, and system—
ization of relations? Instead of emphasizing thatIn mathematics
'1
Mitchell, in Young, Lectures, p. 236.
i

Principles, p. 7.
 ABSTRACTING AND GENERALIZING OPERATIONS 2I

the symbols must be very precise in themselves and also most

meticulously structured into equations, the impression is given
that mathematics is arbitrary and capricious. Let us remember
what rigid limits are set for mathematical operations; for ex—
ample, no division by zero, a zero product of two nonzero factors,
but no zero product of two nonzero factors.
Russell is really implying that logic is a force governing all
other types of intellectual activity. Hence his view:
. in every proposition of pure mathematics, when fully stated,
the variables have an absolutely unrestricted field: any conceivable
entity may be substituted for any one of our variables without impairing
the truth of our proposition.33
Compare this with Hardy’s declaration:
It is impossible to prove, by mathematical reasoning, any proposition
whatsoever concerning the physical world, and only a mathematical
crank would be likely now to imagine it his function to do so.“
Limits. How limited abstracting and generalizing operations
are and howrestricted by criteria is readily observed when con-
sidering such an equation model as .a b ∙−− ∙ It is significant to
+
ask if c gives us something new, whether or not‘it is a combina-
tion of a and b, or nothing at all. Further questions are: Is b
different from a? Or is b a factor which negates a? Finally, do
we have a simple conjoining of a symbolic representation Of an
equation, with the assumption that the terms symbolize some
things from which significant relations can be abstracted?
Moreover, we face the question of the limits of partition and
exclusion. How far Can'we divide a subject matter or a series of
stimulus objects? Illicit abstraction is the result of striving toward
some kind of absolute and final product, such as metaphysical cer-
tainty, irrefragable law or ultimate truth. Any analysis which
yields cosmically tranScendent values involves the injection of
arbitrariness and caprice into our Operations, as illustrated by ab-
stracting from concrete events in order to reach absolute space or
time. For instance, how significant is the’extrapolation indicated
in the following quotation?
3°
Ibid. p.~7.
"Theory, 11. 402. Cf. Russcll’s own strictures upon the Formalists; Principles,-
Introduction.
22 PSYCHOLOGY AND LOGIC

For it seems certain to me that the extent of hidden organization in

our universe is infinite, outside as well as inside of space and time; such
a conviction is very natural to a mathematician, Since the three ordinary
spatial dimensions and the single temporal dimension are for him only
particular instances of infinitely many other conceivable dimensions! 3‘
This is unrestricted and uncritical generalizing par excellence.
Concrete space, as the place or relation of objects, is generalized
into tridimensional abstractions; then, of course, the procedure
can be further continued to any number of dimensions. In the
second step one retreats from original interbehavior with con—
crete events to move on to operations with dimensions substituted
for by numbers or letters. A third step takes one away from Space
and time altogether. Henceforth no limits are set to mystic flights.
Since generalization is essentially more substitutive than is ab—
Straction, the necessity to check the procedureais even more urgent.
Particularly, one needs to place distinct limits upon the creation
of universes. A good remedy is to keep projection within bounds
and to remember that whatever universes one creates must‘be
derived from interbehavior. ‘
If any argument is needed to strengthen the principle of con—
‘trolled abstracting and generalizing, we find it within_the opera-
tions of formalistic logic itself. Logicians have perforce been
sharply checked in their work by such problems as the complete—
ness and consistency of deductive systems. The theory of types
also symbolizes a fundamental problem of generalization limits.
Whatever one may think of the Cretan liar as a problem in logic,
it reinforces the need to specify precisely what situation accom—
modates any particular assertion or proposition. We miss the
point of these problems, however, if we overlook their basic be-
havioral core—namely, the life conditions of logicians.

ILLICIT A. AND G. PRODUCTS

Abstracting and generalizing" operations are appropriate and
Useful in the precise measure in which they correspond to the
things interacted with in specific situations. Once they surpass
such restraining bounds they result1nillicit products.
The leap from counting and enumerating operations to the
'3
Birkhofi, Infuition, p. 608.
 ABSTRACTING AND GENERALIZING OPERATIONS / 23
creation of infinites and transfinites (p. 7) supplies us with a
vivid example. From concrete computational processes to asser-
tions concerning mystic entities is, after all, a very short path.
Calculators and mathematicians constantly find themselves con-
strained to eXpand their operations either toward the large or the
small. But instead of keeping their own activity in the foreground,
and observing themselves in particular frames of action, they
move on toward entities which they fail to recognize as created
in the image of their own behavior.36 Summing integers and frac—
tions leads to sums to infinity.
Similar illicit creation yields such products as infinitesimals.
The mathematician begins with a small abstractive system, for
example, a convergent series. Multiplying his fractions by definite
interbehavior with integers—namely, fractionating them—he
then jumps to the conclusion that an infinite number of fractions
exist. Naturally, by continuing his fractionalizing operation he.
can go on indefinitely. Limits, Obviously, would be placed upon
him by his own continued vigor and life span, but he can be sure
that others can carry on the process over long periods of time.
Observe, nevertheless, that the infinitesimal is really only a
name for a process. When dealing with concrete events this process
is very soon ended. Even if we multiply the factors in an atom
to obtain protons, electrons, neutrons, neutrinos and mesotrons,
the process ‘soon stops. The boundary between actual procedures
and mystical creations is very Sharp, which is not the case when we
operate with abstract relations.
Because mathematicians have long rejected the spurious con~
struct of aninfinite number it is strange they do not ponder more
the techniques which engender such counterfeit products. That is,
they do not observe that such operations as matching, coordinating,
and synthesizing constitute genuine creative acts. To do so would
effectively avoid the interpretation of the infinite as any sort of
ontological essence or pure form of thought. Only one conclusion
is available; the infinite is the name for an iterative, long-con—
tinued action, or a name for the name of that action.
"Such behavior includes linguistic performances. We should not, however, fail
to distinguish between references to, and symbols for, Operations and the basic opera-
tions themselves. Unrestrained operations are not just faulty .uses of language, but
maladjusfrnents.
24. PSYCHOLOGY AND LOGIC

Among the notorious illicit'nonmathematical products stands

brazenly the universe, which, in Hegelian—Bradleyan logic, every
proposition implies. This item furnishes an
excellent example of
uncontrolled reiterative processes (p. 16). The actual operations
are primarily verbal, of course, though the illicit products are re-
puted to be independent existences which, for the- constructor,
afford venerated stimulus Obj ects.
Aside from disregard of constructional limits, illicit products
stem, too, from a neglect of the specificity principle. How else
account for the unqualified-proposition that the whole is greater
than any one of its parts, or that-the whole is greater than all of
its parts? What is meant by greater is of fundamental importance
and can not be neglected. In other words, the operator can not
justifiably abstract from the unique frame of referenCeIn which
he works.
As comparable examples of improper generalization we may
cite the basic theorems of material implication. While these theo-
rems may be satisfactory from the standpoint of the arbitrarily de~ -
fined situation of symbolic logic, the general principles involved
are valid merely by CXpress agreement. In other words, they hold
only in a system constructed and accepted by symbolic logicians.
Deceptive abstractions in scientific and other ‘mor'e'concrete do-
mains than formal logic and mathematics can be traced in large
measure to unrestrained extrapolation. Consider those wild
speCulations concerning the cosmos, based on far—flung extrapola-
tions of data derived within the narrow confines of the laboratory.
Numerous writers37 have protested against extrapolating from the
second law of thermodynamics that the universe is heading toward
its “heat death.” Such extrapolatory behavior we must characterize
as inept and misleading; it indicates moreover\ that the worker
is motivated by sentiment and mystic inclination, and not by an
obj ective estimation of events.
On a general philosophical level writers have frequently
noted the obvious fact that much of our information13 second
hand and not derived from direct contact with things. James dis—
tinguishes knowledge of acquaintance. and knowledge Igaboutf’3
"Dingle, Through; Bridgman, Nature, Cosmical Inquiries, Statistical; Lotka,
Probability-Increase.
” Principles, vol. I, p. 221.
 ABSTRACTING AND GENERALIZING OPERATIONS 2.5
while Russell separates knowledge of things by description from
knowledge by acquaintance or actual eXperience.39 With the aid
of numerological constructions, such as classes having the same
number without reference to any particular number” and pairs
of unknowable integers whose product is known to be greater
than 100, the mystic urge can amply assert itself. It is this urge
which directs all questing for a priori principles, for undiscovered
knowledge, and the Absolute, without which many thinkers find
the intellectual way uncharted and perilous.“1
Also to be numbered among the techniques resulting in unde—
sirable constructional products is improper analogy. The Starting
point may well be authenticated facts. Take the case of malforma-
tions which comprise the one significant and general organic dif-
ference between persons behaving normally and abnormally. By
means of uncritical analogy such malformations are used to jump
to the conclusion that behavior differences in normal persons are
likewise to be accounted for by essential structure variations. And
this regardless of the fact that such structural variations are both
entirely unknown and psychologically irrelevant.42
The examination of illegitimate intellectual products indicates
that they can not always be associated with separate techniques.
Many arise from combined processes. To illustrate this point we
list a sample series of such products, classified as primarily ab-
stractional 'or generalizational.
Primarily Abstractive Products. Prominent here are those quali-
ties, essences, and relations abstracted from objects such as:
Numbers, as autonomous invariant relations.
Sensations, sensa, or sense data, abstractionally created from colors, sounds, etc.
Square circles, griffins, combined from unrelated things.
Average man, economic man.
Speech or reason as basic properties of man.
Abstractions equated with object source or other concrete totality from which
they are derived.

Primarily Generalization Products. Such logical fallacies as non

seguiturs and other improper conclusions Should at least be men-
tioned here, if not placed at the head of the list. Less formal ex—
Problems, pp. 92, I 69 et passim.
Eaton, General, p. 466.
" COhen:‘Reason, p. 14.6.
See \ISantor, Problems.
‘2
26 PSYCHOLOGY AND LOGIC

amples comprise the magnification of a small experiment and its

data into a huge theOry. The following roll illustrates the entire
series:
Color and sound can be correlated with frequency, hence color and sound are
frequencies.
Color is measured with difficulty, hence color is mental.
Personal or private knowledge is subjective, hence mental.
Because some psychological activities constitute conditioning, this is true of all
psychological events. __ 3

Because statistics is a necessary procedure in treating data, it constitutes the scien-

tific method.
Logic deals with deductive systems, 'hence all systems are deductive.
Because mathematical things (relations) are more abstract than other things,
mathematicsis reducible to logic.
Because one theory'15 faulty its opposite is correct.
Because act or action, power to act (instincts, forces).
Begin with characteristics or properties and end Owith principles or powers.

Concluding our discussion of spurious abstracting and, generaliz-

ing operations, we mention that such unrestricted behavior does
not always lead to illicit products. There are striking instances in
which apparently autonomous and inapplicable abstractions turn
out to be not only useful but important in particular situations. 'We
Have referred to the development of negative, irrational, and
imaginary numbers which followed the sheer formalization of
mathematical operations (p. 8). These products of unrestrained
generalization were finally put to essential mathematical uses.
The stock example of assimilating the square root of minus one
into the mathematical theory of electromagnetism not only indi-
cates the close integration of formal and nonformal situations, but
also the regulation and justification of abstractive behavior by con-
crete and specific adjustmental circumstances.

A. AND G. OPERATIONS IN SYSTEM BUILDING

So far we have not stressed the place of abstracting and gen-
eralizing operations as factors in logic or system building. iOn the
contrary, we have been discussing these operations merely as com-
plex forms of intellectual interbehavior. Their location in system
building is a simple matter, however.
In the first place, abstracting and generalizing operations are
such widespread activities that they can hardly be excluded from
the building of even modestly complex systematic structures.
 ABSTRACTING AND GENERAIiIZING OPERATIONS 27
Secondly and conversely, system building, from the simplest to
the most complex, always involves such selection and analysis
of units and groups 'as to demand abstracting and generalizing
processes.
Their systemological aspects, therefore, present no special
problem. Let us briefly point up some of the details. An outstand-
ing question in organizing specific items into structures concerns
the relative concreteness of any item. When concrete things are
manipulated there occurs elementary behavior of discrimination,
selection, and separation, with some generalization when units are
fitted into groups, series, or orders. Abstract things require more
remote handling, and hence substitution by linguistic acts and
things, since abstract things have been produced by prior handlings
of a direct and elementary sort. Then there are such extreme ab-
stracting processes as the'complete exclusion of original properties
and the construction of a new verbally supported unit. All these
operations, we have seen, result in definite products.
Among these products are numbered many simple and com-
plex objects usable in system-building enterprises. On the simpler
levels are concepts, general ideas, universals,'simple categories of
many varieties, all of which are conjoined to produce simple and
complex systems. On the level of more complex structures we in-
clude hypotheses, propositions, and many fictional component sys-
tems of still more comprehensive systems."3 Examples are such
fictions as frictionless surface, perfectly straight line, dimension-
less point, perfectly rigid body, as well as the figure constructions
used in building up geometric demonstrations and mathematical
systems.
'3
Many examples are offered by Vaihinger, Philosophie.
 CHAPTER XIV

DEFINING ANI) CLASSIFYING OPERATIONS

DEFINITION AND CLASSIFICATION IN LOGIC

UNLIKE the classical syllogistic‘ province, the logical sub-

domains of definition and classification do not force upon the
student some dazzling model which, once produced, prevents
him from observing the operations which fashioned it. Quite the
contrary, defining and classifying behavior attracts attention to
objects manipulated, to techniques of operation, and to the work-
er’s motivation. The essential nature and role of such behavior,
therefore, are easily Observed in everyday life and science, as well
as in the formal logical domain.
Unfortunately, this very freedom to study defining and classify—
ing operations engenders a plethora of conflicting views. With
respect to definition especially, the logical scene is asserted to be
not only confused but chaotic. This claim13 made because of the
many disagreements among those who attempt to define the
term definition. A much more serious problem concerns the efiec—
tive description of the Operations involved.
Much of the difliculty may be attributed to the philosophical
assumptions of logical workers. Moved to achieve transcendent
reality, most logicians harness defining and classifying activities
to 'universal and preclusive systems. Whether or not they begin
with actual events, or propositions about events, their goal is to
make one comprehend at least many, and possibly all ,- to make
once, or an instant, include forever,- and to enlarge a minute self-
constructed system to cosmic dimensions.
Even those methodological logicians who do not stress univer—
sality and cosmology seek to develop autistic specifications, and
consequently misconstrue classifying and defining behavior. Bound
by fixed prescription, the actual operations of scientific and logical
work vanish completely. What and how to define and classify are
not determined by the tasks at hand but by authoritative rules.
Cf. Dubislav, Definition; Robinson, Definition; Dewey -and Bentley, “Definition”;
also Knowing, chap. 7.
 DEFINING AND CLASSIFYING OPERATIONS 29
One of the most misleading assumptions made by both onto-
logic and methodistic logicians is that defining operations are
exclusively concerned with words. N o doubt the durability of the
verbal formulae representing definition and classification (p. 3 5)
is an influencing factor. Linguistic entanglements, however, result
in confusing our Own act of defining definition with the defining
operations we warit to describe.
In contrast to such totalitarian views we shall approach defining
and classifying processes as operations. We assume that these terms
represent procedures of locating stimulus objects in particular
situations and for particular ends. In the logical field they con—
stitute specific and heuristic operations for system building.

THE DEFINITIONAL ONE AND MANY

The opposition between the traditional and behavioral views
furnishes an exemplary illustration of the famous problem of the
One and the Many. Though the universalists insist upon a single
definition of definition and classification they agree neither upon
what is defined or classified nor upon the processes involved.2 In
consequence a large number of treatments have accumulated,
meriting the adverse characterizations mentioned above. On the
other hand, those who take definition and classification to be spe-
cific operations suitable for particular situations precipitate .the
question whether these operations can be commonly delineated.
Is it possible, in other words, adequately to describe what persons
do when they define and classify? Our answer is affirmative.
In the present chapter we summarize the common
traits of both
defining and classifying operations, allowing for variations accord-'
ing to the situations in which they occur. Both historical and cur-
rent materials will be considered. Because defining operations are
more complex and more varied than classificatory ones we shall
treat the former far more elaborately.

I. DEFINITION
E ssence: and F orms
Traditionally the treatment of definition has resulted in two
sorts of‘absolutism: one, formalistic, the‘other, methodistic. The
2Cf. Dubislav, Definition; Robinson (Definition) lists 18 names for sorts of
definition.
3o PSYCHOLOGY AND LOGIC

target of methodistic study was the discovery and organization of

the ultimate essences of things, the aim of formalistic investigation
was the organization of systems of equivalent terms. Both these
approaches contrast with the study of concrete defining operations
in which, even when words or signs are concerned, they are not
endowed with absolute properties called meaning or significance.

A. Definition as Pursuit of Essences

Definition, defined as the discovery or presentation of the
essence of that which is defined, has carried diflerent weights in
various cultural periods. In the earliest formal discussion, as in
Plato’s Socratic dialogues, the purpose was merely to achieve a
comfortable resting place for satisfactory argument. At that time
the complexity of situations and the clagh of personal interests
made it necessary to clarify descriptions, distinctions, and even
general references to things. MoreOver, such
definitiohs of
ment were helpful1n isolating similarities1n difierence. state-
The emphasis on conversation and argument soon gave 'way'to
the belief that one should attain the basic character .of things.
Forthwith definitional history progressed from an interest in
stabilized attitudes toward things, to the Specification of the essen-
tial properties and relations of the things themselves.
Though Aristotle made definition the search for abiding es-
sences, he nevertheless, on the one hand, regarded things as actual
items in a naturalistic framework and, on the other, thought of
definition as references or descriptions. In other words, the lingu—
istic factors remained as media for contacts with things.
With the development of medieval cultural conditions the
entire picture changes. Along with the rise of
structions, spiritual realms, domains of
transcendental con-
infinity, etc., logicians
hoped that the quest for absolute essences could be furthered by
definitional processes. In consequence, definition was transfdrmed
into a single absolute and all-inclusive procedure, a specialized
and technical enterprise quite in line with Scholastic metaphysical
logic. Naturally, linguistic techniques were inkaed to achieve
what was not available to persons confined within natural bounds.
The medieval period established a double fetter for logic. Not
only did it crystallize the construction of essence—an absolute set
 DEFINING AND CLASSIFYING OPERATIONS 3I

of traits—but it connected essence with another construction, the

universal.3 ∙

In modern times the search for essences gave rise to the problem
whether they could be located in universals or only in particulars.
Hence logicians divided themselves into nominalists, who made
definition consist of the manipulation of words or terms, and
realists, whose terms reached out toward the essence of things.
Nominal Definition. As a typical example of nominalistic defini-
tion consider Locke’s statement that definition is “nothing but
making another understand by words what idea the term defined
stands for)” The ideas, of course, constitute psychic states pre—
sumed to be the essence of known reality. Though no less psychic
than Locke, Reid builds his definition of definition with greater
emphasis on words.
A definition is nothing else but an explication of the meaning of a
word, by words whose meaning is already known.-Hence it is evident,
that every word cannot be defined; for the definition must consist of'
words; and there could be no definition, if there were not words pre-
viously understood without definition.‘5
By the use of such nominalistic procedures many writers have
hoped permanently to grasp fixed entities. Especially the British
sensationists entertainedthe notion that the reality of an Object
or an event‘could be seized by regarding it as a name. Definition,
therefore, becomes the manipulation of Verbal entities.6 According
to this plan, satisfactory definition is nothing more than the organ-
ization of words to the point of satisfying the organizer. It 'is un—
necessary to reiterate the objections both to the fundamental sen-
sationistic philosophy and to the concrete usage which nominalis-
tic definition implies. On the other hand, taken in a relativistic
framework, nominalistic definition is a useful and important pro—
cedure, its basic Virtue being that it is not equated with the events
defined. The definer, that is, remains aware that definitions consti-
tute constructions. On this basis definition can render service in
r
To be discussed in chap. 18.
Essay, Bk. III, chap. 3, sec. 10. \

“Essays, p. 10.

° Compare‘ Mill’s statement with the above. “The simplest and most correct
notion Of a Definition is a proposition declaratory of the meaning of a word” (Sys-
tem, Bk. I,\\chap. 8, sec. 1).
32 PSYCHOLOGY AND LOGIC

organizing various items of practical or scientific pursuits. A set of

verbal definitions may prove very, convenient in arranging chemi—
cal elements, for‘ instance, or differentiating between various ob-
jects of discourse. In every case, however, the defining procedure
and the final definitions are no more than particular items in spe-
cific classificatory situations.
Further, the process of organizing word equivalences helps to
clarify one’s notion concerning the relationship of particular
events. But never can the adequate naming of things give the
namer a grasp on the nature of. those things.
Real Definition. For those who espOusereal definitions, wOrds
or terms stand for the ultimate essences of things, universes exist—
ing beyond the minds or psychic states of persons. In real, as well
as in nominal, definition, words or terms are made to do heavy
duty because in such instances logicians are dealing with abstrac—
tional products somewhat remote from concrete problems. How—
ever, to define things instead of arranging word equivalents Offers
a more immediate connection ’with actual interbehavior than does
nominal definition, which obviously is a secondary procedure.
Still; one can not regard either operation as a technique for achiev—
ing the ultimate reality of anything. Real definition merely
approximates the practical process of identifying and isolating
types of events. The advantage of such identification is excellently
illustrated by the mathematical symbolization of variables or re—
lations, a process not essentially different from one in which differ-
ent sorts of objects and relations are interacted with.
Concept Definition. A modern version of medieval conceptual-
ism is the current doctrine that what are defined are concepts. The
essence of things is regarded as located in the knower’s mind. The
import of the conCeptual doctrine may be clarified by associating
it with one of its outspoken adherents, Rickert, who vigorously
rejects both nominalistic and realistic doctrines: “Definiert wird
nicht der. Name und nicht die Sache, sondern allein der Begrifi.”7
Keeping before us the writer’s philosophical background, his es—_
pousal Of such spiritistic constructs as transcendental consciousness
and absolute values, we see at once how far this doctrine strays
from the problems and behavior of science. Even when placed1n
7
Lehre, p. 85.
 DEFINING AND CLASSIFYING OPERATIONS 33
a natural setting this conceptualistic theory—building on the cur-
rent psychological doctrine that mental processes produce con-
cepts as psychic products—runs to nonnaturalistic essences. It
therefore bypasses such legitimate Objects for definitional reac-
tions as the attitudes and opinions of persons.

B. Definition as Pursuit of Form

With the rise of logistic or formal symbolic logic and mathe-
matical logic a new Viewpoint has developed concerning terms
and their larger structures called propositions. These elements,
in their various roles as symbols, variables, and logical constants”
are increasingly looked upon as sundered from content of any
sort. Among the symbolic or mathematical logicians the abstrac-
tionistic ideals of pure relation, order, and coherent system fulfill
all logical ambitions, and serve moreover as suflicient criteria for
all system-making processes. Such logicians are guided by the
purely grammatical features of language. Referential‘linguistic
functions are relinquished in favor of the structuralistic processes
of syntax and the interpretative features of semantics.
Syntactical Definition. Mathematical logicians have attempted
to create a special type of nominal or verbal definition, presumed
to be a purely formal system of equivalent symbols designed for
convenience. The best example18 the famous assertion of White—
head and Russell, in their Principia Mat/zematica,8 that a defini—
tion is simply a declaration that a newly introduced symbol13 to
be entirely equivalent to others already familiar. The new symbol
they call the definiendum, the old one the definiens. To carry out
the notion of pureness in the structural organization of the system
the authors declare:
We express a definition by putting the definiendum to the left and
the definiens to the right, with the sign ‘E” between, and the letters
“Df”, to the right of the definiens. It is to be understood that the sign
“E’ and the letters “Df” are to be regarded as together forming one
symbol . . . .

An example of a definition
\
13:

PDQ.=.~PVQDf. \

'Vol. E51? 11.

34 PSYCHOLOGY AND LOGIC

The abstractness of syntactic definitions is indicated by the fact

that they are dissociated from truth or falseness—in other words,
are presumed not to assert anything, and in general to be super-
fluous. They can not be used as premises for deductive or other
purposes.
In characteristic fashion, however, the authors of the Principia
Mathematica do not really abide by their formalistic description
of syntactic definitions. They go on to declare that definitions do
yield important information beyond the original statements or
'terms, and serve especially as means of analogizing the ideas con-
veyed by them. This contradiction simply signalizes that the pre—
tense of dealing with items possessing meaning, without depend-
ence upon a constructor, breaks down. As long as terms, and what
they stand for, are factorsIn interbehavioral situations, problems
and purposes abound1n definition making.
Semantic. Definition. In a special sense, thesemantic type of
formalistic definition corresponds to the concept type Of essence
definition. Instead of at once making words or terms stand for
mental states, the linguistic factors are regarded as immediately
containing or representing signified elements. Semantic definition
is thus presumed to stress not simply equivalences, butadequate
and effective symbolism as well. Proponents of this definitional
doctrine wish to avoid confusions in language by substituting
terms for things or events. However, they are only indirectly
concerned with ultimate essences, and opposed chiefly to mathe—
matical or logistic essences. In this way, semantic and syntactic
definition can be employed by logicians who incline toward at
least a linguistic withdrawal from traditional metaphysical en—
tanglements.
The proponents of both syntactic and semantic definition, even
when they turn to one as a refuge from the other, fail to see that
all definition constitutes descriptions of things for given purposes.
There18 no one definition of definition, as many definitions fire
necessary as there are particular definitorysituations. 1'
Definition as Interbehavioral Operation i~
From an interbehavioral standpoint the 'term definition stands
.for constructive intellectual operations by which one orients and
relates objects for specific purposes. On the basis of variousiin-
 DEFINING AND CLASSIFYING OPERATIONS 35
terests or motives the definer localizes or places objects by accord-
ing them identifying marks. Defining operations are similar to
describing actions when the latter are difierentiating performances
and not references to things or their qualities. Depending upon
the defining situation, the identifying actions are simple or elabo—
rate. In complex situations they may involve an abstractive analy-
sis of the properties and significance of stimulus objects. When the
stimulus Obj ects consist of word things, the identifying background
in which they are localized we call literary contexts instead of
situations.
To treat defining operations interbehaviorally is to stress ac—
tions—ways of operating upon things—not things. Defining acts
are in principle similar to the selective or differential performances
of individuals in perceptual contact with things. Because of the
obvious need to refer to definitional operations, logicians have con—
fused such action with things—namely, words or signs. This sub-
stitution of words and signs for defining acts is undoubtedly the
basis for most of the confusion we have mentioned as existing in ∙
the definitional province of logic (p. 29).
A characteristic feature of defining operations in the logical
domain is that they issue in a product. This fact warrants our de—
fining them as constructive behavior. Unfortunately, however,
the term definition often stands for the product rather than for
the underlying defining performance. Moreover, cultural need has
caused this product to be fixated1n a set of verbal substitutes for
the achieved identifying marks or characters. Thus, as we have in-
dicated, the embodying words and signs are frequently confused
with actual defining behavior.
Range of Defining Operations. Because defining operations
occur both1n system--making and nonsystem——making situations,
each. with its own particular frame of reference, they cover a wide
range, from practical-life situations to the most abstract circum-
stances of the symbolic logics. Rigid defining operations of formal
logiCal systems are no different in\principle, however, from those
occurring in scientific and everyday-life situations. Though the
variations in detail are not unimportant,they are nevertheless
all bounded by criteria localized1n the definitional situation.
Definitional Criteria. Compare definitions developed for formal
systems'and
\ for more casual and ∙1mprec13e ∙
∙ SituatIOns. In the former
36 PSYCHOLOGY AND LOGIC

the system itself constitutes the basis for treating its elements,
whereas in nonformal situations the criteria are less definite and
circumscribed. Whether definitions are formal, rigid, and con-
stant, or informal, loose, and temporary, depends upon the fol-
lowing criteria: (I) the definer’s purpose, or the character of the
problem at hand, (2) the types of stimulus objects defined
(things, qualities such as color, brilliance, etc., words, symbols,
actions) 5 (3) a criterion compounded from (I) and (2), namely,
whether things are operated upon directly” or through the inter-
mediation of such an action as ai'concept or an idea, and, finally,
criterion (4.), the question whether definition18 designed to stress
form vehicle—i. e. the words or signs, Or the product of the
intrinsic definitional operation. Around the fourth criterion are
centered problems of vagueness and ambiguity which we treat in
a later section. _
Analysis of Definitional Situations. We listafour outstanding
factors of definitional situations:
A. Stimulus objects, things defined.
B. 'Defining acts. \

C. Defining product, i.e. definition.

D. Linguistic aspects (statement of, reference to, or substitute for,
C)-
A. Stimulus Oejects. Although the disagreements of traditional
nominalists, conceptualists, and realists turn precisely about the
question of what things are defined, the accepted formal and onto-
logical dogmas remove logicians from actual things. As we have
implied, the vehicular terms are generally regarded as the defini-
tional objects. Most logicians frankly declare that what one de—
fines are words or terms. From an interbehavioral standpoint,
however, every conceivable kind of act, relation, quality, object
or organism may be a stimulus Object in a definitiOnal interbe—
havioral field. Among acts we include concepts, of course, though
construed entirelyIn objective terms. The striking diflierence“1n
VieWpoint here epitomizes the sharp variation between absolutistic _
logics and behavioralssystem buildingk Our View concerns ,the
actual work of persons in whatever situations they find
selves.
them—
'

B. Defining Acts. Preoccupied as writers on\definition are with

terms and what they stand for, they still can not escape the fact
 DEFINING AND CLASSIFYING OPERATIONS 37
that definition involves action. Aside from the everyday apprecia-,
tion that definition requires defining acts, logical writers have
recognized both definition by use or convention,9 and ostensive
definition10 in which the action consists of pointing to, or manipu-
lating, things by way of showing their similarities and difierences.
Acts and Materials. Unfortunately, even writers who insist
most strongly upon action center their discussion around words.
Definitions are thought of as constructed from words, symbols, and
propositions—in short, linguiStic materials. At once the whole
definitional process is reduced to manipulation of Signs and sym-
bols. Such reduction shifts the description of definition, from what
definers do, to the materials used in expressing and referring to
definitions. ∙

Why should defining acts be set aside in favor of signs or

words? In the demand for a fixed verbal formula there is more
than a reminiscence of the age-Old quest to fixate the many and
the fluid by singleness of verbal sign. Definitional operations, on
the other hand, are many and complex, and must not be forced
into a straight jacket of verbal symbolization. They can be de-
scribed, therefOre, only on the basis of observing the elaborate
constructive procedures carried out whenever definitional activities
occur.
C. Definitional Products. Students of logic, we have indicated,
are much more interested in definitional products than in the de-
fining operations which produce them. To logicians the products
doubtless appear more adaptable for system building, and ap-
parently more palpable altogether.
For effectively examining definitional products we present
some typical examples:
Physics is a fundamental science dealing with matter.11
Physics is the science of energy and energy transformations.”
Physics is essentially a system of explanations—answers to the ques-
tion “Why?”—_on the behavior of.inanimate things.13

See Couturat, for example, Principles, p. 182; Church, Definition, in Runes,

Dictionary.
'0
Burks, ‘Empiricism.
" Seinat, Eundamentals, p. 3.
Spinney, Textbook, p. 1.
Websier, Farwell, and Drew, General, p. 6.
38 PSYCHOLOGY AND LOGIC

Notice that the definitions embodied in these statements con-

stitute the results of evaluating and estimating the general physi-
cal3 domain, reviewing the objects dealt with, and then construct-
ing a definitional system corresponding to the essential factors.
Definitions and Propositions. Definitions as interbehavioral
products, it is interesting to note, have much in common with
propositions.14 Since the latter are better known than the former
We can obtain information concerning definitions by comparing
and contrasting them with propositions.
As to similarities, both are subtle objects which, during gesta-
tion, have no independent existence. To begin with, they are acts
of considering, evaluating, comparing, and contrasting things and
subject matter. Because of the importance of the objects inter-
acted with, the nascent definitions and propositions are as de-
pendent on the objects as on the individual’s acts. Independence
as objects is achieved only when they become definitely fixated—
that is, set into linguistic matrices.
But if definitions and propositions are alike in being interbe-
havioral products, they differ widely in use and function. Consider
Situations in which they stand sharpest in opposition—namely, in
science. In that domain, because definitions function forsimple
orientation, they have a lesser import and are, On the whole,
simpler products. They may be likened to first approximations,
premises, even coordination points. The SUperficial and auxiliary
character of definitions in science can not be mistaken for the pro-
found and essential nature of propositions resulting from judg-
ments and conclusions. Outside the scientific domain definitions
and propositions are, of course, not so sharply polarized. In some
nonscientific Situations their functions are indistinguishable. Re-
acting individuals orient themselves with respect to stimulus ob-
jects by means of either product with equal facility. In such cases
the scope and structure of the products are equivalent. ’
D. Linguistic Aspects of Definition. Since definitory behavior
is in itself subtle and intimate, and involves a large amount of
interpersonal relations, it is inseverably connected with linguistic
factors. Both definitional behavior and\the resulting productsrcan
be best observed when represented by diagrams, words, numerical
∙
See Kantor, Interbehavmral. −
 DEFINING AND CLASSIFYING OPERATIONS 39
signs, or other visible evidences. Our task at this point, then, is
not to caution against confusing definitions and their representa-
tives, but also to outline the various linguistic factors foundIn
defining situations.
Diagrams, words, and verbal formulae can, with practice, be
kept distinct from things defined, as well as from the definition
or description of them. Because words or symbols are requisite
tools for making and handling definitions we tend to confuse the
definitional products themselves with the names or descriptions
of those products. This dilemma is avoided by keeping the defi-
nitional product as a description distinct from our secondary de-
scription of it.”
Defining a triangle as a triangular or trilateral figure suggests
the difference between the definiens and the definiendum. The
latter must be distinguished from the act of referring to it and
from the symbols which stand for it. The two sentences written
above and the signs A, ∆ are equivalent representations of the
single definition. Consider another example, the definition of the
law of the inclined plane. In algebraic form the definition is repre-
sented as: Ll/L2/=' Wl/Wg. In words, however, we have:
The weight which can be moved up an inclined plane is to the weight
which would balance it in a vertical. direction as the length of the in-
clined planeis to the vertical height between its ends.
To indicate the proper place of linguistic factors in definitional
Situations we break them down according to their nature and func—
tion.
(a) Language in Definition Construction. Since defining be-
havior is implicit, and not grossly manipulative,-the action must
perforce be linguistic. The definer makes his comparisons and
handlings, his trials and errors, as subvocal and sometimes vocal
references. These references to the things defined. comprise the
materials out of which definitions are made, and are numbered
among the subtle objects referred to in our discussion of definitions
and propositions (p. 38).
(b) Language for Handling
Definition.
Acts and things used
for symbolizing and otherwise substituting for the definition
1”The\problem here is similar to that involved in the hierafchical language situa-
tion and, in general, that situation in which the theory of types plays a. part.
40 PSYCHOLOGY AND LOGIC

product may be regarded as language vehicles. Such are the dia—

grams, elaborate formulae, and other instruments to embody, ex-
hibit, and represent definitional products.
(c) References to Defining Actions and Products. Distinctive
language factors are the acts of referring to definitional products
by way of mentioning, explaining, discoursing upon, and otherwise
interbehaving with them. Reference to the acts of producing the
products are very similar in character. These language factors
are misinterpreted when it is thought that words, and not things
or acts, are defined. The result is that idiomatic and inclusive
formulae are set in the foreground of definitional situations.
Signs and Symbols for Definition. Language things, as signs
and symbols, are special linguistic factors
representatives of definition products. a
which
are counters and

Defining Operations and System Building \

\

Definitions relate to system building in two distinctive, but not

altogether different, ways. In the first place, the definitional
tinuum includes simple designative or descriptive activities, as well
con—
as those productive of small comprehensive definitional systems.
Frequently, to define something is to organize a small local system
of factors and to take into account various relationshipS' newly
discovered or previously known. To set up formal definitions as
discrete units and parts of a larger system displays a maximum of
system building. On the other hand, defining operations have a
distinctive place in general system—building situations. Here the
definer’S operations eventuate in the exclusion of some given stim-
ulus object from a system or its inclusion therein. Euclidean defi—
nitions, for example, indicate what the system is to handle, \what
elements it comprises, etc. But defining operations are
also useful
in the work of classifying, relating, and articulating items in the
systemizing process. It must be added at once that different kinds
of defining operations are serviceable in different system—building
situations.
Definition and Description
Logicians have noticed that defining‘operations are essentially
related to acts of designation or description. Mill16 declares, how—
. 1°
System, chap. 8, sec. 4.
 DEFINING AND CLASSIFYING OPERATIONS 4I

ever, that all logicians reject description from the rank of genuine
definition, on the ground that in description the name of a class
is defined by accidents—namely, attributes not included in its con—
notation. Lotze,17 on the other hand, though an absolutistic writer,
accepts methodological descriptions as definition, but thinks they
require checking by traditional classificatory procedures and defini-
tion rules in order tO control arbitrariness. Basically, Lotze is, of
course, attempting to safeguard his metaphysical interests.
Empirical writers are more favorablyinclined toward descrip—
tion. For example, Pepper18 has recently offered a vigorous plea for
descriptive definition. Adopting a modern and improved interpre—
tation of description, he assumes a position 'vis a vis the symbolic
logicians who, because of their exclusive interest in linguistic
formulations, wish to separate definition completely from proposi-
tions.- Pepper Obj ects to the idea that definitions, by contrast with
propositions, are entirely dissociated from truth reference. To
limit definition to the nominal form which is not responsible to
facts, he holds, prevents attaining the clarity and tolerance that
logic should foster.
Nominal definition, which 'Pepper contrasts with descriptive
definition, has two species called equational and ostensive forms.
Equational definition is described as a dyadic relation in which
S, the symbol, is equated- with, defined by, or substituted for MN,
a combin'atiOn of other symbols. Here we have a simple organiza-
tion of linguistic things.
Ostensive definition is likewise a dyadic relation, but in this
instance the-symbol S indicates, or is ostensibly defined by, 0,
some empirical obj ect. The ostensive feature is a pointing or other
type of operation.
By contrast with these two species of nominal definition Pepper
calls his descriptive definition a triadic relation. S, the symbol,
tentatively indicates, or is OStensibly defined by, O, the S18 tenta—
tively equated with, or descriptively defined by, D, a description;
and finally, D hypothetically describes, or is verified by, O, which
now becomes a field instead of a single object. Despite the fact
that the triadic relation seems to comprise the other two plus a
1'
Logik, I'I, chap. 1.
Descfiptive; Definition.
42 PSYCHOLOGY AND LOGIC

third, Pepper strictly denies it. From our standpoint it is significant

that his denial is made on the basis that the act or intention is al—
together different. Descriptive definitions are formulated through
acts embodied in empirical enquiries.
It is no small merit to enlarge the range of definitions to in—
clude the descriptive type. But we still face the question whether
nominal definitions are not also descriptive, though descriptive of
symbols and words, not of things. However, we are in sympathy
with Pepper as against his critics”, who want to make the word
definition cover only one kind of thing. What seems plainly in—
dicated is that only a genuinely interbehavioral position can take
care of this diversity of detailed actions which still have a com—
mon adj ustmental base.
Highly significant, therefore, is Pepper’s cmultiplication of defi-
nition types on the basis of acts and intentions. His opponents,
nevertheless, have a wedge when he falls back °on actions for dif-
fuse and not for full whole-hearted working material. Admittedly,
Pepper’s theory needs revision to make it entirely defensible.
First, definitions are not to be described as relations, but as
operations resulting in products. Accordingly the contrast be—
tween nominal and several other kinds of definitions will be im-
mediately achieved.
Then, too, definition must not be confined exclusively to reac—
tions to symbols, even if some are indicative of things while others
are not. As we have seen, thedefinition of symbols is undoubtedly
only one kind of defining behavior.

Definition and Meaning

Definitional doctrines center closely around the topic ofmean-
ing. This13 most obvious1n the case of the nominalistic defini-
tions of definition. The lexicological definitions of Reid and
Mill, already quoted (p. 3I), also imply that definitional
lems are problems of meaning. No less closelyIS meaning con—
prob—
nected with definition when concepts are the things defined! In
such instances concepts are Often equated with names .or terms,
or at least concepts are closely linked with them.20
is See Pepper, Definition; also Dewey and Bentley, Definition.
See, for example, Dubs, Definition; Rational, chap. 6.
 DEFINING AND CLASSIFYING OPERATIONS 43
Especially in nominal definition meaning is taken to be either
a commonly accepted usage or the way the user intends the term
to be employed by others. Meaning, in this sense of orienting a
term for one or many persons, or locating a term in a system, is
unobjectionable. Such treatment of meaning, however, is quite
different from the usual notion that it is some sort of mental
process.
Two elucidating tasks confront us: one, to Show the nature of
meaning as an Objective psychological fact, the other, to relate
meaning to definition. .
From an objective or naturalistic standpoint meaning is not a
“state of mind,”21 nor even an Objective status of personal satis—
faction 5 rather, it is an elicited structure, or organization of a situa-
tion or field, which an individual discovers through his interaction
with such fields. For example, when one observes that x—rays
emanate from the fluorescent spot on the glass wall of a discharge
tube, one assumes that there must be a close connection between
x-rays and fluorescence. Indeed, Becquerel observed that a piece
of uranium compound (potassium uranyl sulphate) wrapped in
black paper, that had been exposed to cathode rays, blackened a
photographic plate upon which it was placed. This observation
led him to interpret uranium radiation as a consequence of ex-
posure to cathode rays. ’For uranium to radiate meant, at that
period, not radioactivity, but a fluorescent effect. Later, of course,
it was discovered that uranium and its compounds radiate without
being subjected to bombardment by cathode rays or any kind of
electrical'treatment. What it means for uranium to radiate, there-
fore, is an objective situation elicited by definite operations.
So much, then, for the definition of natural things. In defining
verbal or symbolic things the ascertainment or specification of
meaning follows a similar process. The one great difference be-
tween the two-Situations is accidental. Namely, in verbal or symbol
meaning the various characterizing relations are established
through human agency, sometimes through an elaborate series of
operations.
Meaning, we. repeat, when connected with definition, consti-
tutes particular developments brought‘about through an indi-
21 Vol. I:\p. 20,6.
44. PSYCHOLOGY AND LOGIC

vidual’s operations with specific field factors. If what a thing

means, or how it Should be defined, depends upon a specific frame
of reference, obviously the meanings elicited are specific, not gen-
eral and universal.
Within the conventional framework
that definition'IS concerned
with meaning, the question is sometimes raised whether defini—
tion is capable of achieving meaning. The alternative is offered
that definition is only a special instance of processes designed to
specify meaning.22 It is proposed that meanings may be specified by
providing a set of descriptions, called indicators, concerning the
applicability of a term in various situations. The term species, it
is said, can not be defined, but merely indicated with certain
weights.
Apparently, definition is assumed to besa rigid placing of one
term as equivalent to another, but with the fortunate realization
that such equating is useless for situations in which things termed
are taken into account. In other words, isolated terms are different
from referential terms. We are
constrained to ask, however,
whether it is only terms or concepts that are defined. Furthermore,
when things are taken into consideration, meanings clearly con—
stitute properties and relations in specified fields. Whether or not
we equate specification of meaning with definition, interbehavior
with concrete things comes to the surface. Frequently, meaning
is nothing but a momentary property of an event which iS itself
momentary and nonrecurrent.
The I ndefinable
Because most writings on definition are environed by the logical
doctrine of universality and necessity, it is not a far-fetched sug-
gestion that the problem of indefinables has much in conimon
with, if it does not issue from, negative theology. From what
other source could logic derive the notion of inexpressible being,
whose qualities are indescribable and whose identity18 ineluctable!
All attempts to localize this problem within the confines of simple
qualities fail. Of interest hereIS Johnson’s 23 suggestion that if the
indefinable'IS taken to be that which15' not understood, a cq'rrec-
:2 Kaplan, Definition.
23 Logic, part I, chap. 7.
 DEFINING AND CLASSIFYING OPERATIONS 45
tion is necessary. Indefinable terms, on the contrary, are so well
understood that they need not be defined. Johnson’s point is also
cogent with respect to the indefinability of simple qualities of
things. Realistic philosophers take the view that colors are Simple
unanalyzable qualities.24 Russell does not even allow that the word
yellow as a symbol can be applied to the color yellow, but only
the propositional‘ function ‘x is yellow.’25 Within the frame-
work Of the writers quoted, colors are too simple to require defin—
ing. Johnson asserts, however, that While an adjective name, such
as red, can not be analytically defined, it can be ostensively de-
fined.28 When definition is taken as defining operations there is
nothing that can not be defined. In other words, everything can
be localized in its setting, or the definer can orient himself to any
stimulus object, including colors, odors, etc.27
The indefinable is not always regarded as the simple, but also
as the too complex. Consider the following:
I shall not start with any definition of religion. Religion, like
poetry and most other living things, can not be defined. But one may
perhaps give some description of it, or at least some characteristic
marks.”
Since there is really nothing that can permanently escape brief
intelligible description, suitable for particular purposes, defini-
tion then appears to be a fOrmal substitute for the essence of things,
a sort of 'idol.
Applying our interbehavioral notion of definition to the term
indefinable, we can orient and eXplicate .it by observing where and
how it operates. As an example consider Reid’s definition of defini-
tion quoted earlier (p. 31). The indefinable is merely those unde-
fined elements which constitute the beginnings of a classificatory
or system—building procedure, and in no sense indicates objects
resisting definition. Instructive here are Russell’s “minimum vo-
cabularies”——in other words, systems of words in which none can
be defined in terms of the others.29
2‘
Moore, Principia, p. 7ft.
Logical, p. 3751f.
Logic, part I, p. 94..
'7
Cf.‘Williams, Definition. -
Murray, Stages, p. 19.
'3

”Mental5, p. 14f.5 Knowledge, p. 79f.

46 PSYCHOLOGY AND ’ LOGIC

Operationally, the term indefinable marks a scale of definitory

products on the basis of varying criteria. There are questions con-
cerning how far one wishes to go, or how far it is necessary and
profitable to go. Obviously there are no absolute criteria, and thus
no final and absolute demarcations of things. It is only the urge
to absolutism which leads to the view:‘
No truly fundamental element can be defined, since the definition
would have to be in terms of still more fundamental elements, and
thus the element in question would not really be fundamental.so
Such a View, of course, has no standing when we are concerned
with actual defining problems. Defining Operations, on one level
of interbehavior, are activities with things. This level marks the
identification-qualification level between a lower direct difierential
and discriminative response to things and a higher less direct one
of referential and other verbal linguistic performances.

Definition, Vagueness, and Ambiguity

It is a paradox of the logic of definition that the more rigorous,
binding, and exclusive the formulae, the more empty and abstract
they' become. Hence, definition is hardly a treatment of terms
making for progress in the identification and orientation of things.
Definitional operations, of course, may play a part in clarifying
both things and terms, but these two fields must be kept rigorously
distinct. An excellent example Of how things are confused is pro—
vided by a popular elementary textbook.31 Beginning a discussion
of the purpose and' nature of definition, the authors declare that
everyday language is notoriously vague and- the language of tech-
nical treatises not always much better. To illustrate, they assert
that everyone is familiar with the difliculty of deciding whether
certain organisms are plants or animals.32 Is the difficulty here,
however, one of language, of not knowing how to apply a name
properly, or inadequate orientation of things? The same writers
regard the ambiguity of words as a serious danger for accurate
thinking. Granting the truth of this observation, we still' are
obliged to differentiate between the" ambiguity of words and the
Jones, Kant, p. 140. l
3°

" Cohen and Nagel, Introduction.

”‘Ibid., p. 274..
 DEFINING AND CLASSIFYING OPERATIONS 47
danger to thought arising from improper behavior with respect to
things.
Ambiguity and vagueness are essentially interbehavioral terms,
and, in consequence, relative to a number of coordinates. Since
these coordinates are localized in all sorts of reference frames,
there are no absolute criteria for ambiguity and vagueness. What
counts is the. availability of things, and one’s competency in han—
dling them. Sheer definitional operation can never overcome the
ills of ambiguity and vagueness. This is true for language as well
as for things.33
So far as language is concerned, the criterion of agreement, if
it could be secured, would apply only to specific situations. Basi—
cally this would be a matter of word usage. Actually, the lack of
commonness is no more serious than it is inevitable. Costello points
out:
“constants” are what may be arbitrarily changed, but “vari-
ables” are not arbitrary, being what is really constant in a system.“
The language of “exact science,” this writer asserts, may be curi—
ously inexact and misleading. Obviously this is true only when
someone is attempting to organize an absolutely verbally coherent
and homogeneous system5 perhaps the adjectives ambiguity and
vagueness would have no. cogency in describing its elements.
In science" it is things and events which are basically clarified.
This fact‘in itself guarantees thing as well as word definitions. Ac-
cordingly, the modifications of terms and their usage indicate
changes in total situations. To follow through the efforts of scien-
tists to modify the definitions of force, energy, mass, inertia, heat,
cause, atom,.species, action, electron, both as terms and as. things
or events, is tantamount, to traversing the history of science. The
same observation may be made for all the specific sciences and
.mathematics. In the latter the process of reducing ambiguity and
vagueness involves, of course, operational definitions of acts, both
h'constructional and purely manipulative, in addition to things and
terms. The definition of imaginary-numbers, infinity, and infini-
tesimals as operations exemplifies this pOint.
”In paper entitled “Empiricism and Vagueness,” Burks has argued that
‘a recent
the doctrine the empirical origin of concepts implies inevitable vagueness.
of
“ Naturalism, p. 313.
 48 PSYCHOLOGY AND LOGIC

How the current bias toward language influences writers to

minimize manipulative operations in science is demonstrated by a
recent paper Of Nagel’s on the place of language in the natural
sciences.35 He offers an attractive discussion, with good examples
Of linguistic changes, but indicates a grudging concession to inter-
behavior,36 as the italicized words in ‘the following quotation in—
dicate:
. a critique (of scientific abstractions) must aim to prevent the
hypostatization or reification of instrumental functions into mys-
terious, inaccessible agents, and to show how scientific language, in
conjunction with overt experimental procedure, can render, but with-
out illusion, the pervasive interrelations of natural processes.

If it is true, as is so often asserted, that philosophy is the realm

of the vague and indefinite,37 especially by° contrast with science,
then definitional problems are more acute in the former. Again,
if it is true that philosophy is primarily a domain of language,
particularly bad language, the type of definition useful to it is
verbal definition.38
The neutral student of definition should notice that definitional
operations may consist of word or term clarification when a particu-
lar situation demands. The study of definition implies no rule con-
cerning what should be defined, only that no improper limita-
tions should be set. It may well be that the sole valid definitional
criterion is the organization of a coherent verbal system, as would
be the case when the system is purely constructional. In general,
formal or mathematical situations require mostly verbal or sym-
bolic definition. But in other situations, such as science and every—
day life, the emphasis must be on concrete obj ects.
Up to this point we have not attempted to distinguish between
ambiguity and vagueness. Indeed, there are no standard and tech-
nical criteria for doing so. A practical differentiation13 that am-
biguous things require both radical investigation and definition to
fit them into a system. Vagueness may be relatively cleared up' by
simple definition.
Reflections. \. ’
as
Ibid., p. 624. l‘
”Burks,
:1
Empiricism.
allWeitz (Philosophy) offers an interesting summary of the abuse of language in
philosophy. s
 DEFINING AND CLASSIFYING OPERATIONS 49
Samples of Defining Procedures
To accept the operational interpretation of the definitional Situa—
tion makes room at once for a vast number of defining operations.
Aside from their common orientational function these operations
vary on the basis of things defined, definitional products con-
structed, and the behavioral reference frames in which they occur.
In order to exhibit the specificity of definition and its many varia-
tions, we isolate the following series of definitional classes.
Defining by Naming. The simplest type of definitional opera-
tion is obviously that which attains some measure of orientation by
nominal characterization. Though the definition product is simply
a name word, some of the traits of the definiendum are isolated
for identification.
Examples of naming definition are best observed in mathe—
matical writings. The following is typical:
Definition: The integral part of a logarithm is called the char—
acteristic and the decimal part, when it is written aS a positive number,
is called the mantissa.”
This particular definition product implies a great many activities
when logarithmic things are differentiated from the linguistic fac-
tors.
How can naming definition be set apart from ordinary naming
operations? The answer lies in the criterion of arbitrariness. In
ordinary naming there is no necessary connection between the
characteristics of the object named and the name.
Defining by Delimitation. A superficial form of describing or
characterizing things is simply to isolate or abstract them from
other things. In the definitions of physics quoted above (p. 37),
the interest is to mark off a field in which one plans to work. Such
are the introductory definitions in text books and treatises, which
amount to a simple means oforientation; the descriptive materials
are reduced to elementary boundary marks.
Defining by Classification. When things are somewhat more
closely approached than by delimiting the bounds in which they
are located, we achieve classifying definitions. The most obvious
form is. to locate an object according to its genus and species. For
example}, both a diameter and radius of a circle in Euclid are
\
” Davis and. Nelson, Elements, p. 349.
50 PSYCHOLOGY AND LOGIC

straight lines, the former passing through the center to the cir-
cumference at both ends, while the latter has one end in the
center. _
On the whole, classification definitions do not imply any
thorough penetration into the properties of things defined 5 how—
ever, the defining operations become modified in the face of the
problems the definer sets himself. Genus-species specifications from
ancient Aristotelian logic penetrated to metaphysical ultimates.-
Classificatory definitions also cover any sort of predicative designa—
tion desirable for logical and nonlogical enterprises.
Defining by Postulation. Again, defining procedures may be
differentiated on the basis of whether the objects defined or the
needs and desires of the definer are stressed. When objects are not
stressed the operations and products are naturally somewhat freer
of the objects’ characteristics as ascertained by various contacts
with them. Proceduresemphasizing objects iiivolve, for instance,
the construction of a map or schedule indicating their essential
properties. Postulational definition, on the other hand, implies
that the constructional activity is directed or influenced by assump—
tions brought to bear on the procedure.0 '

Consider Johnson’3 definition of logic:

Logic is most comprehensively and least controversially defined
as the analysis and criticism of thought.“
This definition18 so heavily charged with confidence1n its appro—
priateness that it is recommended as departing1n the least possible
way from the common understanding of. the term logic. And yet it
is the opposite of other definitions, for example, that of Royce:
Logic is the general science of order, the theory‘of the forms of
any orderly Realm of Objects, real or ideal.42
'The wide differences in postulation are apparent not onlyiwhen
defining such complex enterprises as logic, but also when dealing
with things.
l
Analyzing and Synt/zesazmg Definition. When obj ects to be
defined are complex, and the descriptive orientation warranrts,
the

∙
‘° Cf. Campbell, Physics, p. I 22.
↕ Logic, part I, p. xiii.
‘2
Principles, p. 69.
 DEFINING AND CLASSIFYING OPERATIONS 51
operations include the ascertainment of the elements of things and
the organization in which they are synthesized. This procedure
involves a more intimate contact with things than occurs in naming
definitions, or in Simply delimiting their boundaries.
Either analyzing or synthesizing operations may be stressed in
some defining Situations, or both may be equally represented as
in the following example.
Protoplasm constitutes a colloidal complex of water, inorganic
salts, and four organic substances, proteins, fats, carbohydrates, and
extractives.
How'elaborate or compact, how simple or complex a defining op-
eration should be depends upon the orientative circumstances. As
a rule, however, the products of analogy and synthesizing opera-
tions are more elabOrate than those of naming or delimiting situa—'
tions.
Defining by Relating and A nalo gizin g. Though probably no
defining operations can dispense with relating acts, some definitions
are more decidedly relational than others. Mathematical defini—
tions are, on the whole, fine examples of this type. To define a one-
member class the following structure is set up:
If x is a member of a and if y is also a member of a, then y is
identical with x.
'

The form of this definition, which belongs to a particular system,

emphasizes the'identity relation, but this is in no sense the only
kind of relation. When we say the relation azb is equal to there-
lation czd, we may do so on the basis that when we multiply a, b,
c, d by the numbers m, n, we find that whenever:

the same holds for:

ma
> nb, ma = nb, ma < nb

me >nd, me: nd, mc

< nd
This is essentially a definitional ordering by analogy—namely,
eliciting similar properties.
Among analogical definitions may be cited the inclusion of frac-
tions, irrationals, and imaginaries in the field of numbers. Thus
when confronting such processes as x 3 ∙ −∙−−−
9, or x +
3 = o, the
factors involved must be defined as numbers bricause the underly-
ing operations are Similar.
52 PSYCHOLOGY AND LOGIC

Defining by Measuring. A unique form of identifying and

'orienting is achieved through measuring operations. The descrip—
tion of things, either ordinary Obj ects or mathematical figures, fol-
lows the elicitation of specifying marks by comparisons with given
units or standards. Though measuring operations are close to
ordering procedures, we can not limit our notion of measurement
to the conventional correlation of nonnumerical entities with num—
bers. As we shall point out (chap. 23), this correlation idea does
not emphasize sufl'iciently the operational basis of correlation.
As an example of the measuring type of defining procedure we
suggest the definition of the geometric similarities of figures.
Figure A is similar to figure B when there is a constant ratio be-
tween any two points on figure A and the corresponding points on
figure B. o

Definitional equations, as for example, the work equation

W = PS, illustrate arithmetic forms of measuring definitions. The
relationship of work done against resistance, and over a given
distance, has been discovered by measurement. The following
verbal form of definition demonstrates the operation:
The work W, performed by a force P in overcoming a resistance
without acceleration through a distance 8, is measured by the product
of the force P and the distance 8.“3
To define points by rectangular coordinates, or by determining
the modulus and amplitude as in polar coordinates, is another
good illustration of the measurement procedure. Notice that it
displays a legitimate carrying over of the operational term defini-
tion to a rather abstract situation. It also shows a proper form of
the ostensive or denotative type. We might compare the coordinate
definition with that of a function1n which the relationshipIS not
established by measurement but by attribution, as in prescriptive
procedures.
Defining by Prescription. Logicians who adopt an exclusive and
manipulative notion of definition, at variance with formal Or ab-
stract symbolism, as a rule minimize and even reject the claims of
prescriptive or fiat definition. Their groundIS the arbitrary and
inessential nature of such definitions. From an operational stand-
point, however, it is impossible to deny thdorienting function of
Grimsehl, Textbook, vol. I, p. 82.
 DEFINING AND CLASSIFYING OPERATIONS 53
flat operations. Certainly thereIS as much validity1n prescriptive
definition as in naming operations, if, indeed, the former are not
merely extensions and elaborations
of
the latter. For convenience
we differentiate between several subtypes of flat procedures.
(a) Temporal Operation. Mathematical expressions beginning
“Let x = y” are clear-cut examples of the flat type of definition.
Here is a procedure which lays down what a literal number will
do in a given situation, or what the operator can do by thus de—
limiting the value of a number. This form of prescriptive defini—
tion symbolizes the mathematician’s freedom in his manipulation
of fixed relations. ,
Verbal counterparts of fiat procedures are constantly being
pointed out in philosophical writings. Universality is equated with
indestructibility.“ Matter is a tautological equivalent of impene-
trability.‘15
(b) Formal Eguating. Probably the best example of the formal
equating procedure is the substitution of one Sign or symbol for
another. The present Operation finds its best scope and application
in the domain of purely formal Sign manipulation, as in symbolic
logic. An example of such definition we have quoted above from
Principia Mathematica. As Whitehead and Russell say: “A defini—
tion is concerned wholly with the symbols, not with what they
symbolise.”“3 For this reason they regard definition as an unneces-
sary process
which contrasts with descriptive explanation andIs no
part of the subject in which it occurs. Note how Sharply this View
diverges from that of Dewey and Bentley“7 who look upon defini-
tion as the “throbbing heart—both as'pump and as circulation—
of the whole knowledge system.”
Formal equating may be profitably compared with the more or
less prescriptive definition of ∕−−−↕ as i by Gauss and others,
when those mathematicians analogized the referent of the symbol
as a rotational operation through 180° in a counterclock direction.
Though in this case there is considerable discretion exercised, the
emphasis is on the thing defined and not on the marking substi—
tute.
Defining by Specific Operations. Since Bridgman popularized
““Pepper,Categories, p. 548.
“Russell, Analysis, p. 385.
\

”Principia, vol. I, p. 1 I.
“Definition,” p. 287.
‘7
54. PSYCHOLOGY AND LOGIC

the contrast between concepts defined (a) by properties and (b) by

operations,“8 many writers have discussed the subject of opera—
tional definition. But, as Bridgman has indicated,"9 the term opera-
tional as applied to definitions is superfluous, Since all definition
must be operational. It is indeed tautological to say that all de-
fining operations are operations, but the important fact is that
particular defining behavior involves particular operations, as we
have amply‘demonstrated in our exposition of various defining
procedures.
To elaborate this point iS essentially to summarize the entire
subject. This we need not do, since we have sufliciently differenti—
ated between the operations required to define different objects
in Varying Situations. But we might profitably distinguish between
(I) operations for engendering and defining concepts, (2) opera—
tions for genuine and spurious definitions, (3) implicit and ex-
plicit definitional operations, and (4) variousforms of ostensive
definition.
(I) Operations for engendering and describing concepts. To
isolate and describe specific operations is automatically to difieren-
tiate between the activity of engendering ideas Or contepts- and
that of describing them. To the characteristics Of defining opera-
tions already mentioned we might now add that of brevity, limited
scope, and possibly also, auxiliary function. Defining operations
are ancillary to broad and comprehensive orientation processes, as
well as to those for ascertaining the properties Of things.
Definitional products may in no sense be regarded as items in
final and definitive systems of knowledge. At most they are tenta-
tive materials waiting to be worked over and established. All this
applies only to definitions of things. The definitions of terms oc—
cupy still humbler positions in the province Of investigation and
knowledge.
(2) Pseudodefinitional Operations. Activities of simple designa—
tion or proper-name application constitute pseudodefinitional op-
erations. But we must hasten to add that the criterion is the inter-
behavioral one of describing, identifying or orienting something.
It is not diflicult to differentiate betwe\en casual naming and inam-
ing for elementary comparing or distinguishing. “A rose is a
\
48
∙
Logic.
‘3 Some, p. 246.
 DEFINING AND CLASSIFYING OPERATIONS 55
l

rose” is no trivial illustration of the difference between purely

verbal play and the statement that an instance of a rose
a rose in the full organization rose properties.
of flower
Is

Metaphorical references usually are pseudodefinitions, though

here again the criterion is what sort of interaction between an or-
ganism and objects is going on. To describe words or things is
certainly a different act from listing them. A picture or a narrative
may or may not be descriptive, but blue prints, on the whole, must
be regarded as descriptive of a building as a whole and as parts.
The words of a foreign dictionary may be intermediate descrip-
tions or simply word substitutes. In a similar way we can distin-
guish between operations in which words are descriptive and cir—
cularly tautological.
(3) Implicit and Explicit Operations. Formalistic writers dis—
tinguish implicit from so—called eXplicit definitions on the basis
that the former lack organized formulation. Those who know
what “triangle” and “quadrilateral” mean will also know what
“diagonal” means, if told that each of the two diagonals of a
quadrilateral divide it into two triangles."0 It is questionable
whether logicians are concerned with such trivial definitions.
Nevertheless, the above is an illuminating example, Since it indi-
cates that a geometric object or relation is the thing defined, not
just a term, Were Simply a term to be defined, we should not need
to locate’it in a geometric situation.
Those logicians who regard definition as important value the
explicit type more highly. But if we attend to the details of par-
ticular situations the differences hardly allow for a scale of merit.
.Where can we find adequate criteria for preferring Euclid’s ex—
plicit definitions of points, lines, angles, etc., to the modern ge-
ometer’s practice of merely implicitly defining these elements
through the axioms?
Still, from an operational standpoint, it is possible to distinguish
between two very different kinds of operations which produce the
definitional product—namely, manipulative and linguistic. In the
former case the actual drawing of the diagonals constitutes the im-
plicit definition. The clarification of the nature of a diagonal lies
in the performance itself, rather than1n the fact that a productIS
being‘ constructed. Linguistic Operations are more definitely di-
” Enriques, HIstorIcal,p. I 19.
56 PSYCHOLOGY AND LOGIC

rected toward products, in addition to their function as references

and symbolic representations of things.
Consider another illustration revealing the basic continuity in
implicit and eXplicit defining operations. To draw an actual circle
is to be alive to the constancy of the circumferential line. This,
as in the case of the diagonal, maybe regarded as the implicit or
essentially manipulative definition. Explicit definition stresses the
formula which is written as follows:
A circle is the locus of a point moving at a constant distance from
a fixed point.
As always, the word structure must be kept distinct from the defi-
nition product, and notice further that the lack of a completeness
or time factor simulates the absence’ of coordinates in the symbolic
formula r2 x2 + y”. 0

In this connection it is interesting to note Heath’s observation

that the construction'of figures in geometry marks the transition
to Objective definition of things from the subjective definition of
names.51 Acts of all sorts, whether they consist of vocal or graphic
naming, are, of course, objective, but as far as objects are con-
as
cerned it is justifiable to characterize naming acts less descriptive
of Objects and more personal. AS Strong points. out, however,
Euclid certainly did not begin his geometry with entities in his
mind which subsequently became justified."2
Implicit definition may be further distinguished on the basis
that the defining operations are more self-centered than are those
involved in CXplicit definitions. To define for other individuals
involves more formal linguistic interbehavior with the properties
and relations of things.
(4) Varieties of Ostensive Definition. Pointing operations and
casual demonstrative acts have been rightly excluded from the def-
initional domain. For the most part, however, the ground for
doing SO has not been of the firmest. Most frequently, perhaps,
ostensive definition has been condemned because of deviating from
the use of words and symbols; or, when words are used,'only
right—word usage is considered prope‘rf:
InThirteen, p. 1 46.
"Procedures, p.224.
r
"Mill, System, Bk. I, chap. 8, Dewey, Logic, pp. 53f,120, 241f.
I“Feigl, Operationism, p. 252.
 , DEFINING AND CLASSIFYING OPERATIONS 57
From a strictly operational standpoint the problem of ostensive
definition is easily settled Since we possess a definite criterion of
identifying or describing things for orientational purposes. Hence,
the factors in a situati‘On may be analyzed and assessed despite
their variations. The kinds of actions involved depend not only
upon the thing to be defined but also upon the general surround-
ing circumstances.
When words are to be defined, genuine ostensive operations
include the correlation of the words with, or their application to,
things or to other words in context. Those who do not limit defini-
tion to logic, or make logic something else than system building,
are not disturbed by the simplicity of ostensive situations. No
significant basis exists for overlooking the continuity of all defini-‘
tional events. .
Ostensive pointing or similar manipulative performances con—
stitute the operations for defining things. To demonstrate to oneself
or to some one else how a thing operates by provoking it to action
in order to describe it—namely, identifying or locating its prop—
erties—comprises genuine definitional operations.55
Rules for Definition
Considered in their conventional framework the traditional rules
of definition are correct and pertinent. Yet they are remote from
any Significant logical work. For the most part definition rules are
linguistic in character and conform to grammatical principles. The
way most writers expound the rules of adequate or commensurate
definition—circularity, negativity, obscurity, and figurativeness—
is to make them into rules of literary exposition. There is no
serious reaching down to system organization or even a general
handling of things. How to deal with verbal or symbolic formula-
tions is a special logical problem, if any at all.
Probably the most significant point about conventional treat-
ments of rules of definition'18 that most writers in one way or
another shift their ground from words, and even concepts, to
things as the objects of definition. Notice the usual insistence
that the essence of the definiendum must be stated. Paradoxically,
the genes-et—difi‘erentiam rule comes closestto a rule of thing
“Of iconsiderable interest here is Aristotle’s account of\ inductive definition;
Analytica P\osteriora, II, 1 3, 97b.
58’ PSYCHOLOGY AND LOGIC
\

definition, but not all writers offer it, probably on the ground that
it is concerned more with division than definition.
The rule of noncircularity of definition in particular marks the
path from words to things. It is frequently noticed that the repeti—
tion of words is either unimportant or unnecessary, when the
words are employed to refer to, or substitute for, things.“3
For authentic system-building procedures a different set of
rules is required. At the peak stands proper regard for specific
problems and particular reference frames. For effective systemiz-
ing, defining behavior must be articulated with the conditions
under which it is performed. In this way one avoids adopting an
alternative but less adequate procedure.‘
Likewise important is the rule that no particular kind of defini-
tion can serve as a model for all other types. For example, though
mathematical definition is exact and effecti19e for mathematical re—
lations, it can not be employed as a general model for material
which can not be SO rigidly defined (vol. I).
From the standpoint of concrete interbehavioral enterprises the
conventional objection to definition by use is not acceptable. The
question is: Are there specific situations in which such definition is
conducive to the organization of a system? The objection to using
description stems from the neglect Of the psychological aspects of
logic, and the prejudice against concrete behavioral, as over against
abstract formal, procedures.

II. CLASSIFICATION
Classes are essentially systems. Classifying operations, there-
fore, constitute logical or systemizing behavior. They are even
more essentially logical in nature than syllogistic construction op—
erations. Traditional logic, however, has only been Concerned with
the most formal and simple types Of classification.
Here is another instance of the ill effects brought about by the
intrusion Of linguistic factorsInto logical situations. For one thing,
although classification lends itself to the effective arrangements
and organization of actual things, it has been confined to terms and
statements. Nothing has had a 1n trivializing logic.
Consider how the processes of classification have been mishandled
Johnson, Logic, part I, p. 104; Eaton, General, chap.\7; Lewis, Mind. p. 82.
M
 DEFINING AND CLASSIFYING OPERATIONS 59
by contamination with division, especially dichotOmous division!
Even the classical logician in his practical contacts with actual
things57 has had to take account of the great discrepancy between
division and classification.
AS system building, classifying work consists of the construc-
tion of classes as products. These products, like all logical products,
may be constructed linguistically or developed on the basis of han-
dling or describing actual things. In all cases a product class is
not to be taken as an eternally existing autonomous entity. Platonic
classes are, of course, simply reification of terms; as such they il-
lustrate the substitution of a constructed ’description for things
described. When actual things are organized, the criteria may be
selected from their Similar properties. Differences in identity are
arranged to form a coherent system. Thus a class men is achieved.
On the other hand, purely constructed classes are set up by means
of constructed properties. Pure and preferred races, for instance,
are constructed out of the autistically created properties of manu—
factured peoples.
Tradition has it that definition is preliminary to classification.
As we have seen,
however, classification may precede definition.
This apparent ’circularity of operation is not objectionable unless
one adopts an absolutistic notion. When system building concerns
concrete, interbehavior we can go even further and point out that
instead Of a circle we have here a Spiral of procedures. In other
words, there is a continuous ascent from one level to another. On
one level definition precedes classification; on another the opposite.
takes place. The fundamental question is whether the systemizing
operations are progressing.
Classification procedures fOr system building must be differen-
tiated from similar procedures carried on for nonlogical purposes.
Involved are variations in criteria, refinement of procedure, and
intensity or amount of detail. Truth tables of formal logic illus-
trate small systems which include at the same time definition, divi-
sion, and classification. By prescriptive definition such systems are
limited to truth and false categories or statements. Further, they
are defined as P false, and Q true is impossible or excluded. Then
the systems are divided and organized intofa conventional pattern.
\

Jose‘pli, Introduction, p. 133.

5"
 CHAPTER
xv,
SYSTEMS AS CONSTRUCTION PRODUCTS

SELECTION AND ANALYSIS OF SYSTEM PRODUCTS

IN THE two preceding chapters we have discussed outstanding

system—building processes. We now turn to the products of
such constructional interbehavior. The best method, naturally, of
studying productsIs the point—by-point Investigation of the system
builder at work. When this direct methodIs not available, a gOOd
substitute'1S a critical analysis of sample systems.
Because of the immense number of systems and classes of sys-
tems we are immediately confronted by a°problem of selection.
Even the conventional logics constitute complex organizations of
smaller systems. For example, inferences and the syllogism are
more or less independent subsystems of classical deduction. For-
tunately, the very abundance of materials helps to suggest a tech-
nique, one derived from our field-organization principle. This
technique we regard simply as an investigative schema.

INVESTIGATIVE SCHEMA FOR SYSTEM PRODUCTS

Rejecting the assumption of a single universalistic system pat-
tern, we take system products as unique outcomes of given enter—
prises. Thus we proceed to study the product according to the ma-
terials structured and the kinds of operations utilized in manipulat-
ing them. In addition, the plan and use of the system constitute
important features of the total situation.
The field hypothesis enables us (I) to counteract the conVen-
tional View that beyond actual system building itself therelare
unique logical processes such as induction, deduction, and implica-
tion, and (2) to ,bring into prominence the concrete details of
system construction. Frequently it is such details that are \of
lthe
utmost importance for logical theory, Since logicians, like scien—
tists, present only successfully finishedproducts. But even if we
are denied access to the vacillating, and sometimes inept, opTCra-
tional trials and errors leading to the final\ result, we are' not
'obliged to regard logical systems as absolute entities which are
 SYSTEMS AS CONSTRUCTION PRODUCTS 61
simply disCovered and not contingently produced. Taking account
of the great variety of systems, and the enormous number of proc—
esses required to create them, reveals the product Character of all
logic.

FOUR TYPES OF SYSTEM PRODUCTS

By emphasizing certain system—building factors—namely, (1)
materials, (2) operations or techniques, and ( 3) use and function
-—we can arrange system products into four types:
A. Material or Content Systems.
B. Linguistic Systems.
C. Formal Systems.
D. Operational Systems.

A. Material or Content Systems

In 'material or content systems the things dealt with are empha—
sized, rather than the design Or constructional work. Of the exten—
sive range of material systems we shall consider only structures of
things, events,-relations, classes, acts, and sentences.
(1) Things. Systems of things fall between wide limits. They
include any simple or complex arrangement of simple or com-
plex Objects, as well as such intricate structures as Mendeleeff’s
periodic ,‘table or an organizatiOn of atomic constituents. Notice
that amdng such systems actual things are not distinguished from
knowledge and theories about them. Moreover, bOth authentic
things and things merely alleged to exist are included in one and
the same series. In every instance the historical background and
technical details Of the structure suggest the facts entering into
the system’s construction. Frequently also, the sufficiency and in—
sufficiency of systems at different time periods reveal their nature
and the operations employed in building them.
(a) Titius-Bode Planetary System. The organization of the
relative distances between the sun and the planets constitutes an
elementary but interesting scientific system. The basic materials
operated on were the known planetary'distances. For Titius ( I 72 9—
I 796) these distances concerned Mercury, Venus, the Earth, Mars,
Jupiter, and Saturn. Bode also (1747-1 826) knew Uranus, which
.was
diScOvered by Herschel in I 78 I .
62 PSYCHOLOGY AND LOGIC

Now it was easily suggested to Titius that, except for Jupiter

whose distance from the sun is 3.5 times that of Mars, each suc—
ceeding planet iS roughly twice the distance of its predecessor. Ac—
cordingly, the following system was set up. The last row in the
table indicates the approximate distances.

Mercury Venus Earth Mars ? Jupiter Saturn Uranus

4 4 4 4 4 4 4 4
O 3 6 I2
24 48 96 I 92

4 — 7 IO 16 28 “352 100 196

3.9 7.2 10 I5 52 95 I9I

T

0
It is interesting to note that the discovery of Uranus, between
the dates of Titius and Bode, fitted into the Scheme. This. is all
the more remarkable because the original figure for Mercury
should have been 1V2, Since 0 is not V2 of 3. But even mOre ex-
ceptional was the incident of the missing planet or empty space.
Here was an astonishing feat of prediction whereby the planetoid
or asteroid Ceres was discovered on the basis of a crude and un—
workable system. Just how baseless the system was-came into full
view when it broke down with the discovery of Neptune and
Pluto, as the continuation of the above scheme indicates. In the
case of Neptune, however, both Adams and Leverrier assigned
distances to the planet on the basis of the erroneous system which
thus played a part in its discovery.

Neptune Pluto ∙
374
4 SJ
768
4
l
388 779-
i
300 396 ‘
\

(b) Balmer’s Spectral Line System. Among the early stfiking

observations in the spectral analysis of hydrogen was the constancy
and regularity of the four lines in the visible] region. How to ac-
 SYSTEMS AS CONSTRUCTION PRODUCTS 63
count for the relation between them became an exciting problem.
Thus J. J. Balmer (1825—1898), a Swiss mathematics teacher,
constructed a system which he published in I88 51 consisting of
these line things. He began with Angstrom’s careful measurements
of the first four spectral lines, which yielded the following values
in Angstrom, units: 6562, 4860, 4340, and 4101. Dividing the
Angstrom .numbers by an arbitrary constant 3645.6, which he de—
noted by h, he obtained the following fractions:
9/5, 4/3, 25/21, 9/8-
These he put into the forms:
9/5, 16/12, 25/21. 36/32.

obtaining:
32 4’2 52 62

32.....22, 42.__22, 52_22, 62—22.

The completed system consists of the generalized formula:
hm2
m2 _ 22
which can be written as follows:

.= 3645.6 X
m2
2
m
-4
Balm'er, calculating the following series, assigned difierent
values to m as below:
m Calculated Observed
3 6562.08 6562.1
4 4860.8 4860.74
5 4340 4340-1
6 4101.3 4101.2
7 3969 3968-1
8 3888 3887.5
9 . 3834-3 3834
10 3796-9 3795
3769-6 3767-5
12
3749.1, 3745-5
13 3733-3 3730
14
-_ 3720.9 3717-5
15 37II 3707-5
16 3702-9 3699.

Anna-len.
64 PSYCHOLOGY AND LOGIC

The significance of the System lies in the fair agreement of the

Calculated with the observed values of spectral lines, which, of
course, Balmer could not know Since they were only later dis-
covered.
A further important feature of Balmer’s work lay in his realiza—
tion that the 2 denominator in his fOrmula restricts its application
to waves shorter than
6562A,
modified the formula to give:
if m is an integer. He therefore

hm2

When n = 3, the formula becomes appliCable to the Paschen in-

frared series, while 11 = I corresponds to the Lyman ultraviolet
series.
(c) The Bohr Quantum System of Atoiiis. The Bohr quantum-
mechanical interpretation of atomic structure excellently illustrates
scientific system products and their construction. Essentially the
system consists of organizing into a single structure: (I) the
Rutherford planetary atom, (2) the Balmer series of wave
lengths, (3) the Rydberg constant of wave frequencies; and (4)
the Planck—Einstein quanta.
Consonant with the general principle that all'construction iS
reconstruction, the building of this system begins with Bohr’s
improvement of Rutherford’s atomic model. Retaining the gen-
eral planetary idea of an atom, a frequently suggeSted cOnstruc-
tion,2 Bohr altered it to conform to accumulated information fol-
lowing the development of classical electromagnetic theory, es-
pecially the data Obtainedfrom spectroscopic observation
ing radiation emission and absorption.
concern:
Proceeding from simpler to more complex atoms, Bohr accepted
the view that the hydrogen atom constituted a simple Systemlwith
a single positive proton as a nucleus and a single negative electron
revolving about it. At first the orbit was assumed to be circular,
but later, in conformity with Coulomb’s law of electrostatics, the
electronic orbit was described as elliptical; the electron thus
moved abOut the proton with a foréeinversely as the square of
the
distancebetween them. 7‘

aPerrin, 1901-, Nagaoka, 1904; cf. Tucker, in Taylor, Physics, p. 820; also
Poincare; Foundations, p. 317.
 SYSTEMS AS CONSTRUCTION PRODUCTS 65
At this point, further reconstruction of the Rutherford atom
was necessary. The problem was to Show how electrons could
maintain themselves in the outer atom. According to Classic elec—
tromagnetic theory the circulating electron must continuously
radiate energy and describe a spiral path until it collides with the
proton. Since the observed facts of the hydrogen spectrum did
not accord with the collapse of atoms, Bohr had to construct a bold
postulate.
It was no less than the creation of a stationary motion or state.
This consisted of the circulation of an electron in its orbit with a
constant energy. Unless the electron jumps from one orbit to
another there is neither radiation nor loss of energy, nor absorption
increase of energy. But when the electron shifts from one of
the many possible orbits to another, it gains or loses energy and
emits or absorbs radiation. Energy is lost and waves emitted, when
the electron jumps from an outer to an inner orbit. The instantane—
ous change from an inner to an outer orbit marks the absorption of
energy, in which case the atom iS said to be excited and to Show
an absorption line in the spectrum. _
With respect to radiation, the spectral series of hydrogen can
be envisaged as follows: When the electron jumps to the second
orbit from the third, H a is emitted; from the fourth, H B; from
the fifth-\H 7, etc. This is the Balmer series. Shifts from outer to
lowerienergy levels up to n = 3 give the Paschen infrared series,
Up to n = I the Lyman ultraviolet series, and to n = 4 the far
infrared seriesgof Brackett.
Doubtless the most important item of the system is the importa-
tion into it of the quantum theory. Whenever an electron jumps
from one orbit to another it either gains or loses energy in discrete
quanta. For example, in moving into an orbit of higher energy, as
the electron might do by colliding with another electron, by ab-
sorption of radiation or by high temperature collisions, the change
would answer to the formula hv :2 Ez—El. Similarly the loss of.
energy by radiation of light is described by the same equation; It
is Observed that the frequency of the radiation is measured in
terms of Planck’s constant h. .

'SO .far the Bohr system combines the ._.constructs of classical

meChanics with Planck’s quantum theory. We have next to see
66 PSYCHOLOGY AND LOGIC

how the system assimilates the Balmer and Rydberg items. This
assimilation may be regarded as predictable, since the quantum
theory itself is a radiation theory, and because of the close associa-
.tion of the Balmer and Rydberg constants. As early as I900 Haas
had Shown the probability of expressing the Rydberg-Ritz constant
in terms of h and electronic data.3 Certainly, evidence was already
available that the particular structure of an atom or molecule de-
termines its Spectrum. What was needed, then, was to interrelate
the known facts of the hydrogen spectrum with the relations ob-
taining between the hydrogen electron and proton. N ow, the
quantum atom made possible a theoretical basis for the Balmer and
Rydberg work.
We have seen (p. 5) that the Balmer formula for the hydrogen
Series: 0
m2
= 3645.6 X 2 a

m −4
was arbitrary and purely empirical. Yet it was sufficiently descrip—
tiVe of the order and regularity of spectral lines to be predictive.
Rydberg (I 8 54—1 9 I 9), interacting with the same datum—namely,
the hydrogen spectrum—proceeded from the wave length and
the known velocity of light to the calculation of the number of
waves that pass a given point in a second of time. it became ex—
pressed in cms., through the following developments.‘ First, Bal-
mer’s formula was written:

(In21112−− 4)∙
I I

A = 3645 .6
Then, by multiplying the numerator and denominator on the
right side by 4, the following expression was obtained:

)1—3645.6 22
4
(I In2I)
When this formula is eXpressed in cms., it reads:
I I I” I
—= 109,720 −−−− —.
)1
22 m2 1
a Sommerfeld, Atomic, p. 89.
‘The present system illustration is derived in large part from Tucker (Taylor,
.Physics, (p. 82of). \.

5
The factor 109,720 is, of course, the Spectroscopic constant 0r number of Rydberg.
Later determinations of this number gave the more accurate figure of 109,677. 8 cins'1.
 SYSTEMS As CONSTRUCTION PRODUCTS 67
Taking into account the Lyman and Paschen hydrogen-spectral
series, as well as those of Balmer, we obtain the two following

↨ −− − —>
formulae. For the Lyman series:
—
I I 1
I2 m2
3, 4 . . . and forRthe(—Paschen series:
—->
where m = 2,

−−
I I
- I

32 m2
withm=.4, 5,6... .

− —>
Generally, for all spectral lines of hydrogen we have:
—-
I I I

n2 m2
which, when written as follows:

fits in with the Bohr differences in energy levels.

Considering:

i
cha
we have an equation:
12 21r’e‘m I 1

− − eh”
(n12 1122)
which is equivalent to Rydberg’s CXpression for Balmer’s hydrogen
series. The remarkable thing about this is that such an approxima-
tion is obtained with an equation whose constants e, m, c, b have
been obtained independently of the spectrum of hydrogen.°
Widely acclaimed as a superlative product of imaginative
genius, the Bohr quantum system ,was found to be applicable to
more complex atoms than hydrogen, as well as to the light con-
ditions of molecular substances. For‘our study of system products
it is significant that the Bohr atom system was successful insofar
as it-\in‘tegrated and organized the kinds of data With which it was
\
5 Tucker, in‘Taylor, Physics, p. 824.
68 PSYCHOLOGY AND LOGIC

intimately concerned. It is of cardinal importance to keep before

us the enormously detailed operations of Bohr, Sommerfeld, and
numerous others, as they attempted to combine the data obtained
frOm spectroscopy and chemical behaviorIn order to construct the
various arrangements of electrons'1n the atoms of different ele-
ments.
NO leSS important is it to consider the inherent weakness of the
system occasioned by combining classical mechanics with quantum
ideas. In a letter to Bohr, written in 1913, Rutherford speaks of
a grave diflicult‘y in the hypothesis:
“.. .namely, how does an electron decide what frequency it is. go-
ing to vibrate at when it passes from one stationary state to the other?
It seems to me that you would have to assume that the electron knows
beforehand where it is going to stop. ”7
o

It is part of our illustration to suggest that with the discovery

of the Compton effect and the discrepancies of the photoelectric
process physicists had to begin the construction of a wave instead
of a particle system of quantum mechanics.
((1) The System of Chemical Elements. The propositiOn:
The properties of the elements, and therefore, the properties of the
Simple, and of the compound bodies formed from them,'are in periodic
dependence on their atomic weights
marks ,the construction of a scientific thing system of major logical.
importance. The apparent difliculties in organizing the atoms of
all substances, and the exceptions that the arrangement disclosed,
brought to light the work and conditions making for a system
product. Even to N ewlands (1 838-1 898) the correlation of chemi-
cal properties and atomic weights was apparent. To Lothar Meyer
(1830— 1895) and Mendeleeff (1834-1907) the greater
edge of both atomic weights and chemical properties offered ma-
knowl-
terials for setting up a system. ItS success in enabling gallium
(I 875), scandium (I 879), and germanium (1 886) to be predicted
testified to the acumen of the systemizers in choosing a criterion
and collecting information.
Though the Meyer-Mendeleefi periodic table'IS rightlyrre-
garded as an enormous scientific triumph, it is also characterized as
∙ Eve, Rutherford, p. 221.
 SYSTEMS AS CONSTRUCTION PRODUCTS 69
an empirical construction and a comparatively simple taxonomic
structure. The value of the system, with its important role in the
prediction and discovery of new elements, Obviously stems from its
connection with actual, though imperfectly known, things. The
faults of the system—for example, incoordination of pefiodic
chemical properties and atomic weights, variance of position and
property resemblance, fitting elements into groups despite the con-
traindication of their major valences, etc.,—are traceable to the
relative insufficiency of knowledge and the need of increased and
modified operations.
Assuming a continuity in the system of elements, we find a
distinct evolution from the periodic phase, based upon atomic
weights, to Mosley’s (1887-1915) establishment of the criterion
of amount of positive nuclear charge. A more significant and
fertile system was achieved by arranging the elements on the basis
of the number of unit charges of the atom’s inner structure. All
the elements, without exception, fit into their group formulae and
relative table positions. In a sense, too, the table becomes self
validated; between hydrogen and uranium there can be only
92 elements.
The evolution of the chemical—element system presents a lucid
picture of the operations in thing system making. Following an
increased acquaintance 'with things come modification, enlarge—
ment, and general reconstruction. The develOpment of radiation
science, the discovery of isotopes, and, in general, the prevalence
of electrical events in the existence and changes in things all have
their influence. These developments are pointed up in Mosley’s
(1913) observation that Spectral lines are displaced toward the
shorter wave lengths as the atomic number of elements increases.
The achievement of a more significant arrangement of chemical
elements yielded its reward in prediction and discovery. Soon
masurium with atomic number 43, illinium with number 61, haf-
nium with number 72, and rhenium with atomic number 75 were
announced.
(e) Aesthetic Thing Systems; Interesting examples of thing
systems are found in the various ways writers have constructed the
factors or elements of aesthetic objects. To* reduce or analyze an
aesthet\ic object to form, content or subject matter, and meaning
7O PSYCHOLOGY 'AND LOGIC

or Significance illustrates an abstractive type of generalized system

product. Compare different systems of factors alleged to make up
aesthetic products—for instance pure form by itself (one-factor
system), form and content (two-factor system), and form, con-
tent, and meaning (three-factor system). Each of the factors, of
course, as well as the different combinations, represents a special
Creation.
(f) Petrie’s System of Sequence Dating.8 Two maj or problems
in archaeology are the recognition of difierent stages in the de-
velopment of a culture, and the"corre_lation of these stages within'
a chronological framework. In many regions of the Near East-the
remains of various cultural phases were deposited, one on top of
another, in mounds; hence, it is possible to solve the problem of
chronological sequence by the excavation qf ancient settlements.
Very little such stratigraphic evidence is available for predynastic
Egypt. Nearly all the material remains of that time were re—
covered from graves, which in all periods were dug from the
surface of the desert. When Sir Flinders Petrie found, for the
first time, a large number of prehistoric graves, he faced the prob-
lem' of establishing the time relationship between the various in-
dividual burials, each of which was a separate unit physically un-
related to the others. To solve this problem Petrie developed his
ingenious sequence dating system.
He noticed that one general type of pottery vessel—the wavy-
handled class—occurred in many graves, but in different forms.
Some had well-made handles and ovoid bodies; others had rudi-
mentary handles and roughly cylindrical bodies. Certain graves
contained intermediate forms. Petrie correctly assumed that these
various types represent difierent stages in the development of\ the
wavy--handledjars. He assumed further that the functional handles
did not evolve from the rudimentary ones, but that the reverse
was true. This View was confirmed by the occurrence of the cylin-
drical types in graves with pottery of the First Dynasty, and
complete subsequent disappearance of this type. On the basis of Iithe
this evidence Petrie arranged all graves containing wavy—handled
jars in an approximate chronological order.
But not all prehistoric graves contain Wavy-handled pots: In
is
.order to incorporate such graves in his chronological system
3 For this material I amindebted to Dr. Helene J. Kantor.
 SYSTEMS AS CONSTRUCTION PRODUCTS 7I
Petrie assumed that all objects found in a Single undisturbed
grave must be contemporaneous. This means that other types of
pottery (black—topped, polished red, etc.), found in the graves
containing wavy—handled vessels, could also be relatively dated.
Through them, many burials without the wavy—handled class of
pottery could be fitted into the Sequence. This method was, of
course, extended, to other classes of Obj ects (palettes, stone vessels,
etc.).
Petrie observed also that the wavy—handled pots never occur in
the Same graves with white cross-linedivessels, and concluded that
these were earlier than the earliest wavy-handled jars. In this
manner the relative data of still another group of graves was.
established.
Accordingly, Petrie was able to arrange a large number of
graves in a relative chronological order. He then arbitrarily di—
vided this series into fifty groups, each of which represents a par—
ticular stage of change. To allow for possible finds of earlier date
he numbered these groups-from 30 to 80 and referred to them as
sequence dates.
When an object is assigned to a sequence date, it is placed
within a ’particular phase in the development of predynastic cul-
ture, but is not dated in terms of years. A sequence date may repre—
sent either a long or Short period. However, there is no evidence
as to which sequence date was relatively long or short. By means
of the sequence dating system Petrie was able to distinguish clear-
ly two successive predynastic cultures, Amratian and Gerzean.
(2) Event Systems. Everyone sensitive to the specificity of
systems and to the necessity of keeping 'in View the building ma-
terials is alert to the many ways in which events of all sorts are
structured. The following examples are chosen to emphasize the
variety of materials or content.
(a) Chemical Events. Tables of chemical reactions and con-
stants, and handbooks of such tables, certainly illustrate system
building, if only of an elementary and casual sort. The great num-
ber of such events, and the corresponding need to organize them
for effective use, require classifying and ordering operations
(chap. 2). Chemical work demands systems of solubilities, densi-
ties, boiling and freezing constants, and so on.
(b)\Biological Events. Workers concerned with biolOgical
72 .PSYCHOLOGY AND LOGIC

events can not be limited to systems of constants or fixed units, as

in 'the case of chemists, but must deal also with variabilities of all
sorts. Accordingly, systems need to be devised for handling ranges
and limits. Classes are more or less arbitrarily chosen as estab—
lished; frequencies are selected in order that comparisons and
other relations can be set up. On the whole, statistical techniques
must be invoked.
(c) Historical Events. As our third illustration we choose the
enormous mass of events conventionally regarded as outside the
pale of the rigorous and settled‘domain of logic. Still, all those
interestedin intricate political and social events, and in human
happenings in general, ‘Strive to order and “relate them for pur—
poses ~of control and understanding. It is an invalid argument to
deny the systematic character of philosophies,of history because one
can not accept the criteria (working out of spirit, right, etc.) and
the products resulting from such system-buildifig activities. Such
denials simply bespeak a preference for the certainty, simplicity,
and compactness of systems, with abstractive sentences as building
materials.
(3) Relations as System Materials. From the operational point
of View, relations constitute things to be ordered and integrated in
logical work (chap. 15). Operational procedures uflpon- Telations
imply the manipulation of related things, or records of, and sub—
stitutes for, the relations. For example, in material-implication
systems the symbol T stands for the relations FF, TT, FT, but
not for TF. Relations importable into logical situations are already
small-scale system products. For instance, basic and crude natural
relations may be manipulated as a preparation for logical opera-
tions. As manipulable materials, relations are things both for\ the
local systems into which they are Set and the larger systems into
which they are ordered.
(4) Classes as System Materials. More definitely than rela-
tions, classes constitute constructed things, inasmuch as they'al-
ready consist of minor or small—scale systems previously .‘de-
veloped. The possibility is not excluded that the term class may
refer to genuine independent events, such as similarities1n things,
which make them potentially classifiable when the
-operates upon them.
investigator
In system——building situations, however, the Important task18
 SYSTEMS AS CONSTRUCTION PRODUCTS 73‘

organizing Similar things by isolating them from others on the

basis of one or more criteria. The very existence of classes Shows
that a Significant.system—building enterprise has occurred. Again,
on an operational basis classes may exist simply by casual asser-
tional activities. Asserted or alleged classes are also the raw ma-
terials of systems. The procedural and localized character of all
logical situations is something not to be ignored.
And finally, classes may be arbitrarily brought into existence
irrespective of any natural or independent properties. In this sense
classes are more autistic than in the assertional situation, Since in
the latter instance the references may have an existential basis.
(5) Acts as System Materials. Acts or behavior, no less than
any other material, can be structured. The only requirement is
that the Operator be able to interbehave with actions as stimulus
objects. This is no difficult matter as long as the acting thing—w
particle, organism or person—is in observable range. The simplest
system of acts, as in the case with other materials, consists of vari-
ous classifications. Examples are found in every science; for in-
stance, voluntary and involuntary action in physiology and psy-
chology, reversible and irreversible action in chemistry. Through
series of'organizing operations the logician or system maker can
interrelate behavior to form all sorts of systematic arrangements.
These behaviors can themselves be organized into a\hierarchical
structure with a' base in definition and classification.
(6) Linguistic Things as System Materials. A literate universe
is.a fertile field for every possible type of language thing system.
An obvious and inescapable system is represented by the hierarchy
of letter, syllable, word, phrase, and sentence. A better example
indicating the essence and scope of linguistic thing systems is
available in classic grammar. The departments of accidence, lexi-
cology, and syntax comprise innumerable systems of parts and
wholes. Words are constructed from roots, affixes, suffixes, and
infixeS in a rich array; then the words themselves are syntactically
interrelated in series from one-one, through one-many, to many—
many structures of illimitable variety.\

B. Linguistic Systems.
Linguistic systems constitute unique system‘ological products.
Essentially, they are organizations of responses which can be de-
74 PSYCHOWY AND ∞
scribed as references or designations. When building such systems,
the system builder works with referential actions intimately and
inseverably interrelated with the things referred to.
That linguistic systems consist of actions and not things can
scarcely be overstressed. Although the products represent an em—
phasis on language factors, they must not be confused with lin-
guistic content systems of which language things are the ingredi-
ents.’
The constituents of language systems are describable as asser-
tions of every sort and for every purpose. Such assertions may be
magnified to refer to things of cosmic import, or constructed to
designate simple and petty details of events. It is not a reflection
on logic to indicate that organizing a complex court plea is building
a logical system. To distinguish such logical acts from others we
might characterize them as mediate, remote, or nonexecutive re-
sponses. The system builder is handling things assertionally.
At this point we must again call attention to the factors of the
system-building field. In referential acts the system builder’s work
takes precedence over what he interacts with. In this sense linguis-
tic systems are first-degree language systems. Such acts, when re-
ferred to, recorded, or fixated in transcription, become descrip-
tively transformed into second-degree or thing—language systems
(p. 73). Second-degree language systems take their place among
the content structures.
( I) General Reference Systems. The primary distinguishing
feature of these systems is the intimacy of the references. Even if
the referents are regarded as important, there is a personal factor
in the system.
(a) Operational References. Instructions to perform operations
with respect to given things form genuine, and sometimes ex-
tremely important, systems. Their range (p. 84) is enormous
when we take into account references to actions, such as preparing
a prescription, following a recipe, and carrying out complex mathe-
matical operations.
(b) Systems of Argument. The organization of a series of asser-
tions to form arguments is a good illustration of a type of linguistic
system. Consider a lawyer’s series of references to alleged events
For a discussion of language behavior and language things see Kantor, Objective,
chap. 2, and Principles, chap. 23.
 SYSTEMS AS CONSTRUCTION PRODUCTS 75
in order to establish a particular belief or conviction. Perhaps a
more formal, and hence more tightly knit, aggregation of asser-
tions is exemplified by the classic paradoxes of Zeno.
I. Motion is impossible because a body cannot arrive at another
place without passing through infinitely many and infinitely small
separate parts.
2. Achilles cannot overtake the tortoise, since he must first reach
the point simultaneously left by the latter. The tortoise will thus al-
ways be in advance, if only by an interval which constantly decreases
to a minimum.
3. The flying arrow is really at rest since at every instant it occupies
a discrete point of its path. Of such zero movements no genuine mag-
nitude can be summed.
(2) Metizod Systems. Important examples of linguistic systems
are found in various reference organizations for the conduct of
reason and scientific work. Here we may cite such classic systems
as those of Descartes, Newton, and Mill.10 These systems, as or-
ganizations of references to operations, difier from simpler opera-
tional reference systems in their generality and specialization for
achieving important results in comprehensive situations.
(a) Descartes’ precepts for rightly conducting the reason and
seeking for truth in the sciences.11
The first of these was to accept nothing as true which I did not
clearly recognize to be so; that is to say, carefully to avoid precipi-
tation and prejudice in judgments, and to accept in them nothing
more than what was presented to my mind so clearly and distinctly
that I could have no occasion to doubt it.
The second was to divide up each of the dificulties which I ex-
amined into as many parts as possible, and as seemed requisite in
order that it might be resolved in the best manner possible.
The third was to carry on my reflections in due order, commencing
with objects that were the most simple and easy to understand, in
order to rise little by little, or by degrees, to knowledge of the most
complex, assuming an order, even if a fictitious one, among those
which do not follow a natural sequence relatively to one another.
The last was in all cases to make enumerations so complete and
reviews so general that I should be certain of having omited nothing.
'° Mill’s system is treated in chap. 19.
“ Haldane and Ross, Philosophical, vol. I, p. 92.
76 ' PSYCHOLOGY AND LOGIC

(b ‘Newton’s rules of reasoning in Philoso Pby.12

Rule 1. We are to admit no more causes of natural things than
such as are both true and sufficient to explain their appearances.
Rule II. Therefore to the same natural effects we must, as far as
possible, assign the same causes.
Rule III. The qualities of bodies, Which admit neither intensifica-
tion nor remission of degrees, and which are found to belong to all
bodies within the reach of our experiments, are to be esteemed the
universal qualities of all bodies whatsoever.
Rule IV. In experimental philosophy we are to look upon proposi-
tions inferred by general induction from phenomena as accurately or
very nearly true, notwithstanding any contrary hypotheses that may
be imagined, till such time as other phenomena occur, by which they
may either be made more accurate, or liable to exceptions.
b
(3) M etalogical Systems. Metamathematics and metalogics
constitute unique forms of recently developed linguistic systems.
Essentially, these are systems of reference to the rules for, or
operation boundaries of, complex system building and their com—
ponents.
(4) Autological Cosmic Systems. Autological cosmic structures
in science and philosophy provide excellent illustrations of large-
scale assertional systems. Most of the absolute philosophical sys-
tems (Hegel’s dialectic absolutism, Schopenhauer’s will projec-
tion) are determined by traditions and doctrines wittingly and
unwittingly adopted by the system builder. Such assertional struc-
tures are quite remote from things. Contrast such arbitrary
structures with those that are genuinely referential and descriptive,
and which form integral parts of concrete enterprises. For example,
in the mathematical domain, systems of description and explana-
tion refer to the operations and things involved1n the
mathemati-
cal task. Similarly, descriptions and recordings1n scientific work
make up the linguistic structure which'13 an essential feature of the
operations.
∙
Formal Systems
The factor stressed in the formal system product is the structure
or organization, rather than its material or use. The basic specifi—
cations for formalistic structures center about such items as sym-
1”Principia, p. 398.
 SYSTEMS AS
CONSTRUCTION PRODUCTS 77
metry, balance, regularity, order, and completeness. By contrast
with material systems, formal structures depend greatly upon the
system builder’s attitudes and motivation. Thus he and his work
stand out in the system—building situation. Observe how the logi—
cian in choosing the domain in which to localize his system favors
the one Offering least resistance to Shape or structure.
Two primary considerations govern the construction of formal
systems: First, the achievement of certainty, and secondly, the
satisfaction derived from closed and finished products. Seldom
\do logical writers indicate that the achieved certainty and finality
constitute functions of the constructed systems. Obviously, the
intimate and direct path from premises to conclusions is a product
of the logician’s Work. His cOnstructionS, Of course, may be in—
fluenced or determined by the material structured, but in that in—
stance the formal features of the system are secondary in the
enterprise. Also, in such “empirical” situations the structures are
not so rigid and stable as when actual characteristics of things are
abstracted or canceled out. We offer four examples of formal
system products.
(I) Implicatory Systems. Traditional logicians, greatly con—
cerned with valid reasoning, have constructed numerous two—item
subsystems. Frequently these items have been chosen to Show
invariable .reciprocal relations. Such implicatory systems may be
illustrated by immediate inference and opposition squares.
(a) So—celled Immediate Inference. Demonstrations of valid
and stable inference naturally can best be made by setting up, or
isolating, situations involving closed mutual connections, such that
if A suStains a certain relation to B, then B must sustain a similar
relation to A. For example, if Chicago is West of New York, it is
at once inferable that New York is east of Chicago. If 7
> than
5, then 5< than 7. The term inference in this situation con—
notes a sort of game-playing activity. For the most part, it is
merely turning things about verbally. But this process is not to
be condemned as trivial Or remote from reasoning. Consider the
place of such activities in building formal system products, and
observe, moreover, the interrelation\between this type of infer-
ence and other activities referred to by .the\same term.
(b). Square of Opposition. The traditional square of opposition
illustrates a more complicated version of the immediate inference
78 PSYCHOLOGY AND .LOOIC

type of system product. The materials are verbal: alleged sen-

tences or propositions which are so interrelated that any one may
be inferred from the other. Including the diagonals, there are six
pairs of relationships.

A - E
C ontraries

a)
0 c3
E
8 0o,"e E
8
e ∙a:
N
t(b 0/,
N
.0
I.∙ °
(J “ed,
0

Subcontraries
I O
∙
\

The square is constructed to display the following points:

a. Contradictory statements can not both be true or false.
b. Contraries can not both be true, but both can be false.
c. Subcontraries can both be true, but both can not be false.
d. Subalterns are so related that if the universals are true the
particulars are true, but true particulars do not imply true uni—
versals. On the other hand, if particulars are false the universals
are false.
(2) Truth-function (Matrix) Systems. Consider as a typical
achievement of symbolic logic the truth- function table or mat\r1x,
which efiectively exemplifies formal system products. Beginning
with the criterion of material implication, formal system makers
have organized statements and their interrelationships to satisfy
the criteria of simplicity, completeness, and “playability.”
Q P'Q
P-i-‘Q P—>Q' P=Q !

T T T T T T
F T F T
↓ F

∙
T F F T F F
F F F F T
 SYSTEMS AS CONSTRUCTION PRODUCTS 79
Systems involving more than two values require more elabOrate
structures. Suggested modifications become necessary when sys-
tems of simple dichotomy are. opposed to those involving some
content material. On the whole, the two-valued systems can‘ be
kept free from content and SO-called meaning.
(3) Complete Induction. Of all formal situations that of
complete induc‘tion provides the best illustration of the .nature
of system products, chiefly because such systems involve an equal
stress of material and form factors.
The materials consist of relations organized into a stable struc-
ture. N o doubt, emphasis on material has given rise to the histori—
cal term induction. The interbehavioral process involved here iS
primarily the selection of type of relation to be organized. Com-
plete induction products, more than ordinary deductive systems,
stress structurization activity.
Turning to form, complete-induction products consist of systems
in all aspects comparable to the circular structure of deductive
systems. This is well illustrated by the-following classic example.‘
For all integral values of n: '
∙ −
Whoever first developed this system observed the relations
involved, and then set up a formula indicating them. The
formalization procedure consists for the most part of constructing
the formula, a procedure which belongs to the symbolic phase of
system making.
(4.) Mathematical Systems. The acme of formal systems is
presumably attained by constructing a pure mathematical system
or aggregation of axioms—that is, assumptions or postulates—as
in the following example.13
I. If A and B are distinct elements of S, there is at least one m-class
containing both A and B.
2. If A and B are distinct elements of S, there is not more than
one m-class containing both A and B.
. Any two m-classes have at least one element of SIn
There exists at least one m-class.
common.
Every m—class contains at least three elements of S.
All the elements of S do not belong to the same m-class.
No melass contains more than three elements of S.
1”Veblen and Young. Projective, p. 2f.
∞
 8o PSYCHOLOGY AND LOGIC '
N ow, although the synonyms axiom, assumption, and postulate
imply that mathematicians and logicians no longer accept absolute
and self-evident propositions, it is still true that the formality Of
postulate systems leads to notions of pure reason and ultimate
systems Of- objective relations. Many individuals concerned with
formal systems overlook’ the fact that systems can only be con—
structed by' operating upon building materials. Frequently it is
asserted that logical processes may be performed without any
knowledge of concrete objects to which the primitive propositions
or postulates refer,“ or that mathematical deduction can be made
without knowing what one is reasoningxabout.15 More careful
writers, however, go only so far as to say that a pure mathematical
system can be constructed without explicit reference to specific
subject matter.1" Such writers acknowledge, then, that a mathe-
matical system merely intensifies the work of any system which
consists Of abbreviating and abstracting specific complex materials.
Whether or not in presenting the final product any explicit refer—
ence is made to specific subject matter, that reference is definitely
indicated1n the operations performed.
Compare the above abstract system with the following more
concrete point--line system. IS it not obvious that the f_or_m_erwas
derived from the latter, which, aside from the last item, is essen—
tially Veblen and Young’s set of assumptions for a plane proj ec—
tive geometry?
I. If A and B are distinct points, there is at least one line on both

A and B.
If A and B are distinct points, there is not more than one line
3°

on both A and B.
Any two lines have at least one point in common.
There exists at least one line.
Every line contains at least three points of the plane.
All points are not on the same line.
No line contains more than three points of the plane.
In the present situation it is impossible to overlook theI ab—
I
stractor’3 work (chap. I 3,.3), and how that work constitutes
system building. Notice his labor1n mOving from geometry to the
sear

l
1‘
Carmichael, Logic, p. 1 IO.
Veblen, Problems.
Cohen and Nagel, Introduction, p. I 35. '
 SYSTEMS AS CONSTRUCTION PRODUCTS 81
logical system. That- such work with definite materials is indis—
pensable in all system building is amply enough demonstrated
throughout the evolution of all abstruse geometries, which are
based upon the observed relations of concrete things in practical
Situations.
How can. we fail, then, to emphasize the factors governing the
choice and arrangement of axioms? This, we ask, despite the
tremendous force of opinion that we Should disregard all influences
on the constructor, and focus only on the product.17 Does not the
entire modern theory Of postulation demonstrate the necessity to
take into account the constructor’s work and circumstances? Is not
the whole history of non-Euclidean geometry a sharp reminder
that the belief in Euclid’s infallibility long obstructed the evolu-
tion of geOmetry?13
AS to the particular influences on system makers in setting up
their geometric axioms, Why does Hilbert choose his five basic
items, whereas Pieri and Veblen each makes use of only two dif—
fering ones?19 There is no question that all these mathematicians,
no matter how they vary in their results, are influenced by specific
features of the geometric universe of discourse. True, they have
freedom to choose what aspects to stress, but since they are build-
ers in specific situations they are constrained by the subject matter
with which-they work.
Mathematics as Postulation. As we have indicated (Vol. I), the.
employment of postulational methods in mathematics emphasizes
system making and the role of the system maker. But we must
stress further the Specific interbehavioral details of mathematical
work. The whole point to glorifying Pythagoreans, Eudoxus, and
Euclid20 as the first explicit employers of the postulational method
calls attention to the systemizing operations in arithmetic and
geometry. To a certain extent this postulational method consists
of analyzing and stating the processes necessary for organizing a
system of things Or numbers. Incidentally, the mathematician’s
motivations are indicated here—namely, the felt need for order
1'
It is especially those who Work with abstract or formal systems who stress the
genetic fallacy, which they interpret as an undue interference with the study of
finished products by concerning oneself with irrelevances Of origin.

∙
YoDng,\Lectures, p. 28f ;Bell, Development, p. 302.
'3

p. 231.
2°
Bell, Development, 665.
82 PSYCHOLOGY AND LOGIC

and completeness to give rigor and certainty to the demonstra-

tions.
The recent identification of mathematics and logic (vol. I)
through the postulational method of systemization may be inter-
preted as a realization of the system——making operations more or
less common to both disciplines.
Intuition in Mathematics. The present conflict between intuition-
ists and formalists in mathematics sharply points up the need to
consider field factors in mathematics. In the first place, from an
interbehavioral standpoint we' observe here two types of work-
.ers, each stressing different assumptions and attitudes. The two
basic Views represent variant methods of handling the factors1n
the mathematical field. The formalists stress the results of
achieved system, whereas the intuitionists are unable to abstract
completely from the scene of work.
Both the intuitionist and the formalist are describing actual in—
as
terbehavior with relations stimulus Obj ects. It is fatal to assume
that the intuitionist is concerned with the sensuous materials which
he injects into his descriptions. On the other hand, one need not
assUme that the formalist implies any rational mentality which
exists above or beside his work, and which conditions or determines
that work.
Formal Systems as Logically Typical. Doubtless because logi-
cians have historically been concerned with universality and ab-
solutism, with certainty and necessity, they have regarded mathe-
matical and formal systems as typically and essentially logical.
But if logic is an enterprise of system building, a wide gulf exists
between the work of man and the omniscient and omnipotent pow-
ers implied by the absolutely necessary and certain.
Why have modern logicians turned to mathematics, to abstract
and formal systems, in order to satisfy their craving for the‘abso-
lute and the certain? The reverberating answer is that tiirning
away from particulars, from details of events, allowsxfo'r sta-
bility and fixity. To evacuate events, to leave the skeleton of
things, has always appeared as astep toward solidity
petuity.
and per-

Illumination on this point is provided by the classic condemna-

tion of Mill’s allegation that the propositions of mathematics
constitute generalizations from experience. The basic issue is clear
 SYSTEMS AS CONSTRUCTION PRODUCTS 83
in the controversy with Whewell.”l The latter, a Kantian, empha—
sized the permanence and validity Of mathematical principles and
propositions. Hence he argued for a priori factors, for the necessity
and givenness of axioms. As Mill22 put it, for those who hold to
Whewell’s View the truth of axioms is perceived a priori by the
“constitution of the mind itself.”
Mill, on the other hand, rejected the proposition of fixed
mathematical principles or axioms not empirically derived. In
the extreme, then, even 2 —|— 2— 4 may not be true on another
−

planet. The lesson here18 that the opponents in the Mill-Whewell

controversy simply build on two improbable notions Of mind. For
Whewell mind was a unified entity basic to all proposition making,
while for Mill mind consisted simply of a series of associated
states. The latter type of mind did not allow for any a priori
principles—that13, permanent and primary propositions beyond
experience.
What13 often overlooked is that Mill’s logic is just as absolu—
tistic as that of the rationalist or a priorist. The difference lies
simply in the constructs employed. Since Mill, no more than his
opponents, is concerned with concrete propositional products de-
rived from contacts with actual things, there is no merit in his
denying preexperiential principles or axioms as tools for organiz—
ing systems of propositions. On the other hand, we must recognize
the humanistic source of the most abstruse and solidly established
mathematico—deductive system.
The a priori—eXperience controversy is neatly solved by
ering actual system--building procedures. When we look upon
consid-
system products as the outcome of personal and cultural evolu-
tions, we see how systems can be set up with a very tight connec—
tion between premise and conclusion, such that the latter follows
directly from the former. But this is not incompatible with the
activity of constructing the system in the first place. Moreover,
“this View allows for the construction of such systems as imply ob—
jective relations between events, relations which force the cOnstruc—
tion of tightly knit systems. It would
not, however, permit any
absolute or nonpostulational
construction for
mathematics.

↕ ∙ Whewell, History.
I, p. 262.
84. PSYCHOLOGY AND LOGIC

D. Operational Systems
Operational systems emphasize executive action. They consist of
structures or organizations of techniques, methods, and work plans.
Although operational systems are conditioned by the things op—
erated upon, the factor stressed is the interbehavioral feature of
the structured situation.
'Logicians- willing to renounce the exalted character Of abstract
systems need not hesitate to number among operational products
every sort of prescription or recipe, whether the ordering opera—
tions are performed in the kitchen or pharmaceutical laboratory.
To be sure, such structuring of prescriptive operations may be
very .difierendy rated or evaluated from other systems; neverthe-
less, as a system it differs only in detail from the most abstruse
mathematical or logical systemizing. °
(I) Chemical Analysis. The immediate problem here is to dif-
ferentiate five cation groups and then to isolate the three members
of the first group—namely, lead, mercury, and silver.23
Step 1. Adjust acidity of original solution and precipitate with HCl.
The following cations will precipitate as chlorides:
Cations Pb”, Hga”, Ag.
Chlorides PbCla, Hga, AgCl.
Step 2. Acidify solution and precipitate with H28.
The following cations will precipitate as sulfides:
Cations Pb”, Bi+++, C11”, Cd“, Hg“, AS+++, As++++’sb+++, Sn”, SnVH.
Sulfides PbS, BizS:, CuS, CdS, Hg8, A5283, SnS, SnSa, etc.
Step 3. Add the proper amount of ammonia and H28 to the solution.
The following cations will precipitate as sulfides:
Cations Zn”, Ni”, Co”, Mn”, Fe”.
Sulfides ZnS, NiS, COS, MnS, FeS.
The following cations will precipiate as hydroxides:
Cations CH“, AIHI. _
Hydroxides Cr(OH)a, Al(OH):.
Step 4. Adjust alkalinity and add NH4Cl and (NHOaCOa.
The following cations will precipitate as carbonates: l
Cations Ba", Sr”, Ca”.
Carbonates BaCOs, SrCOa, CaCOa. j
The solution will contain the following cations:
Na, K", Mg”.
Step 5. To precipitate of Step I, containing PbCls, ngClz, and AgCl, add boiling"
H20.
Step 6. To confirm Pb++ add KaCr04 to solution and
precipitatePbCrO4. r
\
The materials for this illustration have been drawn primarily from Meldrum and
’3

Daggett, Textbook. ∙
 SYSTEMS AS CONSTRUCTION PRODUCTS 85
Step 7. To residue from solution in Step 5, containing ngCl: and AgCl, add NH:
and precipitate black Hg and HgNH2Cl, this confirms Hga“.
Step 8. Acidify solution from Step 7 with HNOa, the presence of white precipitate
AgCl confirms presence of Ag.

This system is, of course, simply a skeleton of the elaborate

set of directions which imply many other operations that the analyst
may be expected to carry out. It is assumed, for example, that he
is working on a level beyond the point where the system of Opera-
tions includes the elementary directions to keep the utensils clean,
to boil and pour chemicals a certain way, etc., etc.
(2) Solution of a mathematical problem.
To differentiate 3x2
the five following steps.
−
5 the procedure may be structured in

Step 1.
Arrange y 3x:I + 5
The operation here is simply preparing the situation for the
necessary future operations.
Step 2.
Y_-+Ay=3(x+Ax)2+s
“—'31:2 + 6X' Ax + 3(Ax)2 + 5
This step involves another preparatory operation, that of replac—
ing x by x Ax and then a calculative determination of the value
of the function y —|— Ay.
Step 3 .
y+Ay=3x +6' Ax+3(AX)’ +5
y=3x +5
y: 6X°Ax+ 3(Ax)2

Here the given value of the function is subtracted from the new
value to determine the value of y, the increment of the function.
Step 4.
AY —-6
— X + 3 Ax.
Ax
This operation divides the increment of the function by the in-
crement of the independent variable.\
Step 5.
. dy
∙ − 6x
dx
36 PSYCHOLOGY AND LOGIC

The derivative is obtained by finding the limit of the quotient of

step 4, when the increment of the independent variable (Ax)
varies and approaches zero as a limit.

SCIENTIFIC STRUCTURE AND DEDUCTIVE SYSTEMS

To consider the place of deductibn in science aids greatly in
elucidating the nature of system products. Fundamentally, this
problem concerns the efficacy of science in the control and predic-
tion of events. Writers who incline toward a 'rationalistic, and
\even
moderately absolutistic, View“ attribute the rigor and com-
pleteness of scientific structures to a basic deductive core. This
View is generalized into the claim that science is, at bottom, de-
ductive, and that the tradition of inductive science is false.
It is argued that prediction, as an important scientific principle,
constitutes the drawing of a conclusion from rigorously estab—
lished premises. Undoubtedly, the ideal of inva’riable relations and
necessary connections, or abstract implication, is overstressed,
whereas actual interactions with complex recondite events are
minimized. I

Instead of arguing for a deductive faculty or' power—some

autonomous reasoning technique—why not observethat proof and
rigor are achieved by setting up a symmetric or circular structure
of propositions? Such constructed systems we may properly call
deductive. When scientific workers erect a fairly complete and
well articulated system it may be viewed in two ways. First, know—
ing one part of the system, we can deduce the others. On the other
hand, prediction may be described as anticipating that the particu—
lar elements (items) will be found-at their particular points of
articulation in the system. To concern oneself with system con-
struction and system products keeps one’s speculations within the

∙−
domain of natural action, instead of allowing them to fall into
the abyss of transcendentalism.

INTERRELATION OF SYSTEMIZING OPERATIONS,

AND
MATERIALS
If logical work consists of system \building then the processes
or operations must be appropriate to the things worked upon. In
\
2‘
For example, Cohen, Reason.
 SYSTEMS AS CONSTRUCTION PRODUCTS ‘
87
every other field this point is a commonplace, in logic it has been
obscured by several traditional but false views.
Foremost among the latter is the formalistic fallacy, respon-
sible for the view that formalized structures exist independently
of things and of the auspices under which they are generated. 'In
other words, the actual work of system building is completely dis-
regarded. The. formalization of acts, processes, and descriptions is
undoubtedly a necessary and valid technique for organizing the
elements of a system. The importance and use of this technique,
however, is derived precisely from the fact that the elements are
located in a definite set of coordinates. In other words, to formal—
ize events, or our operations upon them, does not allow any re—
treat from specific conditions. On. the contrary, effective formaliza-
tion carries with it an index to the circumstances under which the
formalization occurs (p. 77). Hence it is alogical to regard
formal propositions as holding for a genuine and transcendent uni—
verse of discourse, or to erect general logic as the organization and
transformation of “language.”
What constitutes a logical system in one frame of reference is
illogical in another. In some logical systems, for instance, it is emi-
nently prOper to stress formality on the basis of, say, noncontra-
diction, but even here one must shun any arbitrary acceptance of
a criterion. N on-Euclidean geometry contradicts Euclidean. Surely
this onlymeans that we must take account of the specific conditions
of our systems, even if they are abstract mathematical organiza—
tions of.elements.
To treat the materials of a mechanical object as though they
were sentence things obviously violates the first principle of system
construction. Nothing can be more detrimental, unless it be sum—
marily to declare that logic is an abstractionistic enterprise which
organizes sentences.
Another fallacy concerns projection. Once a system has been
formulated, it is overlooked that'the product has been projected.
The difiiCulty is that one identifies the projected system with some
form of autonomous entity, without regard to the particular things
which have occupied the system builder.
A variant fallacy is confusing cultura‘l-customs—that is, struc—
tures built up through successions oftime—J—with nature. Abstract
logicSare the historical projections of cultural groups.
88 PSYCHOLOGY AND LOGIC

And finally, the universalistic trend in logic favors overlooking

the materials worked upon, since the assumption is made that there
are genefalized rules for organizing systems and that the manipu-
lations are more important than the materials structured.
From the viewpoint of the'present treatise, that logic is an en-
terprise of constructing organized wholes Out of relevant or se-
lected items in a specific situation or in a given field, certain con-
clusions follow.
In the first place, since we are always concerned with a particu-
lar job, we can never place a priori limits to the factors and prOc-
esses involved. Otherwise logic soon turns into illogic. Whatever
processes we isolate as logical operations are only random samples
taken for convenience from the records embodied in logical treat-
ises. To these must be added other processes, both formal and in-
formal. Some of the latter may be regarded as auxiliary, that is,
they are manipulating and experimenting activities which lead to
an acquaintance with ideas, techniques, and relevant methods of
system construction.
In the second place, we need to be alert to the great hiatus exist-
ing between the generalized and fixed descriptions of what we do,
the methods by which we do it, and the actual work. Recall the
frequently mentioned fact that while we can throw ‘Our'reasoning
into a syllogistic form of description our reasoning behavior may
actually be casual, uncertain, and hesitant. From our standpoint,
to ignore the simple and casual details of our materials, and our
approach to them, staticizes system building.

Characteristic Variations of System Products

System products vary on the basis of many factors, such the as
motives of the system creator, his methods of work‘, and the-ele—
ments with which he operates. Out of these variations arise lthe
following system types. ‘
Tight and Loose Systems. Tight systems are typified by .the
absolutistic structures designed to be impenetrable and without
exit. Their builders are motivated by the ideals of rigid and non—
exceptionable rules. Obviously, such itautological systems require
arbitrary techniques of construction, and- are usually formal, as
well as circular. They are built out of self-fashioned words.
In some cases, however, tightly knit systems may be constructed
 SYSTEMS AS CONSTRUCTION PRODUCTS 89
out of concrete things, but in such instances the materials are care-
fully selected and arranged on a preconceived plan. Tight content
systems may be illustrated by an organization of selected proper—
ties of things, as in the classification tree of Porphyry.
systems which stress content are relatively loose and tentative.
But loose systems may be so characterized because the builders
allow for contingence, for factors dependent on local needs and
conditions.
Complete and Expanding Systems. When building systems
which stress the form or product aspect, one may adopt such rules
and. criteria as to produce finished and permanent structures. Con—
trariwise, to stress primarily the materials or building operations
allows for expansion, correction, and development of the finished
products. What stage of completion is attained depends upon the
particular building blocks, as well as upon the persistence and
curiosity of the system builder.
Discovered and Constructed Systems. While every system is
constructed, specific systems vary in the freedom of organization
and according to the resistance of the systemized materials. Thing
or content systems allow great scope for selection and invention.
Still, the various manipulations are conditioned by the type of
materials organized. Descriptive and formal systems afford greater
scope forthe production of system products.
At bottom the issue of discovery and construction comes down
to the ratio of c/d. When working with geometries one is restricted
to three ranges, with similar basic starting points in the relations
with which one works. Only relatively different types of system
can be constructed by beginning with circular or rectilinear rela-
tions. When occupied with generalized features of physiological
and anatomical events, one may choose between (I) a continuum
of one—dimensional time, plus a three dimensional space, and (2)
a four dimensional time--Space continuum. Among other choices we
may mention (I) discontinuous vs. continuous items,and (2) indi—
vidual vs. mass events.
Witting and Unwitting Systemsk No matter how analytic and
searching a systematist may be, the\re are always influencing fac-
tors of\which he may be completely or partially unaware. The for-
malistic~ system builder seeking or assuming absolute knowledge
asserts\ that “we know the correctness of the syllogism in its ab-
9O PSYCHOLOGY AND LOGIC

stract form (i.e., when it is stated in terms of variables) without

needing any appeal to experience.”25 All that really happens is
that he either does not know, or ignores, the fact that he is talking
about a system product.
I

To be oblivious to the history and surrounding circumstances

of syllogistic or other formalistic systems is to risk confusing a
system or assertion product with some sort of Platonic existence,
and ignorance with a priori knowledge.
Whether or not a system builder knows the origin and nature
of a system or its components depends _upon the possibility of trac-
ing the system’s cultural evolution. What evolves as a cultural
product—a set of syllogistic triplets, for inStance—should not be
taken'as some absolute set of relations. Very few logicians stress
the fact that the angle sum of triangles is an evolved system, not\
a Platonic Real capable of being known a priori.
Absolutistic logicians unwittingly foster the Eult of ignorance.
Writers realistically or Platonically inclined argue against the
pertinence of history, and for the potency of analysis, in determin-
ing truth or establishing fact. What this view amounts to is a plea
for ignorance concerning the work of system construction. Validity
or truth established through a cultural process is accepted, along
with the implication that the cultural factors may well remain in
oblivion.
7‘”
Russell, Introduction, p. 204.
 CHAPTER XVI

INSTRUMENTS FOR SYSTEM CONSTRUCTION

MODELS, SCHEMATA, AND FORMULAE As LOGICAL INSTRUMENTS

SYSTEM building requires tools. The investigation of logical

Situations uncovers a formidable array of instruments and
apparatus aids. These tools, so effectively enabling us to carry out
our system-building operations, vary extensively in form, specific
use, and technical efficiency. They are all similar, however, in re—
flecting the fact that an individual is engaged in fashioning a
product.
The very tools we find it necessary to forge in the present chap-
ter are. a good illustration. These instruments, which we construct
in order to systematize our knowledge concerning logical tools,
are really valid for all kinds of systemological situations. We name
our tools: (I) models, (2) schemata, and (3) formulae. Every
system-making enterprise requires one or two of these types, often
all three are used in complex situations.
Models stress primarily the materials and products of the sys-
tem-building situation. They pertain to the properties and opera-
tions of.things and events. Functionally, models are analogically
descriptive: For example, mechanical models duplicate on a small
and manageable scale the machines that at the time are only
designed and projected.
Schemata emphasize the constructive work involved in system
building. Such tools are either employed as scaffolding or as the
essentially operative means of carrying out some project. As such
they facilitate the development of new systems, as well as the en-.
largement of those already set up,
Formulae operate primarily by way of fixing and symbolizing
items used and systems attained. Bracketing propositions with
formulae suggests the enormous number and importance of refer-
ential and calculated procedures in system building.

LOGICAL AND SCIENTIFIC INSTRUMENTS

Pr‘OpOrtional to the importance of logical instruments is the
difficulty of analyzing and describing them. To adopt the three-
92 PSYCHOLOGY AND LOGIC

fold classification mentioned above, we admit, constitutes only a

very elementary step in handling them. An item in our favor, how—
ever, is the fact that owing to similarities in logical and scientific
behavior1 we may profit from our familiarity with scientific models
to illustrate system-building tools. Formulae and schemata such
as propositions, algorithms and theories are employed as aids both
in science and logic. Carnot’s heat cycle, for instance, and Max-
well’s demon—operated sluice gates were constructed in the in-
terest of particular heat theory and general scientific system build-
ing. They are in all respects similar to the instruments
primarily logical purposes.
for
created
.I

LOGICAL INSTRUMENTS PRODUCTS As

The production of logical instruments may be legitimately com-

pared with tool making of any sort. In each case the basic event is
the system builders’ constructive activity, which, when properly
emphasized, restrains us from overstressing the product..N either
the system product nor the description of it must be allowed to
conceal the definite enterprise. To adhere tenaciously to the crude
data of system production finally brings us back to the \construc—
tive act, no matter what approach we first make.
WhatIs true for logical systems in general13 for theIn—
struments by means of which they are constructed. Excluded, then,
true
are all self-sustaining and self-evident assumptions, formulae, and
propositions—in brief, all conventional a priori principles. The
Pythagorean theorem can not be false: but only because it is a
part of a system so constructed as to make it true.
All such system-building tools as logical necessity, the distinc—
tion between matter of fact and relation of ideas, and various im—
plications must be evaluated on the basis of the constructor and the
situation in‘which they are created. To keep the constructional
progess in view—that is, to take seriously the product character
of logical tools—is .to clarify the entire logical enterprise. Like—
wise, the efiect of school and general cultural influences on logical
instruments will assume the prominence it deserves.

INTERBEHAVIORAL SOURCES OF LOGICAL INSTRUMENTS r

Although tools are constructed for the particular task at hand,
We might add everyday behavior, too, in its system-building aspects.
 INSTRUMENTS FOR SYSTEM CONSTRUCTION 93
it is frequently possible to borrow or adapt them, instead of in-
venting them. Among the striking instances of tool borrowing is
the adaptation of conic sections in astronomical science, complex
numbers in electrical engineering,2 matrix algebra in building up
quantum mechanics,3 and tensor calculus in constructing relativity
theory.‘ Creative instruments are illustrated by the definitory fiats
which result iii the existence of dimensionless points and one~
dimension lines.
Whether the system builder exerts himself much or little is the
criterion of his originality. Adaptation and creation actions are
points on a continuum, each representing interbehavior with cer-
tain things within the bounds of a logical situation. Whatever
tools are used constitute constructs for a given purpose.
Any given instance of tool making is illuminated by the ratio
of adaptive to initiative action required for its production. The
simplest type of model or schema is, of course, the reproduction
of some object on a given scale. Here the construction is based pri-
marily on variations in appearance and magnitude. In this case
the a in the a/i (adaptive—initiative) ratio is the smaller magnitude.
At the other extreme is the tool in which the i is at a maximum,
and where one can detect no resemblance between the original
stimulus object initiating the model or schema, and the apparatus
itself. In popular language the tool is asserted to be entirely imag-
inary, as, for instance, when the construction is arbitrary and repre-
sents no existing event. From the standpoint of interbehavioral
psychology, however, there is no dearth Of contact with original
stimulus objects, even thOugh the tool is extremely analogical, or,
as traditional language has it, logical. The most elaborate instances
of this sort Of construction are not regarded as frivolous or im—
possible. For example, need we recoil from the fact that Vassar
graduates reproduce 2.6 children?
Ratios with large i denominators fall into a range of autistic
behavior, and bespeak the assertiveness of system builders. On the
whole, such autistic construction is more in evidence when abstract
materials are being organized.

Dantzig, Number, p. 232.

3
Born, Restless, p. I 32.
Bell,\ Development, p. 194.‘
94 PSYCHOLOGY AND LOGIC

THE SPECIFICITY OF LOGICAL INSTRUMENTS

Logical tools must fit the situation’5 requirements. Models,
schemata, and formulae vary according to the type of system built.
Furthermore, models, on the whole, are better adapted to prac—
tical situations. By comparison, formulae and schemata fit more
closely abstruse and abstractive system building. In general, logical
tools are more eflicacious when they are analogical, representa-
tional, or symbolic according to the problem at hand.
LOGICAL INSTRUMENTS-IN OPERATION
Logical instruments, We have said,
designed for interbehaving with events
are
constructs (products)
in order to achieve prac-
tical and intellectual orientation. We can not regard them,there-
fore, simply as forms in the traditional seiise. Recall that in con-
.ventional logic a distinction is made between the formal and the
material, or the formal and the contentful.5 In other words, models
for the traditional logician are logical or formal as over against
the existential or the ontological.
Another way in which this distinction is made isto differentiate
between concepts, on the one hand, and things, on the other. A tell—
ing illustrationIs Dewey’s view
for example, applies only to the
that
the law of excluded middle,
formal realm, not to the onto-
logical. Dewey does not allow for things being exclusively of one
sort or another, he argues that the transition stages partake of both
extremes. This is certainly reminiscent of Hegelian dialectics,
which reduces ultimacy to some form of spiritualism or verbalism.
According to such logic, universals are formal elements or models
which are not in whole or part the things handled. Such views are
rooted in the theory that logic deals with sentences and/or propo-
sitions, not with ordinary interbehavior with things. Universals are
thus regarded as having merely formal, not existential being or
existence. I

In contrast, our interbehavioral view that logical instruments

are constructs immediately introduces a new factor. The original
situation may be simply some naturalobject, such as a stone,- tree,
or an organism’s action. NextIn the series, this original datumIs
reacted to by some observer. When the reSult of this observation
I \

5
Cf. vol. I, pp. 28f., Iozf., I59f., 294.
 INSTRUMENTS FOR SYSTEM CONSTRUCTION 95
is referred to or "described, we have some sort of construction,
whether or not regarded as a model. A genuine model is an
elaborate construction for some particular purpose—personal ori-
entation, or communication. But the most abstract construction is
always continuous with the crudest contacts of individuals with
things.
Envisaged‘as orientational tools, logical instruments are re—
moved from any contact with metaphysical problems. On the one
hand, we are free from a logic dealing with verbal abstractions re—
mote from actual things. On the other, we avoid the implications.
of Platonic reals, according to which propositions or mathematical
formulae are presumed to constitute realities. Logic on our basis
does not therefore consist of propositions or forms which consti-
tute an autonomous and unique realm, nor is it concerned with
fundamental tautologies—with language, as the logical positivists
have it. Rather, all models, no matter how elaborate, are specific
interbehavioral procedures fitting a particular logician’s purpose.
Because of this specificity, it is inevitable that certain elements
of the construction are not derived from the crude data. Every-
one, depending upon individual experience both professional and
private, has various predilections for certain models. For instance,
one person chooses verbal instruments as over against mathe-
matical. -.Another pov'verful factor is the worker’s cultural back-
ground. A scientist bred in a religious culture sees no incongruity
in using creative and Spiritual ingredients to account for a certain
event, 'whereas another person finds such elements extremely ob-
jectionable. In each case there is a thorough mixture of personal
experience with cultural conditions.

LOGICAL AND NONLOGICAL INSTRUMENTS COMPARED

Although there is no fundamental difierence between the in-
struments of science and logic, such that we can expound the latter
by referring to the former, variations do exist which are neither
unimportant nor negligible. To summarize these differences adds
to our understanding of both.
.The basic variation, of course, is simply that scientific instru-
mehts are forged for immediate investigative purposes. On the
whole, it is informing to regard SUCh instruments as Hilfshegrifi‘e,
96 PSYCHOLOGY AND LOGIC

Gedanhen—Kriichen, und Rechenpfennige, as German writers say.6

Logical instruments, on the other hand, are more intimately in—
volved with system building as such. Naturally these distinctions
are relative, since scientific work may well stress systemization,
while logical work may emphasize precise investigation.
In a brief but informing discussion by Rosenblueth and Wiener
on the role of models in science,7 scientific and logical instruments
are identified. These writers regard models as central necessities
of scientific procedure and divide them into (a) intellectual or
formal and (b) material types. Pitching their discussion on a
definitely operational basis, they consider a series of tools ranging
from commonsense models to scientific theOries. Formal models
are not kept within the bounds of work with things and the prob—
lems they evoke. In the limit.
The ideal formal model would be one which would cover the entire
universe, which would agree with it in complexity, and which would
have a one to one correspondence with it. 3

Theoretical models thus move out to comprehensive system mak-

ing, unrestricted by actual contacts with events.
The limits and character of intellectual tools have been well
demonstrated by the distinction which Rankine9 made between ab-
stractive and hypothetical procedures in science. The-former are.
assumed to start with concrete data; the latter to be more autistic
in constructional technique. Recently, Dingle10 has identified the
extreme hypothetical procedure with quantum mechanics, and the
extreme abstractive method with relativity theory.

LOGICAL INSTRUMENTS As MINIATURE SYSTEMS

We have already indicated thatIn some cases system--making
instruments are really themselves miniature systems. In conven-
tional logic a truth table may be regarded either as a schemallfor-
building a large logical system or looked upon as a small lbut
finished structural system. Similarly, a graph from one vantage
point may be an autonomous system,\whereas from another it is
°Winderlich, Ding, p. 29.
Ro.le ∙ f
Ibid., p. 320.
9 Miscellaneous, p. 245.
Science, chaps. 4., 5. ~.
 INSTRUMENTS FOR SYSTEM CONSTRUCTION 97
only an item in a larger system. Moreover, a theory concerning
certain specific events or some general structurization of data may
be merely an aid in building an extensive thing or theory system.
Thus we achieve a genuine relativity of systems and tools.
Models and schemata which in one situation comprise media-
tional factors in system—building enterprises, in another are end
results or products. The primary differentiating criterion of model
and schema instruments is independence or autonomy. When we
adopt the interbehavioral vieWpoint, there is no difficulty indis-
tinguishing between those situations in which models are smaller
and more immediate units, and those in which they become more
comprehensive, and, in general, systematic.

LOGICAL INSTRUMENTS AS SYSTEM ITEMS

When system building is treated as strictly operational, it is not
surprising that instrument products may serve as items in a system.
For example, the classical points, lines, and surfaces of geometry
were invented as instruments for organizing the facts Of space.
Once these instruments were constructed it became very difficult to
distinguish them sharply from the items in a space system. In
physics, too, we often find that models or formulae, originally de-
signed to aid in systemizing events, come to be entities making up
the system in question.‘

MATERIALS OF SYSTEMOLOGICAL INSTRUMENTS

The 'materials with which system—building instruments are
forged naturally vary with the tool types. We plan to discuss only
the general kinds of available material, irrespective of final prod-
uct. '

Fixated behavior of various sorts, traditionally called concepts

and ideas, provides extremely important materials. To begin with,
concepts and ideas constitute forms of responses to things and
situations. But along with attitudes, beliefs, and conj ectures, they
become fixated and transcribed as tools in systemizing things. To
such fixated behavior we owe the conventional assertion that
imagination is a tremendously important feature of science and
logic.
Words and signs, as constituents of propo‘sitions, of symbols,
and ∫ mathematical patterns, also serve as materials. Mathe-
98 PSYCHOLOGY AND LOGIC

matical patterns are exemplified by various types of curves and

their equations. Frequently, such curves, instead of being regarded
as analogical or descriptive constructs, are mistaken for the originals
they are presumed to represent.
The materials so far considered—namely, fixated acts, signs, and
words—are simple and uncompounded. But as we have already
indicated, they soon become integrated into all sorts of amalgams
and. mixtures, and can therefore be hierarchically ordered. Thus
it is possible to arrange in different series a profuse variation of
curves, icons, diagrams, symbols, schemata (nomog‘raphs, abacs),
tables (of functions, relations), scales, algorithms, etc.
THE ORIENTATIONAL FUNCTION OF LOGICAL INSTRUMENTS
Logical tools, we have seen, vary according to the type of inter—
behavioral situation for which they are develOped. A clear-cut
difference between them13 whether they are used for general or
special systems.1A more intimate differentiation depends upon
Whether the system builder is engaged in intellectual or nonintel-
lectual interbehavior. For intellectual operations four\ general
types of tools may be indicated: (1) representatiOnal, (2) ana-
logically descriptive, (3) explanatory, and (4.) speculative fic—
tional. The amount of autonomous construction involved is a
pivotal criterion of differentiation, and depends upon whether the
model makers are interacting with natural or constructed events.
Representational Tools. Consider those situations in which there
is the least addition of an autonomous sort. As the name repre-
sentational suggests, the operator faces a situation in which he
merely wants to indicate or represent some object, as in the case
of graphing the figures of a table or producing a model of a build-
ing in order to indicate its general form. Interaction at this level
does not depart very far from the first stage of contact with
things, or beyond what we may call the zero
simplest contact with an obj ect.
point—namely, the

Take another example. An individual has just observed

first time the rising tide on a beach. Now order fixate his
In to
fOr
the
Im-

pression, or to tell someone about It, be develops a drawing which

in some form represents the original change\ in the sand. In the
1‘Hempel and Oppenheim, Studies.
 INSTRUMENTS FOR SYSTEM CONSTRUCTION . 99
same way, a map constitutes a model construction of this type. So
far as systems go, we need postulate only that a set of these models
is put together for some purpose. Undoubtedly, such a system may
be regarded as the simplest in the logical domain.
Analogically Descriptive Tools.- When the original objects are
too complicated to be represented simply, they must be referred
to in a more constructive manner. ThatIs, one adds more or less
to them. Samples of such tools are verbal descriptions, which al-
ways involve some metaphorical elements. Mathematical formulae
of all sorts, built on the definite analysis or measurements of the
properties of things, are excellent illustrations of analogically
descriptive tools.
Explanatory Tools. On this level of interaction the general
orientation is one of understanding the events handled. In the
case of natural happenings, instruments are designed to relate ob-
served objects to other objects. Or, when in contact with artifacts,
scaffolds are erected in order to reach the events in question. Fun-
damental illustrations are the graphic representations of numbers,
and, in general, the geometrizing of numerical relations. The dif—
ferentiation between positive and negative numbers as points in
lines either to the right or left of a zero point constitutes such a
tool, as does the construction of the Statement that negative num—
bers are the addition Of what One owes as over against the enumera—
tion of what one actually possesses.
Speculative Fictional Tools. As implied, such tools are con-
structed for interbehaving with imaginary things. The construc—
tional factor, therefore, completely outweighs the direct—contact
factor, since the whole process is substitutional. Whatever the
purpose—whether it is an attempt to account for something, or to
control people and situations—one verbally or graphically creates
objects, processes, forces, etc., without much connection with
actual things-Such are systems‘of demons, universes, and deities.
At best, a constructional procedure of this fashion is guided at long
range by assuming similarities between observed things. .

In no sense is it necessary to exclude from our domain of system

building or instrument construction the most elaborately abstract
systems. Perfectly arbitrary starting points and materials find their
plaCesIn the continuity of system construction. Even such tenuous
 IOO PSYCHOLOGY AND LOGIC

and diaphanous materials as spiritual entities may be the building

blocks. It would be fatal, indeed, to any kind of logical theory to
exclude systems constructed by poets and metaphysicians. In such
cases, of course, it is necessary to indicate the extrapolatory lengths
to which constructors have gone. The_expert observer never fails
to trace step by step the system builder’s starting point, no matter
how far the final construction transcends natural events. In this
sense an objective or naturalistic psychology becomes an essential
factor in all logical theory.
INSTRUMENTS IN VARIOUS ORIENTATIONAL SITUATIONS
The following examples of instrument making and their use
in different fields of study throw into reliefa number of funda-
mental problems. °
Logic. Among outstanding logical instruments we must count
the general and comprehensive syllogism. Historically, of course,
the syllogism has been considered as a process of correct reasoning,
but modern logicians admit that it is simply a sort of construction
into which reasoning, when made into a specialized form (of pro—
cedure, can be placed. Perhaps the basic character Of logical in—
struments is even better revealed by the specialized syllogistic
figures—for example, Barbara and Celarent. Incidentally, the in-
spection of syllogisms indicates precisely how instrument construc-
tions are projected into the world of reasoning, and then made into
the stimulus objects interacted with, instead of being used merely
for certain orientational purposes.
The universe of discourse is another familiar logical instrument.
Whether regarded as a classificatory mechanism or as a set of
boundary stones to delimit a domain, it is a product constructed
for actual system-building enterprises.
Of considerable interest are the recently developed so——called
logical languages, such as those of logical positivism. Syntactic and
semantic characteristics of ordinary speech are employed as a basis
to develop language structures for transforming such systems.'In
this case, as Well as in that of the histOrical syllogism, the scafiold
is confused with the building for the cOnstruction of which
scaffoldlS merely a tool.
Ithe
∙ Additional technical logical instruments are\represented by the
various coherent and symmetrical word or symbOl structures de-
 INSTRUMENTS FOR SYSTEM CONSTRUCTION IOI

signed to orient workers in formal thinking or reasoning situations.

An example is:
Ifevery is‘.... and isa then is 12

Mathematics. As the domain of relations, mathematics is pre-

eminently the field where representational instruments are copi-
ously employed. Innumerable models, schemata, and formulae
have been constructed, beginning with geometric figures and num—
ber ratios, as in the relation of strings in Pythagorean theory.
An interesting type of mathematical instrument is found in
Klein’s description of the a, h, and c plans of mathematical devel—
opment.13 In his a plan the development begins with the formal
theory of equations—that is, operating with rational integral
functions and handling cases in which algebraic equations can be
solved by radicals. Next, is the treatment of the idea of power
and its inverses, which is basic to logarithms. Then follows the
handling of transcendental or trigonometric functions, and, finally,
algebraic analysis. Plan 5 is dominated by the ideas of analytic
geometry 5 it centers around the graphic representation of various
functions and the geometric treatment of curves. Plan c is algo—
rithmic; it concerns systems of calculation, in the sense of calculat-
ing with letters, and, in general, using calculative models for
mathematical development“
Such mathematical constructs as classes, constituting numbers,
or classes, types, and hierarchies of numbers, excellently exemplify
products of the tool-making enterprise. Other samples are signs
and symbols, algorithms, formulae and schemata of many varieties.
Deserving special mention are such instruments as the diagrams
and schemata representing systemizing operations and results.
A comparative gnomon diagram, which clarifies the principle that
the sum of.\any group of consecutive odd numbers, beginning with
one, equals the square of the number of terms in the group, is
shown in the accompanying figure (Fig. I).
No less effective mathematical instruments are the variant con-
structions for demonstrating geometric propositions. A glance at-
two (Figs. 2 and 3) of the 3 5 or more figures constructed to prove
the Pythagorean theorem“ is instructive here.
" See vol. I, p. 314.
Elementary, p. 77f.
" Hofiman, Pythagorischc.
102 I PSYCHOLOGY AND LOGIC

11

O ‘I 2 3 4» 5' 6
Figure I

K
O

D C

F E

G A H B

Figure 2 Figure 3

Physics. In its range from practical technology to all-embracing

theoretical systemization, physics has assiduously fostered every
variety of logical instrument.
For the technological department of physics the important type
of instrument is, of course, the mechanical model. Even today,
with the increasing ascendence of field theory, Kelvin’s assertion
is still valid—namely, that unless one can produce a mechanical
model one does not really understand the processes to be studIed.
On the all-or-none principle that f‘not all physical events are
mechanical” the view developed that ‘mechanical models should
be extruded from physics. Hence, the' model of the electrorriiag—
netic field which stresses not bodies in interbehavior but some-
'thing between them. Problems connected with \the deflection of
 INSTRUMENTs FOR SYSTEM CONSTRUCTION
103
the magnetic needle, and with the ether medium for wave trans-
mission, suggested' the invention of new descriptive and explana-
tory instruments.
Such inventions as tubes and lines of force carried the descrip-
tive constructs of electrical science from things to the more ab-
stractive domain of energies and processes not directly visible. To
construct systems concerning such events called for tools of a
schematic and operational sort. Directions and specifications were
included to provide orientation with respect to particular types of
physical events.
More subtle, naturally, are the various Symbolic instruments
developed on the basis of mathematical techniques. A striking il—
lustration is the use of the calculus to describe and explain motion,
with the result that mathematical models are employed to staticize
motion as a means of controlling dynamic events, as in D’Alem-
bert’s principle.
Another type of physics instrument is the reduction of such
natural events as colors and sounds to_ wave frequencies. In this
instance the instrumental character of models is strikingly evident,
since events like color, sound, and radiation lack any similarity to
the numerical symbols which represent them.
Biology. The consideration of biologically oriented instruments
for system building at- once brings to the fore the distinction be—
tween large— and small—scale systems. Among the former stands
preeminently the construct of the organic character of the cosmos,
a conception already present in Greek thought when the cosmos
was regarded as alive and breathing. In more recent times the
construct of the cosmos as an organism has been presumed to help
greatly in elucidating ontological problems.15
Not far from such cosmic systems is the building of a series of
mechanisms for comprehending biological things and events. To '

make organisms and their parts into bits or portions of machines

is to construct a wholly gratuitous and exceSsive analogy as a sub-
stitute for more or less elaborate system making. We call such ana—
logical constructions cosmic because they depart from specific struc—
ture and operation. These models may be of considerable interest
and value, but they do not enlighten us concerning characteristics
of the Originals. .
Whitehead, Science; Needham, Biologist’s.
I O4. PSYCHOLOGY AND LOGIC

The tools for building fairly small systems, close to biological

events, we designate as intermediate biological instruments. Here
we may cite such constructs as natural selection, adaptation, species,
and functions. Whether or not definitely recognized as such by
biologists, a potent form of instrument construction has been the
various classifications of organisms basic not only to taxonomic
but also to genetic biological investigation. The classification of
Linnaeus provides us our best example of an explanatory form
of schema.
The verbal construction of biological evolution is likewise an
excellent illustration of a schema for the relationship and develop—
ment of all organisms. As one might expect, most biological models
and schemata are unique to the organic field. Prominent among
biological tools is the oil-drop model, which, by its properties of
viscosity and sensitivity to temperature changes, simulates activ—
ities and properties of organisms. The iron wire of neural con-
duction,18 though sometimes severely criticized," has played a
large part in physiological circles. Not to be ignored either are the
many models and schemata concerning brain seats and centers.
Recall the hierarchical model of neural operations'as a telephone
or telegraph system in which the nervous system is made into a
large complex of centers controlling the remainder Of the organism.
The scope of such instruments can be measured by the enormous
substitution of alleged brain and neural control for events in-
volving complex factorial fields.
Immunological literature offers especially interesting biological
instruments. The lock-and—key model of the operation and effec—
tiveness of'toxins, antitoxins, and other antibodies is a paragon of
instrumental construction for system building.
Psychology. The instruments of psychological system making
have been modified again and again to bring the ∏ Into
some connection with events. This history starts with the develop-
ment of a purely fictitious soul, or mind, to set in place of the
person’s behavior as the psychological crude datum., Psychology,
of course, did not begin as a science \On a proper objective basis,
such as the Aristotelian treatises made possible. Instead, it was
completely dominated by theological presuppositions.
Lillie, Transmission of activation, Physical,
Rosenblueth and Wiener, Role, p. 31 8.
Transmission of physiological.
 INSTRUMENTS FOR SYSTEM CONSTRUCTION I05
Through the contacts which psychology gradually developed
with actual events there began the century-old attempts to con-
struct brain mechanisms with which to describe the workings of
the psyche, and thils to account for the existence and efiectiveness
of these tenuous processes in complex behavioral situations. Hence
the contrasting models, cooperatively constructed by psychologists
and biologists, which are designed to differentiate between volun-
tary and conscious action operating on the basis of cerebral func-
tions and involuntary, or unconscious action, working on the basis
of reflex arcs.
Neural models in psychological situations are employed to ef—
fect a transformation of crude data. The procedure is to assume
that neural mechanisms are somehow related to psychic processes.
Accordingly, the confidence which system builders derive from
their preoccupation with the palpable nervous system leads to the
continuation of the view that psychological events are actually im—
plicated with psychic processes.18 ‘

An historically interesting model constructed for building a

scientific psychological system is the “threshold of consciousness.”
This model is a subfeatIIre of a larger system developed by anal—
ogy with various mechanical models of particles crossing from one
region to another. Its character is obvious when one considers the
original event being described—namely, the availability or non-
availability of things for reaction. In other words, an object must
be of a certain size in order to be seen—in order, that is, to cross
the “threshold of consciousness.”
A construction related to the threshold mechanism is sense, dis-
tance, which vacillates between a model and a schema. Essen-
tially, it fixes intervals between things available for reaction.
Other distinctive psychological models consist of apparatuses,
such as a chemical analogy for conditioning. In principle, this type
of instrument resembles the oil-drop model for biological be-
havior.
The history of'psychology afiords us a series of schema created
for the purpose of making plausible the inexistent, yet efiective,
powers of mind. Interesting example‘sare' (I) the Wundtian chem—
ical analogy, and (2) the Jamesian functional analogy from biol-
ogy:
Kantor, Problems.
106 PSYCHOLOGY AND LOGIC

Social Sciences. An' impressive illustration of system-building

tools in this field is the Spencerian invention of the organismic
character of the state or society. That political or other human—
group entities constitute organisms is a construction which plays
an extensive role in descriptive and explanatory systems of political
science. Other examples from the same domain include economic
man, the average man, living wage, and standard of living.
TECHNIQUES IN INSTRUMENT PRODUCTION
How well the techniques of instrument making help to achieve
organization and pattern is a measure of their efiectiveness. Of
the many sources of variation in tool making the operators may
stress either the system product or the work of constructing it.
Again, the system may be more an organization of materials than
a carrying out of systemizing specifications. In every case the
techniques employed reflect the behavior of an isolated system
maker, or a plurality of such workers integrated in schools, logical
traditions or general cultural groups.
Naturally, the particular technique chosen coincides with the
system maker’s interest, as well as with the cultural backgrounds
influencing his work. Certain cultural milieux, for example, pro—
vide a system maker with a relatively large measure of freedom,
whereas another individual is seriously limited by political or reli—
gious controls.
Nor can we overlook the original motivation of the system
maker, as, for instance, his attempt to achieve simplicity or sym-
metry, or, in general, some sort of connectedness. In other cases
his motivation may be the desire to control events or their factors
in systems, or merely to obtain the satisfaction of understanding
or becoming familiar with the materials handled.
Sampling the behavioral techniques we find the following
uniquely representative:
Iconography. In simple cases, and in those involving immedi—
ately manipulative objects, tool making consists to a great extent
of producing an icon or a picture of a thing. A similar process is
that of symbolizing objects in ordef\to manipulate them, as in
drawing curves which rise and fall with the riSe and fall, increase
or decrease of the original objects represented.
Abstraction. We have already noticed that system builders con—
 INSTRUMENTS FOR SYSTEM CONSTRUCTION 107
stantly resort to abstractive processes, for example, in dealing with
periodicities we omit differences and slight variations, or in cer-
tain mathematical Operations we select our populations or attribute
certain class marks to the variables we use.
De N 0120 Construction. While there is no absolute difference
between what we might call free creation and pictorial representa—
tion, there are variations which can be measured in terms of the
distance one retreats from actual contact with the things with
which one interbehaves. Thus the creational process in constructing
logical tools must be regarded as highly distinctive and variant.
 CHAPTER XVII

CATEGORIES AND SYSTEM BUILDING

CATEGORIES AND LOGIC

SUPERFICIAL examination of logical works seems to

indicate that categorial problems are no longer treated in an
explicit manner. Actually there can be no diminution of interest in
categories, for they are inevitably involved with all logical proc—
esses and products. The fashion of neglecting categorial issues
simply reflects the fact that logic is not regarded as system build—
ing. As a result the following historical gnisinterpretations have
predominated: Categories are (I) ontologically abstract character-
istics of being, (2) ultimate elements of thought, or (3) linguistic
or grammatical elements.
Every systemizing operation, either casual and practical classifi—
cation of things or construction of elaborate technical systems of
science or mathematics, reveals the importance and prevalence of
categories. As the enormous range of system building suggests,
categories constitute such widely diflering items as quality, quan—
tity, event, and relation, used by scientific and ontological system
builders, as well as the defined and undefined terms, variables,
and constants of the formalistic or symbolistic logicians. When
categories are operationally described, all the different varieties
can be interrelated, both as items and as criteria and limits for
system building.

NATURE AND OPERATION OF CATEGORIES

Though categories as system-building instruments are not ex-
clusively related to conventional logic or philosophy, it islcon—
venient to illustrate their nature and operation by means of his-
torical category systems. Indeed, traditional category systems ex—
emplify the entire process of system building, as well as the char-
acteristics of logical products. No matter what criterion a system
builder adopts, his logical products reflect such typical categories
as completeness, universality, indispensability, and ultimacy.
What, then, are some of the characteristic“properties of. cate-
 CATEGORIES AND SYSTEM BUILDING 109
gories? From the traditional standpoint at least, they constitute
basic descriptive or evaluative factors for system construction. In
order to build ontological, cosmic, or epistemological systems,
philosophers have had to invent intellectual tools for identifying,
characterizing, relating, and organizing the items comprising their
particular type of system. Logicians such as the mathematical or
symbolic system builders, who are interested not in substances
but in formal operations, still search for such abiding systemic
factors as indefinables and the a priori. Though they do not use
the term category and regard themselves as somewhat independent
of the general categorial tradition, they nevertheless operate with
categories. Think merely of the various symbols and operators
of symbolic logic.
Categories, then, are basic materials of all conventional logics,
whether Aristotelian, Kantian, or Hegelian. In Aristotelian logic
categories assume the form of raw-material terms—namely, sym-
bols and substitutes for classes. In Kantian logic, categories con-
stitute the basic constructs underlying all analytical processes, and,
taken together with the aesthetic forms and dialectical ideas, they
make up the building materials of all intellectual systems. A simi-
lar role is played by the categories of Hegelian logic. Since con—
siderable information concerning categories may be gathered from
their cultural evolution, we glance briefly at the history of cate-
gory systems.

EVOLUTION OF CATEGORIAL SYSTEMS

Pre-Aristotelian Categorization. Philosophical or, better per—
haps, technical categories were first constructed by the pre-Aris-
totelian cosmologists of the Greek tradition (vol. I, chap. 3).
Striving toward a universal and simple summing up of nature’s
essence, early Greek thinkers constructed such categories as water,
fire, air, and the unbounded. That these categories were implicit
constructions for achieving large-scale results in man’s orientation
to reality was not, of course, recognized by the thinkers of the
period. Their procedure, however primitive, was essentially ob—
jective. They pushed toward the primary stuff or action of all
things, \without the elaborate and sophisticated analysis repre—
sented in Aristotle’s organized logical corpus.
Aristotelian Categorization. In Aristotle’s time, with the evolu—
IIO PSYCHOLOGY AND LOGIC

tion of a definite appreciation of expository problems, categories

were set up in a simple linguistic way as an attempt to summarize
all the verbal elements necessary to describe specific things. On
the basis of current grammatical knowledge Aristotle presented
ten categories,1 which subsequently became the paradigms for all
categorization attempts. Categorization for Aristotle was eXplicit
and sophisticated, although from the standpoint of later periods
the whole procedure appeared remarkably simple. It is tradition-
ally recognized that Aristotle merely set down words to repre-
sent the various characteristics of things. Another not less felici-
tous interpretation is that he simply attempted to describe things
in
theirdiverse appearances and relations.
The Ten Aristotelian Categories
1. substance 6. time°
2. quantity 7. situation
3. quality 8. status
4. relation 9. action
5. place 10. passion (being acted on)
Obviously, the categories listed above are elementary ‘obj ective
descriptive terms which would scarcely be diflerent, no matter
what language was spoken by their constructor.
Post-A ristotelian Categorization. Not until the development
of science in modern times were new categorization principlesIn—
troduced. This does not mean, however, that the interpretation
of given categories has remained the same.2 On the contrary, from
the earliest medieval reintroductionof Greek learning into Europe
the distinction between existence and thought, and the superiority
of the latter, have made categories more and more into symbols
for subj ectivistic, or thought, qualities.
Kantian Categorization. Not only did Kant adopt the person-
alistic character of categories, but he modified them to accord with
the Newtonian physics current in his day. Oneeof his most funda-
mental alterations was to difierentiate between the actual, the, pos-
sible, and the necessary. Kant conceived of categories as the most
esseIitial thought forms under which, \Or by means of which, elabo-
According to von Hartmann (Kategorienlehre, p. ix) Aristotle took over his
categories from the Platonic school. .
See vol. I, p. 48f., for discussion of Scholastic transformation of Aristotelian logic.
 CATEGORIES AND SYSTEM BUILDING III

rately abstracted things could be organized or synthesized, rather

than as a straightforward set of predications of existing things.
This is indicated in the accompanying categorial schema:
The Kantian Categories
I II III IV
Quantity Quality Relation Modality '

unity reality inherence

+ substance possible — impossible
plurality negation causality +dependence existence
— nonexistence
totality limitation community (reciproc-
ity between active and
necessity— contingency

passive)
The Categorial System of von Hartmann. Modern categorial
systems reflect the various metaphysical or philosophic viewpoints
of their constructors. Hegel, for example, criticized the Kantian
organization because it was too abstract and formal. For Hegel
categories must be ultimately constitutive of things. Von Hart—
mann, who has treated the category problem most elaborately,
attempts to organize a categorial system by adopting the results of
the Hegelian formalistic criticism, adding the correction that cate-
gories, in addition to being logical, must include the nonlogical,
and even unconscious, characteristics of a world system. His re-
sults are indicated in the accompanying tabular scheme. For the
most part, the represented items are treated as they are presumed
to operate in the subj ective-ideal, obj ective-real, and metaphysical
spheres.
Von Hartmann’s Categories
A. Sense Categories
I. Sensation categories
I.
Sense quality
2. Sense quantity
a. Intensive quantity
b. Extensive quantity or temporality
II." Perception categories
Extensive quantity or spatiality
B. Thought Categories
I. Primitive category of relation
II. Categories of reflective thought
1. Categories of comparative thinking
2. Categories of separating and combining thought
i

3. Categories of'measuring thought

112 PSYCHOLOGY AND LOGIC

4.. Categories of inferring thought

5. Categories of modal thought
III. Categories of speculative thought
I. Causality (etiology)
'2. Finality (teleology)
3. Substantiality (ontology).
Current Attitudes toward Categories. Contemporary philoso-
phers, inclining more toward the unification of science, or the
criticism of language and concepts, than toward elaborate total
systems Of the universe, content themselves with small numbers
of categories. A recent writer3 proposes only three basic categories
for a naturalistic philosophy—namely, event, quality, and rela-
tion. Embedded in this proposal is a shift from the substantive
matter and motion categories of older QphilOSOphical systems
toward the currently more favored functional categories.‘ A‘critic
of the three-category system argues that the’ event category is
not as basic as those of quality and relation, and apparently as-
sumes that the latter are sufficient for all purposes.5
The rejection of a single exclusive metaphysical system radi—
cally” changes one’s attitude toward the categOry problem) For one
thing, there is no longer a need for a unique and inclusive set of
basic categories. In fact, for various nonmetaphysical purposes
many categorial systems are required. Pepper6 has recently pro—
posed that each difierent world hypothesis demands its own type
of category system. World hypotheses, however, are simply se—
lected metaphysical systems. Aside from asking why we need world
hypotheses as metaphysical systems at all, we object to the implied
limitation of the number of categorial systems. Such limitation
sets up a faulty presupposition concerning the nature of categories.

CATEGORIAL SOURCES
Writers of contrasting cultural backgrounds locate the basis of
the universal and necessary thought elementsIn somewhat1dif-
ferent sources. We have just referred to the Kantian notionthat
categories represent a priori forms under which things must be or—
\ ' I

’ Dennes, Categories. P
Cassirer, Substance. ‘

Aiken, Notes.
0 Categories.
 CATEGORIES AND SYSTEM BUILDING 113
ganized in knowledge. Essentially, this source of origin derives
categories from the ultimate character of mind. Mind is thus re—
garded as some sort of cognitive power, which, even if it does not
supply the laws of knowing, does provide the forms in which
knowledge of things is organized. Actually, of course, mind is
made to do more than synthesize objects for experience; by the
categorization technique it “produces” phenomena. The source
of categories, according to this View, lies in the individual—specif—
ically, in his cognitive power.
In contrast stands the View that the mind derives categories
secondarily from cultural sources. In Spencer’s type of evolution-
ism the basic categories for the individual mind are evolved
thrOugh hereditarily transmitted racial habits. Such a theory as
Spencer’s is Often taken to constitute an empirical doctrine. For
the members of the French school of social anthropologists, such
as Durkheim and Lévy—Bruhl, categories are forms of knowledge,
ways in which things and conditions must be cognized, and these
ways are derived from the characteristics of particular social or—
ganizations. The emphasis here shifts from individuals to groups,
from individual to collective mentality. It is part of the French
sociological doctrine that individual mentality reflects the group’s
general mentality. Aside from the difliculties involved in a con—
ception of social mentality, we confront the fallacious idea of es—
sential and‘ fixed forms without flexibility and without the local
functional character which actual categorization entails.
on
Even though writers categories depart considerably from the
traditional emphasis of the synthesizing power of the mind, as
represented, for example, by such a study as Windelband’s" they
do not escape the notion of ultimate and totalitarian analyses of
things. Turning toward the analysis of terminal properties only
partially modifies the interest in an underlying mentality capable
of reaching out to ulterior categorization processes.
To eschew both the direct and indirect connection of categories
with mental powers makes room for the operational enterprises
which persons foster either in furtherance of personal achieve-
ments or as part of large-scale undertakings in which many indi—
viduals participate. At once the sources of category making are
extended'indefinitely.
System.‘
114. PSYCHOLOGY AND LOGIC
CATEGORIES AND CONCEPTS
Since the development of subjectivism and dualism, and their
application to logic, categories have been equated with conceptsf"
In other words, categories have been regarded as mental or psychic
entities which play a part in the operations of judging and reason-
ing. This identification, we have seen, attainedIts full floweringIn
the Kantian theory of categories.
Nothing is more certain than that this identification is based
upon the confusion of our reactions to things with those things
themselves. Notice the sharp transition from the Aristotelian cate—
gories as descriptive elements of things+or the referential ele—
ments of such descriptions—to the Kantian assumption that cate—
gories are the mental processes by which the creative mind organ—
izes objects and events (vol. I, p. 59). The Kantian doctrine is
essentially a rounding out of the spiritualistic principles developed
by Patristic and Medieval thinkers.
Even when concepts are not regarded as psychic states but as
forms of interbehavior, they are, not to be identified with cate—
gories. In operational situations categories are definite construc-
tions—that is, products developed for system-buildingipurposes.
As such they constitute manipulable objects. The range of their
character and use varies as the systems which they help to form
vary from simple classification and description of particular events
to large generalized ontological or epistemological structures.

CATEGORIES, CLASSES, AND UNIVERSALS

How are categories related to classes and universals? Not a
far-fetched interpretation is required to number classes and uni-
versals among the categories. At the very least, we must stress
their great similarity to the latter. Traditionally,classes and uni-
versals (chap. 18) constitute structural elements built tip for
particular purposes by means of verbal or symbolic fixation,
whereas categories have been regarded as more general and per—
vasive. Historically, classes and universals have always been con-
nected with large— scale ultimate systems. From an interbehavioral
standpoint such large—scale systems are merely special cases, and
system building, which constitutes concrete activities in all orienta-
tional situations, is common to all intellectual “work.
Cf. Child, Categories; also Wolff, Unique.
 CATEGORIES AND SYSTEM BUILDING II 5

Classes and universals, therefore, are merely limiting factors

in various descriptive and explanatory enterprises. Since no sharp
line exists between formal and informal system construction, it
is a mistake to restrict classes and universals to formal structures.

CATEGORIES, LANGUAGE, AND SYMBOLS

Both the importance and fixity of categories are owing to their
interrelationship with words and symbols. Categories are the fun—
damental points of contact between logic and language. Recall
that all constructions, whether based upon actual contacts with
things or built out of cultural traditions, must be linguistically re-
ferred to or recorded. Categories, then, are the abstracts and brief
chronicles representing the results obtained by system builders
who point out the essential characteristics of things and the proc—
esses by which they are known.
The technique of categorial creation is extremely simple. It in-
volves two stages. First, categories are developed as descriptive
elements derived from reacting to things and are therefore de—
pendent upon the circumstances of such reaction. Secondly, the
descriptions become encased in words and symbols. Thereafter,
their apparent durability suggests that they are ultimate entities.
This fixation conceals the contingent situationsout of which cate-
gories arise, endows them with an importance superior to that of
the original events quite beyond their descriptive function.

'I
CATEGORIES AS INTERBEHAVIORAL PRODUCTS
From a logical standpoint categories constitute system—building
materials created for some specified enterprise. They become the
end products of interbehavioral processes in particular reference
fields. On such a basis they are never far removed from the opera-
tional situations out of which they arise. Not that categories are
capricious or arbitrary. Their interbehavioral origin permits us
to evaluate them on the basis of how the category constructor has
worked. Fundamentally the criterion is whether categories are de-
rived from contacts with events or simply imposed upon the latter.
Notice that whenever sets of categories are established as exclu-
sive or most basic, in the sense of the ultimate and complete Kant-
ian forms, the category systems are empty and unproductive.
Probably in no case'is it possible to achieve more than a relatively
I I6 PSYCHOLOGY AND LOGIC

satisfying practical set of categories for any particular intellectual

situation. It is highly doubtful whether in any instance alterna—
tive products are unavailable. Witness the various geometric sys-
tems which mathematicians'are constantly setting up. What are the
fundamental categories? Point, line, and plane, point and motion,
-or line and plane? Again, how many fundamental categories are
necessary?
CATEGORIAL CONTINUITY
Interbehaviorally understood, categories are functional factors
in system building. And since system building covers every type of
human activity, a definite continuity runs through the categories
of the most abstruse formal logic, the elements of scientific system-
ization, on down to the most elementary descriptive terms of our
everyday contacts. Even the most abstract categories—for ex-
ample,n n-space
. or relative time—are evolved from everyday life
conditions.
Consider for a moment the hierarchical development of inter-
behavior resulting in products for system construction. Space, for
instance, as a category of formal mathematical relations, repre-
sents a refinement of operation beginning with the basic descrip—
tions of simple place location. The S of a math’ematiCal equation
is an ‘abstracted and refined product developed from the place
or location constructs engendered in everyday contacts with
things. Similarly, T is derived from the most elementary changes
of time. The obvious fact is, however, that the accidents of our
culture have not furnished us with a beginning category for time,
such as the place category for space, unless we take when as the
missing end point for time. We believe that our continuity hypoth-
esis gains, rather than loses, from the need to take account of
such cultural variations in category origin and developinent.
Note, too, that categorization continuity may appear to be'.com—
pletely broken when categories are artifacts developed'In experi-
mental situations. Such eXperiments may engender questions con-
cerning the reversibility or irreversibility of time, and time’s
continuity or discontinuity, as in the dEVelopment of a
time
quantum
or chronon.9 It is our point, however, that, no matter how remote
the experimental situation or how far the \constru'cted categories
°See Lindsay and Margenau, Foundations, chap. 2.
‘
 CATEGORIES AND SYSTEM BUILDING 117
are' extrapolated from everyday events, there is no break in the
interbehavioral continuum.

∙ CATEGORIAL SPECIFICITY
Not only do categories differentiate themselves according to
kind of work—logical, scientific, etc.—but within any one type
the investigations call for particular sets of categories. A classic
scientific illustration is the derivation of work by the Ca-rtesians,
in contrast to the Leibnizian force. Accordingly, for Newton and
others following Descartes mass, force, and momentum are the
original categories, while for Huygens, following the Leibnizian
view, they are work, mass, and vis viva (energy).
' Categorial specificity bears directly upon the hierarchy or re-
duction principle in science. It is frequently, though improperly,
held that the sciences are hierarchical (chap. 13, p. 12), with
mathematics as the most fundamental. A variant of the hierarchical
view discriminates against mathematics as a science, but makes
physics basic and its categories the primary ones. Biological, psy—
chological, and anthropological events are regarded as reducible
to physical and chemical happenings, thus the physical categories,
such as protons, neutrons, and electrons, are made basic to those of
all sciences. ∙ ∙
FrOm'the specificity standpoint, however, each set of categories,
when found sufficient for the needs of a science, is irreducible. The
criteria, therefore, for setting up any set of categories are localized
in the investigative problems of the particular discipline. Two dis-
tinctive enterprises are involved: (a) setting up verbally unifying
systems and (b) constructing categories for systemizing events and
their observation.

SPECIAL CATEGORIAL FUNCTIONS

Aside from their general instrumentality in system building,
categories subserve specific functions in particular situations. The
following four are illustrative of many others.
Diflerentiating Functions. Certain categories mark OE items or
elements. Though they may aid in naming and recording the ma-
terials dealt with, their primary function is to‘ separate and fixate
items.\Excellent examples, on an elaborate scientific level, are the
118 PSYCHOLOGY AND LOGIC

cognate categories of kinetic and potential energy, centrifugal and

centripetal forces, and organism and environment.
Classifying Functions. Since classification18 essentially system
making, classifying functions are naturally very prominent. Taxo-
nomic categories, for instance, loom large in logical situations.
Hierarchical categories in biology illfistrate such classifying func-
tions as exemplified in the series phylum, family, genus, and
species. These categories diflerentiate general and particular char—
acteristics in an organic continuum.
Descriptive Functions. Categorieslike ampere, volt, coulomb,
and gauss, employed in the electrical branch of physics, demon-
strate quantitative descriptive categories. Similar ones are pro-
duced for all science departments on the basis of both mensura-
tional and nonquantitative operations.10 a
Explanatory Functions. Categories subserving elaborate descrip-
tive or relating functions may be regarded aS explanatory. The
biological, medical, and psychological sciences afford such out—
standing category examples as the organic, the functional, the
structural, and the ecological. Similar explanatory categories are
the theoretical constructs watt and ohm, and the practical power
and energy categories of electricity.

RANGE OF CATEGORIAL SYSTEMS

As an indication of the range of categorial systems we consider
a few class types. In a later section categorial family members are
enumerated.
Scientific Departments. While in no department of science is the
ideal of total and inclusive systemization ever achieved, there is
always an attempt to organize data and principles. It is possible,
therefore, to differentiate a set of typical categories for biology,
physics, psychology, sociology, and other disciplines. Within these
various departments, of course, thereIS further categorial speciali—
zation. Physics, for example, requires different category syStems
in mechanics, optics, and electricity. The same may be said for
biology (p. I 27),In which the exigencies of specialization introduce
subsystem categories for anatomy, histology, genetics, and ecology.
1"See an elaborate category system developed for the social sciences by Golden-
weiser, History.
 CATEGORIES AND SYSTEM BUILDING II9
Obviously, a vast number of categories are derived from such mul-
tiplicative processes. .
Specific Researches. Within any one department of scientific
activity specialized researches entail the organization of categorial
systems which bear directly, and to a certain extent exclusively,
upon the particular work in progress. Following the development
of atomic physics unique categories evolved pertaining to these
particular researches. For instance, the study of emitted particles
in nuclear bombardment gave rise to the scattering and nOnscatter—
ing categories and the subcategories of large-angle, small-angle
deflection. The specializations go deeper and deeper as individual
researches demand. .

Property and Dimension Systems. Categorial systems may also

be organized on the basis of a particular property or dimension.
Consider such a category as length. -The choice is ours either to
adopt the conventional attitude that length is a unique and integral
prOperty or that it constitutes a dimensional system. On the latter
basis length is a construction symbolizing not only a certain dimen-
sion or continuum, but a dimension which varies with different in-
vestigations and scientific departments. The fundamental point is
that the unique properties of length-scales depend upon the par—
ticular interbehavioral situation. The centimetre or any other
length unit on a measurement stick is different from the corre-
sponding length on a thermometer scale. Again, in various equa—
tional transformations the length taken in terms of time is a still
different member of this length class.

ILLUSTRATIVE CATEGORIAL OPERATIONS

The nature and development of categories, their multiplicity,
constant change, and specialization for particular systems may all
be summarized by sampling categorial situations and systems.

I .
Mechanics
If we are looking for variation in the fundamental categories
employed to develop a scientific system we find it in the history
of mechanics. Originally the basic\ categories 'Of inertia, velocity,
force, acceleration, energy, and the implied categories of space,
time, and mass were regarded as powers, properties, and existences.
Only recently have they come to be considered as constructions
120 PSYCHOLOGY AND LOGIC

derived from, and tested by, operations upon actual events. A

typical example is the category force, which at one time was pre-
sumed to represent a power inherent in things to bring about cer-
tain effects, but which today is regarded as an index to the char-
acter and result of the impact of one body on another.

2 . Electricity
The electrical branch of physics provides many striking instances
of the origin and increase of categories. First, there were the
simplest categories of positive and negative electricity, later the
systemization of electrical knowledge engendered such categories
as inductance and electrical field. At still later stages the categories
of capacitance, conductance, and resistance elaborated the system.
With the evolution of electrical science, cptegories cross the me-
chanical borders and culminate in such products as electrical mass
and velocity. °

3. Geometry
In this most stable and deductive field there is a perennial shift
and reevaluation of categories as knowledge progresses. Only
when the geometrician is preoccupied with particular data—that
is, the solution of specific problems—can he overlook the evolution-
and multiplication of such basic categories as point, line, surface,
and solid. Such obvious geometrical transformations as in the de—
velopment of non-Euclidean systems is proof enough of cate-
gorial modification. For example, the solid or space of older Eu—
clidean geometry is certainly a different sOrt of space from that
evolved in the investigations of Grassman, Riemann, and others.
Nothing is more striking than the transformation of the categories
point, position, line, and extension into operational space. cate—
gories. In other words, simple absolute position evolves into a
relation, defined by triplets of numbersIn a system of coordihates.
The following statement of Brewster concerning Newton elfec-
tively indicates geometric shifting: -'
In imitation of Cavalieri, he called the momentary increment of a
line a point, though it is not a geometrical point, but an infinitely
short line, and the momentary increment of an area or surface he
called a line, though it is not a geometrical line, but an infinitely
narrow surface.11
x\
Brewster, Memoirs, p. 17.
 CATEGORIES AND sYSTEM BUILDING 121
4. Everyday Categories
The categorizing series constructed by a political candidate
demonstratesIn the best possible manner the nature of categories.
Notice how he builds up a system of “my performarice.” To be
sure, the candidate may sincerely believeIn the political and social
improvements supposed to follow his election. Actually, however,
he is not referring to existing events. What happensIs that by
means of substitute stimulation he13 able to operate with a set of
categories which are products of his interaction with nonexistent
situations—namely, those situations he promises to bring about.
The fundamental import of such categories is that they represent.
modifications in present conditions regarded as unsatisfactory and
capable of improvement.
A similar set of categories may systemize a particular individ—
ual’s attitudes concerning actually existing conditions. As such they
are simply products of immediate interbehavior. In other words,
they do not represent any necessary analysis of things or conditions,
but stem from an individual’s views, reactions, or beliefs. While
the above is not an essential description of everyday categories,
inasmuch as such categories may be precisely like those so well
established in logical and scientific domains, an appreciation of the
nebulous and unstandardized processes of engendering categories
indicatesthat category products are not always formal or ultimate.

5. Religious Categories
Valuable ideas concerning categorization proceSses 'may be
gleaned from the theological problem of positive and neutral
categories centered in the negative theology developed from
Philo, through the Gnostics and Christian apologists, to Plotinus
and beyond. These writers struggled with the question: What
categories can be applied to the deity? Negative theology implies
the recognition that all categories are constructs and thus finite
in character. How, then, can categories, which are simply fixations
of qualities or characteristics observed from concrete interactions:
with things, apply to deities?
Even early theologians observed that the glorification of a deity
is not achieved by ascribing to him qualities which, in so far as
they also apply to mundane things, are clearly not sufficiently ex—
alted. To their credit they discovered that God could not be omni—
I 22 PSYCHOLOGY AND LOGIC

potent and omniscient, since these categories simply extrapolate

from the characteristics and actions of men. Incidentally, the lin—
guistic habit of masculinizing or feminizing deities bespeaks the
powerful influence of cultural institutions upon categorization.
Even the negative theologians of Western Europe could not es-
cape the influence of their cultural traditions in speaking of God
as He or Him.
What better evidence is required for the interbehavioral process
than the entire enterprise of deity creations? The description of
this procedure has not been improved upon since Xenophanes of
Colophon (570-480 3.0.):
But if cattle or lions had hands, so as to paint with their hands and
preduce works of art as men do, they would paint their gods and
give them bodiesIn form like their own—horses like horses,- cattle
like cattle.‘2
a

6. Dimensional Analysis
Dimensional analysis, as a domain of measurement, exemplifies
not only the construction and use of categories but also their in-
tegration into systems. Dimension categories serve as units for
deriving interrelations as functional systems. VelocityIs a structure
deriving from L and T as fundamental dimensions:

[V] = L]:
[T]
L is constructed to serve in whatever capacity is required to deal
with length as distance, interval, etc. V, as a derived unit, is a
local system capable of functioning as an item in more comprehen—
sive systemization. For example, velocity is only a subunit in ac—
celeration systems. When F'symbolizes acceleration, its relation
to V, L, and T is clear from the following:
[F]: [_]=[] l
V L
[Tl [T2]
Assuming that dimensional analyses concern specific types of
mensuration systems, we may ask whether the C.G.S.
unTIxts
or
dimensions are ultimate or basic, either as ontologic, phenomeno—
logic, or epistemic categories. The phenomenologic units are im-
∙
'2
Fairbanks, Philosophers, p. 67. ∙
 CATEGORIES AND SYSTEM BUILDING I 23
mediately derived as a matter of course, since such units are ele-
ments of intolerable spiritistic systems. But, though with proper
interpretation, the ontologic and epistemic systems may be accept—
able, we still face the question whether they are unqualifiedly basic
and indispensable.
Although such a writer as Campbell believes that all funda—
mental measurements belong to physics (267),13 and that the
basic characteristic of physics is its propensity for measurement
(33f.), he does not regard length, mass, and time as basic magni-
tudes, even when to these three are added temperature, dielectric
constant, and permeability. This does not mean, however, that he
does not consider some magnitudes as basic (393), since he realizes
that of such are mensuration systems made. Which are funda—
mental depends upon the mensurational task at hand. Campbell
plainly asserts that measurement is a means toward an end, con—
sequently, the tools and procedures are dependent upon the opera—
tional systems in which they are found.
Dimensional analyses as systemological factors were appreci—
ated by the earliest students of the subject. As Larmor“ points
out, Newton originated the essentials of dimensional doctrine as
a means of comparing the properties of correlated systems, for
example, (a) particles of different systems, (b) particles and
bodies, and (c) resting and moving parts of systems.15 As is well
known, Fourier more explicitly worked out the theory of dimen—
sions and their eXponents in his development of mathematical
systems, or equations, in which the homogeneity of terms must be
maintained. The correlation of heat conductivity with other meas—
urements involved equations eXpressing necessary relations be-
tween common units of length, time, and weight, to which are
added temperature and heat quantity."
So far we have accepted the view that dimensions are quantita-
tive. But we must go further. If we begin with the operational
basis of categories, instead of with conventional notions of onto-
logical or epistemic postulates, we may allow for other sorts of
dimensions. Such dimensions imply nonadditive orders and ar-

The numbers in parentheses refer to pages ianhyiics.‘

1‘
Dimensions.
1”
Principia, Bk. 2, pr0p. 32.
1°
Analytical, chap. 2, sec. 9.
124. PSYCHOLOGY AND LOGIC

rangements, basic comparisons, and, in general, systems difiering

from those of conventional mensurational theory. Further con—
sideration of these problems we postpone to our chapters on
Measurement.
7. Axiological Categories
The comparatively recent development of axiology as a distinc—
tive field also illustrates the fundamental principles of categoriza-
tion, and its place in system building. Because of a deepening in—
terest in the basic problems of ethics, aesthetics, and normative
logic, the category of value has been constructed as a fundamental
core‘in the fields mentioned. Around this Core a systematic set of
categories has been arrayed.
Fundamental among secondary axiological categories are the
objective and subjective, a differentiation based on whether values
are taken as characteristics ‘or properties of things and events, or
simply as evaluations of individuals interested in ethical or aesthetic
situations;
Relation, too, constitutes a distinctive type of axiological cate—
gory. Even scholars who are fairly sympathetic to the‘idea that
values might be objective are unwilling to admit the objectivity
of relation. Axiological constants they consider as relational, and
as depending in part upon the observer or evaluator. The funda-
mental basis for their hesitancy is that evaluative categories are dif-
ficult to establish. But many natural-science categories are also
equally difficult to establish. As for the categories of the medical
sciences, how shaky are such constructs as influenza and inflamma-
tion!" To what system of natural happenings are they related?
Moreover, it is overlooked that all categories are operationalprod-
ucts, thus, even in physical constants there is a selective and cogni-
tive factor. l
A unique feature of axiological categories is that values are
regarded as problematic primarily because they are considered
as exclusively human factors, and thus as categorically different
from materials of the natural sciences. Even those scholars who
assent to the proposition that physicalconstants, or natural--s'cience
values, are derived from the characteristics of things hesitate to
regard human-science values as objective ‘and likewise derived
" See Crookshank, Importance; also Harding, Science.
 CATEGORIES AND SYSTEM BUILDING 125
from things. Obviously, there is a lingering prejudice that human-
istic happenings involve psychic, or nonnatural, elements. Thus
it is overlooked that in both humanistic and natural determinations
all categorization evolves from interaction with things.
Axiological category study provides an important lesson con-
cerning the relativity of logical systems. It is’ conventional to'
assume that wherever we have difficulty in our observations, when—
ever events are recalcitrant, the final categorial products are not
stable, basic, or satisfactory. Actually, this situation simply puts
the diflerent categorial products upon a relative scale. The single
conclusion is that we can have various types of systems 5 that is,
the materials and results of our logical processes are different,
and hence the logics themselves are diverse.
The Value, then, of considering axiological categories lies in
the contrast they afford between the notions (I) that logic is a
universal and inclusive form of action—in other words, that there
are certain absolute logical constants—and (2) that logic con-
stitutes innumerable specific enterprises of system building.
8 . Thermodynamics
I

The evolution of thermodynamics as an area in physics provides:

the systemologist with a splendid instance of novel system con—
struction. As a domain of rapidly changing events, thermodynamics
has had to forsake the all-embracing mechanics of well established
categories. Thermodynamic events could not be put into the frame—
work of enduring particles, which, even in dynamics, moved so
slowly as to be assimilable to selected static systems founded on
geometric constructs localized within definitely closed and
boundary-fixed limits.
By contrast with particle mechanics the study of rapid molecular
changes required a new system of categories. Attempts to throw
these happenings into the old mechanical framework at once
brought to the fore the probability category. Mechanics conse-
quently became statistical mechanics. The events dealt with were
so complex and variable as to make it seem impossible to trace out
paths of single molecules. In a sense, therefore, the categories .of
this area had their sources in operational procedures, rather than
exclusively in ontologic analysis.
Perhaps equally as fundamental as probability is the category
126 PSYCHOLOGY AND LOGIC

ofrconservation, formulated in the first thermodynamic law. The

associated proposition is that, unless bounds are placed on rapidly
occurring events, no system can be achieved. Here is a vital
system-building situation characteristic of all scientific enterprises.
The physicist facing a set of difficult,_but unquestionable, happen—
ings made his first attack, history relates, by constructing a model
such as the Carnot cycle. This model indicated what might happen
if the system were regarded as closed, in line with the conserva—
tion principle.
Next came the development Of the entropy category, an inven—
tion to account for happenings within the system. In any aetual
thermodynamic system—that is, one which can not be thrown
into the ideal form of the Carnot cycle—energy must be re—
garded as dissipated. Entropy is thus a polar category to conserva—
tion. O

Other fundamental categories in thermodynamics are trans-

formation and what may seem to be subcategories: reversibility,
and irreversibility. Thus, all sorts of systems are possible in which
the .processes usually take one direction and at other- times the
reverse. To these categories must be added many others necessary
for describing the events—namely, state, degree of freedom, etc.,
—all of which reveal the qualitative and quantitative changes tak-
ing place.
It should be noted that, in consonance with all scientific work,
categories are interrelated. In a thermodynamic system, for ex—
ample, such categories as molecule, pressure, and temperature are
also elements of other systems. Even so-called common categories
like space, time, and energy, despite linguistic fixation, are really
operationally unique in difierent situations. Moreover, in\ such
an area as thermodynamics the more particularized categories
Comprise units of subsystems within the same general system.
may

9. Biology ‘
In no scientific domain are categorization principles better illus-
trated than in the descriptive and\ interpretative systems con-
structedIn the biological sciences. In \addition to the importation
of categories from other investigative fields, some are constructed
with unique application to biological events: Iéiological categories
 CATEGORIES AND SYSTEM BUILDING 127
are differentiable on the basis of observational derivation or free
construction.
Biologists are undecided whether their categories should in~
cline more toward (1) structure, (2) function, or (3) both. On
the one hand, the organism is emphasized as distinctive and self-
maintaining. Here the fundamental categories are morphological
in type, such' as cell, tissue, and organ. On the other hand, since
biological systems are always. active and changing units, the actional
or physiological categories, such as metabolism, may be stressed.
Ecological events lend themselves best to interrelations of struc-
ture-function categories.
Change and action categories divide themselves into two series,
one of which has to do more with particular organisms, their growth
and coincident anabolic and catabolic processes, and the other
with such general processes as evolution, which is concerned with
the development of types, or individual forms.
Characteristic of the problematic, and perhaps speculative, bio-
logical categories are those which began historically with the cate-
gory of vital force, and which, especially in the modern embryo—
logical branch of biology, come under the heading of organizers,
determiners, etc. A unique form of this type of category is the
gene, which, on the one hand, is deductively constructed from
observations of happenings during reproductive cycles, and, on the
other, may be regarded as a freely constructed category for the
purpose of interpreting changes not easily observed but necessary
in accounting for observed happenings.

∙ Et/tical Categories
The number and type of ethical categories are dependent upon
the views of writers. concerning the nature of ethics. A very gen-
eral category of this field is that of the summum bonum. There
has always been an implicit recognition that the field of ethics
has to do with some sort of norm or value. The innumerable inter-
pretations “of what the summum bonum should be depend upon
various cultural situations with their foci in particular intellectual
formulations.
Obviously, ethical categories as constructs are primarily autistic.
This fact perhaps eXplains why ethicists have differed so radically
128 PSYCHOLOGY AND L0GIC

in their ideas concerning the summum bonum. Ethical categories

are most freely constructed when the supreme norm is asserted
to be given, and to regulate the inclination of man toward the
will of God. It is a paradox that the most absolutistic and trans-
cendent writers are the most arbitrary in their constructional pro-
cedure.
Conventional ethical categories may be grossly distinguished
on the basis of whether they concern (I) essences of various sorts,
that is, are names for ultimate qualities of goodness, or (2) goals
and types of action. When the highest good is regarded as hap-
piness, or self-realization, ethical categories imply goals of action,
though self-realization, like the pursuit Of the ideal, to some
extent may be the name 'of a certain desirable quality. Other
ethical categories are intimately connected with actions themselves,
such as the good, in the sense of good action, the
worthwhile, and so on.
proper, the

11. Logical Categories

Logical categories constitute items regarded as necessary for
building a technical logical theory or system. Since there are many
types of technical logics, despite the prevailing beliefIn a single
cosmorational system, we choose but a few typical categorial illus-
trations.
In classical Formal Logic the three primary categories are term,
proposition, and syllogism. Correlated with these three are the
equally conventional concept, judgment, and inference or reason-
ing.
In symbolic logic similar technical categories are the indefinable,
propositional function (propositional form), primitive proposition,
class, relation, etc. Out of such abstract elements symbolic logicians
build up symbolic systems. I

12. Grammatical Categories I

The categories of formal grammar are widely regarded as' con-
necting links between language and logic. We need
not commit ourselves to such a View in\Order to see that grammati-
cal categories admirably show the process of categorization. Gram-
marians convert actual speech into textual
materials—not deliber-
 CATEGORIES AND SYSTEM BUILDING 129
ately, of course, because this procedure'has a very definite cultural
history—and we may say that grammar arises from an investiga-
tion of forms or symbols. At the present time, however, grammars
constitute just such transformations.
An examination of conventional grammatical categories, espe—
cially parts of speech, indicates the formalization both of lan—
guage events‘ and references to them.13 The resulting products
consist of a series of categories, usually eight—namely, the fa—
miliar noun, Verb, adjective, adverb, preposition, etc. In this con—
nection we may refer again to the widespread notion that the
Aristotelian categories have been derived from the analysis of
parts of speech, or different ways in which things are spoken of.

I3. Philosophical Categories

Throughout its historical development philosophy has centered
about categorization, since philosophers traditionally have at—
tempted to systemize the universe. Accordingly, they have con—
structed such fundamental products as being, essence, substance,
function, and cause. These categories are taken to be ontological
—that is, concerned with the contents of the universe. When epis-
temological, or knowledge, problems came to the front, other
categories, such as ego, knowledge, and mind became prominent.
Categories involving creative powers have extended the list,
though the latter have converged with the religious categories
(p. I 2 I) already examined.
I 4.. A esthetic Categories
Aesthetic categories are frequently connected with general axio—
logical (p. 124f) or more comprehensive philOSOphical systems.
The question here is Whether the aesthetician adopts a general
philosophical standpoint, or whether he undertakes to analyze aes—
thetic events in a more limited frame of reference. Aesthetic cate—
gories are variously. called form, content, significant form, beauty,
ugliness, etc. That many of the essentially aesthetic categories
have names similar to those of other systems is only to be eXpected.
The category form, for instance, is not only a special type of
systemization factor, but is also qualified assignificant.
Kantor, Objective.
‘3
130 PSYCHOLOGY AND LOGIC

I 5.
Mathematical Analysis
Even a casual inspection of the constantly developing field of
mathematical analysis requires an elaborate dictionary merely to
list the categories. Because of the abstract relational nature of
mathematics, its general terminology comprises a unique set of
categories. As is true elsewhere, mathematical categories are de-
rived from operations performed upon crude data, and are thus
specific kinds of constructions. Important mathematical categories
are operation, continuity, field, function, rational, inductive, vari-
able, limit, parameter, constant, infinitesimal, infinite, operator,
magnitude, convergence, divergence, etc.
These categories, which have always impressed thinkers as
ultimates and as independent of particular enterprises, are def-
initely functional 5 they vary with the probléms, types of field, and
mathematical interest of the worker. The many interpretations
put upon such categories as the infinite and the infinitesimal are
sufficient to remind us of their operational and constructional char-
acter.
 CHAPTER XVIII

UNIVERSALS: SYSTEMOLOGICAL COMPONENTS

AND PRODUCTS

THE LOGIC or UNIVERSALS

UNlogics;
IVERSALS occupy an ambiguous position in conventional.
When mind and reasoning are regarded as transcend-
ent powers or ontological essences, disturbing problems inevitably
arise. Are universals concrete or abstract? Granted that they are
abstract, may they be contaminated with qualitative and content
residues, or must they be pure forms? Again, when logic is deeply
concerned with problems of class membership and class inclusion,
must classes be dispensed with as merely linguistic conveniences
because they are “not as genuine objects as their members are if
they are individuals? ”1
No ambiguity attaches to universals in interbehavioral logic.
Universals are products of system-building operations, but they
are something more than end points in the operational procedure,
in addition they constitute the raw materials of further system-
building operations. .This characteristic, it should be noted, uni-
versals share with relations, classes, kinds or species, mathematical
functions, and other constructional forms.
HISTORICAL TREATMENT or UNIVERSALS
A definite cultural continuity relates the current universals
problem to that of earliest antiquity. The natural history of uni-
versals constitutes a fascinating and informing story when con-
sidered as the behavior of persons called logicians who occupy
themselves with the universals issue.
Universals among the Greeks. Greek interest in universals may
be justly stated to have arisen out of a systemological problem.
The development of forms, ideas, definitions, and genera is surely
localized in the work of setting up Systems of knowledge (science)
or logic.
SinCe the Greeks did not Work with nonspatial, or mental, con-
Whitehead and Russell, Principia, p. 72; Russell, Principles, p. x.
I32 PSYCHOLOGY AND LOGIC

structs,2 universals for them constituted definite products of ab-

stracting and generalizing operations. In other words, they ab-
stracted and generalized forms or universals from things in order
to establish science and system—that is, stabilized operations.
These forms the Greeks regarded as patterns or laws for fixing
and organizing the fleeting events of everyday life. They were
harried by no problem concerning the existence or subsistence of
generals or universals as over against particulars. In fact, Aristotle
criticized Plato primarily for separating the Ideas or Forms from
occurring things. What la’ter tOOk the name of Realism was in—
evitable for the objective Greek way of thinking.
The Medieval Concern with Universals: Modern problems of
universals sharply reflect their close articulation with post—Greek
or medieval origins. Psychic realms of existence and process now
take the lead. By contrast with the Greeks, medieval writers were
no longer interested'In systems of things or knowledge, they were
obsessed by the intimate problem of personal salvation, by the
establishment of the continuous existence and value of soul or
intrinsic human nature. Universals at this time. consisted of soul
in that sense. Supported by their special cultural batkgrounds, the
Scholastics divided themselves into distinctive schools.
(I) The Realists. Seeking as the most real that which is least
subject to change and time, the Realists argued that the universal
maintains an independent existence. In this wise they sided more
with Plato than Aristotle. But this doctrine yielded too little to
individuals, and was far from satisfactory to those who stressed the
basic value of individuality.
( 2) The N ominalists. Unsatisfied with Platonic doctrine, many
resorted to Aristotle. The greatest value and significance, they
declared, was attached to individual souls, to particular persons,
and that which is common to all of them, or universal, is the
name.
Only individual things are substance, hence real. Thus names are
used to comprehend a multiplicity of individuals, and to analyze
them for thought and communication.
(3) The Conceptualists. For the conceptualists psychic processes
and existence comprise, besides soul and all essential things; the
mental acts of dealing with them. Universals, then, became inter-
preted as concepts in the mind. Concepts, it‘ was thought, could
See vol. I, pp. 463.
 UNIVERSALS: COMPONENTS AND PRODUCTS I33
mediate between spaceless existents and their names or designa-
tions.
CURRENT TREATMENT OF UNIVERSALS
Present—day discussions of universals, as we have indicated, are
based on a three—point support. The first is post-Greek psychism.
Arguments are bandied back and forth whether universals are
real or whether they exist only in the mind. Secondly, the doctrines
of universalism and necessitarianism influence writers to transform
universals into words or terms. And, finally, writers on universals
do not differentiate materials, work, and products from each other.
All three points can be taken into account by organizing current
treatments under the following headings: (I) spatiotemporality,
(2) character value, and (3) the participation problem.
(I) Spatiotemporality. Two phases of spatiotemporality are
employed to differentiate between universals and particulars. The
first has to do with plurality of spatial location. A universal red
can be simultaneously localized in two objects when each object
can only be in one place at any given time.3
There are several objections to this phase of universal descrip-
tion. For one thing, it is asserted that despite its importance4 this
is not the essence of the matter. So-called physical distinction is
set below logical distinction. Even more significant is the denial
that universals are no more pluralistic than particularistic. For
example, Ramsey begins his discussion of universals by asserting:
The purpose of this paper is to consider whether there is a funda-
mental division of objects into two classes, particulars and universals.
His answer is negative, but later he reconsiders the matter.5
Though he departs somewhat from the spatiotemporal problem
to work more with linguistic materials, he is obliged to reject this
distinction because of the complete lack of difierences in things.
The more crucial phase of spatiotemporal argument interprets
universals as real but not spatially confined, as in the following:
Of course, wisdom and triangularity' exist nowhere, but they-are
as real as triangles and blackboards.6 On this basis one embraces
3
Cf. Ramsey, Foundations, p. I Izfl.
Ibid., p. I I 3.
"Ibid., p. 135f.
Mbore, Why, p. 686.
I34. PSYCHOLOGY AND LOGIC
the absurdity that round squares, triangular circles, griffens, and
the paupers’ millions exist. Of course, those who adopt realism
as a fundamental philosophy argue that serious logic must have a
metaphysics, and that they really are interested in the relation-
ship between triangles and triangularity, wisdom and wise acts,
blueness and blue things. Any doct‘rine, however, that makes
these relationships precisely identical with the relationship be-
tween assertions concerning unicorns and real nonspatiotemporal
unicorns is self condemnatory. Spatial location is, however, not a
fundamental distinction, because'writers on universals immediately
jump to terms—that is, linguistic or verbal items. The problem
of'universals is made into a problem of substantive versus adj ec-
tive, ‘substantives being singular and adj ectives universal.
(2) Character Value. Present-day realists who are willing to
accept eternal forms or ideas as universals are not troubled by the
problem of their character value. This problem, however, is a
poignant one for those who believe that “whatever is, is particu-
lar.” The latter writers deny the existence of so-called Platonic
universals, but, on the other hand, face the issue of associating
blueness or triangularity as general characters with blue or‘triangle
as particular instances. If the universal is an existencein the domain
of particulars, must it not be described as a character like other
abstracted traits of things?
One negative answer is given in terms of classes. For example,
McGilvary7 asserts that the universal blue is the entire class of
blues. Universals comprise classes of prOperties abstracted within
synthetic wholes.8 McGilvary thus regards universals or generals
as analyzed, but inseparable, constituents of complexes or wholes.
By comparison, the common or similar components are made to
stand out. Universals, then, are real existences different
froml
character which they comprise in series or classes, and yet are far
the

from being timeless or spaceless essences. Incidentally, the view

of objectively existing classes impliedIn McGilvary’s theory of
universalsIs not agreeable to other writers. I

(3) Participation. This section heading reminds us of the an—

cient ontological proposal that particulars participate in universals.
Since that era the descriptions of the relationship, regarded as
Relations, p. 30f. \

”Ibid., p. 36.
\
 UNIVERSALS: COMPONENTS AND PRODUCTS I35
either logical or ontological, have centered around words. Such
is the case which Goethe describes:
Denn eben wo Begriffe fehlen
da stellt ein Wort zur rechten Zeit sich ein.
Thus words, in the form of assertions and statements, are em-
ployed to mediate and interrelate universals and particulars. Rus-
sell9 and Johnson” divide off and interrelate universal and par-
ticular terms by making the former into adjectives or predicates
of propositions (sentences), and the latter into substantives. How
little this subj ect-predicate'solution satisfies the participation ques-
tion becomes apparent when considering Johnson’s contention
that, while a substantive can only serve as the subject of a sentence,
adjectives can function either.as predicates or as subjects. For
example, in the sentence: “Punctuality is a fault,” “fault” may be
taken to be a secondary adjective. Russell objects, and asserts
that the proper handling of these words must be as follows: “For
all x, if x is unpunctual, x is reprehensible.”11 A “particular” or
an “individual” can be defined as “anything that can be the sub—
ject of an atomicproposition.” This is not true for “universals”—
that is, a predicable character or relation.
This treatment of participation is a modern linguistic form of
realism. Nominalism, too, mediates verbally between singulars
and generals. Current nominalists declare that universals are
names for properties abstracted from actual objects, events, or
behavior—triangularity from triangles, wisdom from wise be;
havior, blueness from blue hat, etc. In other words, universals are
abstractional products embodied in words or terms. There is justi—
fication for this notion in the sense that the referents for the auton-
omous terms triangularity, wisdom, and blueness are diHerent
from the referents of the terms wisdom (wise action), triangu—
larity (three—sided figure) 5 blueness (a blue hat).
The question now arises: What is the relation between the sub-
sistential references for universal terms (words) and concepts?
Notice that the latter, too, are regarded as nonexistents, repre-
sented by words. It is plainly indicated, therefore, that we should

Problems, chaps. 5, 9.
Logic, part I, p. I I.
1°

Whitehead, Russell, Principia, p. xix.

 36 PSYCHOLOGY AND LOGIC

not look upon universals as names for names, or words, but as

references to, or symbols of, constructive action with respect to
existential Objects and events, such as wisdom, triangularity, and
blue things.

LOGICAL FORMS As THINGS, TERMS, OR THOUGHTS

Given the traditional ideas of things, language, and mind, we
can locate the universals problem, as well as the conflict in solving
It.
The proponents of each of the three Views that forms are
things, terms, and thoughts, use the same general argument to
support their beliefs. A paradoxical situation! The realists, who
are unwilling to relegate universals to the ephemerality of mind,
create a new realm of existence—namelyb subsistence, yet they
assert that the basis must be something in the “real world.”12 They
are unwilling, it seems, to relegate universals to’ mind or to names.
In their favor it may be said that universals are definitely stimulus
objects, in the sense, of course, of constructions formed by abstrac-
tion from concrete things, thus allowing for the Obvious differentia—
tion 'between greenness and the green hat. This theory, however,
is fundamentallyInept in building up a reality independent of any
specifiable space--time.
The nominalists, believing that universals are verbal terms,
argue that the ordinary events of experience and science Show
common characteristics. Russell,13, for example, declares that
words are necessary because of things, that words correspond to
things, and thus the necessity to say certain words is determined
by the existence of corresponding things. Without the word
similar one can not get on, because there just simply are similar
things. In this View that universals are terms we Observe the"\con-
fusion between words as things, and language as action or asser—
tion. Universals are regarded as autonomous words, but the ju'stifi-
cation for this View is that by means of words one refers to ’com—
mon characteristics of things. To be sure, references to common
characteristics may be important aspects of symbological forms, but
can forms, as universals, be made identical with the terms or wOrds?
Dothe difficulties of the first two views lend plausibility to the
12Moore, Why, p. 685. \-

13Inquiry.
 UNIVERSALS: COMPONENTS AND PRODUCTS I37

conceptualist’s idea that logical forms in the final' analysis are

thoughts or concepts?“ Those who believe so stand on good
groundI in their denial both 'Of the name doctrine and the theory
that there are real “things” called universals. However, as- we
have already seen (p. I 32), conceptualism is invalid because based
on traditional mentalism which introduces a new kind of exist—
ence—that is,~concepts which are neither names nor things.

INTERBEHAVIORAL CHARACTER OF UNIVERSALS

Historical and current quarrels about universals stem from the
general dualistic treatment of thought. An objective psychology
might well be substituted for the logistic principle which closely
interrelates,.and even identifies, linguistic products and things.
Our proposal, then, to treat logical forms in general, and univer—
sals and relatiOns in particular, as products of interbehavior, as well
as items in the development of all sorts of systems, means we re-
ject the theory that terms constitute common denominators of
forms. As interbehavioral items, terms are features in an individ—
ual’s concrete operations with stimulus objects. These objects may
be of an immediate or direct sort, or they may require substitute
stimulation.
To illustrate the essential features of our presentation, com-
pare the interbehavioral with a classical mentalistic-psychological
description. For this purpose consider Sellars’ projection descrip-
tion.15 With regard to universals he indicates that the mind,
working with concepts as instruments, projects “the recurrence
of the same meanings in the mind” into the things thought of. In
this way, logical identity—that is, identity of a concept—is trans—
formed into universals and things.
This description requires complete translation and transforma-
tion to make it accord with what actually goes on'. To begin with,
“the recurrence of the same meaning in the mind” refers to a
similar reaction to a similar stimulus object. There is, therefore,
a correspondence between actual objects and responses to them.
Instead of projecting psychic meanings into things in order to
create universals, on an interbehaviOral basis we describe the con,—
struction of a form represented by a term. _
In the scholastic sense of verbum mantis.
'5
Philosophy.
 38 PSYCHOLOGY AND LOGIC

The entire process is exemplified by the situations which give

rise to the successive creation Of the universals blueness, color,
similarity, etc. By simple reference to the color similarity Of dif-
ferent objects such quality universals as blueness and redness are
abstracted and fixated. Then a universal color, or colority, is en-
gendered. Finally, the abstracting and generalizing behavior pro-
duces the apical universal similarity. Such creative processes also
manufacture the order of the universals. Actually, we can not
know what was the precise succession in the cultIIral history of
system making, or whether there was _Such a succession altogether.
These speculations, however, serve the purpose of bringing to the
front the concrete interbehavioral conditions under which the cre-
ation‘of universals occurs.
Light is also thrown upon the problem pf the reality of forms
as urged by realists of all sorts. For example, Cohen,16 maintain—
ing a notion of psychic mind, has argued that universals as ab—
stractions must have some existence aside from the mind that makes
.them. Such a question of the reality, existence, and validity of
forms never arises when we observe how they are derived from
actual problems, solved by persons dealing with Objective things.

SPECIFICITY 0F FORMS
Universals are the specific products of abstracting and generaliz—
ing behavior such as we have described in Chapter I 3. These pro-
cedures take place in definite interbehavioral fields. It is encourag—
ing to find the specificity principle not only recognized, but appre—
ciated, in logical writings. For example, McGilvary17 asserts that
relations, even though generalized to a high degree, are specific.
This View follows from his general realistic philosophic standpoint.
Indeed, pragmatic realism has much in common ‘with an inter—
behavioral View, certainly there is considerable appeal in thel idea
that"universals are really always particulars. The fact that one
generalizes or constructs a universal does not detract from the
specificity of the situation in which the process takes place.,'
Specificity has to do with the particulars of the relations them-
selves. In other words, we have a Check on the conception that
there is a generality or universality of forms which gives them a
16
Reason, p. 2 o4.
Relations.
 UNIVERSALS: COMPONENTS AND PRODUCTS I39
unique status different from particular things. In every case such
generalized or universalized relations constitute simple construc-
tions which are mere generalized referents of verbal propositions.

FORMS AS PRODUCTS
In carrying out all sorts of operations upon things it is neces—
sary to develop or create instruments for performing large-scale
operations in order to summarize, organize, and refer to things
(chap. I 6). In calculating we require a form which will span a class
of values, thus we create a variable. In comparing things we need
an instrument for carrying over from one relation to another.
Hence the construction of a relational form. This is a genuine con-
structional process derived from contacts with things.
Three sources Of raw materials from which to construct forms
or generals are:‘(a) things and events interacted with, (b) acts
themselves, and (c) the products which result from such acts. Let
us characterize each in turn.
(a) Thing or Event Sources
Recall that it was philosophical convention to classify all forms
as universals and relations. But when we start from actual contacts
with things we find that they result in the construction of many
types Of forms. The following list, while not exhaustive, illus-
trates a series of forms constructed from thing and event sources.
I. Qualities. Quality forms, traditionally known as universals,
are definitely derived from things, and for the most part are ab-
stracted from structural characteristics such as color and hardness.
2. Properties. Conditions of things and their changes provide
the basis for abstracting forms or generals such as lastingness, con—
tinuity, and irregularity.
3. Relations. These forms are abstracted either from parts or
aspects of stimulus Obj ects, or from some situation or complex in—
volving several objects. In the former case the complex consists
of the discernible features of a single thing—for example, front or
back, up or down. Out of a configuration involving several things
such relational forms as “between,” “higher,” and “lower” are
derived.
Notice that relations are features likewise Of the connections be—
tween Objects. For example, we interbehave with the relation or
I40 PSYCHOLOGY AND LOGIC

geographic placing of two cities quite as much as with the individ—

ual cities. Their relationship, in other words, is as much a natural
fact as their existence. The eastness or westness of New York with
respect to Chicago comprises a definite fact of land, mountain, etc.
It matters not what frames of reference we adopt, nor is this prin—
ciple interfered with even when we ‘find New York to be west of
Chicago, in the sense that one may start from Chicago, and by
traveling constantly westward reach New York on the east.
From an interbehavioral standpoint the relations between more
abstract things, such as mathematical Objects, constitute similar
definite stimulus objects from which relational forms are- pro-
duced. When we develop such forms as transitivity, intransitivity,
symmetry, asymmetry, and connexity as parts of logical or mathe-
matical systems, we are interbehaving with relationships between
things, even though we have previously constructed those things
and the relations between them. What we fieed to distinguish,
therefore, are merely abstract stimulus objects from concrete ones,
the former being derived from the latter by a long psychological and
cultural history (chap. I3).
4. Classes. ItIs a well-known observation that'classe‘s are con-
structed from similarities and differences between things.“ Cer-
tainly there is no objection to the View that the qualities Of some
things resemble each other more than the qualities of other things.
It is only unique philosophical postulates which make classes into
autistic constructions, developments of “the mind itself.” The fact'
that logicians have drawn theirideas concerning classes chiefly
from mathematics probably accounts for the autistic View. Bi-
ologists, geologists, and astrOnomers, definitely concerned with
things, are most sympathetic to the derivation of class forms from
the actual characteristics Of natural objects.
5. Behavior. The behavior of things as interrelated with similar
or dissimilar things is a fertile source for the production of uni-
versals or forms. Historically, of course, the waxing and whning
of things, and the influence of one thing on another, have .given
rise'to the constructs of functions, pgwers, forces, etc. From such
behavior the mathematical construct bf function has undoubtedly
arisen, an elaborate evolution transforming action into detached
relation.
6. Number. The natural-history observer has no difficultyIn
 ∙
UNIVERSALS: COMPONENTS AND PRODUCTS I4I

seeing how all the number forms or universals, such as unity,

singularity, particularity, plurality, and manifoldness, are by
longer or shorter procedures derived from contacts with concrete
objects. Natural things exist as singles, couples, triples, and higher
multiples. Also, through various conditions they become divided
and thus multiplied. It is only the Platonic bias which excludes
the derivation~of numbers from such processes as counting or
otherwise manipulating things.

(b) Action Sources

Traditionally, writers on universals divided themselves pri-
marily into ontologists andepistemologists. The latter—the con-
ceptualists who pro1 ected universals into things—proceeded on the
basis Of knowledge, rather than of things. To signalize their dif—
ferences from ontologists they worked mostly with terms.
On the 'whole, epistemologists or logisticians operate on the
basis of a relationship between utterances (references) and deno-
tational terms and things. They stress relations which are, as
Dewey13 asserts: “expressed by such words as “is, is-not, if—then,
only (none but) and, or some-all.” Many logicians turn to prob-'
lems of subj ect-predicate, substantive-adj ective, and similar gram-
matical items for instances of universals.19

(c) Product Sources

Many logical forms are higher—order abstractions and generali-
zations. For example, qualities and properties of things are ab-
stracted and generalized to produce such products as sensations
(chap. I 3, p. 12). These products serve in turn as the raw materials
for further abstracting and generalizing processes, with the re-
sulting product sense data. Higher—order products are also illus-
trated by propositional or sentential forms. Abstracting from “the
man is tall” one reaches "such consecutive products as “x is tall,”
“x IS 3”), (CK y),
Higher-order product development may be part Of a short or
long series. In View of the evolution of all psychological events
(vol. I, chap. 6) it is probably safe t\o generalize that all abstract-
ing and generalizing involves a number of successive develop-
1’
LOgic, p. 1.
1°
See p. 6.
I42 PSYCHOLOGY AND LOGIC

ments. In any instance, however, we need only distinguish between

prior interbehavior, in which we originally construct generals or
universals, and the posterior interbehavior Of further abstracting
and generalizing.
N O doubt the highest orders of forms are describable as imagi-
nary relations between imaginary terms (things). Responses to such
higher-order relations may be illustrated by comparing the rela-
tionship between X isx and X ezl with that between X uxl and Tyxl.
All four terms in the preceding sentence refer to planets, each
one larger than the whole Of‘the known galaxies. In this case,
instead of interbehaving with a dyadic relation of concrete or ab-
stract things, we have a triadic relation because we must consider
the tonstructive powers of the interbehaving individual. Keeping
alive the interbehavioral principle, we have definite referents for
our propositions concerning theexistence, reality, and significance
of the terms and their two aspects. °

FORMS AS SYSTEM COMPONENTS

. So far we have studied forms as products by departing from

all historical views of mental faculties and mental objects. Also,

we have not stressed the logical or system—building character of
forms, except to deny, by implication at least, that forms are logi-
cal in the conventional sense of being nonexistential. We turn now
to their system-building aspect.
A preliminary step is to compare forms with things as system-
building components. Forms are products developed for particular
systems, whereas things are any kind of stimulus objects found
ready at hand. Systems can be built of many sorts of blocks. The
following table may help to clarify this distinction.

Farms
Thing constructs
System Components
Things
Objects (natural and artificial)
\
Relation constructs
Property constructs
Relations (discovered and
contrived)
Properties (observed or attributed)l
Number constructs Quantity (singularity and plurality)
Magnitude Dimensional extent
Terms Words, reference acts
Conventional logics are built from forms, though the builders
may not be aware that their materials are constructs produced from
specific manipulations of things. Conventional-(logical building is
carried on in two steps: (I) compounding forms-into propositions
 UNIVERSALS: COMPONENTS AND PRODUCTS I43
and (2) organizing propositions into complexes and sets of propo-
sitions.
Propositions as Form Compounds. As constructs propositions
are compounds developed by elaborate abstractions from events.
The proposition P(Iron is a metal) compounds certain qualities
and certain substances for classificatory or manipulative purposes.
NO one probably would object to such a description of the formula—
tion procedure, nor to the implication that the products constitute
formulae. The question is: Can one, accept the total set Of implica-
tions?
Basically, we are occupied with persons operating upon things
in a remote way. No dualistic version of concepts or ideas is ad-
missible. We have already excluded all notions of mental states
as over against actual things. Nor are we concerned with a num-
ber of individuals (class aggregate) as against the individuals as
members. We are dealing, rather, with products of prior inter-
behavior with things.20 In Short, any logical view harboring uni-
versals, forms, symbols, or propositions which are identical with
things we regard as logistic and derived from historic identifica-
tion of thoughts, things, and words.
Propositional Complexes. Propositions organized into com—
plexes or minor systems are illustrated by syllogisms, implications,
sorites, etc. We must, however, differentiate Sharply between
the sentences which refer to, or stand for, the propositional chains
and the propositional complexes referred to or denoted. All such
formalistic systems take their place among those other systems
built on the basis of actual things, relations, acts, etc. A genuine
operational theory, by keeping clear the situations in which per-
sons are interoperating with their stimulus Obj ects, avoids the pit-
fall of distinguishing between generic and universal propositions
as absolute and inclusive types. Instead we handle interbehavior
with all sorts of things, abstractions, constructs, and so on.

ABSTRACT FORMS AS TOTAL SYSTEMS

To describe total systems as products built out of prior form
products would be tantamount to expanding the classical univer-
mOur view probably has much in common with Dewey’s emphasis of forms
accruing to subject matter by inquiry, but we object to things arising from inquiry
out of a‘vague matrix called “subject matter.”
I44. PSYCHOLOGY AND LOGIC

salistic and exclusive logics. This is not our intention, we merely

want to indicate the trend followed by typical formal logical sys—
tems: the dialectical, analytical, symbolic, and methodological
logics.
Dialectical Logics. By dialectical logic we mean the traditional
ontological systems of Hegel, Bradley, Bosanquet, and others who
identify thoughts and things. Such logic is built mostly of uni-
versals or quality forms. Relations are regarded as internal—that
is, continuous with terms related. As such, relations are presumed
to be qualitative essences. Also, for such logics quantities merge
with qualities.
Whatever representation of fact one can find in such systems
is Vi'tiated by the universalistic and necessitarian character given
them. Precisely because these systems
areo articulated with com-
monsense events, rather than with scientific or mathematical
materials, the fall13 all the greater when the system maker climbs
above the reaches of concrete data.
Analytical Logics. Logical-systems derived from Aristotelian
sources are built with classes or aggregates as building blocks.
The primary problems concern the interrelation of members of
classes. The subsumptionistic features of such logics really con-
cern class inclusion or exclusion.
But what Of the great emphasis upon sentences or propositions?
The answer: Verbal or sentential materials are representational,
and are employed only to interrelate class members and classes.
Symbolic Logics. Symbolic or mathematical logics are con—
structed of relations. The fact that mathematics is the ideal and
prototype of symbolic logic in itself substantiates this point. In
the symbolic domain propositional logic deals only with propo—
sitions as terms in relation. Various truth functions serve to 'relate
sentences or propositions. i
Methodological Logics. Operations, actions, and methods ab-
stracted from various manipulations are the building materials
of those logics claiming to be methodologies of science. We; have
discussed such systemological products in Chapter I 5.
 CHAPTER XIX
INTERBEHAVIORAL LOGIC AND CAUSATION

CAUSATION AS APPLIED LOGIC

THE study: of causal problems brings us to the essentially

applied division of our treatise. This division includes, besides
causation, the consideration of probability, laws of thought, meas—
urement, and scientific method. These are topics not treated by all
logicians, nor, indeed, even regarded as necessary subject matter
of logical investigation. For us, however, their examination is eX—
tremely important, as it enables us to Observe the results we ob—
tain from applying interbehavioral principles to such materials.
The procedure we follow in this work of application is illustrated
by the following questions:
(I) Do such problems as causation, probability, measurement,
and the nature of scientific method belong to the logical domain?
If so, on what basis? And do they belong 'to logic exclusively?
(2) To what extent do answers to the above questions depend
upon cultural presuppositions, as well as upon prior assumptions
concerning psychology, epistemology, and logical theory?
(3) Are there single and unique descriptions of the above top-
ics, or are there many? −

(4) DO any or all of these topics pertain exclusively or pri-

marily to the event, language, or system-construction domains?
In, the present chapter we ask: What light can specificity logic
throw upon the causation problem?

Is CAUSATION A LOGICAL PROBLEM?

Historically, causal problems were not apportioned to the logi—
cal, but rather to the ontolOgical field. Some logicians still declare:
“Cause and effect are not logical categories.”ll They may add that
cause and effect are “recalcitrant metaphysical ideas.”2 In a similar
mood it is asserted that the analysis pf the meaning of causality
is not the logician’s task.3 It is not uncommon, however, to find
Eaton, General, p. 508.
Ibid.
Cohen‘and Nagel, Introduction, p. 245.
I46 PSYCHOLOGY AND LOGIC

logicians firmly insisting that cause is not an ontological but a

logical category concerned with the organization of means and
ends.‘ Some logicians even declare that inquiry into cause is an
inevitable feature of logic.
The question whether causal problems are at home in logic
obviously involves more than a simple adjudication of the boun—
daries set to inquiry. It is a'question of the nature of cause itself.
Moreover, any answer implies a distinctive logical theory. Whether
a logician treats causation, and in what manner, depends also
upon his psychological views—his notions of the way individuals
operate in causal situations. Those who hold that logic—is essen-
tially system building find that _causal _‘situations_provide scope
in a unique way_for logical Work.

EMERGENCE OF CAUSALITY IN NPETHODISTIC LOGIC

To Kant’s Copernican revolution and to Mill’s renovation of
Inductive Logic we may trace the entrance of causal problems into
the logical domain.5 The synthesis of ontological and logical in-
vestigation was initiated when Kant forced Objects to conform to
the knower’s modes of cognition, which, of course, comprise causal
as well as other categories.
But if Kant originated this tradition, Mill established it by
welding logic to the methodology of science. Mill thoroughly
incorporated causal problems into logic when he attempted to con-
vert logic, in whole or part, into a Science of scientific method.“
There is hardly a doubt that Kant and Mill emphasized cause as
a logical problem in order to build up an empirical world system.
In line with their architectonic interests their logics were designed
to shOw how reason operated in ordering actual facts, rather than
to demonstrate the interrelation of inferential propositions, as in
traditional Formal Logic.
The formalistic tradition, however, has never really been sup-
planted. Certainly the tremendous development of mathematics
as a rigid constructional or hypothetico-deductive enterprise has
only served to support and vitalize it. The encouragement which
. I

Cf. Dewey, Logic, pp. 4559i. T‘

We pass over any claims made for the Hegelian system of concrete universals.
°The procedure is to incorporate ampliative induction into logic, and to equate
induction with the search for causes. Cf. Kneale, Probability,Part 2.
 INTERBEHAVIORAL LOGIC AND CAUSATION I47

deductive mathematics gave logicians resulted in the evolution of

logical systems featuring symbol structures, Organizations of re-
lations, types of order, and similar abstractional organizations.
Thus logicians have increasingly inclined toward formulating and
transforming propositions (sentences) and away from observations
concerned with particular existential occurrences. The latter are
even spurned as experiental and intuitive.
Cause, therefore, treated other than as an abstract formal
category, is not welcome in conventional logic. Certainly, cause
envisaged as an interrelationship of particular happenings has a
very precarious place in logical treatises. Methodologians are
formalists even when they are interested in setting up large gen-
eral method systems for scientific work, thus they make cause into
a category belonging to the logical domain. In such a methodologi-
cal treatise as Dewey’s Logic, the adjective logical, employed to
qualify the cause category, removes cause from the domain of
existence and actual things and places it among the constructs.
Conventional logicians apparently can only assimilate causal
problems as long as the latter remaindetached from concrete
events, although sometimes cause is regarded as a leading principle
which directs contact with existing things. This middle course is
adopted because, like species, essence, purpose, end, and simplicity,
ontological cause has been interpreted as an objectionable property
of nature.7 We suggest that the fault of contemporary theories
of causation lies not in their inclination toward ontology or epis-
temology, but in their departure from objective happenings and
their attachment to some culturally influenced attitude. At the
basis of illegitimate interpretations of causation is the absolutistic
logical tradition. It is this same tradition which accounts for un-
satisfactory ontologies and epistemologies.
In order to clarify the causal problem and its relation to logic
we examine briefly the development of ideas concerning causation.

CULTURAL BACKGROUND OF CAUSALITY

It is an appealing convention that the technical notion of cause
originated among the Greek scientists, who developed it as a re-
placement -for ideas of indetermination and chaos. In order here
is the declaration of Leucippus:
Dewey, Logic, p. 4.62.
I48 PSYCHOLOGY AND LOGIC

N 0 thing arises without a cause, but everything through determined

grounds and under the stress of necessity.‘3

The widely held View that when thinkers adopted the causal
notion to account for things they discarded mystic processes and
became exclusively occupied with events as caused by other events
is, however, scarcely true. The evolution of the causality construct
did not once and for all establish a mode of thinking concerned
only with laws derived from observed things and events. Actually,
the interest in cause in no sense excluded the mystical from the
explanations of natural happenings. Worse, from time to time
cause itself was made into a mystical process, for under this rubric
were placed innumerable magical forces which were presumed to
bring things aboutIn some unknown, and sometimes impossible,
manner.
Under simple cultural conditions causes have been regarded as
elementary agencies for producing various obj ects and events,
not infrequently the term cause refers to acts of personal creation.
Again, causes are powers exerted by an agent in carrying out some
purpose. As is well known, even thinkers of the, I7th‘.and early
18th centuries were not emancipated from such views. Hence the
controversy between Leibnizians, who believed the supreme power
so created the universe that it was everlastingly self-Operating, and
Newtonians, who held that the cosmic mechanism required oc-
casional divine intervention to run smoothly;
It is clear, then, that the causal construct has had an uneven
career in the history of thought. What lends credence to the belief
that among the Greeks cause became a definitely naturalistic idea is
the simple but straightforward presentation of Aristotle. He be-
gins with the problem Of knowing a thing and aSSErts that knowl-
edge depends upon cause or primary conditions.°
Thus he enumerates four causes: (I) material, that from which
a thing comes into being—the silver of the bowl, the bronze of a
statue, (2) the formal, in the sense of shape, pattern or formula
-——for example, the ratio 2: I, in the case of the octave, (3) the
primary source of the change, “whattmakes of what13
made,” an
’ Diels, Vorsokratiker, vol. II, p. IO.

Analytica Posteriora, 71b, 9; Physica, 134a, Iz.

 INTERBEHAVIORAL LOGIC AND CAUSATION I49
agent, as an adviser, parent, etc., (4) the end or “that for the sake
of which” a thing is done—health is the end of walking.10
Under more sophisticated conditions cause became an all—em-
bracing impersonal principle or law employed to account for every-
thing that happened, and to order events within a single system,
or set of systems. The scientific problem was construed as the search
for the powers or conditions which gave rise to certain events..
Such ideas were reinforced by the quest for certainty and incon—
testable knowledge.
On a somewhat higher level of cultural evolution cause became
transformed into a series of conditions requisite for explaining Ob-
served events. Causes were accordingly regarded as rules of order
and regularity, or as laws describing, or referring to, events. This
type of construction had its peak development when cause was
finally conjoined with problems of predicting and controlling
future happenings.11
D
CLASSICAL CAUSAL THEORY
Causal theory originates directly from a cultural matrix. Logi-
cians and other writers who espouse a single and universal causal
construct therefore have a large range of items from which to
choose. They may variously characterize cause as some kind of
entity—a category, a power, a relation, or an action.
To facilitate the inspection of classical causal constructs we dif-
ferentiate three types of theory: (a) formal, (b) epistemological,
and (c) ontological. Each type represents a special mode of inter—
behavior with causal situations. The basis for accepting a particu-
lar theory is one’s adherence to a certain philosophical school and
proximity to scientific research. Such factors influence the kind of
problems formulated, the descriptions of cause, and the way the
causal situation is handled.
(a) Formal Cause Theory. The formal theory consists of as-
sertions concerning symbolized relations. Causal events are inter-
acted with as abstruse entities. Even when particular happenings
are alleged to be the sources of constructed theories, the ideas
are pitched on the abstract plane of fOrmal and verbal systems.
The following questions represent typiCalformalIssues:
”Physica, 194b, 16, Metaphysica, 1013a, 24. \

n Cf. Burtt, Value.

 50 PSYCHOLOGY AND LOGIC

IS an exception possible to the law or to the fact of causal necessity?

Are cause and effect absolutely distinct, or can one contain the
other?
Must cause inevitably precede effect, or can cause and effect be
simultaneous?
Can cause continue after the cessation of. effect?
Can effect continue after the cessation of cause?
Can causes produce their effects at a distance, or must they operate
through a contiguous medium?
Are causes different from, or identical with, conditions?

On the whole, interbehaving formally with cause results in

systems of compatible sentences (p. 147). Ample scope is therefore
available for the autistic creation of cause-and-effect things and
relations. 6

(b) Epistemologtcal Cause Theory. Epistemological causal the-

ory concerns processes of knowing the connections existing or al-
leged to exist betweeri things. Kant’s categorial theory to the eflect
that cause is an a priori determiner of invariable and irreversible
succession is an excellent illustration.
The epistemological handling of causal happenings is well
demonstrated by the discussions concerning causal necessity. It is
asked: Is causal necessity an Objective necessity of things, or a
logical necessity of thought or assertion? Again, do assertions of
causal necessity refer to hypothetical and existential things or only
to absolute and mathematical connectiOns?12 Still other questions
concern the exclusive applicability of cause to past events or pre-
dictions of the future.
(c) Ontological Cause Theory. Theoretical constructs of the
ontological type stress relations between things and events.. They
are effectively demonstrated by the following classic examples.
Thing Causes. Though Hume finally settled down to the _belief
that cause basically consists of relations between things, he vacil-
lated between this View and the notion that causes are themselves
things, as the following passages indicate:
The idea, then, of causation must be derived from some
among Objects.13
rIelation
'2
Cf. Pap, Meafiing. ∙
‘3
Treatise, p. 75.
 INTERBEHAVIORAL LOGIC AND -CAUSATION I5I

Whatever Objects are considered as causes are contiguous.“

A cause is an Object precedent and contiguous to another.15
Hume’s choice is influenced, of course, by his doctrine that all
things are reducible to sensation; and cause is not a sensation or
a quality of an obj ect.16
For the more stable prototype of the view that causes are ob-
jects we must. go back to Hobbes’ attempt to establish a natural
philosophy. He opens his discussion of cause and effect with the
statement: −
A body is said to work upon or act, that is to say, do something to
another body, when it either generates or destroys some accident in
it."

Concerning the cause of motion, Hobbes is clear:

There can be no cause ‘of motion, except in a body contiguous and
moved.13

Though statistics are lacking, there is no doubt that a tremen—

dous number of thinkers regard causes as things, as in the case of
chemicals or drugs causing death. The entire traditional formula—
tion of agents and patients, of doers and sufferers, is supported by
the pillar of cause as Object, or property and accident of an 0b--
ject.
Relation Causes. Abstractive and universalistic habits of thought
automatically result in the notion of cause as relation. There are
Hume’s constant conjunction, Kant’s irreversible succession, and
Mill’s uniform and unconditional sequence. Also there are varia-
tions as to constancy of conjunction,19 or absence of recurrence
with a residue of relations between particular conCrete events?o
Again, writers differ as to the terms that are related, some lean
toward Objects instead of events; still others define the causal re—
lation as connecting factors in a single event.21
Ibid., 'p. 75.
15 Ibid., p. 170.
m Ibid., p. 75.
17 Works, vol. I,
p. 120.
19Ibid., p. 124.
Hume, Treatise.
3"
Ducasse, Nature.
2'
Winn, Nature.
152 PSYCHOLOGY AND LOGIC

Event Causes. Change, Wundt declares, is the primary condi—

tion of causality. Cause accordingly concerns not things but events.”2
Obviously, however, even the doctors who agree differ as to the
nature of events. For example, it is asserted that causation'always
involves transmission of energy from place tO place.23 This formu-
lation is opposed to such characterizations of cause as that it is
motion“ or change,25 or, finally, action.“

CAUSALITY: UNIVERSAL AND SPECIFIC

Everyone who approaches causality from the angle of events
must be impressed by the variety of causal situations. Still there
a
is no repressing the great urge to set up single comprehensive
theOry. Conventional doctrine projects the general presupposi-
tion that causality constitutes a basic law of nature or of science.
The assumption is made either that the universe is constructed
on a plan of necessary connection or that scientific explanation de-
mands a universal causal principle.
Once more we see exhibited the enormous hold of abstraction-
ism and generality on logical thought, as well as on the analysis of
science. The result for causal study is such a vaporizationof events
as to allow anything to be said about them. In such a vacuous
causal universe it may be asserted that “Whatever is, may not
be,” “Not everything that could happen does happen,” “The con-
trary of every matter of fact is still possible.”
The universalistic claim that the world is a system of necessary
connections—namely, “every effect must have a cause”—— ——iS not
condemned by us merely because it is a vain apothegm. Not at all.
The gravamen lies in assuming that such an assertion is factually
significant. Even more serious is the fact that it serves to cover
up the inevitable specificity of causal situations. N 0 general, for—
mula can ever satisfactorily symbolize the complicated concoiIrses
Of events comprising causality. l
The assumption that the goal of science is to discover univiersal
absolute causes is to equate science with metaphysics. InsO‘far as
Logik, vol. I, p. 586. −
’3
.
: Winn, Nature, p. 202.
Carus, Problem.‘
−
”Schopenhauer, Fourfold, p. 20, Gotschalk, Nature.
∟
2r'Mercier, Causation. −
 INTERBEHAVIORAL LOGIC AND CAUSATION I
53
Hume and his followers intended to prohibit this kind of equa—
tion they were On the right path. The search for absolute causes
obviously goes counter to any procedure actually practiced in a
scientific laboratory. It is doubtless this type of abstract andre—
mote cause to which Newton referred in connection with gravita-
tion:

Hitherto we have explained the phenomena of the heavens and

of our sea by the power of gravity, but have not yet assigned the
cause of this power.27 . . . Hitherto I have not been able to discover
the cause of those properties of gravity from phenomena, and I frame
no hypotheses, for whatever is not deduced from the phenomena is
to be called an hypothesis, and hypotheses, whether metaphysical
or physical, whether of occult qualities or mechanical, have no place
in experimental philosophy. In this philosophy particular propositions
are inferred from the phenomena, and afterwards rendered general
by induction.28

One might argue that the very number and variety of causal
situations are so great as to accommodate the two grand generaliza-
tions that the universe is a set of necessary connections and that
science demands a universal causal principle. Even if this result
could conceivably be brought about by judicious deflation and
translation of the original assertions, their relevancy at once be—
comes severely limited. Similarly, by proper treatment the more
Specific formulations that we have called formal, epistemological,
and ontological may be adapted to certain phases of causal situa—
tions. In all cases in which generalized formulae are adapted to
specific instances it is the detailed variation of the events under
consideration which becomes emphasized. Causal principles are
the more cogent the closer they approach concrete events.
Keeping close to events makes irrelevant any large-scale formal
assertion. Important only‘are analyses and formulations that serve
to connect one event with others in a class. Specific causal formula-
tions exclusively enable us to make predictions” and in general
achieve some experimental control over certain situations.

Principia, General Scholium, p. 546. ∙

Ibid.,‘p. 547.
Cf. \Schlick, Causality.
154 ‘PSYCHOLOGY AND LOGIC'

LOGICAL N ECESSITY AND FACTUAL CAUSE

Formalistic logicians separate themselves into two. camps. The
strictly mathematical logicians remain completely within the
bounds of mathematical forms. The philosophical formalists in-
cline toward the more contentful systems which articulate with
scientific materials. It is the latter who are confronted with chal-
lenging causal problems. On the one hand, they are concerned
with propositions or sentences implying absolute and universal
necessity. On the other, causality implies events and their. rela—
tions—in short, probabilities, specificities, and, in general, induc—
.tive as over against deductive processes.
The _way out of this discrepancy is found, of course, in the very
fact Of formalism itself—that is, sententiality. Thus arises the
construct Of logical necessity. For the mostcpart, this sort of neCes—
sity consists simply of assertions— —“Every event must have a
cause,” “Order13 the essence of nature.” Formal logic comprises,
in fact, a stock of sentences containing the word cause—sentences
which have no connection with the interrelations of things, their
properties, and relationsIn concrete event systems.
Logical necessity and causation are close to classical metaphysics
despite the fact that formal logicIs presumed to be concerned ex-
clusively with abstruse sentences. Certainly logical necessity is re—
mote from any causal system Of concrete happenings. It has nothing
to do with such factual necessity as providing oxygen for strato—
spheric flying or an adequate diet to maintain growth. The domain
of formal logic allows for such mystic ideas as a deterministic or
indeterministic universe, a causal or an acausal cosmos.
The gulf which linguistic logic creates between logical and fac-
tual necessity is thrown into relief by a comparison recentlymade
between causal and formal implication. Hofstadter
asks his
3°

readers to suppose that the formula “All men are mortal” repre-
sents a formal implication, and at the same time a causal law. N ow,
he says, the real possibility of an immortal man is compatible;l with
the truth of the formal implication. But the very supposition that
under certain conditions a man might live forever signifies the pos-
sibility of violating a causal principle And this goes counter to
the idea of a causal law.31
3°
Hofstadter, Causality.
Ibid., p. 259. ‘
 INTERBEHAVIORAL LOGIC AND CAUSATION I 55

The attempt to diflerentiate logical (analytic) necessity from

causal (real) necessity succeeds only in revealing the irrelation of
formal logic and concrete events. The formal interrelation of
sentences can not be connected with descriptions Of specific hap-
penings. Concrete linguistic events can be handled only by non-
formalistic sentence systems.
Causal events are clearly different from the constructional sys-
tems connected with them. Of the latter there are many different
types. Logicians, accordingly, must choose among these systems.
Every logician sensitive to things, events, and conditions inclines
as far as possible toward sentence systems describing actual inter-
relations of pattern components. But few logicians accept the view
that they are operating with objective happenings, thus they are
inevitably limited to systems of abstruse sentences. It is significant,
however, that logicians attempt to get at least so close to events
as to consider the relationship between formal and causal implica—
tion.
From an event standpoint the logically possible consists of a
set of incomplete actualities (factors). The chemicals are present
but not the catalyzer. No relevance or significance is attached to
any factor or combination except as it is known from observation
and experimentation—that is, through prior interbehavior. .

Unfortunately linguistic logic and its occupation with logical

necessity can only lead to argument and verbal systemology,
which does justice neither to logic nor to causal events. Linguistic
logicians, for example, may entertain the metaphysical aSsump—
tion that the world is such that not everything that could hap—
pen does happen.32 What “could happen” is regarded as a product
of a selective or restrictive action.
What is meant by the logically possible may be illustrated by
considering the number of possible combinations of S and P. A
may be both S and P, S bUt not P, P but not S, and neither. N 0W
the assumption is made that one of these combinations does not
occur. This is interpreted as a causal necessity limiting logical neces—
sity or possibility.
Connected with such a View is the belief that when we seek the

”Hofstadter, Ibid., Oliver, Logic.

The term possibility is used to avoid saying certainty.
Cf. Oliver, Logic, also
3"

Hofstadter, Causality.
I 56 PSYCHOLOGY AND LOGIC

cause of an event we seek the reason “why” this event, and not
some other logically possible one, is occurring. It is chiefly the
linguistic influence which fortifies the belief that a causal investiga-
tion involves anything but the problem '“how” an event occursa4
—namely, how the constituent factors of things, their prOperties
and conditions, are organized in an event situation.

CAUSALITY, FUNCTIONALITY, AND CORRELATION

Sophisticated views of Causality as a feature of nature have de-
veloped under a number of influences. Outstanding has been the
effect of detailed facts such as the sciences uncover. The ideas of
powers and potencies appear unserviceable and even repugnant to
the scientific worker. For this reason causality'In the traditional
sense of potency or productivity has been, replaced by the notion
of functionality. x=f(m), or more elaborately dy/dxz, sym-
bolizes the kind of relationship of variables en’visaged by the scien-
tist. It may well be that such symbolized relations are constructs
that can not be properly descriptive of events, but they neverthe-
less help to avoid occult powers. We are willing to sponsor such
forfnulae if they are descriptive, Since no description can be more
than a referential construct. . _ __
_

Statisticians, too, concerned with concrete data make a sharp

distinction between correlation and cause. Sometimes they gb so
far as to declare that statistics is fundamentally the investigation
of events which, because they can not be causally interrelated, must
therefore be statistically correlated. On the other hand, many stu—
dents admit no cause relation except Hume’s constant temporal
sequence. They believe, then, that cause and effect amount to no
more than a covariance of presence and absence.
Cause and causal relations, therefore, may simply be regarded
as the interrelations of field components. Certainly, causation can
be formulated as correlation. At first, correlationists were prima-
rily interestedIn organizing two-factor systems. Later, the devel-
opment of partial and multiple correlation techniques amplified
the original view. The above paradox'Is easily resolved by indiCat—
ing that cause is, after all, only a type of correlation.
Always there remains, however, dissatisfaction with the isimple
association of factors lacking any eflicacy‘with respect to each
“ Cf. Hempel and Oppenheim, Studies.
 INTERBEHAVIORAL LOGIC AND CAUSATION
I
157
other. Causation, it has been thought, must be an essentially effec-
tive sort of factOrial copresence. On this basis it is argued that
causal factors consist \of actions of things or persons, which, in com—
bination with each other, constitute new fields as compared with
other correlations. In other words, causal correlations are regarded
as a special sort of factorial combination, a coming together of
causal factors which, when together, constitute a different event
system.
To avoid an unwitting return to previously rejected ways of
thinking we must differentiate between effective combinations re-
sulting from a conjunction of “potencies” and effective combina-
tions resulting from changes in event fields. We exclude here all
forms of causal antecedents. NO object, or action of an object,
may be regarded as a causal factor prior to its presence or occur-
rence in a specified field. These antecedent potencies are rejected
when they are presumed to have inhered in any copresent Obj ect,
as in the classical single causal antecedent. Causal changes or fields
are functions of mutual and reciprocal changes in every aspect of
a factorial system.
Causal changes in any field constitute a rearrangement in the
simultaneous coexistence of factors in a unique pattern. When we
say that an inflammable substance is caused to ignite by a spark
which happens to fall upon it, we utter a misleading statement.
No such change of field or event actually occurs without the co-
presence of both the spark and the inflammable substance in a par—
ticular spatiotemporal reference frame. The emphasis here is on
the things, properties, and conditions making up the particular
event field. In some fields the factors constitute a pattern which
may not be intense nor even discernible. Frequently an explosion
is required to stimulate causal investigation, but other event
changes consisting of less _noticeable factorial arrangements are
not different in principle.
The fact that many changes and transformations of events are
as
neglected or regarded inconsequential suggests a differentiation
between three features of causal situations. First, there are the
combinations and recombinations of factors. under specific condi-
tions. Here we must point to events occurring without human in-
tervention—for example, complex reversible chemical reactions.
Next, we have descriptions of such reactions in the form of func-
I 58 PSYCHOLOGY AND LOGIC

tional equations. Here the observations influenced by the observer’s

problems and interests stand out. And finally, there is the more
remote? descriptive interrelating of certain events with other simi—
lar and analogous events. In such instances constructional opera-
tions are emphasized which are only_partially stressed in describ—
ing and recording events.

OPERATIONAL CAUSAL THEORY

Classical cause theories are__ metaphysical—at least, cosmic.
Their proponents interbehave with ~causal problems in a grand
manner. By contrast we propose an operational mode of causal
interbehavior, based on the assumption that causal problems arise
directly from the scientist’s contact with objects, events, and rela-
tions. His operation with causal problems is best examined in a
particular type of situation—a fact which is truc, of course, of any-
thing he investigates. Operational cause'can only be studied in
connection with problems localizable in a certain domain—chem-
istry, biology, physics, and so on. The operations consist of assess-
ing 'events, or interbehaving with their varying factors on the basis
_of developed techniques and available materials in’order to arrive
at some'specific answer to a particular question. .As a rule, the
answers are stated in the form of propositions indicating the inter-
relationship of event factors. Whether or not general causes exist,
and why, are questions falling outside the scientist’s field of in-
terest.
According to our hypothesis that causal processes and relations
constitute factors in event fields, causal elements consist of objects,
their combinations and relations in particular systems. All things
existing as parts or features of a certain pattern of happenings may
be said to participate as factors in that particular causal field. In
some causal events there are few factors, in others many. In case
there are many we find great variations in the proportion of those
factors that appear more prominent than. the remainder. Again,
in some, events there may be no outstanding factors at all. Whether
there are many or few, the factofsxmay be either sequential or
coordinate. in time. Causal Situations'mUSt further be differefitiated
on the basis Of-the relative availability of the factors for Observa—
tion and experimental manipulation. In some.instances the factors
can only be hypothetically named and enumerated.
 INTERBEHAVIORAL LOGIC AND CAUSATION 159
Correspondingly, causal description consists of enumerating
the factors comprising the causal situation. This analysis is ini-
tiated by some problem of immediate achievement or prediction.
Since causation basically amounts to the copresence of a given
number of components, every vestige of ultimacy and absolute-
ness is barred. Furthermore, the only element of universality lies
in the identifying and specifying formula. Beyond that there is no
resemblance to any View that reduces cause to a single exclusive
type of event or construction. Faced with the multiplicity of spe-
cific causal systems one can scarcely remain convinced that there
is but one adequate interpretation of cause.

CLASSICAL PSYCHOLOGY AND CAUSAL THEORY

Causal theories, we have indicated (p. I 49), always originate and
maintain themselves as part of a comprehensive cultural complex.
This means that among the many intellectual counterparts to cause
we find a psychological theory. Outstanding variants of causal
theory are associated with their corresponding psychological doc-
trines of either monistic or monadic mind. For instance, a rational—
istic and a priori causal theory correlates with a doctrine of unitary
mind, while an empirical and a posteriori notion fits into a mona—
distic or atomistic mind theory. To consider some historical ex-
amples, such an adherent of unified mind as Kant naturally enter-
tained a causal theory according to which necessity was a determin—
ing condition for the existence of events, while Hume’s psychic
atomism had no room for necessity among existing things and
their relations. In general, it is psychistic psychology which pro—
vides comfort to those who must diflerentiate between the abstrac-
tion of necessity and the contingency of existence.
Much recent writing on the causal problem centers around the
associationistic psychology of Hume and Mill. Scholars are in-
fluenced by atomistic psychology when they (I) adopt the Hume-
Mill pronouncements on cause, (2) modify them in detail or (3)
oppose them entirely. Discussions of causation, centering around
the possibility of immediate spatialxcontiguity of antecedent and
consequent or necessary connection, as well as inevitable conjunc-
tion, definitely reflect the Hume—Mill reduction of psychological
events to the appearance and reappearance of psychic states as dis—
crete e‘lements tied together by habit and custom, or by the gentle
 I60 PSYCHOLOGY AND LOGIC

forces of association. As we have already intimated, the British

empiricists, for whom both knowledge and the world consisted of
sensations and their interrelations, could not make cause a sub-
stantive quality such as sensations were presumed to be, and there—
fore had to turn to cause as a relation. It followed that for them
causation was a necessary sequence of sensations or their combina—
tions.
From an historical standpoint no one can deny that Hume’s
notion of causation has been of great value in upsetting the inept
rationalism rampant in his time; But it is questionable whether
his own type of sensationism, which, of course, is not altogether
different from rationalism though of another order, should be
allowed to confound the study of causal problems.
Certainly, cause is no quality, as Humels “impressions” are
presumed to be. Nor are objects reducible to psychic processes, SO
that notions of precarious existence or interrelatio‘n are warranted.
No less objectionable are the mentalistic constructions which dif-
ferentiate between experiencing on the basis of impression or sen-
sation as immediate psychic states, and reasoning, which is more or
less independent of such immediately given states.
Rationalistic causal doctrines which, generally speaking, imply
that cause is a necessity of existence, correspond to monistic psy-
chology involving a continuously existing unitary mind. In formu-
lating causal principles this version of monistic psychology is no
more helpful than the monadistic one. No connection can be pos-
tulated between natural events and creative powers, regarded
either as mirrored in a unified mind or as independent of such a
mind. Rationalistic causation may be characterized as why causa—
tion. It certainly harks back to impersonal or personal creative
powers, and, in the latter case, to the projection of human desires
and interests into events. When such a procedure is fortified by
some sort of psychic psychology its weaknessis doubled. No tra-
ditional spiritistic psychology can do justice to any occurrence or
interpretation of natural events.
Such ways of thinking are, of course, completely superseded
by objective psychology. When dealing With concrete situations the
causal agency may be a thing or an event. By thing we may refer
_ to a chemical compound, which, when brought into combination
with other compounds, results
in an CXplOSiOflI‘B‘ut Since in all
 INTERBEHAVIORAL LOGIC AND CAUSATION 161
cases causation has to do with events, causal or agentive things
are themselves events (p. I 57). Admittedly, there is a definite in-
terrelationship of objects in the sense of crude things, such as the
interrelationship of chemical elements or compounds, or when
water is thrown upon a burning house. The reduction of causation
to relationship in the British empirical tradition is, of course,
completely set aside.
On the other hand, interbehavioral logic allows for the specifica—
tion and emphasis of the relational factor as primary in descrip-
tion. This is to be expected, since according to interbehavioral psy-
chology the term relation refers to the combination of concrete
factors in a causal situation. In no case, however, are we dealing
with relations remote from actual things, for, Obviously, from an
objective point of View, relations are only features abstracted from
the relationships of objects or events.

PATTERN OF CAUSAL INVESTIGATION

Conclusions concerning the logic of causation must be drawn
from two sources. First we analyze the causal situation and sec—
ondly we observe the logician interacting with that situation. This
double requirement we can carry out by examining the pattern of
causal investigation.
Following our general paradigm of interbehavioral study We
differentiate between three large classes of factors: (A) data of
various sorts (crude prior to analysis, or refined = analyzed);
(B) investigative operations (experimental, mensurative), and
(C) constructions (hypotheses, theories, laws).
A. Data
There are two sorts of causal data: (I) constructs and (2)
events.
(I) Causal Constructs. Constructs as data consist of reactions or
products of reactions. “Arsenic caused Smith’s death” as descrip-
tive of Jones’ reaction to the event involving the death of Smith
is different from the death event. Smith’s actual death as part of
a system involving arsenic is an event independent of any reaction
to it, though without such a reaction it wouldbe unknown. Causal
stimulus objects as reactions consist of beliefs, utterances, or writing
acts perfOrmed with respect to organizations of systemic factors
I62 PSYCHOLOGY AND LOGIC

which we refer to as Smith’s death by arsenic. As products the

data, or stimulus Objects, comprise propositions, allegations, or
theories concerning events or alleged events.
Because such data are secondary—that is, come to the observer
by means of a person’s (Jones’) intermediating action it may well
happen that no arsenic death has actually transpired. Jones’ belief
may be delusory, his assertion mistaken or false. In such cases we
assume that he is responding to substitute stimulation, not to the
direct Smith~arsenic—death event. Substitute stimuli, of course,
can stand for mystic and magical apparitions as well as for actual
events. Because Of this latter possibility we may regard Jones’
interaction with the substitute stimulus object as an authentic—
though nevertheless fictitious—-causal datum.
On the whole, construct data play a larger role in the causal
investigations conducted by logicians and philosophers than in
those carried on by scientists. The latter are definitely more in—
terested in causal situations in which the 'data consist of events.
We must not, however, lose sight of the fact that scientists may
regard'all causal data not as natural-event data but as construct
data. '
This point is illustrated by Russell’s assertion that there is no
such things as cause and that the classical law of causality is un-
scientific and false.35 Scientists are never interested, Russell asserts,
in necessary succession, but in functional relatiOns among variables.
The causal data of philosophers, or what they interbehave with,
according to him consist of Sentences——in our terms, substitute
stimulus Objects. Pepper, who regards causality as a characteristic
of natural events, asserts that Russell is not talking about causality
at all, but about physicists’ expressions of natural law.“ These ex-
pressions refer to observations of correlations.
That Pepper is probably correct is indicated by the following:
The essential business of physics is the discovery of “causal laws”
by which I mean any principles which, if true, enable us to infer some-
thing about a certain region of space-time from something about some
other region or regions.37
Substitute stimulus objects may correspond to authentic ca‘usa-
35

36
∙
Notlon. \.

How.
37 Russell, Reply, p. 701.
 INTERBEHAVIORAL LOGIC AND CAUSATION 163
tion if one grants the occurrence of such events. Assuming the
correctness of the hypothesis that causation. consists of a pattern
of subevents or event factors we find a continuous series of inter-
behaviors with such events. The simplest level is that of every-
day or “commonsense” situations. Here Simple questions are
asked: the goal of the inquiryis a practical outcome of the inter-
action Of events. On the scientific level more significant questions
are posed, with a correspondingly greater capacity on the part of
the investigator to analyze the interacting factors, hence, a more
sophisticated procedure of interrelating events. The scientist, too,
works out elaborate CXperimental techniques for the dissociation
and technical reassociation of factorsIn order to discover their
variants andInvariants.
In more formal (logical) situations causal interbehavior pro—
ceeds on the basis of contacts with substitutes for highly analytic
data. The direct stimulus objects are kept at a remote distance for
the sake of achieving precise descriptions and interpretations of
causal factors and their interrelationships.
( 2) Causal Events. Causal events as data consist of particular
interrelationships of observed happenings. The essentially causal
feature comprises the copresence of the event components. Events
can be observed to occur through the addition and subtraction, the
separation and contigu‘ation of the factors. Certain chemical com—
ponents are present together, and only a temperature or pressure
change is necessary for a new chemical state to ensue. Again, a
slow chemical reaction may be speeded up by adding another com-
pound Or varying temperature or pressure. If, while describing
our investigation of an event we refer to our Observation, this has
nothing to do with the event itself, its factorial combinations
constitute things independent of our Observation or description.
In psychological terms, these combinations and recombinations of
factors and events are independent stimulus Objects with which
the observer interacts.
Now in some cases it may be legitimate to describe one factor
as agentive, causal, or key, since it may be the factor required to
complete the combination—lfor example, when a person facilitates
or'h‘astens the occurrence of an event by switching on the current
necessary an explosion. His behavior must be regarded as a
for
naturalcomponent of the causal system. It operates like any other
I64 PSYCHOLOGY AND LOGIC

factor necessary to complete a'particular kind of event combina—

tion, and thus constitutes a part of the causal data.
Interbehavioral data of both the construct and event type must
be sharply distinguished from everything pertaining to psycho—
physical philOSOphy. This. admonitionIs necessary because of a
similarity in the use of terms. The terms‘construct and construction
have frequently been employed since Russell formulated “the
supreme maxim in scientific philosophising”33—— ——namely,‘‘wherever
possible, substitute constructions out of known entities for in—
ferences to unknown entities.”39 _
Russell and all the many writers who deal with the contrast
between constructions and inferred entities40 are concerned with
relations between mythical sensations and sensibilia of externally
caused mental states and the products of purely psychic opera—
tions of the mind. This is the case even when writers do not accept
Russell’s View that objects of everyday circumstances and of tech-
nical physics consist of logical constructions made up of the crude
data of sense.
Such problems and solutions belong to a philosophical system
which entertains questions concerning the construction of an ex-
ternal world (p. 206) and its relation to private and individual
,mentality. No contrast could be greater than that marking off
such a philosophy from investigations which exclude “mind,” “sen-
sations,” and “immediate experiences,” either as building ma-
terial or as constructing powers. Constructs as products, from our
standpoint, are always (a) recombinations of things, as the syn-
thesis of water from hydrogen and oxygen, (b) measurements
achieved by definite Operations, (c) descriptions and recordings of
events and of our operations in securing them, and (d) indirect
interrelations of things by means of manual (drawings) and vocal
behavior. In each case the constructing activities 'consist of concrete
interactions of the individual with his stimulus objects on the
basis of problem presented, motivation, expertness, and other
concrete factors.
Central to the constructional problemjs the nature of linguistic
l‘
as Mysticism, p. 155.
anLogical, p. 363.
-"'°See, for example, Feigl, Existential, and Beck,
to whom they refer.
Constructions, and the writers
 INTERBEHAVIORAL LOGIC AND CAUSATION 165
acts and the use of linguistic products. As in so many other situa—
tions, commerce with mythical subject matter is possible by means
of linguistic factors. It is possible to construct real existence, sys—
temic existence, possibilities, and determiners of all sorts without
adequately analyzing acts and stimulus Objects. We suggest that
the elimination of traditional mind constructions clarifies the gen—
eral constructional and inferential situation and enables us to make
a direct approach to events as data.“1

B. Investigative Operations
A study of the pattern of causal investigation demands that we
diflerentiate between the two orders of observer and investigator.
The behavior of the operator mentioned in the agentive illustra—
tion above constitutes a first-order operation. His activity belongs
within the pattern of a causal system.
The second-order observer analyzes the work of the first Op-
erator as part of the causal data before him. He is especially in-
terested in the interaction of the experimenter as he unravels the
interrelated threads he finds among the materials with which he~
works.
Investigative causal operations can be demonstrated by consider-
ing the work of an experimenter as he arranges and rearranges
the factors in his event situations. We select for this purpose an
experiment by Hopkins.42
Beginning with the hypothesis that young animals grow nor-
mally on a diet of protein, fat, carbohydrates, mineral salts, and
water, the investigator discovers that when these substances are
purified and fed to animals they do not produce the assumed re—
sults. To groups of six rats each he fed purified milk casein (pro-
tein), lard (fat), sugar (carbohydrates), and salts obtained from
oat and dog—biscuit ash. In addition, he fed one group 2 c.c. of
milk per day. The milk-fed rats increased in weight normally,
while the control group without milk feeding began to lose
weight after 10 to I 5 days.
Assuming that the operations are well controlled and the re-
sults valid, the next step in the causal investigative pattern is to
‘1
For 5. comprehensive study of construction in logical situations see Kantor, Inter-
behavior. Scientist, p. 291f.
Gedrge,
‘3
I66 PSYCHOLOGY AND LOGIC

construct propositions concerning them. Convenient terms fOr

such constructions are descriptions and interpretations.
\

C. Constructions
Descriptions and interpretations of causal occurrences naturally
differ on the basis of the type of crude datum upon which they are
constructed. Causal theories based on construct data 'are very dif-
ferent from those based on independent event data. Even in the
case of independent event data, theory constructions may Show
divergences. One may set up descriptions and laws of causal events
which are altogether independent of human agency, such as the
tidal relationships between themoon and earth. Other construc—
tions Of causal factors involve the activities of the agent or in—
vestigator who deliberately produces certain ochanges in the crude
data.
Cause and Efect. Faithful to the traditional’ relation concep—
tion, writers on cause have maintained in their descriptions two
related factors, cause and efl‘ect. Critics of such a notion as neces—
sary or inevitable relation have pointed out that it is impossible
to keep these two factors distinct. Since we are dealing with event
patterns, the terms cause and efiect must serve specific descriptive
purposes. Accordingly, when they refer to stimulus Obje'cfs,'these
terms refer to prior and subsequent combinations of factors. Cause
and effect as constructive terms serve only to isolate particular
stages of Observed events.
Causes and Conditions. As we have seen (p. I 52), writers on
causation regard the term cause as standing for a single and unique
process or agency. From the standpoint of field theory, instead of
singularity, we meet with all sorts of combinations of factors.
Some writers‘13 insist upon the distinction between cause and con-
dition on the ground that cause implies a suflicient factor, and
condition a necessary one. This is good as far as it goes, but in
actual investigations a much larger number of diflerentiations
must be made. When we deal with concrete situations we perforce
discover innumerable degrees of availability or participation of
causal factors.
Causes and Causal Interpretation. Both in the event and in
∙
_its interpretation we may specify all sorts of
‘3
Ducasse, Nature, p. 58.
Causal
conditions and
 INTERBEHAVIORAL LOGIC AND CAUSATION 167
relations, depending upon the investigator’s interest. From‘the
standpoint of specificity logic the investigator’s purpose is a definite
objective fact to consider when causes are to be determined. For
instance, there is a vast difference between an investigator’s in-
terbehavior with particular causal conditions—as in his attempt to
understand a certain disease or how X-rays are diffracted—and
that of an individual attempting to set up a generalized definition
for all possible causes. The latter-procedure can not yield any Sig-
nificant result, and the acceptance or rejection of such a formal
schema depends upon the formulator’s caprice.
Here is a convenient place to point out again that interacting
with concrete events is vastly different from interacting with ver-
bal or propositional substitutes for such events. The construction
of formal propositions achieves a certain degree of significance
only if the formulator takes account of concrete situations.
CAUSE AND SYSTEM BUILDING
Causal events as natural occurrences are completely independent
of- any person’s activity. Accordingly such events may be far re—
moved from logical or system-building behavior. Despite this fact
cause and system building may be brought very close together.
In the first place, causation is essentially systematic. Causal situa—
tions, as we have seen, comprise interrelations of factors, and this
in itself is System. Again, human behavior occupies an exceedingly-
large place as an integer in causal systems, as all scientific experi-
mentation testifies. Those who regard logic as system building
are therefore in a favorable position to conj oin logic and causatiOn.-
On'a systemological basis logic can be closely connected with causal
situations in two Ways: (A) as a general structurization en'ter-
prise'and (B) as a local instrumental procedure.

A. General Causal Structurization

Here we observe the behavior of the system builder as he erects
a universalistic causal edifice.
The first thing he does is to develpp or borrow comprehensive
assumptions or postulates as the criteria of his projected system.
We do not imply that he deliberately seizes upon these assump-
tions.'._On the contrary, he is most likely in ‘full possession of
them through the subtle influence of his general cultural milieu.
168 PSYCHOLOGY AND LOGIC

Whatever specific materials he builds upon he is forthwith pre-

pared with a ground plan.
From a number of universalistic systemization enterprises we
abstract the following Series of assumptions. “Everything is ar-
ranged or governed by necessity or law,” “nature is uniform and
recurrent,” “invariant relations exist,” “every event has a cause.”
A prominent feature of general structurization behavior con-
sists in applying these principles to things. This procedure may
be described as defining and interpreting operations. Certain things
are asserted to cause or produce other things as effects or results.
Causes are invariably antecedent to effects. Events which occur
under precisely the same circumstances must cause or produce the
Same' effect.
From this sort of activity issues a product]: which may consist of
systems of actual things and events or systems of arguments, sym-
bols, sentences or propositions, or a combination of these. The
following examples from event, language, and formalistic levels
illustrate the processes employed in the structurization of causal
materials.
(a) Event Level. Things, conditions, and combinations Of things
may be conjoined for many difierent purposes. Naturally, system
builders are motivated in diverse ways, and thus’ employ all sorts
of criteria. Concomitantly, the system products assume every pos—
sible grade of quality.
Generally speaking, causal situations are systemized on either
an analytic or a synthetic plan. In the former case, a situation is
chosen which is regarded as causational, and an organization of
the factors is then made. Synthetic systemization, on the other
hand, stresses the organization of factors for some, purpose. The
features of a situation not immediately apprehended as causal are
creatively connected in order to demonstrate causal relationships.
The aim may be description, explanation, or prediction. Features
of events are ordered as sequential or coexistent, more \or less
proximate or overlapping. Favoring such logical or system-build—
ing procedures is our hypothesis that causation consists of inter-
relationships Of field factors. ∙∙
(b) Linguistic Level. Systemization Of this type is more or less
autistic—namely, independent of events. The emphasisIS on sys—
tem as asserted and alleged, rather than on
events.and their intra-
 INTERBEHAVIORAL LOGIC AND CAUSATION I69
and inter-relations. Here we have organizations of references to,
or symbols of, causal events, such as the structures Of nononto-
logical propositions. While one can hardly put high value on a
logical system so remote from actual events as to exclude singular
and existential happenings, it is quite possible to organize such
systems.
The most famous illustration of linguistic systemization is
Mill’s set of four or five causal methods which he regards as the
only possible modes of experimental inquiry. Because this system
so admirably exemplifies linguistic causal systemization we present
in full the items which he names methods and canons.
I. Method of Agreement
First Canon. If two or more instances of the phenomenon under investigation
have only one circumstance in common, the circumstance in which alone all the
instances agree, is the cause (or effect) of the given phenomenon.

2. Method of Difference
Second Canon. If an instance in which the phenomenon under investigation
occurs, and an instance in which it does not occur, have every circumstance in
common save one, that one occurring only in the former; the“ circumstance in
which alone the two instances differ, is the effect, or the cause, or an indis-
pensable part of the cause, of the phenomenon.
0. Joint Method of Agreement and Difi'erence
Third Canon. If two or more instances in which the phenomenon occurs have
only one circumstance in common, while two or more instances in which it does
not occur. have nothing in common save the absence of that circumstance; the
circumstance in which alone the two sets of instances differ, is the effect, or the
cause, or an indispensable part of the cause, of the phenomenon.
3. Method of Residues
Fourth Canon. Subduct from any phenomenon such part as is known by
previous inductions to be the effect of certain antecedents, and the residue of the
phenomenon is the efiect of the remaining antecedents.

4. Method of Concomitant Variations

Fifth Canon. Whatever phenomenon varies in any manner
whenever another
phenomenon varies in some particular manner, is either a cause or .an effect
of that phenomenon, or is connected with it through some fact of causation.

Mill’s four (5) experimental methods stand out starkly in

their linguistic nudity. All the more so perhaps because he intended
them to be descriptive of actual situations. To begin with, the
whole system is based upon a verbal and general atomistic uni—
verse in which the determination of causes consists merely of in—
terrelating what are said to be antecedents and consequences. But
170 . PSYCHOLOGY AND LOGIC

even the succession of differentiable factors becomes reduced to

the simple abstractive common absence and presence of formalized
circumstances.
Mill, and hosts of logicians after him, have clearly recognized
the many difliculties and insufficiencies of the system of experi-
_mental methods. This fact is illustrated by the following points.
(I) Mill was troubled by the problem of plurality of causes
which negated considerably his simplicity of approach. For this
reason the method of agreement, originally regarded as coordinate
with the method of difference,"had to be devalued. (2) Again,
Mill realized that his system was not as elaborate as it appeared,
since the methods of agreement and difference were basic to the
others. Later writershave charged that treating the formulations
as so many separate methods seriously darkened the Slibject of
induction.“ Furthermore, it has been pointed out that the methods
are essentially negative. Their function is to éliminate combina—
tions of factors which do not meet the causal conjunction criterion.
Conventional criticisms, however, do not really focus on the
fundamental objection that these formulae and systems are essen-
tially incongruous with actual events. To describe methods of dis—
covering causal relations implies taking some account of the vast
number of contacts an investigator must have withithings. SO far
as Mill’s methods went, his idea that they were inductive syllo—
gisms or, at least, similar sorts of schema indicates that he dimly
realized he was concerned with an enormous job of linguistic con—
structionism.
(c) Formal Level. Generalized causal systems on the formal,
like those on the linguistic, level are built primarily out Of words.
The former may be distinguished from the latter, however, on the
basis of linguistic function. Because linguistic systems are reference
systems they still are indexes to something beyond themselves.
Not so formal systems, which in some cases are completely re—
moved from things. Since word or symbol systems have- no re—
ferents, the systems themselves constitute the only things in the
situation. When concrete things are\ forcibly dragged in to il-
lustrate the systems, they are clearly\only descriptive analogies
to actual objects and events.
Formal causal systems are exemplified by‘the classical maxims:
" Joseph, Introduction, p. 4.30.
 INTERBEHAVIORAL LOGIC AND CAUSATION 171
“post hoc propter hoc” or its opposite, “Every eflect has acause,”
“Same cause same effect” and its reciprocal. As we have indicated,
the assumption is made that these maxims are related to data, but
the enormous differences between specific events and generalized
formulae are sufficient to nullify the assumption.
Another excellent example is the causal principle alleged to
exist in identicail structures.45 The causal system here is character-
ized by the presence of the least possible content. An assertion is
made about basic interrelations of things and events on the ab—
stract ground of sameness of structure. Causal connection is as—
serted both on the ground that structural identity in different
things implies a common causal ancestry and that earlier items in
successive identical structures cause or produce the later ones.

B. Systems as Specific Causal Instruments

On the whole, the locale of systems to be used as instruments
in studying causal problems is the field of scientific investigation.
When the goal of causal study is the discovery or explanation of
specific events, local systems are constructed as definite instru—
ments of research. Aside from the fact that instrumental causal
systems are constructed for the solution of Specific problems, they
are kept close to, and influenced by, the events they are designed
to elucidate. They cannot therefore assume the proportions of
abstract grand-scale systems.
In one sense we may say that every specific scientific research
is an example of the construction of a specific causal system. Every
such research involves at some point or. other the organization of
event factors. Even when analytic processes prevail over syn-
thetic, thereIs no less a demonstration of concrete causal system.
The work of dealing with systemic cOmponents and combinations
marks the characteristic causal construction.
It might be argued that the lack of a specified causal interest
in a routine research minimizes somewhat the prominence of a
definite causal element. This point, however, loses validity as soon
as an attempt is made to explain one’s results. Such an explana-
tion is always made in terms of building a system stressing inter—
relations of component factors in an event situation.
Although instrumental and noninstrumental system-building
“Russell, Knowledge, p. 4605.
172 PSYCHOLOGY AND LOGIC

behavior is in general associated with scientific and nonscientific

situations respectively, the correspondence is not at all complete.
It may well be the scientist who attempts to construct universalistic
causal systems, while nonscientists may be keenly interested in
specific causal patterns. Recall, for instance, those rationalistic
scientists who are ambitious to constrflct a single universalistic sys-
tem. Working with mathematical tools they may be interested in
developing systems of equations, all of which are interrelated,
and when taken together may constitute a single system. Even
here, of course, certain criteria 'may be preferred, but the single-
ness of system remains the ideal of achievement. An excellent ex-
ample is Laplace’s classic statement of essential causality:
An intelligence knowing, at a given instant of time, all forces act-
ing in nature, as well as the momentary positions of all things of
which the universe consists, would be able to comprehend the mo-
tions of the largest bodies of the world and those of the smallest
atoms in one single formula, providedit were sufficiently powerful
to subject all data to analysis; to it, nothing would be uncertain,
both future and past would be present before its eyes.“

On the other hand, scientists may employ specific causal prin—

ciples for developing unit systems in any particular scientific field,
such as in chemistry, physics, biology; or within these fields units
may be developed on the basis of the interaction of field factors.
It is also possible to construct instrumental systems outside scien—
tific domains.
So large is the scope for employing specific systems in authentic
causal situations that any one of the many system instruments dis—
cussedIn Chapter I6 may be pointed to as an example of a logical
tool for causal interpretation.

CAUSATION AND INDETERMINISM

The causal paradoxes which have turned up in connection with
quantum mechanics vividly remind us of the curious part which
causality has always played in scientific methodology. The
that quantum—m—echanicalInvestigationshave led to the astounding
rumor
discovery that the universe is bereft of causal principles, of de—
termining relations, of systems of order, and ∂∙ ∙ 18 only a
p ‘
∙
∙ Introduction.
TheorIe, ‘
 INTERBEHAVIORAL LOGIC AND CAUSATION I73
routine episode in causal history. The recent allegation that the
findings of quantum mechanics aflord no stable basis for prediction,
but only for imprecisely fixating random movements by statistical
coefficients, is not the most startling of causal ineptitudes.
The belief that causal law is suspended in the case of microscopic
events reveals, above all, the incongruous views concerning both
facts and laws+in short, events and constructs. N o greater con-
trast can be imagined than exists between (a) the search for an
organizational pattern tying together a series of things as factors
and (b) the metaphysical approaches to causal problems provided
by the various philosophies."7 Obviously, such problems as the
determinism or indeterminism of events are unmitigatedly meta-
physical. If we agree that metaphysics, unlike science, deals with
spurious problems, then the determinism-indeterminism issue is
only a pseudo-problem. Be this as it may, this issue certainly mis-
construes (a) relations between events or partial events, (b) knowl~
edge and event patterns, and (c) descriptions of such knowledge.
Indeterminacy notions in the domain of quantum mechanics
merely add another item to the repetitions of intellectual history.
Recall Hume’s enormous paradox of locating the constancies of
causation, which did not really exist for him, in the customs or
habits of an insubstantial mind, which does not exist for anybody.
Such confusions between events, alleged events, and the way both
are treated (explained) suggest the wildest flights of sentence
making.
Since the problem of quantum indeterminacy arises from a con—
crete scientific Situation in which all sorts of interrelated objects are
under consideration, we can only conclude that the problem is
simply a matter of metaphysical interpretation. In other words, in-
determination is a free construction not from quantum-mechanical
facts but from metaphysical sentences.
Even if it were the caSe that the simultaneous locus and state
of an electron eluded observation or measurement, there would
be no doubt about the existence of electronic events which are de—
scribed in terms of behavioral properties.
Here, as in so many other philosophical situations, the ex-
istence of a thing and its properties are Confused with knowing
and recording them. Then are many operationalists and phenom—
" Cf. fOr example Feigl’s survey of nine viewpoints in Existential.
I74 PSYCHOLOGY AND LOGIC.

enologists who identify observational acts with the things ob-

served. Since most writers who take this position are mentalistic
in their psychological views they reduce the observational formula
to Berkeley’s esse est percipi.
What may well be lacking in a specific causal situation is knowl~
'edge of the pattern of events—that is, causal knowledge. If we
differentiate explanation and reason from ground we may corre—
late the-former with knowledge, the latter with the combination
of factors. These factors are given and are the basis for inquiry
and subsequent knowledge. True, such factors may be missing,
but never the ground about which knOwledge is sought.
The essentially metaphysical interpretation of scientific situa—
tions is strikingly exemplified in the assumption that properties of
things are constituted Of the process or result of measurement.“
The analogy'Is the metaphysical one that since things seen are
createdIn the act of seeing, things measured°are created by acts
of measuring. This analogy is made plausible by the presence of
numbers. Numbers, in such instances, are taken to be either sym-
bols of properties or the properties themselves. Once we take ac-
count, however, of that with which we are interbehaving, we find
no warrant to confuse things with the acts or results of measuring
them, nor constructions with the numbers which represent the di—
mensions of things.
“Cf. Lenzen, Indeterminism.
 CHAPTER XX

THE LAWS OF THOUGHT AND THINGS

LOGIC AND LAW

I—IKE all hUman enterprises, logic is a domain of rules, laws, and

norms. Logical laws, however, envisaged as the conventional
Laws of Thought, or as principles expressing the absolute and
necessary structure of things, are not properly interpreted. Because
the trinity of logical laws—namely, identity, noncontradiction, and
excluded middle or third—have suffered misconstruction, they
have been perennially employed to implement the dogmas of uni-
versality and necessity.
The basic flaw in the" conventional notion of logical laws is that
although few in number they are regarded as universal regulators
of thoughts and things. To make logical laws into fundamental
normative principles reflects the creed of logical omnicompetence.l
To reject such an interpretation means in no sense to disparage
principles, rules, and norms, nor to deny them an important place
in the logical field, it means merely to evaluate more effectively
the rules required for understanding things, acts, and even for
comprehending rules themselves. What, then, are rules and norma—
tive principles?
Above all, they constitute guides to the operations and pro-
cedures comprising the work performed in logic, science, and or—
ganized occupations. At once we must distinguish a priori and
absolute norms from operational principles. The former are well
illustrated by Kant’s definition of logic:
Diese Wissenschaft von den notwendigen Gesetzen des Verstandes
und der Vernunft fiberhaupt, Oder, welches einerlei ist, von der
blossen Form des Denken fiberhaupt, nennen wir nun Logih}
In contrast, scientific and logical laws are definitely constructed
principles based upon observed events; any normative suggestions
are prompted only by the conditions of the enterprise. Principles
or laws so construed constitute a definitive concern of logic. They
are not only to be distinguished from the three, classical principles,
Jfische, Kant’s Logik, p. 14-, see also vol. 1, this treatise, p. 59.
I76 PSYCHOLOGY AND LOGIC

so long presumed to be the sole and exclusive laws of logic, but

also from any finalistic set of prescriptive formulae.
Normative principles on an interbehavioral basis comprise the
theory or significant description of an enterprise. Any enterprise,
logical or otherwise, can go on, however, without a definite and
overt statement of principles. The particular behavior contains
within it its own procedure and justification. In this sense we may
distinguish between logicizing and abiding by formal logical laws.
It is an admirable practice, therefore, to set up clear and forth-
right statements concerning Our' procedures, and especially to de-
rive our principles from ongoing enterprises.
It follows, then, that logical norms must be closely integrated
with ‘the behavior which they describe or regulate. And because
these norms can be generalized and formalized, criteria for or-
ganization can be set up and met, such as consistency, noncontradic-
tion, and precise alternatives. There is a limit,’ however, to gen—
erality, for in each case the criteria are applicable only to particular
systems, or classes of unique systems. Put otherwise, the generality
is specific, and in no sense attains universality. Criteria, therefore,
are not indifferent and absolute regulators of thoughts or things,
but particular reference frames set up for specific systems.
On the ground that people think wrongly or illog'ically2 a num-
ber of logicians have, tacitly at least, objected to deriving the
principles of logic from actual thinking. Thereupon they inflate
the normative character of logical laws to an absolutistic and un-
measurable magnitude transcending behavior.
Normative principles must, of course, be separated from legal
prescriptions which are extraneous to the behavior regulated. Laws
of social organization stem from a source and criterion different
from, and often broader than, the immediate conditions of the
prescribed behavior. Civil and criminal laws may have their origin
in theological beliefs, in social organization and theory, in personal
flats of dominant rulers, economic conditions, etc.3
And, finally, logical principles, either as laws, postulates, norms
or criteria, need not be exclusively concerned with abstruse intel-
lectual operations, nor necessarily
involved
with knowledge, reas-
2Jevons, Principles, p. 7, Bain, Logic, p. 223,
and Nagel, Introduction, p. 182.
Keynes, Studies, p. 4575 Cohen

8Cf. Robson, Civilization. \

u
 THE LAWS OF THOUGHT AND THINGS I77
oning, or scientific activity,,since logic is merely system building.
These considerations may sometimes make logical laws appear to
be trivial rules for games or other simple activities. There is an
advantage in this if we‘ are thereby helped to combat the View that
logical laws are universal regulators of cosmically important ac-
tivities, or organizers Of such fundamental things as scientific
events. °

VARYING FORMULATIONS OF LOGICAL LAWS

Although the clasSical principles of identity, noncontradiction,
and excluded middle or third are generally called laws of thought,
logicians have not agreed that they are exclusively concerned with
thinking. As the following statements from authoritative treatises
indicate, even though logicians vary considerably in their detailed
views concerning the laws, they adhere to the notion that they
are all-inclusive and compelling:
Every A is A.
Every object of thought is conceived as itself.
Nothing can be A and not-A.
No object can be thought under contradictory attributes.
Every possible object is either A or not-A‘
Whatever is true in one form of words, is true in every other form
of words which conveys the same meaning. ,

Contradictory propositions can not both be true.

Of two directly contradictory propositions, one or the other must
be true.5

Whatever is, is.

Nothing can both be and not be.
Everything must either be or not be.

Everything13 identical with itself.

Nothing can have at the same time and at the same place contra-
dictory andInconsistent qualities.

‘Mansel, Prolegomena, chap. 6.

Mill, Examination, vol. II, chap. 21.
178 PSYCHOLOGY AND LOGIC

It is impossible to mention any thing and any quality or circum-

stance, without allowing that the quality or circumstance either
belongs to the thing or does not belong.6
If anything is A it is A.
Nothing can be both A and not A.
Anything must be either A or not A.
If any proposition is true, it is true.
N o proposition can be both true and false.
Any proposition must be either true
or false. 7
Symbolistic formulation yields
the .principle of identity gives: further \variations. FOr example,
A“
— A, or X3 X (Boole).

N oncontradiction is represented by:

∆≠ A, orA.= A, orx(1 — x) (BoOle).
Excluded middle is written:
X A ~ A

The many diflerences in name3 and formulation of‘the classic

laws are exceeded by variety of interpretation: Are they: (I)
applicable to things only, to thoughts only, or to bOth?' (2) inj unc-
tive or not? (3) positive, or simply negative, aids in thinking? (4)
applicable to terms only, to judgments only, or to both? and ( 5)
do they involve truth or falsity, or not?
. Despite all these diversities the tradition surprisingly maintains
itself that these universal laws regulate inferential acts and deter-
mine implication and other relations of things. This15 all the more
remarkable since actual logical work demonstrates the specificIn—
terbehavioral and constructive character of what passes for apriori
and invariant logical laws.
Not that logical laws have been without their critics. Hegel
vigorously attacked them insofar as they were interpreted as ab—
stract principles of understanding.” From his dialectical standpoint,
however, he declared that “contradiction is the Very moving prin—
:Jevons, Elementary,lesson 14..-
Cohen and Nagel, Introduction, chap. 9.
l‘l
Boole called the law of contradiction the law of duality, while
this name to the law of excluded middle.‘
Jevons applied
Wallace, Logic, p. 21 3115 see also Bogoslovsky, Technique.∙
 THE LAWS OF THOUGHT AND THINGS I79
9

ciple of the world.”10 In more

recent times logical laws have been
regarded as tautologies,11 while the constructional mathematicians,
such as Brouwer,12 stage a powerful attack on the law of excluded
middle. In the latter case the laws are presumed to concern in-
variant objective relations rather than statements or propositions.
A study of historical and current variations in logical laws really
constitutes a Survey of the whole of logic. From this survey one
may hope for answers to such questions as: whether and when
logic is normative, regulatory, or descriptive5 whether it deals
with things, thoughts or sentenceS5 whether it is a specific process
of system building or a set of absolute principles governing ulti—
mate reason or necessary existence.

.LOGICAL LAWS As PRODUCTS or CULTURAL DEVELOPMENT

That logic is the study of thought and that the classical laws are
the most pervasive and general features of logic are derivative
notions. A brief glance at the historical origins of these principles
leaves no doubt that the trinity (p. 17 5) of logical laws was not
always regarded as a universal regulator of the rational or onto--
logical domains.
Within the framework of the finite and objective Greek cul-
ture, laws of thought were merely generalized commonsense prin-
ciples. For Aristotle, as the first systemologist, logical laws were
axioms or the most common principles'of demonstration.13 It is
Significant that Aristotle discussed the laws not in his logical books,
but in the Metaphysics, where we find six chapters (3-8) explicitly
dealing with what later came to be known as the Laws of Thought.
In the Analytica Priora, on the other hand, there occurs only a
fleeting passage (1, 32, 47a, 8). Undoubtedly, Aristotle is inter-
ested in things—their organization and arrangement—though, of
course, he puts the matter in the form of being as being. Unless
we read Aristotle through the spectacles of a later age we learn
that Metaphysics is the study of the most general aspects of things.
It is the function of the philosopher or metaphysician, as over
against the special scientist, to investigate all things.“ But such
Wallace, Logic, p. 223.
1°

Wittgenstein, Tractatus.
Hedrick, Tendencies.
Metaphysics, Gamma, chap. 3.
"Ibid., chaps. I, 2.
I80 PSYCHOLOGY AND LOGIC
l
investigation, in 1ts widest and most abstract form, is dialectical in
‘the Sense of abstract reference and description.
In consequence, Aristotle makes no sharp distinction between
investigating things and linguistically organizing them. From our
standpoint this is indeed a naive and commonsense View, but it is
nevertheless objective in keeping clOse to actual things. There is
no sophisticated reduction of things to ideas or symbols, as in
later times. When Aristotle discusses the laws as principles of
being he sees no necessity to distinguish them from principles of
thought. On the contrary, for him.axioms or principles are at
once, in' different aspects, principles of existence and of reflection,
as well as of demonstration or argument.
The practical and commonsense—we may even say the opera—
tional—4character of the Aristotelian principles is demonstrated
by Aristotle’s variance in emphasizing them. The principle of con-
tradiction he regards as the most certain: "
The same attribute cannot at the same time belong and not belong
to the same subject in the same respect.15
Somewhat subordinate is the principle of middle exclusion, formu-
lated at the beginning of Chapter 7.
There cannot be an intermediate between contradictories, but of
one subject we must either affirm or deny any one predicate.16
The identity principle is subordinated to both of these.17 Identity
Aristotle treats as the problem of the meaning, and singleness of
meaning, of words, especially the words being and not being.
First then this at least is obviously true, that the word “be” or
“not be” has a definite meaning.18
For not to have one meaning is to have no meaning.1° There is
no question that he13 concerned with references to actual proper—
ties or attributes of specific things.
Though Aristotle13, no doubt, primarily interested1n common—
sense objects, with ordinary discourse and practical demonstration,
as a philosopher he naturally attempts to attain generality, even
1“Ibid.., 1005b,18fi.
1°
Ibid.., 1011b, 24.
it
'7
Ibid., 1006a, 2955 and 101 1b, 263.
Ibid., 1006a, 31. ‘
∙
19Ibid., 1006b, 9; 1006b, 13.
 THE LAWS OF THOUGHT AND THINGS ' I81
absoluteness of formulation. Still, his formulation of abstract and
general principles has to do with concrete things in specific situa-
tions. Aristotle’s thought was always thought about things, not
a reaching out to transcendent and sophisticated ideas and entities
beyond the borders of spatiotemporal happenings (vol. I, chap. 3).
Not until the post—Aristotelian (psychistic) era of logic were
the practical principles made into regulatory laws for the conduct
of Reason. Only when logic became thoroughly formalized Could
the principles of identity, contradiction, and excluded middle be-
come universal and absolute laws of Reason which not only legis—
lated for things, but also subordinated. them to thought. This
formalization involved a different type of psychology from that
of Aristotle’s. Psychological events were no longer regarded as
functions of biological organisms, but as psychic or spiritual powers
independent of the organism and the things with which it inter—
acted. Since Plotinus, who was the essential transformer of Aris—
totelian Obj ective psychology, mind as Reason or Logos has been
considered not only as self—creative, but also as the source of all
things. '

Because many of the Plotinian doctrines revolve around the

reinterpretation of Logos doctrine, neo-Platonism becomes great-
ly preoccupied with words and symbols. The stage is thus set for
the commentating and interpreting Church Fathers—in short,
for the general precipitation of a gigantic verbological industry.
Verbal symbolsrather than actual things become the materials for
all sorts of theories concerning events. Thought and logic are so
far removed from things as to attain a self—enclosed domain. Logic
not only becomes primarily concerned with names, but the laws
of logic are confined to rules for their manipulation. Here is the
basis for the doctrine of first and second intensions. Logic deals
not with notions of things, but with notions of things as thought,
or with names of the clasSeS of things.
To trace out the thorough formalization—that is, the verbaliza—
tion—of the laws of thought is to encounter the abstruse theologi-
cal arguments of the Middle Ages._‘The laws of identity are em-
phasized to the point of establishing the permanence and un-
changeability of an eternal ground for all things.
The growth of science and technology in the Renaissance period
soon ‘called for a logical reformation. As far as our present problem
1 82 PSYCHOLOGY AND LOGIC

is concerned, logic was no longer confined exclusively to purely

formal principles, and, happily, the question of the application
of laws began once more to concern things in a substantial way.
FrOm this point on we have a genuine differentiation between
types of logic, for example, the logical concern with mathematics,
the new methodology of science, and also the continuation of
formal logic in the sense Of propositional and classical logic.
Significant for this period is the formulation of‘the so-called
fourth law of logic—the principle of Suflicient Reason. Though
this law in its various formulations by Spinoza20 and Leibniz21 is
rooted in theological and general cosmomonistic ground, in Leib-
niz especially it at least indicates that logic is concerned with con-
tingency and fact as over against pure 4 priority and formality of
Reason. Further, its very presence in the speculations of Galileo22
marks it as a link between the pure and remote propositions of the
Scholastics and the scientific occupations of the Scientists and tech—
nologists of the early period of our Own era.
Since the medieval logicians took over and transformed the
rules of argument and demonstration formulated by the Greeks,
these rules have been universalized and instituted as
regulators of thought, despite many disagreements concerning
or canons
their number, nature, and origin. During the 19th century the
notion developed that logical laws are more or less closely con-
nected with things. This view was increasingly fostered when
logic was overtly related to mathematical processes. The incom-
parable illustration is Boole’s notion that logic is a mathematical
procedure (p. 187), and further that the basic purpose of logic in
mathematics is to indicate the fundamental laws of the mind’s
operation when reasoning is performed.23'Were it not for the
Obsession that thought is a psychic process or an exercise of mind,
it might have been clear that the laws of thought are really
criteria of principles designed for organizing systems. In the case
of Boole, and those who agreed with him, the type of systems
capable of regulation by the laws of thought were the abstract
systems of mathematics. For organizing systems of things other
2°
Ethics, I, I 1, proof 2. 7‘

Monadolog'y, secs. 3 I, 32, 33, 36, and in other works..

Enriques, Historic, p. 55.
∙
’2

∙ Vol. I, p. 113. ∙
 THE LAWS OF THOUGHT AND THINGS 183
than mathematical abstractions more resort must be had to the
type of objects dealt with. Laws of logic in the latter case are not
so much concerned with abstract organizations of formal sentences,
and symbols as With specific things.
Current treatments '01? logical laws correspond more or less to
the historical divisions of philosophy—namely, Nominalism, Real-
ism, and Conceptualism (chap. 18, p. I32f., vol. I, p. 37f.). Ac-
cordingly, while all logicians associate logical laws with linguistic
materials on the ground that logic must be formalized, they may
stress (I) linguistic structure, (2) relations, or (3) classes.
(1) Those who emphasize linguistic forms may be identified
with the nominalistic tradition. For them logic ultimately reduces
to organizations of symbols, or to systems of tautologies. Such
logicians, when concerned with mathematical or symbolical materi-
als, become modernized into syntacticists. Logical laws for them
govern the relationship between formal sentences.
(2) Logicians emphasizing relations link themselves with the
realistic tradition in philosophy. They deal with formalized state-
ments concerning things represented by the laws. Thus, logical
laws take on the character of semantic rules. The more scientific
logicians Of this type—the modern extreme ontologists—regard
the things ordered by the laws as invariable relations of existence.
(3) Our third group, the logical epistemologists, emphasize
the formal aspects of knowledge. They want logical'laws to inte—
grate statementsOr formulae with data or relations. Logical

laws, as Well as logical sentences in general, become concep‘tualistic.

Logic on this basis is a discipline for ordering data to thought.
These logicians, who may be called modern Aristotelians,“ are,
on the one hand, merely concerned with commonsense Obj ects. On-
the Other, because they are formalists they are occupied simply
with the truth or falsity of sentences. Hence the stress on classes;
for the most part, classes confined to the true and the false.

UNIVERSALITY AND N ECESSITY OF LOGICAL LAWS

Without exception writers on logical laws take for granted that
they are universal and necessary, even
though they can not claim
a further understanding of their nature. They accept without ques—
\

For\ example, Joseph, Introduction.

3‘
184. ' PSYCHOLOGY AND LOGIC

tion the belief that logic is a universally normative discipline, and

assume that all thinking and reasoning exemplify such laws.25
Naturally, the experts on logical laws have attempted to sup-
port their belief With'as powerful arguments as they could muster.
But the very difierences in argument demonstrate the absence of
any necessary principles.
As We know, the older logicians sought a basis ,for logical laws
in the nature of mind or of things. More recently, with the im-
pact of mathematical theories on logical thought in conjunction
with changing ideas of logic and mind, the source of laWs has be—
come relatively more localized in mathematical operations. Now
in View of the differences in theory concerning (I) the nature and
num‘ber'of'logical laws, (2) their basis or origin, and (3) the
range of their application, it is hardly possible to regard them as
absolute, pervasive, and necessary. Let us look into the procedures
employed for retaining the notion of universality.
(1) Nature and Nnmaer of Logical Principles. If there were
universal and necessary principles of logic they would demand a
single, exclusive description. We find, hoWever, that depending
on the writer’s philosophical persuasion, the formula ranges from
the classic proposition that logical laws are regulatory principles
for all reasoning to the modern view that logical laWs express con-
ditions to be satisfied.""3 We must agree either that only one view
is correct or that logical laws are really not unique pervasive prin—
ciples. Indeed, writers sometimes do not limit logical principles
to the usual three, but make identity, contradiction, and excluded
middle coordinate with the laws of syllogism, tautology, simplifi-
cation, etc.27 Undoubtedly this universalistic theory originates in
verbal creation, that is, through the development of a general
formula the ideaIS engendered that the things referred to \gen-
eral or universal, whatever the particular formula may signify.
(2) Source of Logical Laws. Logicians willing to Surrender
the notion that logical laws have their source in the nature of mind
or in the character Of things stress necessity more than universality.
The self-evident and a priori are concerned with formal criteria of
the genuine, normative, and ,valId The criteria may be derived
from a postulational procedure. l
2°
Avey, Law.
Dewey, Logic,'p. 343E.
Cf. Cohen and Nagel,,Introduction, p. 182. ". I
 THE LAWS OF THOUGHT AND THINGS 185'
AS far back as Mill’s attack on Hamilton’s philosophy it is sug—
gested that the classical laws of thought, although universal, are
postulates of reasoning.213 This postulational view is followed by
Keynes29 who regards the laws as indispensable for consecutive
thought and coherent argument. Schiller, the arch opponent of
formal logic, likewise adopts this conception with the added notion
that, once adopted, such. postulates are absolute.30 Though these
conceptions depart from the traditional notion that logical laws
regulate the operation of a reasoning faculty, can we, even in this
modified form, adopt such a notion of generalized and authorita—
tive postulates? The same question may be asked about Mill’s
conception of logical laws as generalizations from experience. In
all these forms thought or reasoning is reduced, to generalized
processes not easily related to concrete human activities. The basic
assumption is that logic is concerned with abstract verbalistic or
formal systems comprising terms, judgments, and objects of
thought. These abstract systems are then expanded into almonopo—
listic universalism.
Whether logicians adhere to the three classical laws or add vari—
ous modern principles the assumed necessities are founded on ab-
stracting and formalizing procedures. By emptying sentences of
actual referential and symbolic functions they are made to take on
a necessary character. By reducing a situation to a pair of sheer
alternatives it is easy to demand that only one be chosen. But the
necessity here lies clearly in the specific situation, not in any gen-
eral law.
(3) Universality and Necessity of Application. An implied
argument for the universality and necessity of logical principles
is their presumed universality of application. Whether such prin—
ciples are limited to the classic laws of identity, contradiction, and
excluded middle, or broadened to include the law of sufficient
reason, syllogistic laws, rules of inference, etc., the argument is
supported by reducing things to empty forms. By thus reducing
everything to being, substance, quality, essence, to A or B, the
principles are made to cover everything. One might conclude that
everything is nothing. Certainly, there is a shift here from the
23 Examination, II, p. 179.
29 Studies, p. 45of.
3°
Formal, chap. Io.
I86 PSYCHOLOGY_ AND LOGIC

things to which the laws apply to the forms of statements or lin-

guistic systems. '

SPECIFICITY AND CONTINGENCY OF LOGICAL LAws

From the standpoint of specific logical situations it is necessary
to reject every form of absolutism and universality. So far as the
origin .of logical laws is concerned, any principle that may be
legitimately regarded as logical has arisen from some concrete-
situation. Hence there must be a variety of principles. The applica-
bility of logical laws is similarly determined on the basis of the
situations to which they may be adapted.
If conventional logical laws belong to the field of logic at all,
they pertain to organizational or system—building enterprises. Em-
ployed in such situations they concern the organization of things,
words, sentences, and relations. °
As with every other item in the logical domain, the cardinal
criterion with respect to laws is the behavior of logicians. Not the
least important feature is their motivation. Do they seek the self—
evident and a priori? Are logical laws for them the guarantors of
some absolute necessity, the manifestation of universal mental
powers or invariances in things? Or, quite otherwise, are they
actually constructing a definite system, and developing particular
principles on the basis of the specifications and particular materials
of that system?

LOGICAL LAws IN REASONING AND THINKING

Outstanding in the history of logic is the opposition between
those who favor either mind or world order. The former insist
that the conventional laws of thought refer to the capacities, limita-
tions, and necessities of mind. For such logicians the laws of
thought constitute either descriptions of, or prescriptions forl, the
mental operations of reasoning. On the whole, it is impliedthat
thinking and reasoning constitute generalized powers or processes
subject to formal regulating principles.‘
_ Thinking and reasoning, on such a basis, become empty and
i
trivial occupations, rather than modesxof interbehavior involving
particular acts and things.31 On an interbehavioral basis theronly
formalized laws are those. that caution us “against general rules
See vol. I, chap. 7; also Kantor, Principles, vol. I, chaps. {6-2.2.
'31
 THE LAWS OF THOUGHT AND THINGS
187
or regulatory principles. If thinking is to be significant it has to
be adapted to the situation at hand. This means that the laws of
thinking are laws of the problems or conditions about which we
think. ∙
It is extremely important to differentiate between thinking, in
its various forms, and reasoning. On the whole it is the latter, with
its basic inferential processes, that has interested logicians. Now,
inferential behavior can be performed in specific situations with
actual things to be inferred, or in generalized circumstances with
empty forms. It is only to the latter that conventional and formal-
ized rules apply. .

Consistency of thinking, or the application of general and per-

manent principles, can only be interpreted on the basis of personal
habits or cultural conditions. Such descriptive generalizations as,
for example, that we always think or reason by analogy clearly
are imposed upon, and not derived from, situations. Whenever we
follow generalized prescriptions, particular operations are robbed
of their essential character as thinking or reasoning. Indeed, there
is no way of escaping the essential demands of the specific situa-
tion. Our‘ objection, therefore, to the universalistic conception is
chiefly against the fundamental assumption that any prefabricated
principle can be valid in general, as we proceed from one par—
ticular situation to another.
Again, the number of laws required depends upon the situation.
For example, in some situations only consistency may be necessary.
Of the classical laws one, two or three may be needed. When more
,than one general principle is used, theymay or may not be in—
timately related.
Conventional treatments of reasoning and thinking laws really
come down to formalized system making. Though Boole regarded
himself as a discoverer of the fundamental laws of thought,32 he
was, of course, concerned with the construction of systems based
on the procedures. of algebra. As an absolutist and mystic Boole
believed it possible to limit logic to the exclusive and abstract para-
digms of mathematical systems. When, however, all sorts of
system—building procedures are taken into account, the principles
of construction can not be restricted to any variation of the classic
laws of thought, nor to any fixed number and type.
'2
Laws.
I88 PSYCHOLOGY AND LOGIC

At this point. it is appropriate to consider the relation of logical

laws to objects of thought. Is it acceptable to consider logical laws
as especially concerned with the organization of thought objects?
A prompt reply involves the counter—question: What is the nature
of “thought objects? Are they things thought about, or are they
constructions about things in the form of verbal or other lin-
guistic responses? In the latter and more correct case, we are occu—
pied'with descriptions or definitions. In no instance, hOwever, need
we regard thought objects as so related to formal thinking pro-
cedures as to require the application of classical laws of thought.

LOGICAL LAws AND EXISTENCE

Absolutistic logicians with a flair for Platonic ultimates have
interpreted logical laws as laws of things or, at least,.of essences.
Unlike the tautologists, the realistic or ontological writers insist
that inthe final analysis logical laws must be inherently applicable
to existence “because they are concerned with ontological traits
of utmost generality??? We have already quoted an ontologist’s
profession that logical laws represent factors invariant in every
subject matter.“ In other words, they are principles of being, not
merely principles of inference.
What are the ontologist’s motives in constructing logical laws?
In the first place, he aims to achieve wholesale rigorous formulae
for events and relations—in short, to legislate for science. His basic
method is to begin with scientific propositions, and then to extra-
polate and formalize to the point of reaching ultimate and in-
variant existential relations. _
The argument used to justify applying such formalized prin-
ciples to actual things is that logic constitutes the study of the
most general and pervasive character of whatever is and whatever
may be.35 This argument is obj ectionably circular. Only by reduc-
ing events to general characters can one equate them with formal
logical processes and propositions.
Laudable as are the motives to bring logic into
contact
science, we nevertheless ask the following questions: First, does
with
not logic in confining itself to such extrapolations and
formailiza-
∙∙
anCohen and Nagel, Introduction, p. v.
:Vol. I,p. I62.
asCohen and Nagel, Introduction, p. I 85. .
 THE LAWSIOF THOUGHT AND THINGS
189
tions renounce the benefits derived from its connection with sci—
ence? Secondly, do not the very processes of extrapolation remove
science too far from actual events? '

True, science thrives on formal descriptions. The scientist’s in—-

terest, however, is not in the formality, but in the scope such
formulae aflord
for interacting with the events which are the
grounds for the system constructed. Sheer abstraction1n science
helps not at all"In analyzing problems,'In formulating hypotheses,
in carrying on experimentation, and in verifying-theories. On the
contrary, abstractionism belongs to the domain of argument, in
which, for example, no thief can be proved to be a religious man,
since to be areligious man is tantamount to not stealing.
There remains, then, the contention that by means of logical
principles an abstractive systemization of science can be efiected.
But is it possible to achieve a unity of science through abstraCt
formulae? Here we must differentiate sharply between verbal
assertion and operational achievement, a difierence excellently
illustrated by the following incident. At the head of his treatise,
On the Equilibria of Heterogeneous Substances, Gibbs sets the
following propositions from Clausius:
Die Energie der Welt ist konstant.
Die Entropie der Welt strebt einem Maximum zu.
In the body of his work, however, Gibbs abstains from all refer-'
ence to the world, and, as is inevitable and proper, works out his
laws on the basis of specific isolated physiochemical systems.
Interesting in this connection is Bridgman’s protest against
3°

attempting to apply the second law of thermodynamics to the uni-

verse. In general, he objects to the extrapolational use of a sta-
tistical model for reaching into remote epochs of time and remote
reaches of space. Would it not be even more fatal, to deal with a
universe if extrapolating. sentences are replaced by completely self-
constructed propositions?
We are now prepared to consider the general problem of uni—
versality or unity in science. It ispossible, of course, to regard
science as unitary, as essentially systemic, since there is a con:
tinuity in natural events, or at least» in our operations upon, and
our knowledge concerning, them. But this Work of systemizing
\

a"
Statistical.
190 PSYCHOLOGY AND LOGIC

all of science at once can only proceed to the point of specifying

the Upper and lower boundaries of a comprehensive scientific sys—
tem. The logical or systemological laws to be invoked here are
exclusively concerned with these boundaries.
What may be designated as the upper boundary is determined
by the rule that the scientific field of operations has a spatio-
temporal frame of reference beyond which it is impossible for any
scientist to go. Whenever propositions transcend these spatiotem-
poral limits, we may be certain__that someone is indulging in fan-
tastic verbal construction.
The corresponding rule applicable to the lower boundary of a
scientific system stresses the limits of specific happenings. How-
ever well justified the insistence upon a continuum in events, and
in our scientific operations upon them, va‘lid knowledge consists
of propositions concerning the specific character and mode of oc-
currence of obj ects and processes. This fact places distinct limita-
tions upon carrying over conclusions reached in one particular
scientific domain to other fields. It is simply futile to reduce one
science to another, or to overlook the differences between the ob-
served factors in the physical, biological, and psychological sci—
ences. These scientific specificities apply not only. to. particular
techniques, methods, and apparatus, but also to conclusions. What
peculiar form of scientific uneasiness lies at the basis of our re—
pugnance toward specificity and particularity?

LOGICAL LAws AND DATA

Even the most formalistic logicians do not really wish to con—
fine logic strictly to formal structures. No, they insist that logical
forms reach out toward something more significant,‘if not to actual
existence. It is therefore best to regard our triplicate of doctrines
(p. 183) as varying in degree. If the extreme tautologists stress
sentences and sentence systems it is because they tie up most
closely with mathematical materials, which, if only at long re—
move, are related to things. The realists connect with ultimate re-
lations linking together mathematical statements and concrete
events. Our third group, the epistemologists, wishing to artibulate
logic still closer with actual things, relates logic to knowledge
problems, at least to the extent of connecting, logicallaws with
 THE LAWS OF THOUGHT AND THINGS 19I

processes of verifiability. It is assumed, then, that the truth values

of statements are interrelated with data.
In conventional logic this connection is assertional only. It re-
sults more in formalizing knowledge than concretizing statements.
The general technique is to begin with a formula—say, S is P.
The assumption is made that this formula has a definite referent,
and the questibn is whether the sentence is true—in other words,
whether S really has the property P or not. Apparently, this
sentence connects directly with certain definite things; thus its
truth or falsity can be determined—namely, verified.
Such verification, of course, is logical—that is, formalistic. There
is no checking of assertion or of formal statement with actual
things. Truth and falsity signify purely dichotomous classifica-
tion. Thus the conventional abstract laws of contradiction and ex-
cluded middle are presumed to underlie all logical truth and fal-
sity. An interesting illustration is to Show that the logical constants
or relations such as implication, disjunction, and conjunction can
be defined in terms of truth and falsity, but truth and falsity are
made to depend upon the logical laws.
As McGill points out:
Implication can be defined as holding for every pair of propositions,
p q, except when p is true and q false, disjunction, when they are not
both false, while conjunction only holds when p and q are both true.
With the legical constants defined in terms of true and false possi-
bilities, we can test the truth of all of the elementary propositions.37
When the truth tables are originally constructed, the laws of con-
tradiction and excluded middle prevent sentences from having
both true and false values.
Epistemological logicians do not overlook, however, the vacu—
ous nature of logical laws, and, indeed, try to overcome this fault.
The laws, they argue, are objective because we must not confuse
the significance of a proposition with the means of knowing it. For
example, if one knows that S is either P or not P it does not follow
that S is not P. Otherwise put, 'to \be unable to know the truth
of a proposition is not the same thing as the proposition being
not true. This argument about the object or datum as a basis for
\

u - p. 200.
Concernmg, '
192 PSYCHOLOGY AND LOGIC

theverification of a proposition is reduced to bare alternatives in

order to 'support the law of,excluded middle. In other words, it is
argued that whether or not you can verify the proposition it is
either true or-false,33 and hence the law of excluded middle is
established. _
That logical laws are concerned with data or events is a matter
of assumption only. Epistemological logicians assume that when
they assert S is P the symbols represent something definite. This,
in spite of the fact that, as we'have just seen, the simple-alterna-
tive argument precludes us from even attempting to discover what
the original event is. The statement, therefore, remains a simple
assertion 5 and despite the belief in its objective and referential
character it is merely a sentence. Secondly, it is implied that once
the sentence is formulated there must be some actual referent. This
is the familiar smoke—fire fallacy, which presupposes such a tre-
mendous faith in texts that the existence of the sentence is pre-
sumed to guarantee an existential referent.
Epistemological and semantic logicians have recently developed
a general technique for supporting their faith1n formal and au-
tonomous statements by stressing the contrast between verifiability
and verification. The basis of this technique13 the eageracceptance
of the antimetaphysical proposition that unverifiable statements are
meaningless. This proposition signalizes the simple recognition
that verification must deal with contents or things in order to yield
genuine knowledge. In other words, the truth of a statement must
'involve, in addition to simple coherence of structure, some definite
relationship to data.
But the'logician’s absolutistic and formalistic attitudes remain
paramount. To begin with, logical statements are traditionally not
designed to operate directly with actual events. To be formal
'and universal they must tend toward the simple and tautological.
On the other hand, most knowledge propositions can not bel abso—
lutely verified. To be occupied with actual events means that many
lacunae in the conditions necessary for verification are often' pres-
ent. Logicians seek a way out by enlarging the concept of verifica-
tion to include indirect as well as direct, and incomplete as well
as decisive, verification.“9 Thus13 constructed the notion of veri—.
as Cf. Toms, Law.
See, for example, Lewis, Logical.
 THE LAWS OF THOUGHT AND THINGS I93
fiability. This notion is similar to that of prObable confirmation“o
which is centered in empirical knowledge and inductive logic.
To illustrate their principle of verifiability logicians naturally
turn to number. For example, it is asserted that though there is
no whole number which can not be counted, the counting of whole
numbers can not be completed.41 Is this other than a case of
formalizing a sentence instead of concretizing it by the verification
technique? We ask: In what sense is number a thing which is not
counted? In what sense do we expect to complete the counting of
whole numbers? Notice that given numbers are products of a
counting process, and that naturally we can set up limits to the
counting procedure. Such argument is not only remote from the
application of logical laws to data, it is another instance of employ—
ing an arbitrary conception of language Or propositions.
Epistemological formulations of logical laws therefore aim
primarily at establishing truth without evidence. A typical ex—
ample is the employment of the law of excluded middle to estab-
lish:
. . . the logical theory of truth, involving the possibility of events
that no one experiences and of pr0positions that are true although
there can never be any evidence in their favour.‘2

The technique is to invoke the great power of disjunction, the

abstract either-or. We are brought headlong to a forked road,
forced to go either one way or the other. If we reject one way, it
is declared we have accepted the other. If one is false, the other
must be true, if one does not exist, then the other exists. The tech-
nique is presumed to become completely compelling when the
“two forks of the road” are allegedly connected with concrete
events by analogical description, as in the case of Aristotle’s sea—
fight example.43 Even Dewey agrees that while this necessity does
not apply to existence it does apply to the obj ect of thought. He
believes the disjunction is necessary. This is certainly questionable.
Actually, the necessity lies simply in the form of the sentence, but
patterns of sentences must be related either to events or reactions
4°
Reichenbach, Experience.
Lewis, Logical.
”Russell, Inquiry, p. 383. _
Vol. I, p. 160.
194. PSYCHOLOGY AND LOGIC

to events. Where is the necessity for a sea fight to occur alto-

gether? Why the question or assertion?
If, in order to point out that the disjunction is potent, this
alternative is set up after an event, our comment would be that we
are no longer concerned with formal sentences but with a simple
statement of facts. What point is there in translating occurrences
into, statements of logical necessity?

LOGICAL LAWS AND LANGUAGE

Language plays a double part in relation to logical laws. First,
of course, it serves as a means to formulate and organize logical
principles by way of symbolizing the systematic features referred
to, secondly, language in the form of sentences or‘linguistic forms
constitutes a field for applying the formulae:
Language serves admirably to absolutize things (vol. I, p. 41).
Because of the willingness to accept finalistic assertions there'IS un-
limited scope for establishing what one wishes by declaration and
proclamation. Who can dispute: “Whatever is, is,” “N 0 thing
can both be and not be,” “NO thing can escape either being or not
being.” Language is so pliable a medium that one can do remark-
able things with it, even use it to achieve. universality, ultimacy,
and a priority. The procedures vary. Logicians may make language
identical with an invariant correlate of, or an absolute substitute
for, things.
How language can be used to establish the a priori is Well illus—
trated by Kant’s separation of dichotomy from polytomy.“ Only
the former involves primitive and a priori division, since the mem—
bers are set over against each other such that the opposite of A is
non-A. The real basis for the division is the exclusive verbal-han—
dling of the alleged members. Ostensibly things or their prOper—
ties are referred to. However, the principle of contradiction is op-'
erative here because no actual things aside from the words are in—
volved, unless we assume that words sUbstitute for things. The
principle of noncontradiction can not apply to polytomy because
that type of division involves perceptipn of objects.“
The writers referred to above are the ontologists.
Thosei‘who
“Jiische, Kant’s Logik, p. 161.
“For Kant polytomy in mathematics, as in dividing the conics, can involve
a priori intuition. .' '
 THE LAWS OF THOUGHT AND THINGS I95
retreat from this position, but who still hold that logic is the au-
thority and legislator directly for thought and indirectly for things,
make language the basis for necessary knowledge and truth. Such
formalists hold that logic is primarily, if not exclusively, concerned
with language or symbols. The final goal of logic is a series of
self-contained sentences. Formalistic logicians, then, are concerned
only with ultimate linguistic products of description and mathe-
matical formulation, that is, with tautologies.
Here is an erroneous conception of the nature of language and
symbols, Descriptions and formulations are abstracted from their
setting in interbehavioralsituations. For example, accurate descrip—
tions are properly difierentiated fr'om inaccurate references to
events, and then the former are given an autonomous place in
the domain of knowledge. In this connection, a mathematical
formula is substituted for, and made into a law of, events. The
ultimate result is to separate the truth of statements from the
reality of events in order to attain a domain of forms as against a
domain of contents. It is as though one were to study scientific
protocols and records in order to check the syntax of sentences,
rather than the correspondence of the protocols, with observed
facts. ~
Logical laws for linguistic logicians consist of rules for regulat-
ing the abstract properties of symbol systems. Logic is thus re-
garded as a relatively simple and effective instrument for the easy
and permanent accession of ultimate reality and absolute truth.
Since Leibniz logicians have hoped to deveIOp a strict language
for the precise ordering of all knowledge. Leibniz’s ambition Was
to evolve a universal formal language to serve as the commodious
capsule for storing all knowledge. Actually, this is a 17th century
resuscitation of the Pythagorean notion that number is the essence
of things, and that everything can be reduced to numerical sym-
bols. Especially inept, however, is the assumption that we can
build up linguistic structures which unify all scientific findings.
Our observations and experiments convince us that we can not
even sum up the nature of a single~~event, such as. light, within the
bounds of one type of equation or verbal proposition. We have
done with such views as those of Jevons: “Science arises from the
discoyery of Identity amidst Diversity.”46 Maxims and proverbs
“ Principles, p. I.
196' PSYCHOLOGY ‘AND LOGIC

may be intereSting, even instructive, but often have nothing to do

with the substance of science or logic.
From a study of sentential logic we may well conclude that
not only can linguistic systems be organized, but that there may
also be value in doing so. It is definitely implied, however, that
the systems built are specific and Compresent with many others.
Whoever wishes to construct systems of sentences Can discover
System-building laws to further this work. Such laws must difier-
entiate between referential and symbol language, as well as be—
tween language which constitutes autonomous things and language
which describes and represents things.
There are logical laws or principles forithe organization of sen-
tences without‘any regard to existences. Similarly, there are logical
laws for structuring sentences. concerning things. In the latter case,
further specifications may be made as to the existence and nature
of things with respect to particular times, places, and conditions.
Instead of assuming that language consists of sentences, which
in their elements or totals “express” thought or things, we need
to establish a critique of language (p. 194.) within the large matrix
of actual personal behavior. Such a critique, concerned both with
systems of sentences as things and as acts referring to. things, can
be useful in guiding scientists in the right direction. The proper
use of sentential formulation or precise reference, however, does
not compete with, but is only ancillary to, observation and ex-
perimentation.

LOGICAL LAWS AND MATHEMATICS

Throughout the history of logic the source of necessity and cer-
tainty lay close to numbers. But not until the 19th century did
logicians conclude that logical and mathematical laws are identical,
and theIewith attempt to make logical laws independent of thought
and thinking. The identifying process mentioned was carried on
under medieval realistic auspices. Logical laws as ultimately
mathematical were presumed to deal with what must absolutely
be, because transhuman reality15 abSolute.
Mathematics, too, has undergone its evolution. It has likewise
been unable to escape its human origin, its\ specific operatiOns in
a particular behavior domain. Let us note a few high spots in its
transformation. First, there was the freeing ofgeometry from the
 THE LAWS OF THOUGHT AND 197
THINGS
absolutistic trammels of Euclidean assumptions. Out of the de-
velopment of non-Euclidean geometry may be traced the germs
of the postulational principles of mathematics. Then came Hamil—
ton’s and Grassmann’s discovery that algebra, too, need not be
eternally bound by the commutative law. Mathematics thus be—
came reasonably free to build its own unique and specialized
systems.
∙
The emancipation of mathematics soon exerted an effect upon
logic. The discovery that mathematics could renounce the numer—
ology of Pythagoras, and that by investigating groups, point sets,
and mathematical fields one cOuld move on to the study of rela—
tion systems and their structures,“ was exploited by the logisticians
in the interest of a Platonic absolutism. Mathematics became
formalistic and transhuman logic. Of the formalist and logistic
mathematician no question may be asked concerning the essen—
tiality and validity of classical logical laws. Because the mathe—
matician is intimately concerned with abstract things—with simple
and complex relations—he may regard himself as bound by a logic
of simple alternatives—yes or no, true or false, existence or non—
existence—without regard to the specific details of the systems in—
volved. Is it absolutely essential, as Russell and other logistic
writers assert, that P V P’ is always the case?48 Does x or y ex—
clusively exist; or is it possible to have a three or multi-valued
logic?
Brouwer stands out as the most vigorous recent assailant of the
law of excluded middle (LEM). This attack has been made in
connection with the intuitionist’s assault upon the logistic and
formalistic schools of mathematics. From our standpoint, however,
we regard Brouwer’s position as operational or constructible, as
recent writers have indicated. His point is that in no mathematical
Situation is it proper to assert that a certain number exists, when
no technique of operation is available for exhibiting this number
in a finite number of steps. In point here is Brouwer’s famous
question: Does a number K exist expressing the number of digits
in the decimal representation of It at which, for the first time, the
series 0123456789 begins? Brouwer. insists that in the absence of
calculative operations one can neither answer yes or no exclusively.
‘7
Bell, Development, p. I77.
‘8
Cf. Russell, Reply, p. 682f.; Inquiry, chaps. 20, 21.
198 PSYCHOLOGY AND LOGIC

A similar problem is posed by Becker in the following form. Take

the series of numbers x −−− 2“
+
I for increasing values of n, and
define another series k such that if x is prime the number I is as—
signed to k, if composite the number 2. The series k, then, has
the appearance 1, I, 2, I, 2.-Now, if we ask whether the number
I is contained in k after the 16th place, we can neither, according
to Becker, answer “yes” or “no,” nor even “yes or no.” The
reason is that our knowledge does not extend beyond 11 = 16 for
primes of this form.‘9
If the universalized LEM breaks down in so abstract a domain
as.mathematics we not only can draw our own conclusions con-
cerning its place in other more concrete situations, but assess its
general value as well. The whole matter revolves around the in—
terpretation we make of logical laws. Those who insist upon their
absolute character maintain that the problems we have indicated do
not interfere with
logical principles. For example, one writer de-
clares:
Brouwer indeed is not denying the tertium non datur in the gen-
erally accepted interpretation of that logical principle, but rather
emphasizing that existence in mathematics is synonymous“ with con-
structibility.‘50
Other attempts to save the law are based on the interpretation
of truth. Church,51 for instance, declares that the truth of a propo-
sition isan indefinable, not depending upon a possible operation of
proving it. He does admit, however, that such an interpretation is
a matter of choice rather than necessity. Russell assumes an un-
compromising position against Brouwer’s view that truth should
be subject to verification. Russell’s View of verification is some
limited personal observation such as perceiving or remembering.”
As an adherent to an ancient psychophysical dualism he also re-
gards personal psychological processes as psychic. His argument,
on this basis, is well taken. There are unverifiable truths—that is,
events unknown to particular persons as internal states‘ of con-
sciousness. But to go on from this point to true propositionS'(sen—
4°
Mathematische, p. 67.
5"
Black, Nature, p. 196.
Hedrick, Tendencies, p. 339.
Inquiry, chaps. 20, 2.1.
 THE LAWS OF THOUGHT AND THINGS 199
tences) that can never be confronted with evidence in their favor
opens wide the door to obscurantism.
Truth interpretation Ofier's an interesting item for specificity
logic. It is appealing, of course, to believe that truth is independent
of actual demonstration, but this belief, though rational in the
world of cOncrete events, hardly has a place in mathematics. While
analogous demonstration of relations is legitimate in mathematics,
the analogical relation must be demonstrated. Not so in the case
of concrete events, in which instance analogous situations are more
easily observed and accepted.
It turns out that in any rigorous thinking we can not appeal to
traditions, however firmly entrenched. Certainly, we can not
universalize and hypostatize observed conditions into general prin—
ciples as a means of avoiding further observations or demonstra-
tions. We conclude, therefore, that truth can not be guaranteed by
resorting to such a universal law as excluded middle, but must be
determined upon the specific conditions dealt with. Nor can a gen-
eralized and arbitrary interpretatiOn of truth establish the LEM.
Of course, by reducing it to a single alternative it may be accept-
able, but while such a logical law is one of many\possible ones, it
does not on this assumptional basis constitute a factor in any uni-
versal logic.
Significant here is Lewis’ assertion that the LEM is not a uni-
versal law, but that it “reflects our stubborn adherence to the
simplest of all possible modes of division.”53 We go even further
than his negative proposition that “there are no laws of logic
which can be attributed to the universe or to human reason in the
traditional form.”“ We point out the specific and operational
character of logical principles in all system-building enterprises.55

LOGICAL LAWS As FACTORS 1N SYSTEM BUILDING

A summary of the difl'iculties of the traditional treatments of
logical laws includes the following assumptions: (a) there are
universal binding principles for the conduct of thought or for the
character of existence, (b) linguistic\or verbal formulae substitute
M
Alternative, p. 505.
5‘
Ibid., p. 483; see also Waismann, Alternative.‘
“VOl.I,p.2II.
200 PSYCHOLOGY. AND LOGIC

for'both thoughts and things, (c) thinking or reasoning consists

of mental acts of persons more or less remote from things, which
raises the question whether it is binding or not with respect to
those things, and (d) things are objects of thought either inde-
pendent of, different from, or conjoint with thought.
In contrast to traditional discussions are the following points:
(a) logical laws are constructions made by students of logic on the
basis of observing organizational activities, (b) thinking and reas—
oning are specified forms of interbehavior with stimulus objects,
which may be events, objects, acts of things and persons, or prod-
ucts of these acts—so—called imaginary things. In general discus-
sions of logical laws the concrete human character of logical events
is passed over. Writers on logic are slowly, if at all, becoming
cognizant of an objective and scientific psychology.
An analysis of the formulation and employment of logical laws
promptly leads to the conclusion that what are called logical laws
are not especially or exclusively connected with logic. Hence,
their employment in other situations is entirely legitimate. Many
are the instances in which logical laws serve as regulatory guides
to various kinds of intellectual work not describable as system
building. To stay by the original issue (identity), to avoid con-
fusion of one thing with another (contradiction), and to maintain
a point (excluded middle) are cardinal rules for significant in-
tellectual behavior.
Indeed, if logical laws are not treated as Platonic reals with an
autonomous and remote ontological existence, they stand out as
concrete functional principles, as criteria for putting things to-
gether. For example, numerically different things may be quali-
tatively identified, qualitatively difierent things, in whole or part,
may be numerically classified or put into a singleiclass.
Our aim is to place principles, whether or not rigidly formal-
ized, in their proper interbehavioral setting. Are we concerned
with ways of citing, equating, and identifying things or processes
that might or might not be similar? Are we setting up a criterion
for equating more or less similar things? Are we equating ex—
pressions or making x equal to y Or~.y2? Again, are we asking
whether ptomaine poison is like some other poison? Is there a
substance or chemical element in this case, as, in others? Is a virus
a living substance? Are vitamins capable of‘--identification with
 THE LAWS OF THOUGHT AND 'THINGS 201
Bios? Such questions illustrate our operations with constructions
based upon our contacts with things. We may also identify pure
constructions—that is, free creations sustaining no discernible
analogies with known things. In every instance the emphasis is
upon particular events, not upon cosmic schemes which escape our
control and depart, presumably, from the scene of specific human
affairs. ‘
In logical system building the classic laws of thought serve as
definite operational tools. When necessary or convenient the laws
of identity, contradiction, and excluded middle are employed as
limiting principles for organizing particular kinds of systems.
They may function as references to qualities of things or to the
existence of events, they may concern propositions or other sys-
tem—building materials.
In such enterprises the principle of identity is a scheme for
ordering and relating events for a certain purpose, putting them
into classes or groups which can then be manipulated. An excellent
example is the "development of class marks in statistics. Items are
regarded as practically the same or similar. Our educational
systems Show the presence of so many children in each of so many
grades, so many teachers, etc. For administrative purpOses a uni-
versity may list auditors, curators, librarians, and others among
its professors. ∙

The principle of contradiction, or consistency of organization, is

used for building systems of things, as well as of propositions. In
other words, if we apply a principle of action to a certain indi—
vidual we do not make an exception, if we are consistent, in the
case of another person. In general, the‘ principle supplies a rule
of consistent action.
What is conventionally called the LEM becomes in specific
system building a principle of alternativity. If x is put into a certain
class y, then it is excluded from class 2. Whenever the situation
calls for such alternativity the principle is fairly rigorously ap—
plied. But there are orders of alternativity, such that in some
cases a dichotomy is made between those situations in which x is
either y or z, or x is y and 2, as compared with other_systems in
which a different rule operates. In some instances x is not to be
connected with y or z at all, but with l and In.
Depending upon particular enterprises the rules of system
202 PSYCHOLOGY AND LOGIC

building may apply to things, manipulations, actions, judgments,

and propositional arrangement. All these enterprises must be kept
strictly apart in order to avoid confusion. In the case of identity
we may be interested exclusively in the sheer Verbal formula 11 is
a, and more generally in the arrangement of the words is and is
not, or in making things similar, as in counting, classifying, order-
ing or mathematically equating things. Again, we may be interested
in merely asserting that there is or ought to be constancy or cer-
tainty.

LOGICAL LAWS AND SYSTEMOLOGICAL BEHAVIOR

The difliculties which current logiciahS face with respect to
logical laws arise directly from an excess of knowledge and so-
phistication. Though dominated by the cultural heritage that logic
is the discipline of principles, of the a priori, the basic, the neces-
sary, the stable, and the ultimate, logicians discover, after all, that
these principles do not stand up in the face of scientific and mathe—
matical analysis. To their credit they are unwilling to cleave to
necessary systems which are remote from developments in mathe—
matics and science.
Glance at one of the many excellent examples showing the effect
of the assumption that logic is prior to, and regulateS,‘Other dis-
ciplines—namely, the problem whether logic is or is not onto-
logical. Logic is ontological, it is declared, and reveals the in-
variant traits and relations of things. Then, when a closer approach
to scientific investigations demonstrates that the Realistic identifica—
tion of a set of sentences or propositions with events in nature is
untenable, it is asserted that logic is free from ontology.56 Why
such leaps from ontological to nonontological logic except for the
assumption that logic regulates? The solution usually oflered for
the incompatibility of logical principles and the nature of things
is to relegate the governing powers of logic to sentences. But to
make logical laws into instructions for establishing necessary con-
nections between statements either reduces statements to innocu—
ous and empty things or continues the ontological procedure. All
this confusionlS avoided by holdingfirmly to the behavior sys—
tem builders. ItIs not suflicient to show that language does not
of
MCompare Nagel’s shift of view from Can logic to Logick.‘
 THE LAWS OF THOUGHT AND THINGS. 203
constitute things, or that words and terms should be precisely em—
ployed.
What is required is an analysis of the behavior enterprise under
investigation. Above all, it is necessary to observe that rules belong
to the job at hand, and that all rules have been derived from spe-
cific enterprises.
Significant here is Whitehead’s remark:
In formal logic, a contradiction is the signal of a defeat: but in the
evolution of real knowledge it marks the first step in progress toward
a victory.57
To maintain and revere any set of absolute principles, to attempt
to make all things and events match them is to behave reverentially,
instead of logically, to be concerned with religious, instead of with
logical, laws.
'7
Science, p. 267.
 CHAPTER XXI
PROBABILITY AND INTERBEHAVIORAL LOGIC

PROBABILITY: GUIDE TO LOGIC

LOGICIAN S as professional system builders inevitably turn to

probability study. Howsoever imbued they may be with the
ideal of certainty, with absolute systems, they are invariably
brought sharply to the dead end of abstract and formal structures.
Hence the unique place which the probability field oCcupies in
logic. Every alert approach to probabilities—that is, to actual
events—brings to the surface the necessity to organize or Struc—
ture things—in brief, system making. Even if the system maker
fails to appreciate that he is actually structuring problematic ob—
jects and conditions, or his reactions to them, his work can not
escape a system-building description. Depending upon his be—
havioral circumstances he produces system products by way of
(a) intellectually adjusting to elusive things, (b) ealculating
chances, (c) playing winning games, (d) predicting events, and
(e) performing manipulations efliectively.
There exists, to be sure, the paradox of probability. Though
inevitably dealing with contingencies, with partial and changing
facts, with complex and indeterminate events, logicians neverthe-
less like to set up univocal descriptions and fixed rules. This Situa-
tion is well illustrated by Bertrand’s remark in'beginning his
Calcul des ↑ ∂ ∂↑ ∫∙ “How dare we speak of the laws of
chance? Is not chance the antithesis of all law? ”1
This paradox plumbs the depths of logic. ThereIs a domain
'of chance. ItIs not only the ancients, as Poincaré says, who were
forced to face the uncertainties of prediction, the of es-
tablishing laws for recondite events. As he himself at once observes,
his definition of chance‘‘as only the measure of our ignorance”
will not do. For so-called fortuitous events are constantly predicted
and calculated. Moreover, the predietions and calculations appear
plausible, Since they resultIn the distribution of dividends. i“
And so the logician is obliged to enlarge\his traditional domain

Poincare, Foundations, p. 395.

 PROBABILITY AND INTERBEHAVIORAL LOGIC 205
to assimilate, if possible, those elements which resist his urge to
achieve totality and certainty. The systemological difliculties which
he could pass over while confining himself to formal linguistic
structures stand out Sharply when actual events obtrude them—
selves. The logician’s task, then, has become that of developing
a logic of induction to complement his venerable deductive logic,
a logic of errors and estimates to match his logic of truth and
reality.
Nothing, however, stands1n the way of thus enlarging the
scope of system building except the strictures imposed by the his—
torical philosophy of mind. It is only the unified mind of the
rationalist which afliliates with rigid realities; only the difiuse
subjectivism of the empiricist which devalues the continuity of
concrete events. Probability logic calls into play naturalistic psy—'
chological resources. Problems of chance and fortuity as objec—
tive fact, problems of belief, expectation, and prediction among
contingencies can only. be solved by a logical theory based on an
objective psychology.‘
For the interbehavioral logician the prObability problem pre—
sents no difficulty. All types of situations ofier occasion for sys-
tem construction. His horizon is broad enough to encompass the
most rigid and certain systems, as well as those remaining close
to problematic facts, to events which will not bear hammering
into the shape of empty abstractions.

THE DOMAIN OF PROBABILITY

Upon examining source material, students of probability logic
immediately discover that system makers approach probability
situations with presuppositions foreshadowing the pattern of their
systems. In other words, instead of producing system products on
the basis of their contacts with probability materials, they impose
upon the latter such binding specifications as to affect the system
product. The resulting description of the probability situation is
then erroneously presumed to be identical with the original situa-
tion.
Two factors influence this circumstance. In the first place, the
probability domain consists of a vast and heterogeneous series of
relative and unstable events. Consequently thereIs free play for
the workers’ presuppositions on the basis of their diverse back-
206 PSYCHOLOGY AND LOGIC

grounds and interests. In the secondlplace, the historical origins

of probability studies are rooted in the calculation of chances and
in the philosophical attempts to justify nonrational knowledge.
In order, therefore, to arrive at the fountainhead of probability
situations we shall examine some of the presuppositions blocking
the way. For convenience we separate them into two types: first,
the philosophical, stemming from a more general cultural back-
ground than immediate probability situations; secondly, the his—
torical presuppositions growing out of the accident that probability
studies were originally concerned with calculating chances.
Philosophical Presuppositiom. Logicians who so" frequently
quote Bishop Butler’s maxim. “To us probabilityIS the very guide
of life,”" do not point out that Butler13 making a grand conces-
sion to beings of limited capacities. For an “infinite intelligence,’
“nothing which is the possible Obj ect of knowledge, whether past,
present or future, can be probable.” In othe’i‘ words, probability
is not the guide of life because it is characteristic of actual living
conditions, but because men live only in a pale reflection of abso—
luteness and universality. Though logicians may regard them-
selVes as remote from theological situations, they fail tO‘ recognize
the similarity between theology and speculative cosmology. For
instance, they see no disharmony between working on the basis
of sound scientific method and accepting the belief that thereIs
a problem concerning the existence of an external world.
The point is that the prevailing presuppositions of universality
and absoluteness of probability systems have little to do with the
situations out of which probability problems arise and upon which
valid systems must be built. We have already mentioned the
tendency to regard probability systems as conforming to the cer—
tainty specifications of deductive structures. By reducing proba—
bility to situations satisfied by a single univocal system, it is as—
sumed that their respective systems are exclusive and absOlute.3
Actually, any probability formulation, even when most solidly
grounded, constitutes only one example of many possible types of
construction.‘ In other words, there .are many kinds of probability
situations, each demanding a specific sOrtuof investigation.
:Analogy, p. 73. .
Greenwood, in Runes’ Dictionary, describes seven ihterpretations of probability.
This statement holds for general mathematical or calculative probability, as well
as for any probability theory applied'1n practical statistics.
 PROBABILITY AND INTERBEHAVIORAL LOGIC 207
Both the neglect of the concrete situation and the formulation
of unsatisfactory probability theories can be traced directly to tra-
ditional ontological and epistemological presuppositions. Histori-
cal ontologism lies at the basis of the problem whether the world
is one of ultimate chance and contingency, or of irrevocable deter-
mination. Ontologies, coupled with rationalistic epistemology,
foster the- notion that probability is a form, or degree, of certainty,
while the sensationistic subjectivist inclines toward a view of hap—
hazardness that can only be corrected by calculating ignorance
and indifference.
Calculative Presuppositions. Because the first quantitative prob—
ability Studies were initiated by problems derived from calculat—
ing gambling chances, probability has become identified with sim—
plified things and events which can be treated by a calculus with
fixed and narrow limits. In earlier days probability systems cen—
tered around the symmetrical ratio I/ 2, and the definite possibility
of a die’s 1 / 6. When probability came to be applied to vital sta—
tistics, insurance problems, and scientific data, the calculation of
limits themselves became a problEm.

PROBABILITY EVENTS AND SITUATIONs

Assuming that we can clear away the presuppositions blocking
the path to probability events, we still face the problem of isolat—
ing them within their wide domain. We start with an interbehav-
ioral field. An individual is interbehaving with things, events, or
relations, but his intellectual orientation, his manipulatory action,
and attempts at prediction can not be rounded out and completed.
Either he has insufficient access to data or the data themselves are
neither established nor stable. In fact, this description applies to
every complex human situation. Keeping close to concrete hap—
penings, we discover that problematic situations are not uncommon,
nor are they unique, rather, they are instances of what happens
when we face a situation diflerent from one to which we habitually
and automatically adjust ourselves.
Several important implications follow. In the first place, we can
distinguish between (a) the scientifically useful admission that all
complex situations are problematic, or that all knowledge is less
than certain, and (b) the metaphysical presupposition that the ex-
istence‘of an external world is doubtful, problematic, or requires
208 PSYCHOLOGY AND LOGIC

proof. Writers on probability set up the prOposition that truth

is a form of probability, not, however, on the observation of human
behavior, but on the untenable ground of spiritistic epistemology
and fallacious ideas of the psychology of perception, knowledge,
and intellectual construction. _
A second implication is that we can welcome any, and all, fac-
tors of a given situation. From the interbehavioral standpoint a
probability situation involves the actions both of persons and of
the various problematic objects upon which they operate. One
factor can not be separated from the other. The history of prob—
ability theory is in part a record of writers who erroneously iden—
tify probability either with the stimulus events, which they some-
times reify in a Platonic manner, or with the reactions to those
events. The latter instance occurs when probability is defined as
the degree or quantity of knowledge, belief, or ignorance. But
whenever equations or propositions are set up “to symbolize some-
thing concerning the probability of events, the interbehavioral
process is immediately acknowledged.
In the third place, the interbehavioral approach implies that
stimulus objects within problematic situations are- objective. For
example, the problem may concern the adequacy of a sample, the
weight of a thing, the velocity of an object—in s'um,‘any of the
manifold conditions to which persons must respondIn their every-
day activities or in scientific laboratories.

PROBABILITY RANGE AND CONTINUITY

As genuine happenings, all probability situations constitute
points on a continuum ranging from ordinary life situations to the
A
abstruse entities of mathematical and conventional logic. similar
continuity pervades probability operations. This view, however,
in no sense implies that contingency always characterizes events.
We must not violate the principle of specificity. Our characteriza-
tion of a situation follows our contacts with it. There are stable
and relatively fixed things and events whose nature and operation
we need not doubt. Contingency here18 at a minimum.
Of considerableImportance is the enlargement of the probability
range to include Single events as a matter of principle. This is con-
trary to the opinions of mathematical writers. who approach prob-
ability investigation exclusively from the standpoint of series or
 PROBABILITY AND INTERBEHAVIORAL LOGIC 209
collections. We include in the probability range both past events
and those to occur. In the latter case anticipation and prediction
operations come into play. These operations imply that the task
of the investigator in contact with a \number of factors is to fore-
see the probable course of their interconnection. It goes without
saying that his operations involve a complement of knowledge
derived from previous observations, without which he could not
arrive at a prognosis. Can one deny, for instance, the probability
problem in situations complicated by extremely important condi-
tions, such as prediCting the outcome of a disease?
From the interbehavioral standpoint, it is futile to deny that
probability has to do with degree of belief. Belief consists. of a
form of interbehavior with particular conditions 3 it is an objec-
tive and definite situation in which we attempt to determine
whether a certain outcome will be x instead of y. Even in this
type of case, probability is'figured on the basis of the character
of events and the knowledge and expertness of the individual
interbehaving with such events. Thus, while the emphasis may be
on one or another factor, probability as a degree of belief is an inte-
gral part of all probability Situations.
Behaviorally speaking, there is no ultimate diflerence between
probability as degree of belief and as frequency. In the case of
frequency the difference in detail may be summed up as an ac-
cumulation of records, whereas in the case of belief there are not
so many data upon which to base a prediction. A fundamental dif-
ference is the amount of possible manipulation of data on the part
of the operator.
Adherence to specificity principles obviates the exclusion from
probability situations of relations between propositions. But here
we must differentiate between propositions taken as records of
analogous events and as isolated empty sentences or word com—
binations. In each case we construct a different kind of probability
system. In the second or word-combination case we have a system
based upon prior or primary constructions, whereas in the former
the constructions are founded on more direct cOntacts with things.
Sentential records of an indirect sort are often necessary in order
to operate efiectively in direct probability "Situations, though this
does not mean that the observer is not in some sense concerned
with objective events.
210 PSYCHOLOGY AND LOGIC

PROBABILITY: UNIVOCAL OR MULTIVOCAL,P

So powerful is the influence of,universalistic logical tradition
that, on the whole, probability is univocally defined. There are, to
be sure, suggestions" that the term prohahility refers to different
types of situations, but1n the end writers do not go much further
than a differentiation between prObability occurrences and proba-
bility judgments or estimates. The following observations illus-
trate the point.
Reichenbach.6 This writer adopts the uncompromising univocal
position that all probability situations sum up to a pattern of an
unknown infinite series of items which require precise determina-
tion or at least relative delimitation. While Reichenbach is will-
ing to consider a distinction between (a) mathematical and (b)
logical types of probability, his systematic constructions are de-
signed on a strictly mathematical basis.
Russell.7 Writing as though he were analyzing situations, but
really building on historical constructions, Russell distinguishes
between (a) mathematical probability, and (b) degree of credi-
bility. The former, he asserts, can be numerically measured, and
hence satisfy the axioms of the probability calculus.
Credibility
probability applies to single propositions, whether. or not based
on available relevant evidence. It is this sort of probability which
underlies the View that all knowledge is probable. What Russell
says, in effect, is that, whereas different probability interpretations
are inescapable, actually the important systems of propositions are
those of the probability calculus.
Nagel.8 Many writers on probability differentiate sharply be-
tween the probability calculus and various interpretations of it.
The calculus is taken as a neutral objective instrument for evaluat-
ing probability events. This fact does not, however, prevent cer—
tain writers from interpreting the calculus in a manner which
helps to assimilate it to their own ways of thinking. Nagel, for
example, differentiates between (a) series of recurring events
capable of treatment by a frequency calculus and (b) nonquantita-
tive happenings. In the latter
‘subdomain
fall single instances
°Hawkins, Existential, Kemble, Probability,
°Experience- Wahrscheinlichkeitslehre.
Ryle, Induction.
\
' Knowledge, part 5, chap. 1.
3
Principles,
 PROBABILITY AND INTERBEHAVIORAL LOGIC ZII

which require us to ascertain the adequacy of whatever evidence

is presented in their favor. AS a supporter of a frequency theory of
probability he then identifies his own interpretation of the cal—
culus with the quantifying instrument. Probability events do not
therefore ordinarily include single instances. But by making suit-
able verbal adjustments the frequency calculus can be made to
cover single inStances as well as series.
Carnap.9 Without differentiating between single-instance situa-
tions and those involving series, and by keeping in view future
action, Carnap distinguishes between two concepts of probability:
(a) frequency and (b) confirmation. The former covers relations
between events or series of events, whereas the latter concerns
relations between events and the evidence accepted for them. Prob-
abilities of the first type lead to propositional systems of the ana-
lytic and deductive sort, the latter involve sentential structures
of the synthetic and inductive order.
Haw/aim.10 This writer proposes to set off (a) epistemic from
(b)existential probability. Existential probability raises the ques-
tion whether events occur or whether we may speak only of the
truth of propositions aboutevents. The propositions concern physi-
cal relations—the metrical properties and causal relations of chance
mechanisms. Epistemic probability involves propositions about
weight of evidence and the reliability and correctness of beliefs
or inferences.
Kemble.11 In order to make room for what he calls an opera-
tional concept of probability this author distinguishes three prob-
ability concepts. The first is the Laplacian a priori type based on
a judgment of equal likelihood. Inductive probability is the sta-
tistically based calculation of past eXperiments and trials. Both of
these are primary. The third is the secondary, or theoretical, con—
cept of probability, consisting of a construct in a universe of con-
structs. Not completely unrelated to these is the operational con-
cept which describes probabilities as numbers computed from data
according to standardized rules. Kemble bases his conception on
the Machian notion that the external world is created or invented
as an interpretation of invariant patterns of sensation in the stream
Testability; Probability, Two Concepts.
Existential.
\
1°

11 Probability.
212 PSYCHOLOGY 'AND LOGIC

of consciousness. Since this view combines the mental and the

physical, it can reconcile frequencies with eXpectancies.
Ryle.12 Assuming that all probability concepts are concerned
with relations between “premises” and “conclusions,” Ryle dif-
ferentiates from each other (I) the plausibility of abstract theories,
(2) numerical odds, and (3) reliability of inductions. Only the last
two he regards as normally employed in science. Numerical odds
may be correlated with the Laplacian tradition, while inductive re-
liabilities belong to the statistical frequency area.
Polya.13 Anextremely interestingattitude is exhibited in Polya’s
coupling of probability and plausibility.gAs a mathematician this
writer is wholeheartedly committed to an exclusively objective
frequency interpretation. Still, he is sympathetic toward a subj ec-
tive or a degree—of—belief principle, for the latter, which belongs
to the domain /of heuristic reasoning, possesses—like deductive
reasoning—a certain pertinence in human affairs. The counterpart
of probability Polya calls plausibility. Although the difference is
great between plausibility and probability, a calculus can be for—
mulated for it. But unlike probability, which is measured by a de-
ter'minate number between 0 and 1, plausibility c‘orresp‘onds to an
indeterminate number with the open interval (0, 1_)_ as its do-
main.
All these examples reveal the influence of philosophic and tech—
nical theory. Of course, since we are examining systematic pro-
nouncements, a certain amount of formalism is to be expected.
What is lacking, however, is the recognition that systems are con—
structed within a domain of concrete events, and that because con—
struction is different from the ground materials it should therefore
be placed in perspective with reference to them.
I

GENERALITY AND SPECIFICITY 0F PROBABILITY SYSTEMS

Probability logic is hampered in its development, even more
than general logic, by the totalitarian and monopolistic‘ attitudes
of logicians. System builders clash with respect to motivation, pre-
supposition, material selection, andoperational techniques.
Probability systems should be difiérentiated on the basis~of the
criteria of generality and specificity. We have already mentioned
Induction.
3
Heunstlc.
I I I
 PROBABILITY AND INTERBEHAVIORAL LOGIC 213
(p. 206) that from the standpoint of general and exclusive systems
the number has been limited to seven. Interbehavioral logic de-
mands that we take into consideration the factors confronted in
probability situations—the eventualities and conditions which are
to occur or have occurred. We are therefore opposed to any view
that limits the number of probability systems, or excludes certain
Situations from the probability domain on whatever ground.
To indicate the difference between general and exclusive sys-
tems and practical structures concerning actual life situations we
consider briefly the system of von Mises.“ This author argues that
probability must be regarded as a scientific concept, and like other
terms, such as work in mechanics, it must be separated from any
popular usage. He goes on to argue that probability concerns only
three types of happenings—namely, games of chance, life insur-
ance, and the treatment of certain mechanical and physical events—
for example, the movements of gas molecules or the random mo—
tion of colloidal particles. What he would exclude is the probability
of one country’s going to war with another, the historical existence
of certain persons or events, the reliability of witnesses, the cor-
rectness of judicial verdicts, and so on.
Whether or not von Mises has the right to include games of
chance and life insurance in the scientific field, he ison solid ground
when he asserts that there are three fundamental purposes of
science; namely, bringing order into events, predicting their de—
velopment, and bringing about particular happenings in which we
are interested. Is it the case, however, that these three are exclu—
sively the interests of science? Or that science alone can accomplish
such purposes? 'These assertions are merely arguments for limit—
ing the definition of probability in a way acceptable to the author.
The statistical study of kinetic-gas theory is certainly an important
type of probability investigation, but it does not rule out other
kinds of probability inquiries.
Von Mises may well ask whether the usefulness of probability
in life insurance and mechanics is owing to the adoption of a single
correct probability definition (frequency), or whether it happens
that in these situations classifiable instances are available and al-
low one to set up series or isolate repetitive elements. But why
must One limit one’s interest to such types of situation? Certainly
“ Probability; see also Popper, Logik.
2I4. PSYCHOLOGY AND LOGIC

it is hardly fair to regard von Mises’ frequency theory as rational

by contrasting it with all competitive vieWs, with the implication
that the latter are mystical andIntuitive.
Interbehavioral theory, on the other hand, allows for the prob—
ability of theories in science and everyday life, as well as for the
classical head-or—tail tosses or the appearance of a die face. It is
only the adherence to a frequency theory of some form which
draws the line at single events. The test of the opposing View is,
of course, whether it fits problematic situations.
Von Mises’ conception is susceptible to two interpretations. In—
sofar as it indicates that science consists of a rigid analysis of events
we should indeed agree that refined intellectual operations are
designed to clear away the superficialities and irrelevancies of or—
dinary contacts with things. All science Gonsists of more analytic
and more exact investigations than simple everyday interbehavior
permits. On the other hand, if von Mises implies a generic dif-
ference between scientific and everyday interbehavior it is not ac—
ceptable, Since there is no break in the continuity of our contacts
with things and events. To be sure, the constructions of science can
not at every point represent, nor be integrated With, actual events,
nevertheless, science is bound to such events. Probably no one
would disagree with this statement were it not for the influence of
philosophical traditions which make us forget that our scientific
constructions are in no sense arbitrary, and that no matter how
complicated or abstract they become they are designed for the
purpose of describing and interpreting observable events. Even
though scientific work may become so channeled that much of its
activity consists of the production of artifacts for purposesof rigid
control, we still can not get far away from interbehavioral situa—
tlons.

RELATIVE VALUE OF PROBABILITY SYSTEMS

Faced with large numbers of possible probability systems we
must ask whether there are any evaluative scales with Which to
weigh or measure them. Are there criteria for allotting a higher
or lower value to a system involved with drawing balls out. of an
urn, as compared with one based on therelative frequencies of
car loadings or dice gaming?
∙
 PROBABILITY AND INTERBEHAVIORAL LOGIC 215
From a strictly systemological standpoint, the only 'criteriaare
those pertaining to the effectiveness of the system products them-
selves. In view, then, of the fact that the urn situation can fUrnish
rigid rules for a probability system, we may accord it high value.
On the other hand, though drawing balls out Of an urn provides
calculative events with fixed conditions, such as the total number
of balls, number of kinds, and relative number of each kind, do
we want to tie» ourselves to absolute relations at the sacrifice of
richness and complexity of events? Wherever we can employ
mathematical techniques we are in possession of invaluable cal-
culative instruments for achieving prediction. But to value mathe—
matics as calculative devices is one thing, to make calculation the
exclusive systemological criterion is another. The latter has led
to the futile conception of absolute and intuitive probability de—
terminations, and to the idea that probability'Is an irreducible and
indefinable property.
We conclude, then, that a place must be made for many criteria,
based on all sorts of human interests and types of probability ma—
terial. The acceptance of a variety of criteria argues nothing against
their value. We need only justify choosing such criteria, and make
sure of the evidence for the actuality of their sources.

HISTORICAL EVOLUTION OF PROBABILITY SYSTEMS

Students of logic divide themselves sharply into those who find
profit in the teachings of history, and those Who are sure that only
by timeless analysis can one achieve significant systems. Anyone
sensitive to the records of history may easily detect a progressive
evolution of probability systems toward an operational goal. The
evidences are of two sorts: first, changes in intellectual culture, sec—
ondly, structural transformations in probability systems.
Of the cultural shifts one of the most important is the reception
by science of probability ideas and techniques. When scientists
gave up the search for absolute laws and exceptionless proposi-
tions, in favor of statistical data and significance tests for observa—
tions, it became possible to interpret probability systems as sets
of interbehavioral factorsIn contingential situations.
Operational progress in the probability\ field has also been
furthered by the development of the postulational conception of
216 PSYCHOLOGY AND LOGIC

mathematics. To move away from the absoluteness of Euclidean

geometry to the relativity of premises and'conclusions bespeaks a
growing appreciation that even mathematics is the home of op-
erational principles. Howsoever far mathematicians and logicians
have been motivated by their zeal in searching for rigor, they have
learned that accuracy and precision
tional procedures.
can
only be achieved by opera-

The cultural changes mentioned, along with the acceptance of

the postulational principle, in the final accounting have only
hastened a trend. The goal is yet to be reached. The full realiza-
tion of effective probability systems must wait upon further re-
laxation of the powerful hand of cultural tradition. Still needed
is a separation of probability logic from dualistic psychology. The
lag of psychological ideas behind mathematical operations pre—
vents logicians from seeing that operations are
interbehavior, and
that the analysis of probability and probability calculation must
have better foundations. ItIs strange that what the historical evo-
lution of probability studies shows so plainly is not yet put to
work.
Pragmatic Beginnings. Laplace finds it remarkable that a
science which began with the consideration of games .of chance
should'have become the most important object of human knowl-
edge.15 That was because the founder of the modern phase of
probability theory was most at home in the lofty study of celestial
mechanics and the system of the world. Actually, the situation
marking “the humble beginnings” of probability study possessed
the same factors as any study more highly assessed.
That games are under consideration plainly indicates that we
are observing interbehavior with things. Let them be a pair of
dice which Cardan throws to meet a certain criterion18 incidental
to winning the game. Dice throwing is the basic interbehavior and
may be described as the lowest-level operation. The next level is
the analysis of the play situation, the ascertainment of the results
given the original conditions. There is only one way in which two
or twelve dots can appear. Eleven can\be thrown in two ways, and
ten in three ways. Whether it is Card‘an, the player, who makes

, .
TheorIe, . p. CXll..
Introduction,
1°
Todhunter, History.
 PROBABILITY AND INTERBEHAVIORAL LOGIC 217
the analyses, or Kepler or Galileo, all are operating by way of a
simple ordering and counting of results bounded by the character
of the devices employed under the rules of the game.
There are two important arguments for accepting the conven—
tional View that the beginnings of probability study lie in the work
of Pascal and Fermat. In the first place, by their time analytic
operations had .become elaborate enough to be dignified by the
term theory. Workers responded to themselves as factors in the
probability situation. The question had been raised whether the
propositions of arithmetic were inconsistent. Moreover, Pascal
criticized Fermat’s solution of the Problem of Points. In general,
probability situations had become enormously complex. In the
second place, Pascal and Fermat formulated a theory that the en—
tire probability situation can be handled by the operational tech-
niques of permutation and combination.
Response-Centered Systems. Following a long developmental
period in which the calculation of chances was assiduously cul—
tivated by mathematicians interested not only in games but
also in mortality and life insurance, there appeared a system
pitched on a highly sophisticated plane. The work of combina¥
torial analysis and probability theory, carried on by Huygens,
Leibniz, the Bernoullis, Montmort, DeMoivre, Euler, D’Alem-
bert, Bayes, Lagrange, and Condorcet, culminated in the technical
construction of Laplace.
The outstanding characteristic of' Laplace’s system of proba—
bility is that it is response centered. Not only is probability re-
garded as a determination of things by a. calculating individual,
but personal psychological processes are emphasized. For this
reason Laplacian probability has been seriously condemned as
subjective. Laplace declared:
The theory of chances consists in reducing all events of the same
kind to a certain number of equipossible cases, that is to say, cases
such that we are equally undecided about their existence, and deter-
mining the number of cases favourable to the event of which- the
probability is sought. The ratio of this number to that of all the
possible cases is the measure of the probability which is no more than
a fraction whose numerator is the number. of favourable cases and
whose denominator is the number of all possible\cases."
1“Theorie, Introduction, p. iv.
218 PSYCHOLOGY AND LOGIC

This excerpt makes clear that, for Laplace, the power of intel-
ligence is such as to penetrate the mysteries of nature, even if it
has to begin with nothing but symmetrical ignorance. The difficulty
which critics find in the Laplacian View is that states of mind can
neither be measured nor calculated.
But ignorance and belief need not be regarded as states of mind.
Such an interpretation stands in complete disregard of the newer
objective psychological theory. After all, the situation must be
described as one in which Laplace, a mathematician, is operating
upon certain data. And there are degrees of efl'iciency of contact
with the events of which the probability or possibility is to be cal-
culated, or otherwise determined.
A: more substantial criticism than the charge of subjectivism
is that Laplace does not explicitly indicateohis actual contacts with
things. He is determining the ratio of favorable cases to the sum
of both favorable and unfavorable cases, and that means he is ob-
serving or otherwise interbehaving with events and Situations.
Event—Centered Systems. Thinkers who harbor mental states
in their intellectual households finally escape from subjective
probability systems by turning completely to Obj ective things. A
comprehensive~system was thus built up on the pattern of a ma—
terial or frequency theory. The frequency theory was proposed
by Ellis18 and Cournot,19 and established by Venn.20 An extreme
empiricist, Venn regarded his logical interest of attaining certainty
as satisfied only by turning away from subjective belief to the ob—
jectivity of series or frequencies.
The frequency theory of Ellis, Cournot, and Venn is an in-
comparable example of the discrepancy between a logician’s op-
erational procedures and his formal structurizatiQn, between his
practice and his theory. Venn assumed that he was avoiding sub—
jectivity, that he was dealing exclusively with external things or
conditions. Actually, of course, he was concerned with system
building, which offered him confidence in the data with which he
was interacting. His practice, however, belied his theory that he
was. not dealing with belief, conditions of ignorance, etc.
Such theory-practice discrepancy stems directly from the under—
“ Foundations.
Exposition.
Logic.
 PROBABILITY AND INTERBEHAVIORAL LOGIC 219
lying mentalistic psychology. Had not Venn, like the frequentists
since his time, been dominated by mentalistic psychology, he
would have realized (1) that the a priorists whom he was attack-
ing were systemizing calculations and schemas having to do with
objective events, as DeMorgan and other belief proponents held,
and (2) that centering on objects is arbitrarily Selecting only one
interbehavioral factor. Accordingly, both the empirical frequent—
ists and the opposing rationalists ignore the fact that probability
situations may be of different kinds. What they could not know
at their time of historical development was that since there is no
mental, but only interbehavior, each type of probability system
really goes back to contacts of individuals with stimulus objects.
Because the empirical frequency system is so vigorously recom—
mended as obj ective, the paradox exists that objects and their fre—
quencies become dissipated into nebulous states of mind. Some de—
vice, therefore, must be invented to keep close to the basic situa—
tions giving rise to probability problems. Venn’s formulation that
probability is a body of rules for drawing inferences about classes
of events which combine individual irregularity with aggregate
regularity is an appealing one.
Propositional Probability. A decidedly new level of probability
systems emerged when probability problems became centered in
propositions instead of in responses or Objects. For one thing,
propositional probability mirrors a more complicated set of situa—
tions. It surpasses the relatively unsophisticated inclination toward
one pole or the other in what is essentially a bipolar situation. The
superior sophistication of propositional probability is measurable
in its greater or lesser articulation with operational processes.
While it does not renounce the psychology of mind, it still con-
stitutes a step toward an objective analysis of probability interbe—
havior. Such is the case when propositional relations are regarded
as unique and rational, and“ as referring to the conditions of things.
(a) Probability Propositions and Things. This form of prop—
ositional system, for which we choose Peirce as- an illustrative
proponent, is based squarely on Venn’s formula stressing the. ob—
servation of recurrences of events. Peirce modified the system of
emphasizing the quantitative precision which mathematical calcula—
tion supplies. Propositions are best interpreted as equations. This
is exemplified by his statement:
220 PSYCHOLOGY AND LOGIC

Probability is a kind of relative number, namely it is the ratio of

the number of arguments of a certain genus which carry truth with
them to the total number of arguments of that genus.21

We may call this propositional system the truth-frequency the-

ory; it implies that propositions must be observationally confirmed,
and thus reveals its connection with-scientific problems. Confirm-
ing a proposition is tantamount to contact with things. A high point
is reached here. But the system stops short of a thoroughgoing in-
terbehavioral position. To a certain extent the emphasis of propo-
sitions neutralizes the underlying philosophy and psychology of
sensations. When propositions are taken to be equations they
appear strongly supported. Still, the lack Of a basic obj ective psy-
chOlogy, coupled with the influence of formal and monopolistic
logical theory, does not allow a full—fledged operational inter-
pretation.
Completely carried out, an interbehavioral View emphasizes the
operational character of calculations. Moreover, calculations are
not regarded as detached actions leading to products which are
later applied to situations; instead they are treated as interactions
with substituted—for relations—in other words, interactions with
things handled by means of substituting symbols. On this basis the
calculus is necessarily bound by the calculated things, and proba-
bility events are not shied at because they can not be measured
or ordered—at least, in conventional ways.
(b) Probability Propositions and Rationality. Of the many
forms in which this probability theory can be described we choose
the one sponsored by Keynes.22 He moves on from happenings or
events to propositions in order to avoid the subjectivity of belief
or knowledge doctrines. The theory’s primary design, then, is to
favor operations instead of things, but with a stable and calculative
basis. Probability has to do with a relation of propositions.
Let our premisses consist of any set of propositions h, and our con-
knowledge
clusions consist of any set of propositions a, then, if a
h justifies a rational beliefIn a of degree a, we say that there is a
of

probability-relation of degree a between \aand h. This will be written

−−
a/h— a.23 ∙
n C. P. (Collected Papers) 2,657.
22 Treatise.
'3
Ibid., p. 4.
 PROBABILITY AND INTERBEHAVIORAL LOGIC 221
Under the influence of mentalistic psychology and realistic
philosophy the rationalistic probability View turns squarely away
from an interbehavioral situation, though it connects with it at
long range. To assume that premise propositions must be weighted
by evidence is an excellent provision. Again, the application of the
system to particular cases, instead of exclusively to frequencies,
a
carries with it good principle. All the advantages, however, are
cut short by a rationalistic and intuitive philosophy.
To avoid subjectivism Keynes holds to a Platonic version of
probability—namely, probability is indefinable. In other words,
since empirical evidence or frequencies of occurrence do not lead
to the determination of probability, he resorts to mystical intuition.

PROBABILITY AS STATISTICAL OPERATIONS

Statistical operations in the various scientific domains constitute
the presystematic basis for a probability system in line with ob—
jective psychology. The point is that particular systems are or—
ganized as adapted to specific situations. Such systems are con-
structed by way of overcoming the inevitable contingencies of tech—
nological research and scientific investigation.
Even if it is true that only a negativistic approach is made to
an interbehavioral type of probability, there is the advantage of not
involving oneself in a disserviceable system based on philosophical
rationalism or empiricism.
Such a statistical probability system as we have been suggesting
is not to be confused with the conventional frequency structure.
The basic idea consists of the concrete activities of individuals in
making measurements and estimates in order to carry on the
various operations of science and technology.

BASIC ORIENTATION 0F PROBABILITY SYSTEMS

So numerous and varied are the sources for probability systems
that it is advisable to make clear whether we are concerned with
( I) the reliability of knowledge, (2) the existence of things, (3)
the adequacy of statements, (4.) the eflectiveness of a method, or
(5) some combination of these. Furthermore, it is of considerable
importance to make sure that we properly locate probability situa-
tions preliminary to organizing probability structures. Because of
epistemological and ontological influences such orientation has
22.2 PSYCHOLOGY AND LOGIC

'been lacking. In the following paragraphs we consider some of

the conventional landmarks for isolating and comparing proba-
bility situations.
Probability, Certainty, and Necessity. Keeping close to specific
systems means that we incline toward a critical isolation and com-
prehension of probability structures as over against large-scale
philosophical systems. In addition, we achieve workable criteria
to differentiate between the probable, the certain, and the neces-
sary. This is no less a task than the separation from each other of
linguistic, philosophical, and scientific problems. The situation
can be well illustrated by a quotation from Jevons:
In nature the happening of an event has been pre-determined from
"the first fashioning of the universe. . . . A steam-vessel, for instance,
is missing and some persons believe that she has sunk in mid-ocean,
others think differently. In the event itself there can be no such un-
certainty, the steam-vessel either has sunk or has not sunk, and no
subsequent discussion of the probable nature of the event can alter
the fact.24
At once we can Sift out and discard the metaphysical remarks about
predetermination, unless, indeed, we are interested1n ontological
speculations rather than actual events. Next we Should clear up a
terminological problem. While it is true that terms such as cer—
tainty and necessity have no standard and fixed usage, certainty
really refers to response rather than to stimulus object. Jevons
appears to equate certainty and necessity as matters .of existence,
as over against probability concerned with thought or belief. Not
much can be said in favor of the term necessity, for it belongs to
the metaphysical domain of determination and predetermination,
but if it is used it belongs more to Objects than\to reactions to
them. The criterion invoked here is altogether a practical one.
Whether or not a steamer sinks depends upon a number of fac-
tors. For example, a certain combination of meteorological circum-
stances, the materials out of which the vessel is built, the nature
and disposition of its cargo, the lack of cargo or ballast, the num—
ber and competence of'ofl'icers and meg
aboard, and so on. In case
of war, there are the conditions Of pursuit by destroyers, whether
or not they are able to overtake the vessel before reaching the stip-
ulated limitsIn their orders, whether they
Principles, p. I98.
have
run out of fuel,
\
 PROBABILITY AND INTERBEHAVIORAL LOGIC 223
etc. In such cases, of course, knowledge of the event should never
be confused with the conditions of the event._There are no abso—
lutely isolated events, but combinations and interrelations of events
which make room for probability. It is precisely the estimation of
such probabilities which constitutes the work of probability system
building. .
However, on‘ce the ship has sunk there is no longer a proba-
bility in the event. All probability systems now have to be built
upon the basis of knowledge factors. Concerning a repetition of the
sinking one can build alternative systems based upon available
data. We must weigh our evidence upon a variety of scales possess-
ing a greater or lesser basis in combinations of factors.
Probability, as knowledge or reactional attitudes, is an approxi-
mation based upon definite contact conditions—that is, availability
of data or techniques of securing data. Among objective events
we also have a set of approximations, but here the conditions are
matters of interrelations of natural conditions. For instance, how
much reaction between chemicals takes place depends upon pres-
sure, temperature, the nature and amount of material, and the
presence or absence of catalyzers.
Probability, Truth, and Error. There is a factual basis for the
conventional diflerentiation between truth and existence. Truth
as a term refers to knowledge and belief—in short, to reactions to
things as over against the existence of things. This statement may
be extended also to the conventional view that 'truth as compared
with existence concerns a reference to things or a statement about
things. How far do sentences conform to events? Statements or
propositions are true and definitive when they match things to
which they refer or which they symbolize.
Complete and satisfactory agreement between statements or be-
liefs and things is not always attainable. We need only mention
the limitations imposed by complexity of things and events and
the need to meet criteria of precision and accuracy. The reactor
in such situations faces the problem of relative and approximate
conformity. In science we fully appreciate the approximate charac-
ter of our propositions. We must be fully aware of the specific
conditions under which the propositions hold. Scientific proposi—
tions always carry with them the realization that they are not ab-
solutely true.
224 PSYCHOLOGY AND LOGIC

We conclude, then, that probability systems pertaining to knowl-

edge and belief can be clearly separated from systems stressing
things or events. But we must next insist that the distinction is a
relative one when we deal with actual situations as compared with
epistemological or ontological theories. Notice, therefore, that
truth and falsehood depend upon many specific conditions, and
not upon some general metaphysical criterion. Probability does not
differ from truth or certainty (I) because all certainty is meta—
physically regarded as probable, or (2) because probability has to
do only with “matters of fact” different from “matters of reason.”
That probability propositions are never true or false is set up
on a generalized metaphysical basis that there is no truth, only
probability.25 Taken in a specific context the view that differenti—
ates truth from probability is agreeable. cNamely, there are true
prOpositions and there are probable ones. In each case the situation
must be described in terms of the things and events involved, and
of what the responding individual does to them. We have already
seen that the reduction of all truth and existence situations to
probability situations is based upon elusive constructions concern—
ing'mental states and the creation of external worlds:
The distinction between “matters of fact” and “matters of
reason” goés back to an earlier phase of the doctrine that psychic
states must be taken into account. We must keep in View that this
distinction was developed by Locke, Hume, and other rationalistic
empiricists because they were committed to an identification of
things with states of mind.
Probability, Possibility, and Impossibility.
When
problems center about a knowledge criterion, the question is
probability
asked: How reliable is knowledge? The degrees of knowledge, or
probability, cover the range from the possible to the impossible.
Basically, the criterion focuses on the range of conformity of prop-
ositions with corresponding facts. If the proposition refers to an
impossible situation it is totally unreliable, if to a possible one, its
reliability is a function of coverage, or of degree of correlation.
Systems of knowledge reliabilityi‘should, of course, be confined
to concrete situations, where alone they can have significance.
Nevertheless, the attempt is frequently made to establish systems
2 ∙
Re1chenbach, ∙
ExperIence.
i
'
 PROBABILITY AND INTERBEHAVIORAL LOGIC 225
of absolute possibility or impossibility. In such cases the criteria
are useful only in estimating the credulity of the system maker,
rather than in determining any correspondence of his knowledge
with evidence. -
More significant systems are developed in concrete Situations
when reliabilities can be checked in terms of events, their meas-
ures and ranks. In scientific investigation the possibility or impos-
sibility of either an event or a response to it can be established in-
terms of probable errors, of standard deviations centering around
means, and by other devices. And these checks and balances are
added to whatever other tests are used.
Probability, Potentiality, and Actuality. Despite the fact that
logicians do nbt think much in terms of behavior, they can not
escape it. When dealing with inductive systems they are forced
to take activities into account. Whenever they consider questions
of expectation, reasonableness, and belief or evidence, activities
surge to the surface. Again, when logicians are confronted by prob-
lems of risk and play, by problems of making wagers and posits,
developing sampling techniques and organizing predictive systems,
they enter with full force into behavior situations.
Activities already performed, along with the present general
status of the behavior field, constitute an actuality factor from
which future potentialities may be estimated or calculated. Past
or present operations become premises for inferential structures.
On the stimulus—obj ect side, sampling techniques are designed to
ascertain what the potentialities are, so that the mass from which
the sample is taken will measure up to a certain standard.

PROBABILITY IN SCIENCE
The hypothesis that scientific enterprises are inevitably occu-
pied with probability situations has become established only at the
cost of considerable struggle. Even now many writers accept the
proposition grudgingly, at best. Though science is obviously in—
vestigation—namely, interbehavior with events that are difficult
and changeable—scientists still cling to the notion that they are
seeking absolute knowledge. Hence their bewilderment when
relativity, discontinuity, and indeterminacy-are forced upon them.“
2°
Cf. Darwin, Logic.
226 PSYCHOLOGY AND LOGIC

Probabilities appear strange and disconcerting. Let us glance

briefly at the course of this absolutistic tendency.
To go back only to the Renaissance, the establishment of science,
so extravagantly celebrated as the turning away from authority
toward investigation, did not, in fact, release the scientist from the
fetters of the absolute. On the contrary, faith in scientific discovery
was established upon an absolutistic basis. For example, the three
laws of Kepler were achieved by his ability to correlate Brahe’s
observations with Euclidean geometry. By splendid acts of crea—
tive imagination Kepler transformedthe sun into a focal point
of an ellipse, which itself was the locus of the successive points
described by the earthIn its motion about the sun, and thus es—
tablished astronomy. Though the orbit of the earthIs not actually
an ellipse, this fact did not interfere with the development of a
thoroughly dependable deductive principle. From this, and other
expressions of loyalty to Euclidean absolutis,m arose the brilliant
achievements of celestial mechanics.
From this time, to beyond the middle of the nineteenth cen—
tury, the reign of deduction and Euclidean geometry and the con—
ception of a fixed universe operating according to absolLIte laws
supported the a priori powers of the scientist, and stabilized faith
and certainty. Investigation and experimentation only appeared
to confirm the alleged certainty that facts cOuld be subordinated
to the scientist’s rational powers. Hence, the overweening faith
in cause as the antecedent of effect in a closed system (chap. 19),
and the facetious assertion that absolute laws were only substitutes
for deity.
Science, however, is, after all, investigation. And, despite the
fact that the cultural conditions of the time demanded and re—
ceived 'their due, changes inevitably took place in the general
scientific structure. In biological nature, for example, it was dis-
covered that fixity is not the law, but that changes are constantly
occurring in animal forms. In physics itself, contacts with thermal
and electrical happenings proved that the old closed system, with
its Euclidean background, was not a universal system. Work with
thermal events indicated a break, in the Sense that events werevnot
reversible. Problems of entropy revealed that the old particles,
.which were really Euclidean points, could not be traced in their
 PROBABILITY AND INTERBEHAVIORAL LOGIC 227
movements as though they conformed .to the lines of geometric
figures.
Statistical mechanics was the result. It meant the introduction
of probability into physics. At first, this probability notion di-
verged but slightly from the old conceptions dominated by Eu—
clidean geometry and the power of absolute numbers. Maxwell
still thought 'Of the distribution of molecular velocities as deter-
mined by the Gaussian law of errors. Quite lacking in this so'rt of
statistics is the derivation of variables and laws from actual observa-
tion. Facts are still fitted to a priori principles. For this reason
statistics does'not necessarily articulate with genuine probability,
nevertheless, an entering wedge is available. Unfortunately, how—
ever, in the sense of statistical data, probability is still regarded
as inharmonious with experimentally derived facts, although, as a
method, probability techniques are safely ensconced. How long
deductive science will remain the dominant form is hard to say,
but non'deductive, nonexact, or statistical materials can no longer
be excluded from science.
In addition to the general deductive tradition, science is often
handicapped by the cost and stability of apparatus
findings appear inevitable. which make

A definite impact of probability upon scientific thinking has its

source in the Heisenberg Indeterminacy principle. Briefly, the
simultaneous observation of the exact position and velocity of a
particle can not be achieved because the use of light necessary for
the observation modifies the character of the thing observed. Un~
doubtedly, this intractability of events, with a corre3ponding dif~
'ficulty of mastering them, is really common in every investigative
field, but it has influenced scientists tremendously, simply because
it occurs in the traditionally most exact science. The law of er—
ror distribution, the obvious fact that all measurements vary, and
the ample use of hypotheses in all science have not hitherto ap-
peared very impressive.
All these factors can no longer be interpreted as Obstructions to
knowledge, nor can they be used to\ support ideas of cosmic chance
or fortuity. What they confirm is the fact, so frequently empha—
sized throughout this volume, that all creditable laws are construc—
tions resulting from the scientist’s operatiVe contacts with the
228 PSYCHOLOGY AND LOGIC

things and events which set him his problems. In other words, to
determine probabilities on the basis of frequencies of occurrence is
to interbehave with probability circumstances. The tools we use,
such as propositions and various calculations, are only tools. Some
particular probability task may consist of sorting out and arranging
certain propositions concerning the presence, absence, and fre-
quency of various participating factors, but the probability enter—
prise itself is a more inclusive type of event.27
It is necessary to recognize that statistical or experimental
studies are only designed to discover. the natural properties and
relations of data. This means that scientists can not reject things
and events not subject to calculation or measurement. Such quali—
ties are as genuine and as real as mathematical relations. Indeed,
it is precisely the lesser availability of these cmeasurable properties
and relations that makes for probability enterprises.

PROBABILITY AND INDUCTION

The area of induction is the locus of the technical aspects of
probability system building. In fact, in this area the basic issues
of prObability logic come to the fore. We have already mentioned
that inductive logic developed as the correlate of the logic of cer-
tainty.
Logic and Events. The relation between probability and in—
duction at once brings up the general question: How can logic be
,related to events? The development of science and the upsurge of
investigative manipulations, including measurement and calcula—
tion, promoted a need for a system to correlate with traditional
deductive logic. Influenced by the historical convention of universal
systemization, logicians felt obliged to assimilate scientific facts
to their cosmic structures. This they did by means of the processes
of reduction and substitution. Inductive processes, which should
have suggested specific content systems for particular contacts
with events, became subordinated to the universalizing process of
cosmic system building. The need for absolute criteria engendered
the problem of how inductive logic is‘ related to deductive logic.
Relation of Induction and Deduction; Uneasy about induétive
processes, logicians pursue their futile search for ways and means
r
∙
Cf. Kantor, Operational.
 PROBABILITY AND INTERBEHAVIORAL LOGIC 229
to transform induction into deduction. Induction has been charac-
terized as disguised deduction, a mere method of making plausible
guesses.” This View has been severely criticized;29 in the interest
of science it is declared that much deduction is disguised induc-
tion.30
One thing is certain: both the opponents and defenders of in—
ductive logic stand upon metaphysical foundations. Neither group
takes the position of the actual system builder. Often those who
support the inductive principle do so in the interest of a sensation—
istic metaphysics. Thus the inductionists attempt to approach a
stable system. Induction is set up with bracing straps and made
into a law for scientific work. To approach the inductive situation
with an interest in system building requires no strait jacketing
of science, nor of any other domain of human action. System build—
ing is a type of labor performed for its own sake, or in conjunction
with some scientific, technological, or even more practical pursuit.
From this standpoint there is no problem of induction, no need to
justify any sort of preoccupation with metaphysical principles.
Above all, there is no problem of either-or, as between deductive
and inductive system making. System builders are many, and they
build all sorts of systems. The problems, then, concerning their
activities and products should be examined in their local habitats.
Probability and Inductive Principles. LOgicians, We have in—
dicated, inevitably feel the pressure of events which disturb totality
and absoluteness. Events somehow force to the front inconstan—
cies, irregularities, and inconsistencies. But, on the other hand,
system implies structure and this, in turn, suggests stability. Con—
sequently, in the history of culture various notions have arisen
concerning the existence of principles which guarantee proposi-
tions, or systems of propositions, about contingent matters.
Principles in this sense are directly related to all the historical
varieties of a priori constructions. In every case such principles con-
stitute in the final analysis a verbal structure referring to some
feature abstracted from contacts with stimulus objects. When
stabilities and recurrences in events\are noticed, a principle is set
23Russell, Principles, p. 360; Jeflreys (Theory, p. 5) suggests that Russell Objects
to being too frequently quoted on this point, but Cohen (Preface, p. 19) endorses
the statement.
Bread,
Relation.
3"
Campbell, Physics, p. 9.
230 PSYCHOLOGY AND LOGIC

up that things sustain invariant relations. To account for the ob—

server’s ability to discover these relations, writers invent a priori
powers in the human mind, or necessary principles abstracted from
all knowing. In general, processes and operations are hypostatized
by verbal description. _
A very instructive a priori notion is‘ located in the recently de-
veloped argument from perceptiOn. Curiously enough, this argu-
ment begins with the everyday observatioh that in our perceptual
behavior we respond to things on the basis of past responses to
them, instead of purely on the ground of immediate contacts with
available properties. But at once this fact isclifted out of its native
domain of interbehavior with things into the empyrean of the
mind’knowing something without experience.31 Here is an ex-
cellent illustration of imposing ↓ ∏ interpretations upon
events, while erroneously assuming that the events point to such
interpretations. The a priori is established by eyents which really
show the complete antithesis to all a prioris.
Probability and Prediction. Prediction problems constitute a
bridge between probability and Inductive Logic. Historically,
inductive problems have been concerned with the interrelation of
past, present, and future events, by way of anticipating what may
occur. Now probability events can be selected for-their bearing
upon the future. Can we assert that any future event is probable,
or will probably occur?
Future events are as yet nonexistent. How can we say anything
about them? This is a sample of the purely syntactically formu-
lated question. Notice how greatly it differs from a concrete inter-
behavioral question. Future events taken as actual life situations
may beregarded as no different from any event notyet observed.
What we call future event situations consist of a particular colloca—
tion of factors of the same general sort as those with which we
have had considerable experience. Thus, we must rely upon the
frequencies with which the type of anticipated event has previously
occurred. In other words, our determination of probability value
is made on the basis of interbehavior with similar events. Very
little significance can be accorded any kind of prediction which is
not so grounded. Naturally, interbehavior with future events is
’1
Kneale, Probability. ∙
 PROBABILITY AND INTERBEHAVIORAL LOGIC 231
more hazardous and open to error than the determination of what
factors have contributed to an Observed event. Still, this circum-
stance presents no serious difliculty when we operate with specific
situations. Only when we deal with purely verbal questions, such
as what right we haveto believe that conditions in the future.
will be similar to past ones, do we run into the philosophical and,
we might add, futile problems of Hume.
 CHAPTER XXII
ANALYSIS OF PROBABILITY SYSTEMS
DESIGN FOR PROBABILITY SYSTEM ANALYSIS

THE most effective method for studying probability logic is

the critical examination of probability systems. To pursue this
method we follow the study design already prefigured in the
interbehavioral fields in which persons interact with contingent
and indefinite things.
We feature five outstanding components of this investigative
pattern: (1) the builder, (2) the work he performs, (3) his build—
ing materials, (4) the conditions influencing his work, and (5)
the product
he completes. These are mostefiectively structured
in the following set of questions and specifications.
(1) The System Builder. Concerning the worker we ask:
(a) Is be building a general system—say, an inductive
logic? OrIs he constructing a localized system for a more particular
purpose?
(b) Is he motivated by the principle of system
sake, or system designed for some further end?
forsystem’s

(c) Is his guiding preconception absolutistic or relativistic,

leading respectively to rigorous results or to those valid for par-
ticular situations?
(d) Is he so dominated by traditional notions of mind that
he builds an epistemologically rational or empirical system, or is
he erecting a nonmetaphysical system for carrying on some specific
enterprise?
(2) Methods and PrOcedures.
(a) Linguistic: constructional work performed by setting
up references and symbols.
(b) Mathematical: constructional work reduced to cal—
culating and computing.
(c) Observational: what part is played in system building
by recording and structuring the resultsof observation?
((1) Manipulative: what place have manipulations among
the systemizing procedures? ∙
 ANALYSIS OF PROBABILITY SYSTEMS 233
(3) Materials.
(a) Events (chemical, physiological, astronomical, bio—
logical or humanistic).
'

(b) Propositions (sentences" and relations between them).

(c) Statistical frequencies and ratios (or similar subject
matter). '

(d) Aets or processes of various sorts.

(4) Working Auspices.
(a) Philosophical.
(b) Scientific.
(c) General systemological.
(d) Programs of everyday activities.
(5) Work Products.
(a) Theoretical (abstract systems).
(b) Practical (statistical tables, lists of constants, other
tools).
(c) Actual processes (recipes for calculation, etc.).
This design for system analysis sets aside all conceptions of
philosophy and psychology built on the traditional views of mind.
The mind theory basic to both historical rationalism and empiri—
cism is replaced by hypotheses derived directly from observations
of system-building situations.
Once more we refer to one of the basic postulates of the inter-
behavioral—field theory pointed out in Volume I—namely, that all
psychological activity, including that of the scientist, mathema—
tician, and logician, constitutes the interaction of persons with
stimulus objects, whether natural things (trees, land, water), or
creations such as round squares, souls, and psychic states which
can only substitutively exist through assertions or symbols. ∙
Because the interbehavioral-field construction is based upon
actual observations of persons at work under particular conditions
our study design enables us to see whether things and events or
traditional theories are stressed. To illustrate, how muchwere the
constructors of electrical field theory influenced by the need to op-
pose action at a distance? Again, how much has theological thinking
influenced the development of the infinite in mathematics?
Ourxstudy design constitutes a mirror image of probability
situations. We set up interbehavioral—field postulates on the basis
of interbehavioral situations in order to derive interbehavioral
234 PSYCHOLOGY AND LOGIC

construction products. Interbehavioral fields are therefore based

upon interbehavior. Implied, then, is a limited regress. Is this
regression objectionable? Hardly, for all the levels are definitely
set in human events, from Which all theories are derived and
beyond which none, in any significant direction, can go.

TYPES OF PROBABILITY SYSTEMS

Despite their belief that probability systems are exclusive and
authoritative, logicians build different types of systems. In this
manner they unwittingly endorse the.interbehavioral view that it
is impossible to confine probability theory to a single type of
formula or situation. By reviewing a number of typical probability
formulations we shall find that the arguments for exclusivistic
theories are based on overemphases of particular items. Hence
the numerous conflictsIn probability theory.
3

(1) Probability as Rational Responses

Laplacian Probability Theory. Historically, the earliest theory
of probability was formulated under the auspices of an excessively
rationalistic culture. Recall the classical formulation that proba—
bility is a fraction whose numerator is the number of_ favorable
cases, and whose denominator is the total number of equally likely
favorable and unfavorable alternatives. This was a development,
surely, from the Renaissance awakening to the powers of mathe-
matics. The source of probability studies is the same that gave rise
to man’s confidence in his deductive powers, and which was en-
couraged by the pursuits of geometry. An effective symbol of the
period was Galileo’s and Kepler’s faith that‘authority could be
replaced by an a priori reading of the numbersIn which the book
of nature was written.
We have seen in Chapter 21 how probability theory originated
in the application 'of mathematics to the practical problems of
chance. Recall the effects of Pascal’s and Fermat’s assistance to the
Chevalier de Méré in his gaming problems. The rationalistic em—
phasis, with a recession of stimulus objects, is favored by the Situa—
tion in which one desires to gamble sutcessfully when knowledge
of the eventualities are not, to be sure, cOmpletely unknown, but
certainly not sufliciently kno.wn
When the vast intelligence can not embrace 1n .the same certain
 ANALYSIS OF PROBABILITY SYSTEMS 235
formula all the past and the future, it must, according to Laplace,
fall back upon the principle of suflicient reason for help in formu-
lating judgments concerning equally likely eventualities. In direct
line with this general View an. equivalent statement could be de-.
veloped in the form of an equal distribution of ignorance with
a fraction of one-half.
Unless one is alive to the postulational character of mathematics
one may be surprised that Laplace’s essentially mathematical
theory of probability should result in an impotency of probability
calculation and a consequent subjectivity of viewpoint. Opponents
of the classical a priori theory find it easy to discover its faults,
though they fail to produce a satisfactory substitute. The
ing examples indicate the mode of attack. follow-
Equally Likely Cases do not Always Exist. Suppose the face of a
die is somewhat imperfect; thenit is impossible to apply the
equally—possible or equally-likely Criterion. Again, assume that 'a
biased coin is estimated to turn up head with a probability of .63;
it eludes the equipossible ratio.1
The critics go even further in the case of more complicated hap-
penings. When mortality records show that a thirty—year-old man
has a probability of .950 of living another year it
seemsabsurd
to say that, of a thousand possible events, the ratio of favorable
to unfavorable cases is 9 50/50+ 9 50. Similarly, should mortality
tables indicate that the chance of a man of forty dying within the
next year is .01 I, what are the equally likely alternatives? Should
the ratio be 11/989+I I or 33/2967—l—33?2
Perhaps it is too much to expectja theory to be applicable to con—
ditions from which it was not derived and which it was not the
author’s intention to handle. Of course, what the critics point out
is entirely valid—namely, the classical theory is not a universal
one. Nevertheless, they are mistaken in assuming that the classical
theory alone lacks universality; all theories do. By all means, of
course, when frequencies are involved, a frequency theory has to
be resorted to, but will all probability problems inVolve num-
bers of cases? Frequency theorists \naturally contend that they
can always construct series or collectives. Must they not, however,
consider that at least sometimes the frequencies are not essentially
Nagel, Principles, p.45.
2 von Mises, Probability, p. 102.
236 PSYCHOLOGY 'AND LOGIC

numerical, but
are really based upon the abstraction of Specific
qualities or characters?
Unimportance of Equally Likely Cases. The classical theory
is also attacked because in the final analysis it consists only of a
mathematical calculation of permutations and combinations—
that is, the interrelationship of numbers. In simple cases of penny
tossing the probability V2 does not indicate whether the head or
tail actually will fall up. Or, in a more complex situation, knowing
that a geometrically symmetrical cube is also kinetically sym-
metrical—that is, possesses equal statical moments and moments
of inertia, yielding a probability of 1/6 of falling on any side—
does not enable one to know that it will aetually fall on any one
particular side.
Again, equally likely cases are said to beoinsignificant as against
the type of probability concerned with averages, dispersions, aver-
age error, probability function, and the law of 3error. The greater
importance of the latter is taken for granted because of its appli-
cation to scientific work in physics (kinetic theory, thermodynam—
ics), astronomy (star origin and distribution),
netic’s).
and biology (ge-

E quiprobabzlity Involves Trivialities and Contradictions. As a

mathematically originated theory the classical formulation of
probabilityIs presumed to be a priori because of its independence
of, and potency concerning, actual events. Its fundamental ratio
with respect to, say, the tossing of coins is V2. From whence is this
fraction derived? The answer: Equiprobability means one can not
say beforehand what will happen except in terms of the alterna-
tives. In other words, the only basis is the presence of two factors,
each of which is equally likely. Despite the circularity in the def—
inition classical Writers have, of course, persisted in applying this
type of calculus.
In support of this issue Bernoulli invoked the Principle of In-
sufiicient Reason, which Keynes has renamed The Principle of I n—
difference,- it is also called The Principle of Equal Distribution
of Ignorance. According to this principle two events are equi-
probable if thereIs no reason for supposing either one or the other
will occur.
Keynes,3 on the basis of some suggestions of yen Kries, indicates
' Treatise, _p. 4 3.
∙
 ANALYSIS OF PROBABILITY SYSTEMS 237
the inherent contradiction. For example, knowing nothing con—.
cerning the color of a book we could say that the proposition “This
book is red” has a probability V2, but the same thing could be.
said of any other color. Accordingly, we would have the impossible
case of 3, 4 or more equally likely exclusive cases. In other
words, the sum of probabilities would be greater than 1.
It is questionable whether the criticism is justified, since a‘tcord-
ing to the proponents of this theory the emphasis is upon the ratio
and not upon the concrete events involved. On the other hand,
one may argue that there is always something known—namely,
there are two sides to a coin.
To correct the classical theory, and to confine it to its proper
domain, we need only consider that it was proposed for problems
involving dice throwing and for the selection of hands in card
games. Though we may regard such situations as trifling, there is
good reason for defending the work done upon them. In the first
place, aside from the fact that such intellectual heroes as Cardan,
Galileo, Fermat, Pascal, Leibniz, Huygens, the Bernoullis, La—
place, and Poisson were concerned with suchsituations, they stimu—
lated the development of a probability calculus—one of the most
significant mathematical building stones. Indeed, it would be
cultural myopia on our part to be insensitive to the origins and
ramifications of cultural factors. In addition, there is considerable
intellectual profit in observing the specificity of situations and keep-
ing Our formulae thoroughly articulated within their framework.
Despite the almost universal condemnation of the classical or
rational theory—and the impetus given to this condemnation by
the absorption into probability situations of scientific problems,
such as those of statistical mechanics—one may still find justifica-
tion for it. For one thing, the condemnation is ameliorated by the
fact that the critics themselves are merely using a certain kind of
situation for the basis of their criticism, one which, moreover,
favors acceptance of their own formulations. Howsoever correct.
they may be in cleaving to more important sorts of probability
events, they are still universalizing g principle. The result: They
exclude all other kinds of probability situations.
(2) Probability as Knowledge Responses
Probability as Quantity of Knowledge. British writers on prob—
ability, in contrast to those on the Continent, are concerned with
238 PSYCHOLOGY AND LOGIC

individualistic rather than with general cosmic happenings. Even

when they accept probability as an a priori factor, as Jevons does
when he declares:
The Laws of Probability rest upon the fundamental principles of
reasoning, and cannot be really negatived by any possible experience,‘
they prefer to put it in the form of belief or knowledge, rather
than of reasoning. For Jevons, probability theory: ∙
defines rational eXpectation by measuring the comparative amounts
of knowledge and ignorance.5

The value of probability is its efficacy ini‘regulating action with

regard to future events, in a manner leading to least disappoint-
ment in the long run.
Jevons develops his specific formula on the basis of a criticism
of DeMorgan’s description of probability as degree of belief and
Donkin’s interpretation of probability as quantity of belief. Jevons
finds the conception of belief too diflicult. He rejects Donkin’s
emendation of DeMorgan on' the ground that probability is not
a quantity of an entertained belief, but rather what one ought to
believe. Probability for Jevons, then, is the quantity of knowledge.
The line of British writers, including DeMorgan, Donkin,
Jevons, and others, is referred to as subjective. Certainly they
stress reactions instead of things. Jevons specifically asserts that
“Probability belongs wholly to the mind.”6 Chance, he believes
with Laplace, is only an expression for ignorance. Since he accepts
a complete determinism there is no such thing for him as chance
in nature.
Subj ectivistic theories like those of the writers quoted are almost
universally condemned, even by the British writers of the time.
Certainly, as universal theory they are unacceptable. The question
remains: Do they represent any facts whatever? By way of answer-
ing let us relate them to underlying interbehavioral principles.
It is a significant fact that Jevons is the man who endorses
Herschel’s dictum that “Numerical precision is the soul of science,”
and who was certain that “Every science as it progresses will be—

Principles, p. 206.
' Ibid., p. 200.
°Ibid., p. 198.
 ANALYSIS OF PROBABILITY SYSTEMS. 239
come gradually more and more quantitative.”7 Again, Jevons is
counted among the forerunners of those who applied mathematics
even to economic data. Penetrating below the surface of his doc—
trine, we find it is concerned with much more than psychic or sub—
jective phenomena in the historical sense. In short, We should
not dismiss a probability doctrine simply because it resembles an
unacceptable theory, or is itself unacceptable in some form.
We can not reject this type of probability interpretation on the
ground that it is impossible to measure belief as a subjective
event, although, as we have indicated, the stress here is Upon the
response phase. The assumption is that events are not probable
but certain, and that our reactions to events, in the sense of knowl-
edge or belief alone, are probable. Since there can be no proba-
bility reaction without reference to an eVent, and since degree-of—
knowledge situations are not only authentic but occur frequently,
we must regard them as a genuine type of probability Situation.
Moreover, it would be very diflicult to support the idea that such
probability situations do not occasion various system—building op-
erations.
From the interbehavioral or field standpoint there is nothing
mysterious about the connection of “subjective states” or beliefs
with calculation. The earliest probability studies have been inter-
related with calculation, especially the processes of combination
and permutation. We need but recall that all probability theory
consists of constructions. derived in some sense from crude data.
True, such 'constructions do not always adhere closely to the data,
in the sense that the final theory adequately represents the original
events. This is scarcely surprising, since probability data them-
selves are problematic or inadequate. For the most part, then, the
theories represent considerable autistic construction based upon
the manipulation of mathematical processes. The result is a tend-
ency to abstract from event factors, which nevertheless may still
be available despite their general problematic character.
One more point. All mathematical or calculative processes are
interbehavioral and in the final analysis go back to contacts with
events. The fact that these events maybe relations—and therefore
permit a formal treatment irrespective of objects upon which
to act—makes possible the traditional purely formal manipulations.
'..Ibid.,'\p. 273.
240 PSYCHOLOGY AND LOGIC

or the recent tautological and purely syntactic processes. Precisely

because scholars haVe historically regarded believing and knowing
as subjective states these psychological processes have inevitably
been condemned as a basis for probability determination. Hence
arose the powerful tradition that probability must be reduced to
statistical frequency. On the other hand, it is difficult to displace
the proposition that probability situations are limited to events
which occur frequently, or in general that probability is exclu-
sively concerned with collectives or series. This being the case, it is
no anomaly that such competent logicians and mathematicians as
De Finetti and Ramsey, clinging to a so-called subjective theory,
still contend that probability has to do with belief, and that we
are Obliged to measure such belief.
Important for the View of these writegs is the connection of
beliefs with readiness to act; that is, belief is regarded as meas—
ured by the readiness with which those entertaining beliefs act
upon them. Ramsey also proposes the wager conception, according
to which the willingness to wager serves as an index to the strength
of a belief or the intensity with which it is held. A favorable aspect
of this subjective View is that it does attempt to bring probability
down to actual human situations, an attempt
sey’s rejection of the convergence principle.
substantiatedby Ram-

The convergence principle may be described as a statement of

possible results in throwing coins. Thus, for example, it would
have to be established that “In n consecutive throws the number
of heads lies between n/: 2 €(n).”3 Ramsey, however, argues
that aside from the fact that no such proposition has been estab-
lished, it would be broken if enough instances were tested. He
goes on to say:
Nor is there any fact established empirically about infinite series of
throws; this formulation 'is only adopted to avoid contradiction by
experience; and what no experience can contradict, none can confirm,
let alone establish.
Unfortunately, Ramsey’s argument against abstract logical rela—
tions in favor of concrete events remains only an argument. How-
soever valid the strictures of Ramsey and other proponents Of the
subjective theory may be with respect to ap opponent’s theory,
3
Ramsey, Foundations, p. 206.
 ANALYSIS OF PROBABILITY SYSTEMS 24.1
they fail to support their own argument because'they are not really
interested in actual occurrences—namely, beliefs as concrete inter-
behavior with specific objects. Adherence to a single probability
formula is sufficient evidence. Moreover, even on mentalistic
ground there is a powerful objection to limiting probability ex-
clusively to inferential reactions, whether belief, judgment or any
sort of expertness.

(3) Probability as Laws of Things

Probability as Frequency of Occurrence. Frequency theorists
insist upon stressing the stimulus-Obj ect phases of probability
events. This fact is evident from the essential frequency formula—
namely, probability is the limiting value of the relative frequency
with which an event occurs in a class or series. Especially the early
frequency theorists—for example, Ellis,9 Cournot,10 and Venn”—
emphasized the necessity of keeping in contact with actual events.
Subjective theories were rejected; the restriction of probability cal—
culation to permutations and combinations was opposed. Authentic
probability problems were claimed to concern actual things ob—
served to happen in established series. Similarly, probability calcu—
lations were based upon the observed recurrences of events.
Because of its emphasis on things the frequency form of theory
is by far the most important and the most widely held. In fact, it
fits eminently into the culture of the present century, and sympto—
mizes the developments and achievements of modern science.
With the transition of science from the mechanical and historical
to the statistical phase, as Maxwell put it, this has indeed become
the century of statistics,12 and hence of probability. Still, we must
not overlook the differences between frequency theorists; for they
are animated by different attitudes concerning probability prob-
lems. Accordingly, we now consider three types of frequency
theory, leaving for later examination a fourth distinct form called
by the same name.
(a) Probability as Empirical Frequency. One of the earliest fre—'
quency formulations, that associated with Venn, was developed as
9
Foundations.
Exposition.
"1: Logic.
Merz, History, vol. II.
242 PSYCHOLOGY AND LOGIC

a criticism of the intuitive or rational mathematical theory. Em-

phasis was to be. placed upon experience or empirical records. The
theory is definitely an expression of the British Empirical philoso—
phy in opposition to Continental Rationalism. Instead of reac-
tions objects are stressed, and all probabilities are presumed to be
calculated on the basis of actual observations with similar ma-
terials.
Although this theory presumes to discover how things really
are, it attempts to avoid subjectivism. In this connection Venn
asserts:
Probability is not so extensive as that over which variation of belief
might be observed. Probability only considers the cases in which
this variation is brought about in a certain definite statistical way.13

As a generalized theory the frequencies involved are not spe-

cific and limited, such as those having to do with particular enter-
prises of individuals. Though the theory ostensibly takes its de~
parture from actual statistical frequencies it does not feature
specifically determined data, but rather a general explanation of
uniformities and divergences. In other words, it is an empirical
philosophical, rather than a scientific statistical, formulation.
Central to this theory is the conception of series. Probabilities
as over against certainties have to do with attributes, or classes of
attributes, the causes of which are numerous but not evident.
Probability series are defined as a combination of individual
irregularities and aggregate regularities. From an indefinitely
large number of terms representing an attribute a, one can differ-
entiate a smaller class of which the members possess a more re~
stricted attribute b. Now, choosing m members, in all of which l
belong to the less class, the probability of b assuming a consists
of the limit l/m, when m becomes indefinitely large.
General Series are Certain. Critics have objected that this is not
a probability, but a certainty, theory. Indeed, since the essential
background of Venn’s theory is the uniformity of nature, it misses
actual probability. The latter has to do not with ascertained uni-
formity or certainty, but merely with‘samples.
Neglect of Unique Occurrences. AnOther equally valid objection
to Venn’s formulation is that it does not allow for specific occur-
1’
Logic, p. 143; see also Edgeworth, Probability.
 ANALYSIS OF PROBABILITY SYSTEMS 24.3
rences, which are certainly among the most characteristically prob-
able events. It is in connection with such events that the notion
of probability is most valuable. Frequency theories which deal
with classes or repetitions of events are dbfectively exclusive when
nonrepetitive happenings are in question; as, for example:
. . . when we speak of the probability that the solar system is formed
by the disruptixie approach of two suns, or that the stellar universe
is symmetrical."
Others” who oppose Venn on the same ground admit that he
realizes the restriction.
The Paradox of Measurement. Aside from avoiding subjectiv-
ity, Venn argues that his theory allows for exact measurement. But
critics16 point out that on Venn’s basis even a judgment concerning
the tossing of a penny can not be made until it has been tossed
an infinite number of times. Accordingly, there is no case where
the value of a probability is known.
(b) Probability as Statistical Frequency. This theory is an out—
growth of the original empirical frequency formulation. Essen—
tially, however, it has developed out of the application of a proba-
bility calculus to the fields of life expectancy and to various scien—
tific problems such as thermodynamics, biological relations, and,
more recently, quantum mechanics. According to this view proba-
bility is'defined as the limiting value of the relative frequency
of a given attribute within a given collective.17 Furthermore:
The purpose of the theory of probability is exclusively the calcula-
tion of probability distributions in collectives derived by means of
given distributions in the initial collective."8
This type of theory frankly makes probability synonymous with
statistics. The bridge, of course, is the theory of error. As Shep-
pard puts it, the theory of error is the main link between the
theory of probability and the theory of statistical frequency:
It forms a large part of the former, and is a particular case of
the latter.19
Jeffreys, Scientific, p. 219.
Keynes, Treatise, p. 96.
'° Jeffreys, Scientific, p. 219.
" Cf. Von Mises, Probability, p. 38 et passim.
1"Ibid., p. 308.
'° Probability.
24.4 PSYCHOLOGY AND LOGIC

In contrast to the empirical theory the present version is based

upon the actual manipulation of data for ,specific scientific pur-
poses. Accordingly, the frequency limits are not known prior to
handling the data, nor are they rigidly set. Frequencies are pre-
sumably observed or arranged on the basis of the recurrence of
events.
Intellectual Probability a Form of Certainty. Like general em-
pirical probability this theory is subject to the criticism of aiming
at certainty—though a different certainty, to be sure. Here the
series or cOllective is derived from .actual calculations. But even
so, the development of series or frequencies is accomplished by
invoking. the normal law. If such a limit is really placed upon them
the charge of translating probability into certainty appears sup—
ported. Again, if the proponents of this view believe that prob-
ability can never be given a true value, but only an approximate
one based upon actual past findings—in other words, that prob-
ability is always inverse probability—then the pragmatic cer-
tainty charge is indeed substantiated.”
Statistical Probability Restrictive. It is perhaps valid criticism
Of this theory that by hypothesis it is restricted in Seope, and there—
fore not a theory of probability but only a type of probability con-
fined _to particular situations. Howsoever powerful this criticism
may be, it only holds against the proponents of the statistical fre-
quency theory, because they, like their opponents, re‘gard prob—
ability as a general theory, and not as one applicable to specific
operational fields.
Statistical Probability Only a Technique. Connected with the
above is the criticism that this theory is really only a technique for
the manipulation of certain events. In other words, it is no more
than a generalized principle of calculative processes—Of great im—
portance, indeed, in science—but not a general theory.
Statistical Probability Points Backward. Another argument is
that such a theory deals too much with recorded past events and
has slight, if any, bearing upon future actions or knowledge. The
counter to this criticism is that thereis no other‘way of predicting
or applying knowledge than on a frequency basis. But the question
remains whether the frequency theory does not set aside experi-
mental and other investigations which maybe much more effica-
See Sheppard, Probability.
 ANALYSIS OF PROBABILITY SYSTEMS- 245
cious for any or all scientific purposes, and which require neither
numbers large or small nor series of any sort.
(c) Probability as Mathematical Frequency. When at the outset
fixed limits are assigned to frequencies (whether or not such fre-
quencies are regarded as concerned with concrete occurrences), or
to factors in general mathematical systems, we have a type of
theory which reduces probability to mathematical manipulations—
either simple arithmetical, algebraic, or more complex analytic
processes. Indeed the proponents of this theory have adopted the
calculus of classical theory and added frequency as a substitute
for the subjective interpretations. In other words, coin tossing,
dice throwing, or some sort of fixed collective is used as a paragon
of, if not the exclusive guide to, probabilities. Though the pro-
ponents of this theory may believe that only by trial can proba-
bility ratios be ascertained, the fundamental principle is that the
factors involved approximate fixed limits, whether finite or in—
finite.
An illustration is von Mises’ argument that probability involves
stochastic, random, or casual variations. In coin tossing, for ex—
ample, if a certain Characteristic, say, heads, occurs with definite
regularity, no probability is involved. On the other hand, every
sample selected from the series must Show the same probability.
The probability is, of course, one half, and the larger the num-
ber of throws the nearer the frequency approaches this probability
fraction. Despite the randomness, the same result is obtained
when we number or order the throws. We may select (a) all
even or odd numbered throws, (b) all those the number of which
is a prime, (c) all throws after a head, or ((1) those following
two tails. _
The mathematical theory consists of two subtypes. The first
begins with‘ natural happenings, say, biological data as in the case
of much of Pearson’s Work, or economic materials, which are then
connected with the mathematical theory. The technique is, of
course, to apply a probability calculus. The second type is framed
by a mathematical setting. Examples are all those probabilities
stated in terms of collectives or infinite series.
Calculation Rather than Probability. An objection to a theory
of this sort is that at most it merely recognizes the existence of'
problematic events. Its main emphasis is on numerical manipula-
246 PSYCHOLOGY AND LOGIC

tion, which is believed to be best achieved by the probability calcu-

lus. But since this is a rigid instrument, based upon given premises,
the total result is not completely satisfactory. Even the stress of
series or numbers becomes an emphasis upon numbers, rather than
upon characteristics of events.
To overstress the repetitive and Serial character of happenings
may lead to the emasculation of a good principle. It may be carried
so far as to conceal genuine events and make probability entirely
a matter of the manipulation of mathematical variables. Accord-
ingly, the best check is to ask: Whence are the variables derived,
and how closely do our probability calculations keep in touch with
the data calculated and with the original problems basic to the
calculation.?
Certainly, empiricists and positivists like Venn, Jevons, and
Pearson adhere to a psychological and philosophical view that
enables them to reduce the stimulus objects with which presumably
they are dealing to their own calculations. Experience of objects
in any case reduces to states of mind. It is therefore only a step
from the experience of the numbers of things to
selves.
the things them—
(4) Probability as Propositional Relations~
Stemming from Leibniz and other rationalists of the 17th cen-
tury is the theory that makes probability a step in a rational
approach to certainty. Probability situations are inconclusive con—
ditions, as compared with mathematical ones where reason holds
full demonstrative sway. To probability arguments it is rational to
attach some weight, but not as much as to those arguments in which
our knowledge is certain. Such situations, however, come within
the province of logic which investigates the general principles of
valid thought.
Keynes, one of the main proponents of this type of theory, con—
fines his treatment of probability to a “general theory of argu—
ments from premises leading to conclusions which are reasonable
but not,certain.”21 As we have seen (p. 220), a symbol of this funda—
mental relation between premises and conclusive propositions is
a/h, representing the probability of a ronclusion a derived "from
the premise or hypothesis h.
2'
Treatise, p. 98.
 ANALYSIS OF PROBABILITY SYSTEMS 247
Keynes proposes the theory of propositional relations as a com-
plete theory, presumed to deal with unique events as well as with
repetitive ones. It derives this power from its emphasis of propo-
sitions, rather than of things or beliefs. According to the logical
theory, probability is independent of actual events; it is a unique
relation, an indefinable concept. Probability, then, must be treated
as an absolute relationship between propositions.
On the same basis Keynes regards probability relations as
thoroughly objective. This means that the relations are independ—
ent of any observed events. To justify this peculiar doctrine
Keynes falls back upon the notion that knowledge is of two sorts:
(I) direct—that is, based upon contacts with things, and (2) logi-
cal—namely, argument. Though Keynes frequently refers to.
the occupation of workers with things and with knowledge of
things, and though he discusses evidence for propositions, in the
final analysis he retreats to a Platonic position.
Since no emphasis is placed upon things, and frequencies of
actual occurrences are minimized, the primary technique of deal-
ing with probabilities is a process of intuition or a priori reason-
ing. Furthermore, Keynes forgoes exact numerical statements of
probability, satisfying himself merely with the absolute relation-
ship between propositions which state the conclusion and, presum-
ably, represent the evidence. The measurability of probability for
Keynes, therefore, has to do with the numerical relation between
the degrees of rational belief in one of a pair of conclusions rela—
tive to the other. Similarly, the simple comparisons equal to,
greater than, and less than are concerned more with comparable
rational beliefs than with definitive relations of conclusions and
hypotheses.
Logically Intuitive. A fundamental criticism of the intuitive
type of probability formulation is that it is impossible to deal with
such unavailable processes as intuition. It may be argued, however,
that such criticism applies only to the form of Keynes’ statements,
for in actual practice he himself asserts that what he means by
intuition is merely direct recognition of the truth or validity of
propositions. These direct intuitions are used as means toward
some cognitive end. For example, he uses the illustration of em—
ploying trigonometry to determine the position of an object, not
by direct observation of it, but by observing some other object
248 PSYCHOLOGY AND LOGIC

together with certain relations. This he regards as a type of in-

direct method.
Furthermore, one might say for this theory -._that it does place
considerable emphasis upon the expertness of the judge. In other
words, more than other theories it gives a place to the reactor in
a problematic situation. Of course, to grant this point means to
'modify its Platonic character and bring it closer to actual problem-
'atic situations. ∙
Nonmeasurability. The View that probabilities are not measur-
able provokes serious critical attacks. In the first place, it has been
argued—by Jeffreys,22 for example—that one may believe proba-
bility is a unique indefinable relationship, and yet give each
probability a numerical value. Other objectors point out that if
probabilities are not numerical or measurable there is a complete
hiatus between probability and scientific situations—in other words,
such probability theory is said to be philosophically abstractive
and unrelated to scientific situations, as well as disconnected from
statistics.
N oneventual. 'Another serious Obj ection to the propositional re-
lations theory is that it stands remote from concrete events. For
one thing, relations between propositions are independent of any
observation, and are presumed to hold regardless of any actual
state of affairs.23
Propositional Relations are Linguistic. Again, this theory is
clearly a system of linguistic items. Presumably the propositions
refer to, or stand for, actual events, but there is lacking a safe
bridge between the interrelated propositions and events.
A final comment. Despite all the weakness of the proposition~
relation system, we must not overlook its merit as a check on other
theories. Certainly we may grant its value in demonstrating that
the probability domain is large enough to accommodatemore than
one system. Furthermore, it is of some importance to question the
wisdom of exclusively occupying oneself with a frequency system.

(5) . The Truth-Frequency Theory

The present theory, which may be traced
back to Peirce, origi-
nally was an outgrowth of Venn’s so-called mathematical or sta-
∙
I
22 Scientific;Theory.
See, Nisbet, Foundations.
 ANALYSIS or PROBABILITY SYSTEMS 249
tistical doctrine.24 The claim is made for it that it interrelates
objects with propositions concerning objects. Peirce may well be
expected to think in terms of scientific situations, even though he
occupies himself with abstract systems rather than concrete enter—
prises. The following represents his more concrete attitude:
The inference from the premiss, A, to the conclusion, B, depends
. on the guiding principle, that if a fact of the class A is true, a
fact of the class B is true. The probability consists of the fraction
whose numerator is the number of times in which both A and B are
true, and whose denominator is the total number of times in which
A is true, whether B is so-or not. Instead of speaking of this as the
probability of the inference, there is not the slightest objection to
calling it the probability that, if A happens, B happens.25
Peirce, however, was essentially a mathematical logician. Accord-
ingly, we do not find him cleaving too closely to events, or even
to their numerical representation.” For this reason he constantly
strives to preserve the rational or mathematical factors in proba-
bility theory. Both Peirce’s empiricism and rationalism appear in
the View that probability relates to the future27 and that great
emphasis must be placed Upon propositions in the probability
situation. His ideas are excellently combined in the definition that
probability is the limiting value of the truth frequencies arising
from the application of leading principles. Probability propositions
are compared on the basis of the relative frequency with which
they actually do turn out to be true, and “truth consists in the
existence of a real fact corresponding to the true proposition.”28
Though the truth—frequency theory is undoubtedly an improve—
ment over its Vennian prototype, it has lately been worked over
in an appealing way. Nagel29 especially has attempted to articulate
it with recent scientific situatiOns, as well as to free it from some
of the objections made against the earlier, or material—frequency,
theory.
In particular, it is claimed that the improved truth-frequency
2’
C.P. 2, 651.
2‘
Ibid.
C.P. 2, 101.
2'
C.P. 2, 66.
2"
C.P. 2, 652.
29 Nagel, Frequency.
250 PSYCHOLOGY AND LOGIC

theory is less restricted than other frequency theories. While it can

take over the mathematical method and apparatus of the statistical
and mathematical frequency theories it can also deal with inde-
terminate cases. Again, it is not restricted to the future, but can
also treat of. the past. And, finally, because the emphasis is upon
propositions instead of events, it can handle the probability of
single events on the basis of the relative truth—frequency of propo-
sitions concerning the evidence for the original proposition. On
the whole, the truth-frequency theory approximates an opera-
tional vieWpoint.
Not Operational Enough. N evertheless it does not go far enough
in an operational direction. Even though it assimilates several
originally independent theories, it does not escape the stigma that
attaches to a doctrine presumed to be universal and complete. To
be sufficiently operational a theory must be able to handle specific
situations as they occur. N o matter how inclusive and comprehen-
sive the description of a process must be, it is not permissible to
dissipate the original referents in the descriptive or referential
process.
The criticism, it will be said, is admitted and hence ofsno force.
For at bottom the truth—frequency theory deals with inferences.
But how are we to understand inferences? It is our contention that
inferences are forms of interbehavior with stimulus objects; con-
sequently, one can not get away from specific events. Inferences
are not detached propositions, though they must always be lin-
guistically referred to or described.
Why Stress Propositions? After all, why Should we make proba-
bility a matter of propositions rather than of events? Even if we
properly distinguish between (a) propositions as intellectual con-
structs and (b) sentences as symbolizing or referring instruments
there is no advantage in completely bypassing events. And if we
are dealing with propositions we are in touch with events, albeit
indirectly. Whether or not this is suflicient depends upon specific
problems.
When, on the other hand, we are operating with sentences, we
run into two different sorts of problems. We may be merely fol—
lowing the conventional logical tradition which substitutes lin-
guistic factors for various psychic states—for example, (a) con-
cepts or ideas (terms), (b) judgments (“propositions”), and (c)
 ANALYSIS OF PROBABILITY SYSTEMS 251
inferences (syllogisms). Thus we confuse our propositions with
our materials, for the handling of Which we construct propositions
to begin with. '

By no means do we object to the construction of a probability

theory covering the testing and comparing of propositions. Such
a theory, however, should not be permitted to run' over into situa—
tions for which it was not designed. In general, it would be con—
cerned with conditions for formulating propositions and applying
them to data.
Incidentally, a theory of this nature would be brought heavily
to bear on the crippling doctrine that a sentence is a basis for an
inference without regard to how it was constructed. This deducti-
bility property of sentences, tied up as it is with syntactistic doctrine,
can have no meaning whatever except in mathematical situations.
And there the manipulation of symbols in given systems is only
possible because the symbols are known to stand for strict relations.
Do Propositions Overcome Subjectivity? In favor of the propo—
sitional formulation is the argument that it attempts to overcome
subjectivity. But can we by this method hope to avoid the difli—
culty of subjective probability theory? In the first place, does not
the propositional or linguistic theory misinterpret the subject mat—
ter of probability by making it exclusively into truth-frequencies,
while at the same time it confuses a person’s operations upon ob-
jects with. mental states? Insofar as this is true the theory serves
to perpetuate the mentalistic doctrine. In'the second place, this
theory is contradictory in rejecting psychological processes and at
the same time standing as a theory of inference. Only when an
interbehavioral view is adopted can this contradiction be resolved.
In that case, however, the theory would not constitute exclusively
either a frequency or a propositional system.
Inapplicable to Unique Events. The claim that the truth-fre-
quency theory can deal 'with unique events is not really sub—
stantiated. If we are concerned with a specific situation—say, a
medical prognosis—in what sense can we test the frequency of
application of certain propositions? The attempt to avoid this difli—
culty involves one or all of three procedures.
First, resort is had to the covert definition of frequency as con-
tact with actual conditions. While frequency is described as the
statistics of science and as an objective ground for measurement
252 PSYCHOLOGY AND LOGIC

and prediction, it is still assumed that particular contacts with

events are being dealt with. To interbehave with single-occurrence
situations is in no scientific sense objectionable, but why stress the
frequency construction? There is danger, on the other hand, in
emphasizing numbers and repetitions if by so doing one interferes
with successive investigative approximations.
A second procedure is to shift from frequency of occurrence to
frequency of evidence of a single occurrence. Here, frequency
signifies testimony, which may or may not be a record of occur-
rence. Such conditions are sometimes useful, but note that we
are dealing with different kinds of frequency in the two cases.
It is an advantage, however, if we observe. the necessity of con—
sidering different kinds of probability events.
The third procedure is. to shift to phenomenal classes so that
we can carry over from one event to another. This means an im-
mediate shift from a frequency to an operational theory. Natu-
rally, if there are resemblances we can draw inferences concerning
event a from observing b, c, etc. The emphasis, however, is on the
possibility and validity of the inferences. On the other hand, when
we cease to strive for numerical ratios we are able to deal with
various situations as they actually occur, rather than revert to
some modified form of rational or a priori theory.
Nonnumerical Frequency. A serious question concerning the
truth-frequency theory is how it can square with the lack of nu—
merical coeflicients, as in the caseof propositions concerning events
without numerically significant confirmation. This query is espe—
cially pertinent when it is admitted that probability propositions
are only theoretically capable of being numerically formulated.
In some cases it is even granted that only a general confidence in
the. value of truth frequencies is available.
For this theory numbers, then, turn out to be goals not always
attainable. It simply yields to the convention that numbers are
important and powerful. But this is to forget that the first elabo-
rate number or calculation techniques were developed by means
of the classical theory. Indeed, numbers or calculations are merely
techniques and are only necessary and \applicable where they can
be used. Moreover, numbers are frequencies only by definition.
The numbers 0 and I, or even 2, can hardly be connected with
 ANALYSIS OF PROBABILITY SYSTEMS 253
frequencies. Numbers may well be regarded as characteristics, if
not qualities, of events or things.
We Should not conclude that the lack of numerical probability
weakens the theory. On the contrary, if that lack helps to direct
it toward the goal of concrete investigation it distinctly gains in
value.
(6) iProbability as
Factorial Coordination
A unique probability theory concerns combinations of field
factors without regard to the reactor’s observation or formulation
of them. This theory deals only with the interbehavior of factors
of the sort we may presume to occur in such singular events as
the formation of a solar system, the birth or death of an individual,
the eventuation of an electron, etc. The emphasis is upon the
collocation of factOrs which in one combination (absence, presence,
or pattern)'constitutes 'a different state of affairs from another.
Many varying combinations may occur. Probability situations from
the present standpoint involve comparative or relative possibilities,
and may be associated with experiential and properly interpreted
causal circumstances.30
A simple illustration of two situations sharing some features
and not others, with different resu1ts, may be found in life ex-
pectancies of different individuals. It may be true that a'has a
greater prbbability of living to be 75 years old than has b. The
basis? Given equally favorable biological constitutions, as well as
comparable living habits, a appears to be heading toward a profes—
sion more favorable to a longer life than b. At least so the insur—
ance company decides.
Probability Requires Response. The factorial-coordination con-
struction immediately arouses the anomaly objection. Probability
theorists are extreme in their predilection for some. phase of re—
sponse as over against the stimulus object. They stress the making
of sentences or propositions, counting, evaluating, and believing.
This predilection in turn is based upon traditional doctrines con—
cerning the power of knowing to engender things. At least, exist—
ence is made to depend upon knowing (p. 308f).
Because probability as an obj ective‘set of possibility conditions
3°
Cf. chap. 19; also Ducasse, Neglected, and Struik, Foundations.
254 PSYCHOLOGY AND LOGIC

requires someone to know, count, or evaluate the probability—if

it is to be known, counted, or evaluated—does it! follow that proba—
bility does not exist without the responder? Does not our very
approach to things and events in order to enumerate, estimate, and
evaluate them imply their independent and prior existence? Inde-
pendent probability events can not be denied on the basis that
assertions about them can only be made on the ground of how
frequently certain signs have been observed to connect with par—
ticular eventualities.31 _
Factorial Coordination not Reducible. To refer to a probability
situation suggests that one is reSponding to an independent oc—
currence which in no wise depends upon anyone’s manipulation
or knowledge. Moreover, one’s later acquaintance with some-
thing happening or not happening need not involve frequencies
or even classification, but simply an analysis of field factors. For
two reasons we regard as futile the attempt to equate knowledge
of, and ability to analyze, a situation with a frequency formula.
First, it is intellectually reprehensible to assert that one has
either.frequency or nothing, it is quite possible that an individual
might have a general acquaintance with events bearing upon a
unique case. Even'in the form of frequency assertions it is im-
proper to make probability situations into absolutes or certainties.
From an objective standpoint, reactions, whether beliefs, asser-
tions, _or predictions, constitute definitely occurring events even if
they happen only once. As a matter of fact, a probability situation
may be similar to other cases and still be unique. One does not
attempt to predict a long or short life for an individual upon no
basis at all, and this basis can be regarded as an accumulation of
observations; thus any present probability situation‘ is part of a
collective or series.
In the second place, we must acknowledge the hypothetical
character of probability judgments. Because of the uniqueness of
the situation the predicter makes a hypothetical statement, in
other words, he asserts his belief on the basis of the assumed
copresence of certain factors'1n the situation. He may be wrong.
This very fact implies that he'13 attempting to react to events
independent of himself, rather than on the basis that the events,
their probability, exist because of his knowledge or assertion.
Nagel, Frequency, p. 546\f.
'1
 ANALYSIS OF PROBABILITY SYSTEMS 255
Factorial Coordination Specific. Not the least appealing feature
of this construction is its lack of comprehensiveness. Because it is
a formulation with a genuine emphasis of specific events it con-
trasts with inclusive formulations implying universality of appli-
cation. We regard it as a distinctive merit of this construction that
it stresses probability events which owe their fortuity to their
complexity and fluidity. While no probability theory can be in-
clusive and final it would be a distinct deprivation to probability
logic not to have a place for this formulation among the others.

PROBABILITY SYSTEMS AND LOGICAL MULTIPLEXITY

Our analysis of probability systems bears out the View that
probability is a multiplex domain of system building. Even if we
confine ourselves to general formal systems, to the exclusion of
specific operational enterprises, there are too many irreducible
varieties to warrant any universal formula.
We conclude further that probability systems can not be as—
similated to the older conventional formalistic logics. Nor can
we construct universal probability systems to replace those logics.
In other words, we can not, for example, replace a two—valued
logic—true—false—by a single many-valued probability system.
A unique form of logical universalization is to assert that there
is a general‘ type of probability logic, though it'has a limited field
—for example, science. The proponents of this view agree that the
logic of science is of the probability sort, in contrast to the logic of
truth which reigns in the domain of mathematics and other ab-
stractionistic locales. While this partially universal View is prefer—
able to a completely totalitarian one, it still stops short of a speci—
ficity system in that it verbally erases the lines marking off differ-
ent problems, different subject matter, and different modes of
investigation. The defect is not mitigated by proposing the idea
of probability deduction. Doubtless, probability may be inferred if
not deduced, but neither deduction nor inference constitutes a
criterion of logic.
Can probability be assimilated to a universalistic logic on the
ground that systems in themselves are problematic? No: In the
first place, not all systems are problematic. In an immense number
of cases, either because systems are arbitrary or practical, no such
question arises. In the second place, the question whether systems
2 56 PSYCHOLOGY AND LOGIC

are problematic really refers to the investigation of system build—

ing rather than to system building itself. The investigation,'then,
is the stimulus object reacted to, and the probability question con—
cerns work upon" systems. Such an investigation may well be a
nonlOgical enterprise. Logic as system building may be sharply
distinguished from (I) the investigation of the degree of know-
ing, (2) the basis for calculation, or (3) the character of events
entitling them to a place in a system.

PROBABILITY, EPISTEMOLOGY, AND ONTOLOGY

Every probability theory or System carries within it the philo-
sophical presuppositions influencing the constructor. This is a fact
we can not deplore, unless we immediately assume that all phi-
losophy is objectionable. Such an assumption is wholly unwar—
ranted, since the specifications for an acceptable philosophy are
available. They may be summarized as follows. (1) Philosophy
constitutes a set of attitudes developed by an individual in contact
with a considerable area of things and events. (2) Its propositions
depart radically from uncontrolled magnification and multiplica-
tion. (3) The philosophical system remains rOoted in the actual
affairs from which the promulgator builds it, and inhis measured
extrapolations from them.
AS it happens, however, most writers who attempt to explain
the purpose, power, and scope of probability indicate their en—
tanglement with traditional ontological and epistemological prob—
lemsaz—for instance, when they adopt finite and infinite collec-
tions as their operational fields. We have already referred to the
metaphysical determinism implied in the Laplacian formula—
tions (p. 218).
Acceptable philosophical presuppositions for structuring proba—
bility favor systems for handling specific situations and exclude
world views such as are implied in the following:
If I have seen just three crows and they were all black, then it
is certain not only that my next crow \will be black but that all the
crows I shall ever see will be black.33
3'
See, in this connection, A Symposium on Probability, in Philosophy and Flu-
_nomenological Research, Part I, 1944-45, 5, 449-532; Part II, 1945-46, 6, 11-86;
Part III, ibid., 590-622.
Williams, Ground, p._ 54.
 ANALYSIS OF PROBABILITY SYSTEMS
257
A probability system based on such specifications as we have men-
tioned above might be set up in the crow situation, but it would
lack completely any absoluteness concerning either all the possible
animals to be met or the validity duration of the system.
General Philosophical Problems. A few samples of general
philosophical problems inherent in probability discussions will be
suflicient. We have already had occasion to notice how seriously
probability students consider the question of the a priori or' 'a
posteriori character of probability inferences. There are also prob-
lems concerning faith in belief and judgment. Such principles as
nonsuflicient Reason or Indifference are certainly epistemological,
while such laws as those of Large Numbers are not only onto—
logico—metaphySical, but may be mystical as well. Here, too, are
located such puzzles as Euler’s attempt to account for his firm
confidence that the stones in the Magdeburg Church were heavy
and would fall unless supported. Jevons“ quotes with approval
Euler’s assertion that it would be impossible to fix on any one
thing really existing of which we could have so perfect a knowl—
edge as to put us beyond the reach of mistakes.
Special Philosophical Problems. Among the more special issues
we number all the attempts to relate probability situations with
traditional nominalism, conceptualism, and realism. We include
here also those efforts to bring independently developed proba—
bility calculations into relation with actual events.
The procedure of philosophizing probability encounters the
serious indictment of” intellectual treason. All writers pay lip serv-
ice to the principle that one should above all investigate and
systemize events. But as soon as they encounter probability hap—
penings they smother them in philosophical principles. Thus,
instead of making authentic probability studies available for philo-
sophical reflection such writers impose illegitimate philosophical
notions upon probability events. The result: Events are distorted
and misinterpreted.

INTERBEHAVIORAL PROBABILITY VERSUS EPISTEMOLOGY

When a naturalistic probability view is adopted, such problems
as how one can have confidence in one’s judgments, how to be

Prinbiples, p. 2 38.
3‘
258 PSYCHOLOGY AND LOGIC

satisfied with one’s beliefs, and how to develop expertness in pre-

diction are all easily solved in terms of an individual’s operation
with stimulus objects. Preoccupation with concrete interbehavior
makes it easy to see which knowledge events have been trans-
formed into epistemological issues. Is not the construct of an
a
organizing and unifying mind merely bad description of an indi-
vidual carrying over judgments from one situation to another?
Is Keynes’ intuition anything more than the practical expertness
of one who can estimate the relation of hypothesis to evidence on
the basis of his prior contacts With similar and dissimilar situa-
tions?
are.
On.an interbehavioral basis inferences made on the ground
of contacts with things, even though a propositional or linguistic
construction may be necessary in the process. From this stand-
point the a priori is merely a term for practical expectations em-
ployed as a ground for inferring from one behavioral situation
to another.

INTERBEHAVIORAL PROBABILITY VERSUS ONTOLOGY

as
Similarly, when we eliminate such illicit constructions meta—
physical chance and determinism, ontological questions are
brought down to events. Probability formulae and calculi, for
example, either refer to events or are employed as tools for ma—
nipulating them. 'When we deal with Specific happenings we
lack no criteria for distinguishing between (I) stimulus objects,
(2) responses to them, and (3) the interbehavior from which
both are abstracted.

PROBABILITY PROPOSITIONS VERSUS PROBABILITY EVENTS

Like all intellectual enterprises probability studies have borne
their Share of damage inflicted by the recent symbolistic revival.
Not only philosophers but scientists have heeded Goethe’s injunc-
tion:
. haltet Euch an Worte!
∙ ∙ ∙ ∙
Mit Worten lasst sich trefflich streiten,
Mit Worten ein System bereiten.\
And in consequence they have disguised sentences, syntactic sym—
 ANALYSIS OF PROBABILITY SYSTEMS 259
bols, and symbol systems in an efiort to' make them look like
things and events.
Specifically, probability students have been influenced to build
ultimate probability systems in order to eliminate events from
the probability field. Events are replaced by propositions.
Keynes?5 traces back to Ancillon36 the first suggestion to depart
from events. This view is stressed also by Boole,37 and shared alike
by Czuber38 and Stumpf.39 Boole, however, regarded the nominal-
istic substitution of propositions for events as only a procedure for
logical purposes. Other writers insist, however, that questions of
probability always involve factual, in addition to purely logical,
considerations.‘10
Probability studies have suflered greatly from symbology by
being thrown into the epistemological arena of N eo-Nominalistic
metaphysics. According to linguistic (semantic, syntactic) meta-.
physics, one transforms things, events, and relations into proposi-
tions or sentences. This movement has gone so far as to absorb
not only the stimulus events with which scientists operate, but
also the scientists’ operations upon (interbehavior with) them.
Symbolic philosophers overlook the fact that in the. end all
knowledge and all description of knowledge depend upon the
existence of things with which to interact.41
Though linguistic philosophers do distinguish between thing
language (semantics) and pure language (syntactic formations
and transformations), they still substitute descriptions for things
described, worse, they overlook that things exist independently of
descriptions. To determine the probability or improbability of
events is not an epistemological or a linguistic problem, but one of
observing and estimating the conditions under which certain
events do or do not occur.
At this point we must distinguish sharply between (a) actual
probability operations and (b) the linguistic descriptions of, and
references to, them. This distinction is tellingly illustrated by a
a"Treatise, p.
"Doutes.
5..
Laws, 7, 167.
3'
Warscheinlichkeitsrechnung, vol. 1, p. 5.
'9
Ueber.
‘° Nagel, Frequency, p. 541.
The sliift away from linguism is taking on momentum. See Feigl, Existential.
26o PSYCHOLOGY AND LOGIC

controversial exchange between two of the participants in the

probability symposium .mentioned earlier. Whereas Carnap"2
characterizes von Mises’ frequency theory as purely mathematical,
not physical, the latter43 is surprised that Carnap, whose whole
recent stress has been on empty syntactic sentences, should add
an inductive logic to his extremely deductive System.
The importance of the above distinction merits another illustra-
tion. When linguistic logicians face probability problems, as they
inevitably do, they naturally run into problems of scientific
methodology. They make a plate for themselves by assuming that
their task is to make “rational reconstructions” of scientific pro—
cedures. But that they soon identify their rational reconstructions
with ‘those scientific procedures is exemplified by their assertions
that the constructions they develop are of importance to the
methodology of empirical science“ and to practical life.45 Observe
carefully, however, that no description of, or reference to, a thing
absorbs thatthing (p. 259). Also we can avoid a “rational recon-
struction” when no actual thing is being worked on.46
We can not urge too strongly the full consideration of statisti-
cal inference and basic elementary manipulations'as mOdelS of
probability situations. A considerable literature is developing
which illustrates correct probability operations, though, unfortu-
nately, the metaphysical curse is not entirely lifted.“7 The way to
avoid this metaphysics may be to keep close to the kind of sta-
tistical operations employed, for example, in quality control or
other concrete situations.“
Inevitably We return to interbehavioral operations. An excellent
analogy is D’Alembert’s organization of dynamics by introducing
a factor of inertia. This is tantamount to reducing dynamics to
statics. The question here is whether some useful scientific pur-
pose is served by this procedure. Since the mathematician does
Remarks.
*3
Comment.
“ Hempel and Oppenheim, Definition.
‘5
Carnap, Probability. ∙
'0
Bergmann, Comments.
‘1
Churchman, Theory, Probability.
Butterbaugh, Bibliography; Neyman and Pearson, Contributions; Shewhart, Eco-
‘3

nomic, Statistical; Wald, Principles. \

 ANALYSIS OF PROBABILITY SYSTEMS 261
not interfere with the dynamic processes, but simply develops a
form of reference to them which 'is useful in further concrete or
referential reactions to the original data, a satisfactory relationship
exists between the formulation and the thing formalized. In this
argument lies justification for Setting up probability formulations
on the basis of infinite series: one does not thereby interfere with
problematic events, nor with the descriptions designed to handle
them.

PROBABILITY CALCULATION: GOAL OR OPERATION?

The undesirable authoritarianism and universalism which
characterizes traditional logic has cast its net over probability logic
by means of the probability calculus. Numeralization and calcula—
tion have become touchstones, even ends in themselves. Only
where there is a calculus is there authentic probability. Not only
the proponents of frequency theories, but also those who hold to
the classical or Laplacian view, practically define probability in
numerical terms. On this basis probability becomes closely inte—
grated with statistics.
Granting that nothing escapes statistics where numbers of things
are dealt with, and also that the subject is highly important where
it is applicable, it is still necessary to consider whether statistics
is identical with probability or merely an essential feature of it.
But before going further into this question let us look into the
nature of statistics and calculation.
Statistics and Probability. Statistical systems have evolved from
something very unlike the calculative manipulation of numerical
data. As an enterprise, statistics began as political arithmetic, as
the enumeration of Vital records. In the form of enumerated items
statistics involved no probability problem; the primary purpose
was to specify the changing conditions of life. With the accumu—
lation of such records questions of significance inevitably arose.
Purely calculative questions concerning the relation and organiza-
tion of data came to the front. Later, more fundamental problems
relating to prediction were precipitated, the outcome of which was
the construction and application of probability theories. Thus,
influences, correlations, and, if possible, cause-and—effect conditions,
have been established as the significant features of statistical data,
262 PSYCHOLOGY AND LOGIC

We might say that originally statistical data were concerned with

the present and past, Whereas an interest in the future introduced
probability.
With the development of a semitechnical interest in building
systems involving, on the one hand, present and past events and,
.on the other, future contingencies, probability and statistics come
together. This coupling of the two disciplines forces to the front
a number of their common properties.
Starting from the side of probability as a general enterprise
for operating with particular happenings, it is inevitable that the
statistical technique in the form of calculation processes should be
utilized. The tendency is to depart immediately from actual
events and to jump to numerical abstractions as units and quanti—
ties. Thus statistics and probability areidegtified, and probability
definitions are made to center around frequencies of various sorts.
Probability then becomes a measure of frequ'éncy, or, as in the
most popular definition, probability is the limit of the relative
frequency of events in series.
Undoubtedly, the possibility of being able to calculate more or
less accurately the probability of an event, or the probability value
of a proposition, adds definiteness and value to a probability situa—
tion. But are we interested in the probability, whether or not it
yields calculative results, or are we limiting our probabilities to
calculative situations? Possibly we can learn more about proba-
bility by adhering to probability fields than by restricting our—
selves tO any sort of defining condition, no matter how attractive
the latter condition may be.
Again, we ask that most crucial of questions: Is the operator
keeping in touch with actual happenings? Is he expertly summing
up and predicting conditions on the basis of his prior contacts
with similar situations? A secondary question is whether the situa—
tions involved lend themselves to calculative treatment, or
whether the probability concerns merely a relationship between
a certain type of event and other similar happenings.
The Nature of Calculation. Calculation consists of the manipu-
lation of relations. Now it is possible\to. emphasize the relations
of observed facts or of abstracted and formal relations. In general,
calculation consists of such interbehavior withrelations as to indi-
cate their similarity or equation. Or it may
ini‘folve enumerating
 ANALYSIS OF PROBABILITY SYSTEMS 263
items or elements in a field—that is, ordering elements on the
basis of given criteria—or ascertaining the quantitative determina-
tions of elements.
The question arises, then, whether our calculations are features
of concrete situations or whether calculations are the exclusively
formal ordering of relations. Calculation is fundamentally a type
of interbehavior: for particular and definite purposes. Can we,
therefore, make calculation a criterion of our interbehavior, or
should we regard interbehavior as bounded by certain conditions?
To make calculation the criterion of probability immediately
prej udices the case. Probability is a condition of facts discoverable
by observation, not a result of calculation.
Above all, it is illegitimate to confuse calculative constructions
with obj ects or relations. The latter may be, entirely independent
of calculative operations. The specificity principle forewarns us
to keep such situations distinct from those in which the data or
relata are themselves constructions. In the latter type of situation,
of course, we may regard calculation as the probability criterion.
Those who make the calculative procedure the criterion of
probability frequently rely upon the actuarial situation as their
guide, and point out the truth that statistics can not predict any-
thing about a single case. But must we take the insurance or
tistical situation as the probability model? In other words, are we
sta—

required always to deal with collections or series? If we are inter—

ested1n the nature of probability we shall have to consider thatIn
important circumstances we may want to know the life expectancy
of a particular individual—for example, a prospective employer or
employee, a company or institution.
Even insurance companies do not limit their investigations to
the actuarial department. Nor do they pay attention only to
generalized numbers or averages, except possibly to calculate
mathematical expectations. For the most part, the acceptance of a
risk may be regarded as a scientific investigation of a particular
person. The medical examination serves as a biological investi—
gation, an economic and sociological one is made by inquiring into
the habits and occupation of the insured individual.
The extensive development of statistical methods suggests their
indispensability for achieving certain ends—in other words, for
interbehaving with certain happenings. The psychologist, physi—
264. PSYCHOLOGY AND LOGIC

cist, biologist, or economist who concerns himself with statistics

and the laws of probability is primarily interested in psychological,
physical, biological, or economic events, and it is an accepted
canon‘ that the calculative procedures he employs must be suitable
for the ascertainment of the kind of information demanded by his
scientific specialty.
Collectives or series, therefore, are constructed, and are not
metaphysically unique and autonomous. They may be summaries
or organizations of particular kinds of objects, or they may be
constructional in the sense of creations more or less arbitrarily
and analogically designed.
 CHAPTER XXIII

LOGICAL ASPECTS OF MEASUREMENT

THE PLACE OF MEASUREMENT IN LOGIC

EASUREMENT implies system. Measuring - operations
typify the entire system—building enterprise. At the very
least, mensurational situations require operations to interrelate:
(a) things measured, (b) measuring behavior, (c) measuring in—
struments, and (d) standards of measurement.
Thus the topic of measurement fits perfectly into the logical
domain. In fact, metrology as the science of measurement is a
phase of systemology, while mensurational practice implies men-
surational systems. Even those logicians who make use of meas—
urement materials to justify a metaphysical rather than a system—
building theory acknowledge this fact efiectively, if not explicitly.
Indeed, for many metaphysical logicians mensurational operations
provide the only link between their arbitrary constructions and the
objective events which alone lend significance to their labors. How
an author treats the topic of measurement may well provide a
criterion for the appraisal of his logical doctrines.

WHAT IS MEASUREMENT?
Although mensurational operations are readily inspected, few
attempts have been made to describe them. Even these few Show
disagreements and contradictions which can only be explained by
the greater influence which traditional philosophical ideas have
exerted upon measurement theory, as compared with mensura-
tional practice. Measurement described on the basis of what is
done in measuring situatiOns consists essentially of a series of sys—
temizing operations for ascertaining precisely the properties and
relations of things and events. Formally, measuring operations
constitute the organization of criteria, units, andstandards, and
the observation of the relative positions occupied by items with
respect to the scaling instruments. The emphasis must be placed
on thessystemizing- of, or structuring operations upon, things.1
In the Stinse of Stimulus objects, cf. vol. I, chap. 5.
266 PSYCHOLOGY AND LOGIC

We reject both of the following views: Through measuring opera-

tions one can (a) discover the “basic” (quantitative) properties of
things, and (b) interact with transcendental magnitudes.

IDEOLOGICAL TRENDS IN MEASUREMENT THEORY

Current measurement theories manifest a series of unsatisfac-
tory constructional trends.
First, measurement is assumed to be a single type of fact. As a
consequence, whatever things are to be measured are forced to
fit one exclusive set of mensuration principles. That this is an un—
satisfactory assumption has long been appreciated, as indicated
by the metrologist’s struggle with the problem of intensive and
extensive properties. More recently, the development of micro-
scopic or atomic physics, with its indeterminacies (pp. 31 If.,3I9f.),
has precipitated theIssue concerning radical differencesIn meas-
uring situations.
Secondly, measurement theories by being closely assimilated to
philosophical views are decidedly colored by Idealistic, Realistic,
Conventionalistic, or Pragmatistic ways of thinking. It is not un—
usual to find measurement theory complicated with the'problem
of the existence of an external world, and with the belief in the
existence of invariable relations independent of those wishing to
measure or approximate them. In general, philosophical construc—
tions tend to supplant the actual measuring operations performed in
both nonscientific and scientific situations. Philosophic theories,
therefore, simply analogize and rationalize measuring operations.
Thirdly, measurement theories for the most part are confused
with the descriptive formulations of particular measuring opera-
tions (p. 307). Generally, metrological ideas are limited to classical
dynamics, even kinematics. Although it is realized that larger—
event domains must be taken into account, the ideal pursued is
derived from early sources of mensurational ideas. Accordingly,
there is little relationship between (I) the theoretical construc—
tions and (2) the measurement operations of everyday life and
science. And so, even if the theory dOes articulate with some fea-
tures of mensurational behavior, it hardly Offers a sufficient cov—
erage of the facts involved.
Finally, most, if not all, mensurational‘ systems imply some
sort of dualistic psychological view. When a
scientist writes that
 LOGICAL ASPECTS OF MEASUREMENT 267
“our primary data are sensations” and that when we say “we ob-
serve an obj ect” we are really saying that we have a series of sen-
sations coordinated by forming the concept of an Obj ect, he holds
views that carry within them more than mere potentiality for dis-
torting the events handled.2

MEASUREMENT AS APPLIED MATHEMATICS

The universalistic and philosophical trends just mentioned, as
well as the whole procrustean enterprise of forcing mensurational
facts to fit a theory, are well illustrated by the popular doctrine
that measurement is applied mathematics. Measurement is thus
regarded as a process of interrelating numerals or numbers with
things. This formulation has a definite historical background in the
triumphal achievements of mechanics during the interval from the
Renaissance to the I 9th century.
Since Helmholtz was an early formulator of the essential prin—
ciples of this theory, let us briefly review his attitude toward meas—
urement.3 As an empirical philosopher and eXperimental scientist
he regarded calculation and measurement as the most exact, fruit-
ful, and certain of scientific methods. But these methods them-
selves must be logically or epistemologically founded. Thus he
demanded some metamensurational justification for mensurating
operations.
Though to a certain extent, Helmholtz was willing to eschew
transcendental intuition, his empiricism did not go so far as to al-
low him to depart completely from a priori laws. Nor could he re-
sist founding measurement on axioms concerning “the fundamental
nature of mind.” He was an empiricist, relative only to the ab-
solute idealist. For Helmholtz mathematics still comprised a
transcendental source of certainty. Consequently he could not
really assimilate the idea concerning the cultural development of
numbers and number systems.
With the replacement of Kantian intuitive mathematics by the
Realistic variety the applied-mathematics doctrine of measure—
ment took on the latter’s complexion: Since the revival of Pla-
tonic Realism in mathematics, measurement has been defined in

Cf. Jeffreys, Scientific, p. 206.

a Zéihlen}
'268 PSYCHOLOGY AND LOGIC

terms of abstract magnitudes and numbers. For example, Russell

writes:
Measurement of magnitude is, in its most general sense, any
method by which a unique and reciprocal correspondence is established
between all or some of the magnitudes of a kind and all or some of
of the numbers, integral, rational or real, as the case may be.4
With a lessening of emphasis on abstract, preexisting numbers
and magnitudes the applied—mathematics view takes the form
that measurement consists Of interrelating numbers with things.
What these things are considered to be depends upon the auspices
under which the proposition is formulated}. The philosophical Op—
ponents of extreme realism refer to some eicperiential factor. We
quote from Spaier:
0
La mesure est l’établissement d’un correspondance uni-univoque
entre les nombres et les données évaluées.5 ∙
In connection with experimental investigations Campbell as—
serts that measurement iS' the process of assigning numerals to
represent qualities6 or properties7 in accordance with scientific laws.
In debate, however, he varies his position to make measurement
“the assignment of numerals to things, so as to represent facts
and conventions about them.”8 Writers who adOpt Campbell’s
general viewpoint accept the following phrasing:
Measurement, in the broadest sense, is defined as the assignment of
numerals to Objects or events according to rules.9
To the critical Observer it is clear that the applied-mathematics
theory of measurement does scant justice to the complex facts in
question. Howsoever important numerals and numbers are in
measurement Situations, they hardly seem to be the center and
focus of all the complicated activity comprised under the heading
of measurement. True as it may be that the goal of certain men—
surational operations is to ascertain the quantity Of some thing
relative to another, it is not appropriate to reduce all of measure-
‘Principles, p. 1 76.
“La Pensée, p. 34.
° Physics, p. 267.

1Ibid., p. 295.
i

a Quantitative, p. 34.0.
° Stevens, Theory, p. 1577.
 LOGICAL ASPECTS OF MEASUREMENT 269
ment to this feature. It is a reasonable hypothesis that the applied—
mathematics theory stems from the Pythagorean tradition that
somehow numbers are the heart of reality. There are definite im-
plications here that require examination if they are not to pre—
judice measurement theory.
The Power of Mathematical Reasoning. Although the number—
assignment theory is really concerned with the concrete conditions
of applied mathematics or arithmetic, it harks back to the View Of
the great power of mathematical reasoning in science. The basic
idea is that mathematics purveys certainty and achievement.
Against this unwitting projection of mathematical reasoning-
into science there are powerful protests. We have quoted Hardy’s
declaration concerning the inapplicability of mathematical reason-
ing to the physical world.10 Similarly, Einstein raises a serious
question concerning the relationship between mathematical pro—
positions and events.
Insofern sich die Satze der Mathematik auf die Wirklichkeit be-
ziehen Sind Sie nicht Sicher, und insofern sie sicher sind, beziehen sie
sich nicht auf die Wirklichkeit.11
Furthermore, what of the recent developments of mathematical
thought which make plain that number is not the basis of mathe—
matics? Mathematics may perhaps be better regarded as a science
of order or, even more to the point, as a science Of relations. On
the latter bases we must modify our
of numberIn measurement.
views
concerning the place

Numbers as Supreme Universals. Deeply imbedded in our

scientific and logical culture is' the tradition that numbers con-
stitute the supreme universals. Not only are other universals,
such as things and qualities, placed lower in the scale, but num—
bers are regarded as the means of transcending things altogether.
Even more, through numbers a transcendent realm of certainty
and accuracy is considered attainable. As a matter of fact, those
who put so much stress on numbers in measurement have in view
those transcendentals of infinity, absolute identity, and difference
conventionally represented by discrete integers. As superior uni-
versals, numbers are abstracted and extrapolated from various
\

Chap. 13,‘p. 21.

‘1
Geonietrie.
270 PSYCHOLOGY AND LOGIC

practical relations of things and made into metaphysical invariants

of nature, or, even more frankly, into cosmic essences.
Numbers as Philosophical Abstractions. Most writers who glor-
ify numbers, thus making measurement into the essence of science,
do so on the ground that numbers somehow pertain to a higher
order of being. Whatever is connected with number is presumed to
transcend the humble qualities of things. A paradox found in the
writings on measurement is that while the ascertainment of the
metrical properties of things is justifiably regarded aS an essential
feature of scientific method, the characteristics ascribed to metric
principles project them out to unreachable. metaphysical domains.
Numbers and External Reality. The metaphysics of the “ex-
ternal world” is frankly implied in the stress of number and quan-
tity as basic factors in measurement. To a ggeat extent the empha-
sis of numerical and metrical properties is predicated on the as-
sumption that numbers as superior entities guarantee the existence
of the “external world.” Here we run into the perennial ideologi-
cal convention. Reality can not be found in the evanescent occur-
rences of sensory experience, but must be sought in the general
formulae of abstract relations. Thus numbers and: mathematical
equations provide the perennially desired tongs with which to
grasp and hold Reality.
Numbers and Mental States. It is evident that the tradition of
psychic processes is basic to the historical quest for reality and for
a guarantee of the existence of the external world. The need for
guarantees and absolute formulae of relations of things is inti-
mately tied up with theconcept of “sensations,” and of other
psychic elements, implied by traditional notions of experience.

MENSURATIONAL OPERATIONS AND N UMERAL ASSIGNMENT

Measurement involves a complex handling of things and events
on the basis of particular problems in distinct situations. How-
ever, because of the ubiquity of measuring instruments, such as
balances, rulers, calipers, thermometers, and galvanometers, meas-
urement is taken to be simply the manipulation of these tools. The
nature of the instruments, their derivation and calibration; are
usually taken for granted. The assertion that measurement con-
sists of matching the ends of things measured with those of the
measuring instrument is a case in point.
 LOGICAL ASPECTS OF MEASUREMENT 271
Actually, the processes of measurement constitute complex
forms Of interbehavior in situations involving many things and
conditions. We can not neglect What is being measured, nor the
use to which the results will be put. Measurement envisaged as
a series of systemizing operations does full justice to the process
of number assignment. Moreover, when we regard measurement
as system making rather than as practical manipulation, we find
that number assignment plays two important, but still auxiliary,
parts in measurement situations.
(1) Of the two numeral assignment processes the more elaborate
and useful consists of calculation and relation structuring, ex-
ecuted by means of various mathematical techniques. However
essential these structuring processes may be, they are always
subsidiary to the measuring operations which involve definite
contacts with things. Even so strong an adherent of the number—
assignment theory as Campbell12 takes account of all sorts of ma-
nipulative operations anterior to number assignment.
(2) A less prominent, though still important, numeral-assign-
ment process consists of recording the results of measuring ma-
nipulations. By statistically treating mensurational records these
second auxiliary operations may take on an importance hardly
subsidiary to the mensurational manipulations themselves. The
priority of manipulative operations is clear from a number of facts.
First, no one would say that any assignment of numbers is meas-
urement.13 On this ground writers distinguish between counting
and measurement. Again, the emphasis upon additive units and
processes bespeaks a bias toward things and their qualities as over
against the abstract relations of formal arithmetic.
Those theorists who adopt the number-assignment formula do
so mainly because they are influenced by symbolic and mathematical
logic systems. And this, despite the fact that number assignment
may be more closely allied with practical arithmetical operations
than with the more abstruse theory of number relations.

SPECIFICITY 0F MEASUREMENT
In line with the prevailing universality of logical theories metro-
logical systems are naturally patterned as general and inclusive
"Physics, Account, Measurement.
Campbell, Physics, et passim ; Einstein, Meaning, p.
’3
I.
272 PSYCHOLOGY AND LOGIC

as possible. Minor theoretical variations are incorporated into a

single comprehensive formula.
From an interbehavioral standpoint there is no single law or
logic of measurement. On the contrary, there is a range of sys—
tems corresponding, to: (I) variety of things and events to be
measured, (2) available operations—for example, (a) isolating
particular materials and fields of operation, (b) constructively
analyzing applicable units, and (C) organizing a scale relative to
situations more or less distantly removed from the one immedi-
ately1n view—and, finally, (3) needs and purposes Of those per—
forming the measuring operations.
When we describe or formalize measurements we,’ of course,
pounce upon an abstraction, and construct some sort of idol. Either
we develop a verbal model very difierent from our actual opera-
tions, or We elevate a part of the measuring situation to the dig-
nity of the whole. By such means anyone can transform all meas-
uring, even all science, into pointer reading. It is such analogical
thinking that stimulated the I9th century idea that physics is the
basic science because it is essentially the mensurational discipline.
TO overlook the contingent and specific character of measure-
ment is to depart from concrete measurement operations. Above
all, it is extremely hazardous, as well as scientifically objection-
able, to withdraw from the actual things and processes measured.
Significant measurement can neither go On in a vacuum nor be
reduced to the formal application of instruments to obj ects. Meas-
urement conceived independently of the things to be measured in
specific situations is nothing more than a fetish, whereas indifferent
manipulation of instruments glorifies the instruments and the
manipulative movements.
(I) Specificity of Measured Objects. The stresS of operations
in measurement is surely justified when one considers the large
number of Objects requiringprecise evaluation. We can do no
more at present than list classes of measurable things calling for
specific sorts of operative procedures. For convenience we shall
diflerentiate from each other three grand classes of measurable
things: (a) Objects, qualities, and rElations, (b) interactions of
things, and (c) interactions of judges and experts with such things
as are comprisedIn (a) and (b).
(a) Things and relations. Here we include actual things
 LOGICAL ASPECTS OF MEASUREMENT 2'73
and their properties, both relations and qualities. Among the quali—
tative properties some are of the first order—for example, color,
length, area, volume, density, hardness, fragility, malleability, etc.
Second—order qualitative properties comprise Similarities, contrasts,
identities, equalities and equivalences.
(b) Interactions of things. In this class there are various
sub—classes. For .example, measurable interactions of inorganic
things require difierent manipulations and scales from those de-
manded by organic things. The characteristics to be elicited in the
inorganic subclass are relative rates, motions, frequencies, and
products. What interests us with respect to organic interactions are
effects on organisms, measurable as interferences of behavior, in-
hibition of growth, immunological warding off of pathological con-
ditions, etc.
(c) Judging and Evaluating Behavior. A general interest
in measuring systems demands that we consider ways and means of
measuring the behavior of judges and experts. Considerable at-
tention has recently been given such mensurational interactions
(attitudes) with things. The measurement of judging and evaluat—
ing behavior implies no retreat from standards and ideals of ac-
curacy and precision, deSpite the difference between such events
and the things treated by means of the conventional C.G.S. sys-
tems. ∙
(2) Specificity of Operations and Techniques. Because there are
many types of Objects to be measured it follows that different
mensurational operations must be performed. The following
types are selected as illustrative.
(a) Direct and Indirect Measurements. The operations
here consist in comparing and matching properties, either by apply—
ing some direct standard or by means of some indirect device.
Stretching a chain between two given points is essentially a direct
technique. A good example of an indirect method is Faraday’s tech-
nique of weighing gold leaves to Obtain their thickness (p. 291),
or Galileo’s weighing water to measure time.
(b) Simple and Complex Measurements. Even when
operations and instruments have been perfected for mensurational
situations their mode of application may be relatively simple or
complex. To illustrate, simple measurements require either (I) the
application of rules or scales to the objects measured, or ( 2) some
274. PSYCHOLOGY AND LOGIC

computational procedure. Complex measurements, on' the other

hand, require both forms of Operation.
(c) Fundamental and derived measurements. It has be-
come conventional to refer to linear and weight measurements as
fundamental. While this is an arbitrary designation, it is presumed
to imply some basic mensurational difference from temperature or
density determinations. If we regard mensurative operations as
processes for achieving a more precise determination of the proper-
ties of things, we look upon the “fundamental” and “derived”
classes only as items in a series.” This is not to deny that from the
physicist’s standpoint the distinction may be useful and even im-
portant. However, the assumptions implied by the distinction
shOIild be carefully noted. Unquestionably, for sOme purposes
the conventional fundamental operations gre properly named,In
other instances this is not the case.
((1) Status and Construction Measurements. Most men-
surational operations and techniques are bounded by relatively
stable circumstances. The status of the things to be measured and
the instruments employed are conventional and standardized. But
there are other situations in which considerable fluidity prevails.
No more striking example can be cited than the circumstances sur-
rounding
‘ Here
mensuration processes in the domain of q1iantum mechan-
the mensurational work requires building a practically
cosmplete metrological system. One must set up criteria of ob—
servability and of scale making and using.
Many are the circumstances leading to constructive procedures
and uniquely original instruments. In our quantum-mechanics
example we have obvious difliculties in the microscopic character
of the things to be measured and in the unavailability of close
contact with the units of energy we aim to determine. The con-
structional nature of the Operations is excellently indicated in the
development of a wave and a particle model and of the formulae
for their successive and simultaneous application in measuring
situations. And when the measuring operations lead to probability
functions, rather than to numeralsrepresenting concrete cardinal
magnitudes, we have an even bettef‘illustration of constructional
meaSUrement.
(3) Specificity of Mensurative Situations.
1‘See next chapter.
When we assert that
 LOGICAL ASPECTS OF MEASUREMENT 275
measurement is secondary to some need or purpose we are referring
to a specific mensurational situation. Of course, this purpose may,
be highly abstractive and theoretical, rather than immediately
practical. Still, the results achieved, or the numbers attained, serve
the needs of a-particular investigation. Numbers, in other words,
are instruments for measurement.
The basic conclusion is that no type of mensurative product
need be rejected as a member of the classes of mensurative sys—
tems. Numbers representing a person’s reactions in the form of
estimations and judgments constitute objective mensurative fea—
tures in the same degree as quantities of length or weight. Judg—
ments or estimates as events consisting of interactions with things
are just as free of arbitrary construction as are interactions between
nonhuman objects.

LEVELS OF MEASUREMENT
The mensurational data so far treated may be arranged on a
horizontal plane. Weighing, length measuring, scaling, and order-
ing activities may be set side by side. We have still to consider the
variations of measuring situations organized in the vertical dimen-
sion. Thus, we survey briefly the continuum ranging from the
practical Situations of everyday life, through the intervals of tech-
nological operations, to the most abstruse ordering procedures of
special and ‘general mathematical situations.
Practical Measurement. On this level the operations constitute
comparatively simple, sometimes crude, manipulations for ascer-
taining the values of the things handled. The qualities and prop-.
erties of objects are measured for some immediate purpose. On
the whole, practical measurement implies the mere application of
an already—existing measurement system or the simple coordina-
tion of means and ends.
The direct handling of obj ects characterizing the simplest meas-
uring situations may be illustrated by the following process of
fitting boards to a surface. The edge of one board is laid against
one boundary of the place to be covered, and the length required
is marked by matching with the other boundary. Then the strip
cut first is used as a standard for sawing as many more strips as
are required. When a ruler is used the process is somewhat indirect,
and thus the level becomes Slightly raised.
276 PSYCHOLOGY AND LOGIC

How high the practical level rises from the gross handling
indicated may be gauged by the amount of labor and ingenuity
required to make scales and standards for comparing things. The
primary operational emphasis, however, is on the things and
their manipulation. Since the things measured, together with their
circumstances, are the focal points, we add the following list to
suggest the range of practical measurement.
1) Things and properties (Size, length, area, volume).
2) Relations (number, quantity, magnitude).
3) Strength or intensity in relation to (a) function (electrical
resistance) or (b) efiect on other things.
4.) Rates (frequency according to various criteria): period,
time interval, speed, velocity, acceleration.
5) Behavior (what a thing does under c‘various circumstances).
The primary purpose of this list is to indicate the concrete oper-
ations useful or necessary for handling given "objects. The opera-
tions are as specific and as local as the things and Situations demand.
Techno-logical Measurement. When the obj ects to be measured
are small, complex, or extremely important, we reach a measure-
ment level in which the whole situation is no longer focused on
the objects. The operations take on a unique importance; the goal
of precision assumes a somewhat autonomous character. Techno—
logical measurement requires elaborate manipulative processes in—
volving specially designed apparatus. The operations include cal—
culation and sometimes actual research work. The most intricate
technological measurements may be said to establish the identity
and properties of things whose scientific existence is problematic.
Technological measurement operations are best observed in the
behavior of scientists as they construct their mensurational instru-
ments. While devising units and scales considerable research may
be'needed to establish the existence and value of things and events,
as well as ways of dealing with them. AS our examples we choose
the developments in measuring temperature, and electricity and
magnetism.
From the beginning of thermometer making by' Galileo and
Drebbel in the early years of the 17th century,15 workers desiring
to compare warmth differences had to contend with the problem
of bringing into system a Series of factors.“On the Side of things
Grimsehl, vol. II, p. 2, et passim.
 LOGICAL ASPECTS OF MEASUREMENT 277
measured there were warmth changes to be correlated with ex-
pansions of substances like alcohol, air, mercury, and hydrogen
gas. Then scale points had to be determined, naturally, the melt-
ing point of ice and the boiling point of water were taken as stand-
ards. Magnitude and denomination of intervals required the de-
cision whether zero or 32 should mark the melting point of ice.
Whether 80, 100, or 212 degrees should be the boiling point
became a matter of specific system making.
More complex and more refined temperature systems were
developed with the evolution of knowledge concerning light,
energy, and the spectrum. Coincident with Ritter’s and Herschel’s
discovery of the ultraviolet and infrared extensions of the energy
spectrum was the development of Nobili’s thermopile and Lang-
ley’s bolometer. With the latter instrument:
Measurement of the charge in resistance could be made so pre-
cisely that temperature changes of one ten-millionth of a centigrade
degree were identifiable.16
Perhaps even more informative examples of the development
of technological measuring systems are found in the evolution of
electrical units and scales. Here we can only mention the complex
labors of Ohm which finally resulted in the system for measuring
electrical current, resistance, and potential difference, and the
work of Oersted, Ampere, Gauss, Weber, and others in establish—
ing the units called ampere, volt, coulomb, farad, etc.
Formal Metrology. When measurement as operation and sys—
tem becomes interesting in itself, and is thus made an object of
study, it attains the level of abstract constructional ordering.
Measurement on this level is quite remote from the practical and
technological circumstances of the other two levels. Formal me-
trology stresses a theoretical system-building interest. Accordingly,
speculative and general philosophical views are allowed free scope.
Stressing measurement systems as the objects of their study, metro—
logists concern themselves with abstracted relations of formal
geometric and analytic consistency structures. More particularly,
formal metrology centers about invafiant relations which can be
comprehended and fixated in deductive. systems.
The best illustrations Of systems on the present level are found
1°
Tayloi', Physics, p. 542.
278 PSYCHOLOGY AND LOGIC

in the‘abstruse geometries based on the criteria of congruence and

coherence. The upper 'ranges of this level include mensurational
speculations concerning subsistential magnitudes and modes of
reaching them. The field of formal metrology may be described
as the construction of generalizations about measurement as re-
moved from all the concrete factors of specific mensurational Situ—
atIOns.
Interrelation of Levels. AS We should expect, the numerous
gradations in each level extend so far as to interrelate all three
types. True, an individual may use_ rulers, balances, ammeters,
even more complicated measuring instruments, without knowing
anything about their nature or origin. But. the fact that cultural
practice extends beyond technical system does not permit an
absolute barrier between the manipulations and the creative organ—
ization of a system. Even if we reject the notion that measurement
presupposes geometry,17 because only through’ geometry can the
necessary determinations of rigid bodies and straight lines be
made, the interrelationship between the use and creation of a
mensurational system can not be set aside.
That mensurational levels are interrelated follows from the fact
that they all consist of interoperations with things. There is‘a
continuum in human behavior on any point of which mensura-
tional behavior may be represented. Certainly, those who construct
measuring systems may proceed either on a humble manipulative
level or on one involving elaborate abstractive formulations.
Despite the dense-point continuum, the theory of measurement
must be kept distinct from particular mensurational operations. To
do so is to respect the differences between diverse kinds of systems,
especially those concerned with (a) the relationshipbetween meas—-
urable things and (b) the structure of factors in purely descriptive
systems.
QUALITY AND QUANTITY 1

ItIs a paradox of measuring theory that even when one’s pri-

mary interest is in quality one comes to stress quantity. on
the
whole, this paradox is inherent in \the fact that discussions of
measuring are usually pitched on an extremely abstract
actual mensurational Situations the paradox\does not exist.
level.
In

Russell, Knowledge, p. 282.

 LOGICAL ASPECTS OF MEASUREMENT 279
Metaphysical writers range themselves on opposite Sides of the
quality-quantity problem. Some adopt the ancient Pythagorean:
Platonic idea of the superiority Of numbers as eternal essences,
but others look upon the two categories quality and quantity as at
bottom one, or as transformable one into the other.
The sharp differentiation between quantity and quality is, of
course, a purely historical accident. Out of the 17th-century
dichotomy of nature into thought and extension the conception
has evolved that quantity iS more fundamental than quality. It is
still a prominent feature Of our dualistic culture to regard quality
as a diaphandus manifestation of quantity.
However, we need no longer be bound by this tradition. We
have already achieved an obj ective psychological View which not
only obviates the thought-extension dichotomy, but enables us to
consider scientific as well as other kinds of magnitudes as definite
products of interbehavioral enterprises. It is interesting to note
that the persistence of dualistic dogma has resulted in overlooking
the fact that in our century accurate measurements of such events
as color, heat, and sound have been achieved. No longer need we
abide by the notion that reality is quantitative and that number
rules the world.
Though few, there are some writers who protest against the
subservience of quality to quantity.” The most favorable result of
such protest would be a modification of view concerning measure-
ment. The entire enterprise of abstracting qualities and quantities,
and using them as elements in the construction of systems, is
either a private or a group operation.
Today it is clear that the distinction of primary and secondary
properties is only an outcome of operational situations. What were
called primary qualities were those that could easily be measured
with the Simple tools and techniques available. This fact is apparent
when we analyze the various aspects or properties of things. On the
whole, metrical properties are more stable and simple, or at least
more fixable, than qualitative ones. But the point to emphasize is
that they are aspects of things. When we break a stick we can put
the pieces end to end and regard the Operation as adding the parts
to make up the total length. The same process does not apply to
the hardness of the glue which holds the pieces together. Similar
See, for example, Spaier, La Pensée; Nagel, Measurement; Benjamin, Logic.
28o PSYCHOLOGY AND LOGIC

variations in operating upon different things yield constructions

concerning qualitative and quantitative properties. In each case
the basis is type of available manipulation.
Notice how influential the selective operation is in distinguishing
between quality and quantity. Such selection is rooted in the proc-
esses of describing or referring to things. We may or may not
emphasize the fact that the organism is two—sided or bilaterally
symmetrical, is two—handed, two—legged, 200 boned, and so on.
When we are concerned with these characteristics, however, we
surely need not separate quality from quantity. Similarly, the
organism’s height and weight are metric properties, but where18
the sharp line between one kind of metric property and another?
All properties are, or can be made, metric.
The relation between qualitative and quantitative aspects of
things for measurement theory suggests a differentiation between
quantities as: (1) properties of things, (2) prdducts of Operating
upon things, and ( 3) more or less independent constructions con—
cerning things
(1) Quantity as property of things consists of all the aspects
of any quality or dimension which allow for diversity. The quali—
tative aspect constitutes primarily that which is uniform and con—
tinuous. To look to the limit of spread or relation of color,
extension, hardness, or cohesiveness is to select out the quantitative
property. In some cases, of course, quality and quantity are so
interdependent that we have an indifference point in description
or scaling. Analysis of volume illustrates the inseparability of
mathematical quantity and any relevant form of quality. When
we ask how old a person or thing is, how many revolutions per
second or minute, how long a person or thing will live—that is,
number of years or time units—we can not distinguish quality from
quantity, though we do so in formalized description.19
In concrete measurement Situations recall that we are always
dealing with things, not with abstract quantities or magnitudes.
Nor are we interested primarily, certainly not exclusively, in quan-
tity. The extension, difference, and relation of things exist within
the'limits of what we ordinarily call OIie thing or within the-con-
nection between several. In the latter
properties of things as relatively extrinsic.
casewe
may describe the
\

Concerning some identities of quality and quantity, see Benjamin, Logic.

 LOGICAL ASPECTS OF MEASUREMENT 281
(2) Quantity as measure constitutes a highly selective aspect
of things complicated by measuring processes and instruments. The
intrinsically quantitative and relationally numerative properties
have been rendered a bit more remote from the things of which
they are a part. Measured, recorded, and described quantities in-
clude effects based upon (a) the interaction of the measurer, as
well as of (b) the measuring conditions; in this sense quantities are
contrived and are thus more or less artificial.
The contrived character of quantity is best illustrated by the
varying qualities of numbers in the sense of relations, as these
qualities depend upon measuring or other situations. For example,
the conventional over-evaluation of Simple additive quantities re—
flects the Operator’s domination of the things he operates upon. By
contrast, Operation with vector or tensor quantities suggests a sen-
sitivity to the qualities and properties of the things.
( 3) Quantity as constructions are best illustrated by various
averages, coefficients, and other products such as numerical laws or
equations. The important criterion here is the initiative and degree
of independence in manipulating the results of measuring opera—
tions. The types of product developed reflect the influence of the
operator’s adherence to some sort of theory or philosophy of num—
ber, as well as his proficiency in handling calculative techniques.
' ~.VARIETY SPECIFITY OF MAGNITUDES
AND

The fact that measuring operations are specific with respect to

objects measured means there must be as many different kinds of
magnitudes as there are kinds of measurable things. The tradition
that dichotomizes all magnitudes into intensive and extensive types
appears therefore to be wide of the mark. This bipartition of mag—
nitudes is not, however, without some advantage. Since extensive
magnitude is presumed to be based upon the prototypical operation
of comparing straight lines, it has historically been regarded as the
basic, even exclusive, type. Thus, to insist that there is also a non~
additive type indicates a need for further descriptive terms.
The most serious Objection to the twofold classification lies in
overstressing additive properties. Why should we be overwhelmed
by such things as can be heaped up or laid end to end, as in the case
of primary additive things yielding weights and linear measure.P20
2°
Stevens, Theory; Churchman, Materialist; Comrey, Operational.
282 PSYcHoLOGY AND LOGIC.

In what sense are hardness, density, color, elasticity, and tempera—

ture less important or less fundamental?
Aside from harking back to the dichotomy of primary and sec-
ondary qualities the two—fold classificatiOn stresses illegitimate
analogies in the sense of reducing measuring operations to the
manipulation of rulers and balances; Moreover, the assumption
that there are certain fundamental properties results in overlooking
the variety of things to be measured. We ask: Should the syste-
matic consideration of measurement stress abstract structures, or
rather the process of systemizing operational procedures?
The farther one departs from metaphysical constructions the
greater the number and variety of magnitIIdes. The-very purpose
of measurement operations is to work out as many magnitudes and
dimensions as are necessary to provide the required information
for the task at hand. ItIS fatal, therefore, to measurement theory
to make metaphysical or mathematical systems the criteria for
measurement systems. Is it proper to regard a thing’s length as
more basic than its hardness or compressibility merely because one
has available an obvious and simple tool to facilitate mensurational
operations? .
That some things and events are difficult to appraise constitutes
no more than a challenge to those who regard inVestigation and
measurement as techniques for discovery and experimentation.
Surely it is contrary to scientific behavior to make the novelties of
inquiry conform to established rules. Such behavior is inimical to
intellectual advancement. It is the most appealing feature of scien-
tific history that the unknown becomes discovered and the un—
measurable becomes determined and subsumed under rule through
the effective action of the unbiased worker.

MEASURING SYSTEMS
Measuring systems constitute the most practical and manipula—
tive of logical structures. On the whole, they are not constructed
except for some purpose beyond themselves, a fact which probably
accoisnts for the paucity of attempts‘to describe them explicitly.
Still, a formalizing description is of considerable importance, even
though for. the moment it
means disregardin‘gkthe fact that measur—
ing operations are always specific. According to. our plan, therefore,
 LOGICAL ASPECTS or MEASUREMENT 283
of treating measuring systems by means of a generalized model21
we propose the following five-factor description, based on many
simple and complex Situations.
A. Acts of assumption and postulation.
B. Choosing and refining units.
C. Instruments and manipulations.
How many scales?
Scales and scale construction.
Bias in Scale making.
Scale application (mensuration).
Scale fitting operations.
Extrapolation and interpolation.
D. Recording. .
E. Statistical treatment.
A. Assumptions and Postulates of Measurement. In practical
situations the assumptions usually concern immediate operational
details—for example, the sufficiency and trueness of the devices
used, whether levers, rulers, galvanometers or other electrical in-’
struments.
When, as in the case of more theoretical measuring systems,
assumptionsare examined and evaluated, they become authentic
and sometimes significant postulates. Such is the assumption that
mathematical relations are somehow ultimate features not only of
measuring processes, but of the nature of nature itself. Because all
measurement involves comparison of relations, we expect numbers
to play a part in all measuring systems. It is a crucial question, how—
ever, whether to regard such numbers as metaphysical elements,-
rather than as operational factors in actual situations. We must not
overlook the inevitable postulate that the development of number
systems, as well as of mathematics in general, consists of a complex
cultural evolution, and consequently involves the activities of indi—
viduals. Among the interbehavioral implications we list the rela—
tively arbitrary nature of scales. In other words, units. and degrees,
or their variation, are selectively constructed. A competing postu—
late regards all measuring devices as ihstruments for achieving cer—
The effectiveness of this model is determined'by the amount of necessary detail
that can be included without overcrowding the exposition or overemphasizing any
element.
284. PSYCHOLOGY AND LOGIC

tain purposes; thus the kinds of things worked with are kept well
in view. Whether or not mathematical relations constitute the
fundamental nature of these things13 a problem to be determined,
not an absolute assumption from which we must make our depart-
ure.
Though mensurational postulates are factors in a system, they
are nonetheless susceptible to scrutiny and estimation. When rank-
ing assumptions we place at the metaphysical pole the view men-
tioned above concerning the relational essence of things. At a some-
what lower point is the popular attitude that science consists of
measuring behavior, and that scientific data therefore comprise
nothing but the products of such behaviorlirrespective of what is
originally measured. On still lower levels, which approach more
closely the interbehavioral operations themselves, are the assump-
tions (a) that what13 measured13 a fair sample of the total sample,
and (b) that there13 or is not a population or a universe beyond the
larger sample.
B. Choosing and Refining Units. Within specific systems units
vary on the basis of the sort of thing measured. Also, the refine--
ment and improvement of units are differentially variable accord-
ing to the mensurative situations. Probably the simplest and most
satisfying systems are those in which the units conStitute parts of
the things measured, as in the so-called fundamental units—for
example, proportions of length and weight. The things measured
in these instances possess definite metrical properties which are
divided into convenient units.
More analogical are the essentially constructive units; for ex-
ample, all units derived and compounded from SO-called basic
dimensionsw—units of density, temperature, velocity, etc. The
as
units for all derived23 magnitudes may be regarded tools of com—
parison, their primary function being to indicate relative differences
of change or proportion between properties of things.
Those whose mensurational horizon- is not bounded by the addi-
tive constructs of physics, and who therefore have room for esti-
mating, assessing and judging systems, are free to construct other

2“As the most important of the properties that can be fundamentally measured,
Campbell (Measurement, p. 127) lists numbers, mass, volume, length, angle, period,
force, electrical resistance, current, and voltage.
:1 Campbell, Physics,
p. 346f.
 LOGICAL ASPECTS OF MEASUREMENT 285
types of units. Such freedom is limited only by the exigencies of
the problem of ascertaining properties and relations of things and
events. In other words, the guiding principle is to carry out a piece
of work instead of applying formal and rigid specifications devel-
oped and suitable only for another situation.“
C. Instruments and Operations. Mensurative operations so fre-
quently consistiof handling instruments that the occupation with
apparatus and things measured constitutes a single complex factor
in mensuration systems. The primary instruments comprise scales
of which there are many sorts. Measuring involves the problem of
making a scale to fit the objects to be measured. Should we em-
phasize the comparison of things or merely fit Obj ects to arbitrarily
chosen scales? Some of these issues, as we have already implied,
hark back to philosophical views.
How Many Scales? The general trend of opinion is that there
are only a few kinds of scales. The tradition of primary and sec-
ondary qualities has influenced scientific writers to divide qualities
into additive and nonadditive types. Only the primary are consid-
ered fundamentally measurable. Alternate names for these quali-
ties are extensive and intensive. Writers so influenced allow only
four Scales. On the whole, physicists, in whose province one locates
the things capable of addition and subtraction, permit a limited
number'of-scales, while psychologists and social scientists are will-
ing to increase the list to accommodate less palpable items than
those Of physics.
The physicists’ scales, on the whole, are of two types. The first,
dealing with extensive or additive properties, is called a ratio
scale; the second, concerned with intensive or nonadditive mag-
nitudes, is named an interval scale.25 Psychologists and sociologists
are inclined to add two others—namely, a nominal and an ordinal
scale. The former is employed for identifying and classifying
things, the latter for arranging things in series?6
The standard for setting up this quadruple scale system is the
conventional opinion that measurement is the application of nu-
merals to things. Such a scale systems can only be justified when

this sentence suggests that we question the assumption that any science,
including physics, is the scientia scientiarunz.
2’
There is no uniformity, however, in term usage. Cf. Stevens, Theory.
Stevens, Classification.
286 PSYCHOLOGY AND LOGIC

writers acknowledge that they have set up a range of mensura-

tional scales on the basis of their own fields of investigation. But
there are two problems here. First, the use of scales must be dif—
ferentiated from theories of scale making. Using certain scales
is a process dictated, for instance, by the kind of magnitudes with
which one works. It is another matter to make one’s own scale
the exclusive model of measurements, and thus to limit scale
variation. Secondly, from a logical standpoint we are obliged to
make room for as many scales as mensurational Operations indicate.
To fix an arbitrary limit tO' their numberIs to adopt an authori-
tarian attitude. ∙
Scales and Scale Construction. Scales as tOols for ascertaining the
nature of things must maintain a workable harmony with the
things measured. In order to clarify thlSOpOlnt We contrast our
view that (a) scales must conform to the things measured and also
to the problem underInquiry with (b) the numbers conformity
theory. As a basis for this discussion we examine Stevens’ four
scales.
I. Nominal Scale. Defining measurement as “assignment of
numerals according to rules,” Stevens includes the nominal type
among his scales. The Operation consists merely of applying nu-
merals to items in a series for differentiating and identifying them.
The numerals serve as Ieady labels or names, as in the marking
of policemen, prisoners, box cars, football players, and race horses.
The rule here is simply not to assign the same numeral to dif-
ferent items at the same time, or different numerals to members
of the same group when classes are to be distinguished. Stevens
himself questions the nominal scale as a measuring device, but is
influenced to retain it because it conforms to the given definition.
The basic objection to this scale, which is really no measurement
scale at all, is that the operations in no sense yield any information
about‘ the qualities or characteristics of the labeled things. Even
the ordered numefals on state highways are scantily informing un-
less we already know the marking system.
Despite these criticisms, however, thereIS justification forIn-
cluding a nominal scaleIn a scale systeIn. First, the heavy hand of
quantity is lifted. The difference between numerals and numbers
is recognized. But more important is the implIed View that meas—
uring scales should not be limited to a few classic types. Tradi-
 LOGICAL ASPECTS OF MEASUREMENT
_ 287
tional measurement discussions take no account of the tremendous
development of technological and scientific domains and of the
tools and operations invented to cope with new events.
11. Ordinal Scale. Ordinal scales are ranking devices for plac—
ing things and properties in relative positions on- the basis of a
given criterion. The Operations are performed to determine the
asymmetric relation of greater or less of the compared objects or
qualities. Because they are not strictly quantitative, the measure—
ments are not subject to statistical treatment involving means and
standard deviations. The classic example of an ordinal scale is
Mohs’ instrument for differentiating and ordering minerals ac-
cording to hardness, which is determined by the scratches one
mineral makes upon another. This test results in the following
order:
(I) talc, (2) gypsum, ( 3) calcite, (4) fluorite, (5)- apatite,
(6) feldspar, (7) quartz, (8) topaz, (9) corundum, (10)
diamond.
Physicists, for obvious reasons, are less tolerant of such scales
than are psychologists. The physicist, interested in and influenced
by things that can be handled as independent and additive, can not
admit hardness as a measurable property at all. In fact, Campbell
admits only money value as equally measurable with the phe-
nomena! of physics.""7 To be sure, things vary in their capability
for precise determination; it is equally certain that the function Of
science is better served when arithmetical quantity is not made into
a fetish, but maintained as a working ideal. However faulty an
ordinal scale like Mohs’ may be in the inequality of its intervals,28
it is still true that mensurational principles call for a multiplication
of measuring scales as features of investigative systems.
III. Interval Scale. Interval scales, such as the Centigrade and
Fahrenheit temperature scales, are regarded by Stevens as gen-
uinely quantitative. Accordingly, the interval scale is presumed
to exemplify the isomorphism between what can be done with the
aspects of objects and the properties of. the numeral series. While
the additive principle is absent here, the equality of expansion

Mea‘snrement, p. 1 3 5.
3
Spencer, Abrasives.
288 PSYCHOLOGY AND LOGIC

volume can be accurately matched with equal intervals on the

temperature scale. _
In many ways the interval scale is an excellent representative
of the entire series of mensuration constructs and operations. Not
only are many important statistical treatments available—compu-
tation of mean, standard deviation, correlation coeflicients—but the
range of applicability is larger.
IV. Ratio Scale. The authority of the physical sciences and the
'power of numbers are responsible for an irresistible tendency to
accord the greatest weight to ratio scales. Such scales are con-
structed for things that can be added together and grouped in
various ways, as well as divided and otherwise manipulated.
Further, there is ascertainable a definite zero point to help in
comparing things. 0

An interesting suggestion of Stevens is that the cardinal number

itself constitutes a ratio scale. This brings Up the important ques-
tion of the relation between counting and measuring (p. 296). At
the least, such a suggestion prompts the consideration of the large
range of scales necessary to discover the significant properties of
things.
Bias in Scale Making. The impact Of philosophical thinking 011
the logic of measurement is illustrated by the arguments concern-
ing the nature of psychological scales. Since their earliest attempts
to measure things psychologists have been assiduously engaged
in finding means of scaling psychic as over against physical mag-
nitudes. Because psychic magnitudes were not available for ma-
nipulation, the scales devised were presumed to measure relations-
between “sensations,” for example, and the magnitude of the
stimulus or the action of‘the organism. The vigorous discussion
concerning the possibility of psychological measurement which
started when such measurement was first attempted continues to
the present moment.29
Doubtless the solution of such problems lies in determining pre-.
ciselywhat is to" be" measured before scales are made and applied.
One important suggestion is that psychological scales measure ac-
tions of organisms in relation to stimulus objects. Since these ac-
The gravitational center of this discussion is thexquestion whether psychic
qualities can be measured. Cf. Ferguson, Quantitative; Reese,\Application; McGregor,
Scientific; Scates, How; Bartlett, Measurement; Wiener, New Theory.
 LOGICAL ASPECTS OF MEASUREMENT 289
tions vary,‘ the authentic scale problem is to devise instruments with
which to assess the values of such mensurative objects for the mode
of inquiry.
Scale Application. Juxtaposing a rigid rod with an object is con-
ventionally regarded as the primary mensurational operation. No
one can deny that the application of a ruler scale typifies the pro-
cedural factor Of a measuring system. It is frequently remarked
that there are comparatively few measurements which do not re-
solve themselves into reading off a length on a scale.30 Still, it is
inexcusable to single out linear measurement as illustrating more
than a general method. Scale applications vary enormously and al—
ways directly with the situations for which scales are designed.
Added to the influence of the above physics—measurement situa-
tion is the overemphasis of the fact that most measuring situations
constitute scale application, not scale construction. This leads di-
rectly to overstressing scales as instruments. Hence one clings to
the linear-scale model and disregards the fact that linear values
constitute some, but not all, of the results sought and obtained.
The reductio ad absurdum of scale application is to assume that
mensurational operations may be performed without definitely
knowing what is being measured. Mass, it is asserted, was meas-
ured before it was clearly appreciated as an inertial property. The
development of the formula F M X A did not differentiate
between force defined in terms of mass and mass defined in terms
of force.31 As this example clearly indicates, all that is involved
here are different levels of measurement. It is indeed true that
measurement on one level may be very different from measure-
ment on another, and that the earlier may be basic to that per-
formed later. In the present case the measurements were first made
on a simpler'level by means of observations on bodies differently
accelerated. With increased knowledge concerning the factors in—
volved-one can better deScribe the system. We may well assume
that all complex mensurational systems are evolved by a step—by—
step evolution.
Similar assertions are made by psychologists who accept the view
that numerical values are obtainable without knowing what is being
measured. A generalized statement of this sort has it that if a
Ritchie, Scientific, p. 125.
3°

3’
Ibid.,p. 14.8.
290 PSYCHOLOGY AND LOGIC

“structure” is decomposable into countless elements or units, the

number obtained by adding such units gives the measure of the
structure’s magnitude.32 This attitude has recently been developed
in connection with the measurement of intelligence, which has been
regarded as an intangible and unobservable entity determining per-
formances. While measuring the performances it was. assumed that
the imaginary entity intelligence was being measured. But why
interpret this situation as measuring an unknown entity when one
is actually measuring performance? As it happens, when what is
measured once becomes known, it turns out that what was orig-
inally measured was the performance. Surely this iS the case when
the measured performances consist of answers given by different
individuals to “series of conundrums.”33
To such an extent have writers emphasizoed measuring instru—
ments that. they have regarded measurement as a guarantee of the
existence of a thing. Misconstruing Kelvin’s (Thomson’s) famous
assertion in his lecture on Electrical Units of Measurement:
In physical science a first essential step in the direction of learning
any subject is to find principles of numerical reckoning and methods
for pr'acticably measuring some quality connected with it. I often say
that when you can measure what you are speaking about, and express
it in numbers, you know something about it; but when you can not
measure it, when you cannot eXpress it in numbers, your knowledge
is of a meager and unsatisfactory kind: it 'may be the beginning of
knowledge, but you have scarcely, in your thoughts, advanced to the
.stage of science, whatever the matter may be."
they have translated knowing into being, and misplaced Kelvin’s
view which was excellent for the circumstances of which he spoke.
How strange that it should ever be necessary to insist upon tak-
ing properly into account the quantitative or numerical properties
of things! But such is the case whenever one departs from spatio—
temporal objects to talk about magnitudes as important abstract en-
tities. Why avoid any quantitative property of things, as long as it
is aVailable? As a matter of fact, when we deal with actual things
\ .
3’
Horst, Measurement, p. 63 2f. \
33 Ibid.,
p. 635.
3'
Thomson, Popular, vol. I, p. 73. Kelvin’s statement\typifies the attitude that
-science is measurement and stands in contrast to the view that science is not based
upon measurement. See George, Scientist, p. 332.
 LOGICAL ASPECTS OF MEASUREMENT 291
we find that they may have more than one kind of quantitative
property. When liquids can be differentiated as drops they may 're—
quire counting; when not, they can be treated by weight ‘or volume.
Scale Fitting. Complex mensurational situations usually call for
more ingenious procedures than the mere application of scales.
Analogizing and transforming operations are required to achieve
accurate specifications of properties and relations. An' excellent
example is the substitution of weighing for linear measurement
when the problem concerns the precise determination of the magni-
tude of a very thin object.
Faraday weighed 2000 leaves of gold, each 3%; inches square, and
found them equal to 384. grains. From the known. Specific gravity
of gold it was easy to calculate that the average thickness of the leaves
was 1/282.000 of an inch.’5

Jevons Offers a wide range of illustrations of what he calls

indirect measurement. The most striking instances consist Of
substituting forces in equilibrium; in other words, Substituting
static conditions for dynamic or kinematic events, as in the case of
D’Alembert’s principle.
The latter example, illustrating the more analogical form of
transformation, has relatively more to do with constructional op—
erations. _Transformations developed during the manipulative
processes Of measuring may be described as interchanging the ascer-
tainable properties of the measured things with attributes deemed
necessary or eXpedient. The most radical transformations belong to
the interpretative phase. For calculative purposes, events or ob-
jects become points in a field of coordinates. In more remote cases,
numerals representing items become parts of curves representing
trend lines.
Interpolation and Extrapolation. Complex measurements in-
volve both calculative and manipulative operations (p. 276). The
calculative operations may be equated with judging and estimating
activities. In general we may separate measuring situations on the
basis of direct or indirect manipulation. In the former the objects
to be measured are stressed; in theilatter the behavioral strategy
necessary to assay the thing in question is emphasized. Prominent
”Jevons, Principles, p. 296, from material derived from Faraday, Chemical Re—
searches,_p. 393.
292 PSYCHOLOGY AND LOGIC

among the calculative schemes employed to control obj ects beyond

the range of direct manipulation are the mathematical processes of
interpolation and extrapolation.
The substitution, in whole or part, of calculative devices for
direct manipulation is analogous to making indirect estimates on the
basis of samples. Obviously, the extreme case iS that in which no
application of a definite scale or other instrument is possible; thus
from casual or partial observation one is obliged to build up a com-
plete system of speculative estimates.
D. Recording. NO inconsiderable part of mensurative operations
consists of recording the results obtained. This point is obvious in
view of the importance of holding to the basic information secured.
Planck36 asserts that measurement is the recording of answers which
nature gives to questions put to it in the form of experiments.
Within the scope of recording lie such operatjons as tabulating,
ordering, and graphing data. The expertness required for dealing
with this material in complex situations is directly proportional to
the care demanded in building and applying scales and standards.
Not only are recording operations integral parts of the total meas—
uring'behavior, but they influence the other operational procedures
and for this reason are highly important in the all—around measur-
ing situation.
The importance of recording operations reaches down' to the
very marks and schemes employed. Notice the part played by Signs,
graphs, and tables in recording results for future treatment. Recall
the role of proper notation in the development of effective calcula—
tion and theoretical mathematics.37 On a higher plane we discern at
once the advantages and disadvantages of particular techniques in
organizing data for summing, averaging, and integratingQMerely
to consider the extensive place that tabulating and graphing occupy
in the establishment of trends and constants as a basis for scientific
laws is to acknowledge the proper value of recording operations in
mensurative enterprises.”
E. Statistical Operations. The statistical factors of the total men—
surational system ai'e performed primarily in order to verify’and
3°
Meaning, p. 235. .
3:Cf. vol. I, p. 213.
∙
38We are reminded of Whewell’s (Novum) inclusion of ‘the method of curves
among his special inductive, methods. \
 LOGICAL ASPECTS OF MEASUREMENT
293
test such features as obtaining, identifying, organizing, and com-
paring data. Whenever computations are included in the system,
checks must be selected and applied. To illustrate controls for cal-
culations and correlations we mention merely the method of least
squares and similar statistical tests.39
A word of caution: The interrelation of the various factors in a
mensurative system is symmetrical. While it is true that statistical
operations constitute essential checks on the other operations, it is
likewise true that statistical operations are conditioned by the other
features of the enterprise. The total set of the system’s factors con—
stitutes the criterion for the component operations.

How ARBITRARY ARE UNITS OR STANDARDS?

Expert metrological opinion vacillates between two views: (1)
that units and standards are, or ought to be, absolute and (2) that
they are completely arbitrary. It is tempting to follow the conven-
tional procedure of declaring that the truth lies somewhere in the
mean, but this is not the case. Both views are based on oversimpli-
fied premises concerning measurement; and from two wrong no—
tions it is hardly possible to extract a correct one. In neither instance
are the many complex features of mensurational system building
taken into account. _
Although it is Obvious that units and standards can not be ab-
stracted sO‘ far from the obj ects and events interacted with as to be
made absolute, the attempt is nevertheless made. It is probable that
Gauss40 did not intend his time, length, and mass units to be more
than relatively absolute and fundamental. Even though metrolo—
gists have taken pains to reject the absolutistic construct,41 there
seems to be some notion that such a’standard can be approximated.
For one thing, observe the attitude of metrologists in the follow-
ing statement:
The science of measurement is in practice restricted to mean meas-
urement of the three fundamental quantities, mass, length, and time,
from which all other quantities, such, for example, as volume, density,
velocity, acceleration, force and power, are derived.42
Cf. Fisher, Statistical.
4o
∙
IntensItas.
Cf.‘for example, Runge, Maass; Sears, Measurements.
’1

’3
Sears, Measurements, p. 841.
294. PSYCHOLOGY AND LOGIC

This not merely reflects the physicist’s bias or the influence of

mathematical ideals, but implies something beyond the urge to
rigor and precision. We might trace it to the cultural assumption of
the essentially quantitative character of reality and to the influence
of the mechanical domination of science. The latter influence is
clearly visible in Gauss’ desire to carry‘ electric measurements back
to basic mechanical standards.
An interesting speculation is how close the correlation might be
between the absolutistic idea and the palpable character of “funda-
mental” and “absolute” standards. Obviously, units that can be
embodied and conserved in platinum and iridium must appear as
somewhat ultimate. The physicist’s attempt .to substitute the wave
length of the red cadmium line as a standard‘13 may seem to modify
this motive; on the other hand, it emphasizes how basic the factor
of length is among metrologists.
The importance attached to restricted and r‘elatively absolute
standards is evident in the admirable development of dimensional
theory. It is undeniable that this theory indicates how well organ-
ized mensurational systems can be; nevertheless, there are objec-
tions 'to overstressing certain standards and units, or larger systems
made from them.
Length, for example, is important, but it is no more basic than
any other dimension. Moreover, we must hold it to the situations
where it is appropriate. We have already suggested the cultural in—
fluence operating here. Indeed, linear mensuration is historically
the first type to be developed. Nor need we question its most pre~
cise procedures. Still, by overlooking other important mensurative
forms we drift into a false analogy.
Furthermore,'we need not be overly influenced by pragmatic
conditions. While in practice we are limited by the particular in—
struments we used and by the practical difliculties in varying them,
such circumstances Should not be too binding. Because of the funda-
mental integration of our culture and its impedimenta we are re—
stricted by the gradual evolution of instruments and processes.
Automobiles still Show their descent from the horse—drawn vehicle,
despite the fact that the basic function of‘instruments of transporta—
tion allows for tremendous variation in Shape and pattern. The
Michelson, Light.
 LOGICAL ASPECTS OF MEASUREMENT 295
same thing may be said, of course, for the roads upon which they
travel.
Turning now to the notion that measuring units and standards
are arbitrary, we observe at once that this is only true within narrow
limits. Even though units as features of mensuration systems are
constructional products, they are derived from specific situations
and are modified by self—correcting operations upon actual things.
N O scientist accepts as satisfactory for any serious purpose arbi-
trary operations with indifferent objects as instruments or tools.
Even for historically early units we must presuppose the insight
necessary to fit measuring operations to the needs of the situation.
The growth of scientific sophistication and of particular kinds of
measurement demonstrates an increasing precision in the con—
formity of measuring instruments and operations to things meas-
ured.
As factors in systems, units can not be more arbitrary than the
total system. Even writers like Campbell,“ who regard units as
arbitrary, allow for no arbitrariness in any other measuring ele-
ment. But even the unit’s apparent arbitrariness disappears when
the measuring system is not taken to be, as in Campbell’s case, an
arithmetical system, but looked upon as a means to a specific end.
Is it possible that the idea of the arbitrariness of units stems from
confusing .units with their names or descriptions? The history of
the development and adoption of modern electrical units provides
some basis for this suggestion. 3

THE NATURE OF QUANTITY, NUMBER, AND MAGNITUDE

As inveterate apostles of precision, writers on the logic of meas—
urement are confused concerning the referents of such terms as
quantity, number, and magnitude. This situation is not simply one
of semantic deficiency. It indicates the depth _to which such prob-
lems penetrate metaphysical foundations.
We have already suggested the orientation of Helmholtz’s ideas
in philosophical territory. To take a more recent example, Rus-
sell,“5 building on an absolute Platonic foundation, contrasts magni-
tude and quantity on the basis that the former has to do with some-
“ Physics, p. 290ff.
’5
Principles, chap. 19.
296 PSYCHOLOGY AND LOGIC

thing greater or less but never equal (p. 164.). Quantities, how-
ever, as the less abstract entities, can be equal in that they possess
the same magnitude. Russell illustrates his point by remarking
that “an actual footrule is a quantity; its length is a magnitude” (p.
159). Lenzen,46 who adopts this View, asserts that quantity is a
particular, while magnitude is a universal.
Less theoretical writers use the terms quantity and magnitude
interchangeably, but, inflUenced by mathematical abstractionism,
regard them as standing for some entity different from the actual
Objects measured. Both practical-and theoretical writers consider
magnitudes as some abstract property of extension and speak of
measuring magnitude or quantity. Such a paradox can easily be
avoided by taking into account that the measurer is trying to ascer-
tain the size of something, the amount ofosome quality, or the
character of some relation an object bears to another object. One
may then use the term magnitude for Size factors, and quantity for
amount units. Similarly, the term number, also used synonymously
with these two, might be reserved for relationship determination.
For number, order may also be important. In every case these
terms, singly and together, must be associated with dimensions of
obj ects as ascertained through mensurational procedures.
Since upon an operational basis measurement is'not Simply a
matter of constructing an abstract or logical mathematical system,
but rather an enterprise carried on for specific purposes, there is a
limit to the amount of reduction and elimination of that which is
measured.
Admittedly, there exists no accepted standard for the terms
quantity, number, and magnitude. Their use depends upon the
measuring job. When we abstract from actual things, hOwever,
number, quantity, and magnitude become indifferent or equivalent
terms, and can be adapted to any sort of metaphysical position.
MEASURING AND COUNTING
To clarify the relation between measuring and counting is to
cast considerable light upon measurement principles.
For Helmholtz measuring is a coiInting procedure when the
results are denominate or concrete numbers expressing the values
of magnitudes. The latter constitute objects,\_or attributes of ob-
Nature, p. 22.
 LOGICAL ASPECTS OF MEASUREMENT 297
jects, which upon comparison permit the distinction of greater,
equal, or smaller.“7 Here the absorption of measurement by arith—
metic is plain, but the empiricistic atititude exhibited does full jus—
tice to the auspices under which measuring procedures Operate.
On more abstruse grounds rationalistic writers who assimilate
measurement to counting do so because of an adherence to some
absolutistic logit of numberJThe numerals attached to things in
measuring operations are regarded as symbols for the numbers be—
longing to a transcendental system of relations. It is for this fea—
'Son that the properties of addition and independence are made basic
to measurement.48 The statement that “The primary purpose of
numerals is to express facts about number’”9 is a clear indication of
the great influence abstract relations exert upon ideas concerning
measurement. This point is further enforced by the statement:
In order that the measurement shall be really satisfactory, the prop—
erty measured in this way must Obey the rules of arithmetic.50
When measurement is approached from a more concrete scien—
tific viewpoint, especially when the difficulties of precise determina-
tion of properties are apparent, measuring and counting do not
appear so closely related. In fact, the two are sharply separated on
the ground that, whereas counting involves cardinal numbers, and
no manipulation, measurement involves ratios and manipulatory
operations.51 On the surface this differentiation appears reasonable,
but unfortunately it is still based upon an insuflicient handling of
number problems. There is no consideration of actual mensura-
tional work and its relation to counting in concrete situations.
If we turn to specific interbehavior we find that in some cases
counting is measuring, while in others it is not. Counting is found
to comprise various kinds of operation when we depart from the
standpoint of mathematics and approach interbehavior with things.
Granting the interpretation that measuring is a process of ascer—
taining the properties of things, we must include counting and
calculating as measuring actions. Situations demanding knowledge
concerning numerical or quantitative properties imply that such
’7
Ziihlen, p. 35.
Campbell, Measurement, p. 127.
’9
Ibid., p. 122.
5°
Ibidi‘,\p. 1 27.
’1
Ritchie, Scientific, chap. 5.
298 PSYCHOLOGY AND LOGIC

properties are features of things in the same degree as color, den-

sity, length, and area. On the other hand, counting may be merely
reciting the progression of the II —|— 1 relation by means of its con-
ventionally named items. Such counting can be connected with
measuring only in a very indirect way. Counting properly associ-
ated with measuring may be distinguished by the criterion of facili-
tating an adjustment to some difficult situation. When dealing with
recalcitrant materials measuring may require much complex calcu-
lation. .
Mensurational operations cover a vast range of specific perform-
ances. Counting operations are definitely included. But counting
may never be separated from manipulations.‘ In complex situations
counting may require moving things about and instrumentally ex-
amining small objects. A relevant example is supplied by the work
and apparatus required to count blood cells accurately.
Herodotus Offers an interesting account of the manner in which
Xerxes ascertained that his army at Doriscus had the numerical
value ,of 1,700,000:
A body of ten thousand men was brought to a certain place, and
the men were made to Stand as close together as possible; after which
a circle was drawn around them, and the men were let go: then where
the circle had been, a fence was built about the height of a man’s mid-
dle; and the enclosure was filled continually with fresh troops, till
the whole army had in this way been numbered.‘52
"2
History, p. 377.
 CHAPTER XXIV

METROLOGY AND METASCIENCE

PROBLEMS OF MENSURATIONAL LOGIC

0 UR study of measurement, we believe, has amply justified the

suggestion made at the beginning of the preceding chapter
that measuring situations illumine every feature of the system-
building enterprise. Measurement may thus be regarded as
thoroughly integrated with systemology.
This does not mean, however, that mensurational logic is in any
sense a static domain. Indeed, it constantly engenders problems just
as any other logic does, and in no smaller degree than science itself
—the primary environment of mensurational theory.
In the present chapter we are concerned with several problems
provoked by the apparent disharmony between modern scientific
investigations and traditional mensurational practice. As we shall
see, the issues break through the conventional bounds of science to
the neighboring metascientific environment. The problems we are
to consider arise out of the three following situations.
(1) The influence of modern science, especially microscopic
physics (quantology), upon metrology.
Interest in measurement has been keenly intensified by the im—
portant issues arising in relativity theory and quantum mechanics.
Especially has the Heisenberg indeterminacy principle quickened
the study of metrological foundations (p. 309E.) . A significant out~
come of the newer developments in physics is the freedom they
afford from the fetters Of traditional logic of measurement.
(2) The bearing of symbolic logic on the relations between
metrology and general systemology.
The deveIOpment of modern science coincides with that of for—
mal logic. Naturally, then, logicians attempt to harmonize their
discipline with the newer scientific developments in measurement.
Their success at coordination, we shall discover, depends upon the
competence of logic to match concretemensurational events.
(3) The consequences of microscopic measurement for scientific
foundations.
Such‘fundamental issues as are raised by the mensurational prob—
300 PSYCI—I OLOGY AND LOGIC

lems of microscopic physics have gone way beyond the bounds of

classical science. Measurement problems are intermixed with ques-
tions concerning the nature of scientific observation, knowledge,
cause, objectivity, even reality itself.
The bridge between metrology and such metascientific problems
Obviously consists of mensurational systems. To examine briefly a
typical series of such systems prepares the way for the analysis of
the metrological issues developed in the train of quantum me—
chanics.
SYSTEM TYPE AND METROLOGICAL EVOLUTION
Our investigation has certainly demonstrated that when we start
from measurement situations, rather than from the applied-mathe—
matics theory, we encounter a great numbercand variety of metro-
logical systems. Up to this point we have stressed chiefly the cor-
respondence between measurement systems and the situations to
which they belong. N ow we must emphasize the fact that measure—
ment systems occupy distinctive chronological positions in the evo-
lution of science. On the whole, measurement principles, like the
term measurement, comprise a heritage from the time when geo-
metric mensuration dominated the scientific scene. Today we know
that historical mensurational principles belong to but one of many
kinds of metrological situations. A comparison of metrological sys-
tems throws light on the evolution Of science and its mensurational
techniques.
As a basis for our comparison we use four variant criteria—
namely, (a) things measured, (b) instruments employed, (c)
standards used, and (d) operations involved.
(1) Measuring Systems. Influenced in part by traditional lin-
guistic usage we place under this heading all mensurational systems
involving linear relations. Included are all activities to ascertain
lengths or distances between points. Points constitute actual places
on or between objects. Though we cannot avoid linguistic habits in
referring to the values ascertained and recorded, we repeat that we
are concerned not with abstruse geometrical magnitudes but with
concrete things and relations. We may\say, then, that the values
here comprise all lengths, areas, volumes, and distances.
SO great is the number and variety of thiIi‘gs belonging to the
present class of measuring situations that listing is impossible,
 METROLOGY AND METASCIENCE
I
301
also unnecessary. Its range extends from everyday objects to astro-
nomical bodies, to radiation of every variety, and to abstract geo-
metric relations as represented by drawn figures and analytic for-
mulae. The instruments used indicate the specific character of
mensurating situations. To the conventional rigid body may be
added all sorts of optical instruments required for triangulation
and other long—distance procedures. In the simplest measuring
situations the pertinent operations consist merely of juxtaposing a
Standard measuring instrument as a workable unit. Sometimes an
object to be fitted into a given place is juxtaposed with the place to
be covered.
(2) Weighing Systems. Weighing systems involve things that
can be heaped together or separated. The primary comparisOn is
between (a) one or more bodies and (b) distinct heaps. The typi-
cal metrological weighing instrument is the beam balance con-
structed with equal lever arms, so that an attached pointer in an
equilibrium position indicates the symmetry of the system on the
basis of the bodies hung from the arms. Approximations to the re-
quired results are obtained from various spring arrangements.
Aside from the manipulation of balances, weighing operations also
include various procedures for determining the homogeneity of the‘
material to be weighed. Scales can be balanced by adding salt or
sand to sugar, but the results can not be carried over to Situations
demanding a more discriminating result than bulk equivalence.
Weighing information can also be Obtained by an equivalent
form of‘measuring. Assume that the homogeneous material is
available in large heaps; then by arranging a rectangular container
the right amount of, say, sand can be determined by linear measure.—
ment.
Interesting here is Maxwell’s suggestion of measuring sub—
stances by the effects they produce:
. . . if we are dealing with sulphuric acid of uniform strength, we
may estimate the quantity of a given portion of it in several ways. We
may weigh it, we may pour it into a graduated vessel, and SO measure
its volume, or we. may ascertain how much of a standard solution of
potash it will neutralize.’
(3) Counting Systems. When the metrological question is how
\

1Matter, p. 33.
302 PSYCHOLOGY AND LOGIC

many rather than how far, or how long in duration, or how much,
we develop a counting system. Accordingly, the things dealt with
consist of unit objects which can be enumerated and added. The
operations constitute the behavior of the operator himself, unless
some recording machine (manual or electrical registration device—
“Geiger pulse counter”) is utilized.
In situations in which no apparatus is employed we must still
distinguish between the direct interaction with objects and the more
or less long-range operations performed in dealing with abstract
relations. Behavior such as uttering number names illustrates long-
range contacts with things.
(4) Calculating Systems. Generally speaking, calculative as-
pects Of metrology concern auxiliary features of gathering informa-
tion. Operations upon things are substitutive in character, and the
instruments consist of mathematical processes applicable to every
type of Obj ect.
The rules of proper identification and classification of items are
determined by the specific interbehavioral fields in which they oc-
cur. As substitution Operations calculation processes are involved
with recorded data specifying lengths, areas, volumes, numbers of
things, and so on. More specifically, the operations consist of mul-
tiplying, dividing, and the computing of averages, rates, correla-
tions, etc. Perhaps the most definite features of such systems are
various ratios—rates of production, motion, and changes of all
sorts.
( 5) Ordering Systems. TOO numerous to mention are the kinds
of things subject to this type of systemization. In fact there is
hardly any sort of object for which an ordering rule or standard
can not be contrived. Moving things into certain positions on the
basis of previously ascertained information concerning their quali-
ties and prOperties is the fundamental operation. When things are
not sufliciently tangible to be handled in this way they can be in-
directly ordered. Recall that among the enormous range of things
subject to ordering are traits of every variety—merits, achieve-
ments, actions, as well as social and cultural properties.
(6) Estimating Systems. Objects belonging to estimating sys-
tems are generally characterized as less definite and less manage-
able than objectsIn other systems. As a matter of fact, writers who
regard the physicist’s arithmetical metrology aSthe only genuine
 METROLOGY AND METASCIENCE 303
type discriminate against estimating activities as measuring alto-
gether.’
The measurer’s Operations occupy a unique place, indeed a focal
position, in the metrological system. The ability to set up a system
depends to a great extent upon his eXpertness. In a genuine sense,
therefore, the operator himself constitutes the metrological instru-
ment. The diagnostician in medical practice, for instance, is a more
or less eflicient instrument qperating like a ruler, galvanometer, or
balance. One is reminded here of the early days of electrical science
when Cavendish used himself as a more or less sensitive galva-l
nometer. Concerning Cavendish’s work, Maxwell wrote:
Cavendish is the first verifier of Ohm’s law, for he finds by succes-
sive series of experiments that the resistance is as the following power
of the velocity, 1.08, 1.03, .980, and concludes that it is as the first
power. All this by the physiological galvanometer.3
Such incidents in the historical development of science suggest a
tolerance toward various as yet unstandardized metrological sys—
tems.
A distinctive type of estimating measurement system is that by
which the mass of the meson (mesotron) is determined. While
cosmic-ray mesons are too energetic to be treated directly by mag-
netic fields, the mass can be indirectly estimated.‘
Becausei‘of the interrelation of the charge, mass, and velocity
values of particles, the cloud-chamber technique furnishes photo—
graphs yielding the following quantities which can be employed in
a measuring system: (I) curvature of the track, (2) range of the
particle, ( 3) ionization per centimeter of path, and (4.) rates of
change of these quantities during the passage of the particle.
The character of the measuring system is illustrated by the fact
that while ionization does not depend greatly upon the mass it
does depend upon the velocity of a particle. To estimate the mass,
then, one must invoke the relationship between mass and velocity.
(7) Evaluating Systems. The measured things comprised
within the present type of system consist of second or higher-order
relations. The measurer’s interest is in the significance of things,
2'Cf. Johnson, Pseudomathematics.
’ Campbell and Garnett, Life, p. 402.
’ Stran‘athan, “Particles,” p. 520.
304. PSYCHOLOGY AND LOGIC

rather than in such crude surface properties as length and weight.

To a great extent his operations amount to imposing a value upon
an object, rather than drawing it out of the thing itself. Obviously,
the criteria are not as rigid as in other metrological systems, but en—
tail much more viable acts of assumption and postulation. Another
way of saying this is that the mensurative standards are not as close
to the things measured as to the behavior and behavioral history of
theevaluating individual.
A prominent factor here is the influence of the cultural condi—
tions which comprise the immediate milieu of the measurer’s ac-
tivity. Excellent examples are the various, axiological techniques
(chap. 17) organized in an attempt to
Despite the fact that the builders of suchmeasureaesthetic values.
systems are excessively
imbued by the ideals of measurement in physics, their work plainly
shows the large place that the measurer and his cultural traits oc-
cupy in the completed structure. These points are well illustrated
by the formula M=O/ C developed for measurement in the arts.5
M as aesthetic measure or quantitative index is defined as the ratio
of aesthetic association to motor adjustments Of the person reacting
to or appreciating the comparative pleasantness-producing qualities
of aesthetic obj ects.
(8) E guating Systems. To a great extent equating situations,
along with the included equating behavior, may be regarded as
generalized metrological types. E’quating one thing with another
appears to be the fundamental comparing activity of all measure-
ment. Nevertheless, under this designation it is possible to isolate
a technical mensurating system.
Basically, measurement in equating situations consists of discov-
ering equivalence relations. Mensurational obj ects in these in-
stances are not directly manipulable. They consist of abstractive
and constructive things such as intricately interrelated geometric
objects. Metrological systems contrived for such Obj ects lead to the
ascertainment of'various sorts of congruity relations. In general,
equating measurements are involved with a variety of nonmetric
mathematical systems and situations.~Moreover, there is a decided
emphasis on the standards constructed‘for producing the matches
and equivalences.
(9) Probability Systems. The most important examples of prob-
5
Birkhoff, Aesthetic.
 M ETROLOGY AND M ETASCIENCE 305
ability measurements are localizable in the domain of atomic phys—
ics. What is dealt with are nonvisible things, such as energy dis—
turbances and energy interchanges. Here the greatest resources of
analogy and complementarity must be exploited to accomplish de—
sirable results.G Atomic or microscopic measurements must be built
up from such elementary and indexical observables as cloud cham-
ber tracks, oil—drop behavior in magnetic fields, and the activation
of silver grains on photographic plates.7
To measure the behavior of photons or electrons requires the
observer to invoke the statistical probability that particles or cor—
puscles passing through a slit will distribute themselves at cal—-
culable terminal points On a screen. In one sense such corpuscular
behavior is comparable to statistical or group activities plus the
variations provided by the specific field conditions in which the par-
ticular group items occur. Among the differences we must note
that the probabilities dealt with in quantum mechanics are not de-
rived from generalizations concerning individuals, but directly and
immediately from groups and aggregations as such.8
Probability measurement differs greatly from other sorts in ac-
cordance with the general principle that each situation is unique. It
is an error, however, to enlarge the differences between Situations
to the dimensions of kind rather than of degree. Probability meas—
urements merely involve less directly manipulative operations than
other types.
Since probability measurements are most intimately at home in
microscopic physics, what are the basic variations from macrOSCOpic
situations? Certainly we should disagree that in the former situa-
tions causality in the proper sense is excluded. True it is, however,
that hierarchies of construction must be relatively more empha-
sized. The intensification and broadening of knowledge concerning
complex and recondite events call for courageous initiative in ana-
lytical and constructional work. This implies that the operations in
probability measurement are refined and original acts of system
building leading to an understanding of things and events on the
basis of more elaborate analogical treatment than is given to ob-
jects immediately observable and manipulable.
Bohr, Causality.
Margenau, Critical; Jordan, Process.
3 Einstein-and Infeld, Evolution, p. 299f.
306 PSYCHOLOGY AND LOGIC

Coincident with the growth and transformation of Scientific prob-

lems, measuring situations also evolve and thus may become enor-
mously complex. Unfortunately, this fact leads those scientists who
entertain traditional presuppositions to translate metrological prob—
lems into metascientific issues. For this reason clarification of. the
following measurement problems promises a gain for both science
and logic.

MEASUREMENT: ABSTRACT AND CONCRETE

Since measuring is a procedure for comparing at least two things
—that is, ascertaining relations—it is easy to overstress the rela—
tional feature. Moreover, it has become conventional that relations
are abstract or formal. The peak of abstractionism is attained when
the relation is represented by a numeral. 6
From an interbehavioral standpoint no numeral or number is
apriorily an empty abstraction. This does not mean that we may not
differentiate between a numeral and the mensurational operations
which engender it. Still, in the interest of effective analysis the op—
eration should be kept in the foreground.
While discussing symbols9 we pointed out the interdependence
of numbers and things. This interdependence can not be over-
looked either when we deal with number symbols or with numbers
as (a) ascertained relations and (b) numerical qualities.
Jeffreys10 asks: Is an angle a number in terms of a certain meas—
ure? Is a right angle /2 :n: and the angle of a complete circum—
ference of a circle 211:? In each case it must be specified that the
numbers of radians are in question. The same writer goes on to
Show that while 2 sheep
− −
3 sheep— 5 sheep,2 sheep 3 houses
=l= 5 of anything. AS a matter of fact,- both the concreteness of num-
bers and their instrumental character can be Observed in the fact
that 2 sheep —I— 3 houses do make 5 things. Of course, it depends
on the situation and on the mensurational postulates how one in-
terprets things. Why exclude even an equation like 2 sheep —|— 3
houses = 5x in which x symbolizes trouble? To increase one’s pos-
sessions so far is to increase one’5 trouble by a factor of 5 as com-
pared with possessing only 1 or no animal. ∙
In addition, we must check the range Of allowed abstractiOnism.
"Vol. I, p. 213f. \

10 Scientific, p. 138.
 METROLOGY AND METASCIENCE 307
In order to balance a weight on one pan we can heap sugar plus
sand on the other. From an abstract standpoint the materials placed
on the pan are sufliciently homogeneous. Not SO when a specific
quantity of Sugar is called for. Only extremely abstract situations
can be satisfied by simple equality or inequality.
Abstractionizing measurement results from confusing concrete
measuring operations with generalized descriptions of them. Meas-
uring operations may be legitimately described as work done to ob-
tain numerals, but then the other features of the measuring situa-
tion must not be slurred over. Descriptions constitute verbal analo—
gies to measuring manipulation. However, generalizing and ab—
stracting must not be carried so far as to allow the equating of lin—
ear measurement with the measurement of energy, of strength of
materials, and of other processes leading to coeflicients of expan-
sion, torques, etc.
Thus, when describing measuring situations, it becomes illicit
abstraction to overstress the relational factors, so that arithmetic be-
comes the center and the model. Events are not arithmetic, though
arithmetical operations are events. Conventional descriptions of
measurement carry relations far beyond actual measurement situa-
tions.
The basic point here may be formulated as that oft—repeated
warning not to absorb events in descriptions. When quantum me-
chanical measurements are under consideration the question
whether we are dealing with waves or particles suggests that, un-
like ordinary physical descriptions made with direct reference to
Space and time, it is now expedient to resort to the relatively more
constructional description in terms of probabilities.11

MEASUREMENT: CONSTRUCTION AND OBSERVATION

Problems of construction and observation have become intensi-
fied in the scientific measurement situation as a consequence of the
evolution of relativity and quantum physics. _
Bohr puts the matter concisely. Einstein’s relativistic emphasis
that:
Every Observation or measurement ultimately rests on the co-
incidence of two independent events at the same- Space-time point12
1'Einstein p.and53f.;
“ Ato’rnic, Infeld, Evolution, p. 3o8f.
also see, p. 97f.
308 PSYCHOLOGY AND LOGIC

he thinks conflicts with the findings of atomic research showing an

interaction of Object and measuring instrument: ∙
An independent reality in the ordinary physical sense can neither
be ascribed to the phenomena nor to the agencies of Observation."3
Hence the question has been posed whether measurement in
general does not involve a constructive operation creating both the
measurement and thing measured.“ This has led to the juxtapo-
sition of metaphysical and physical problems, and to the physicist’s
occupation with questions concerning Objectivity and reality. A
thorough clarification Of the nature of Observation and construction
is called for, with respect not only to measurement but to scientific
workin general.
Observation. In its lowest terms observing consists of perform-
ing a differential response to something exi§ting independently of
the immediate field. The simplest observation involves no instru—
mental aid (microscope, telescope), although it may require assum-
ing a facilitating position with regard to the object, aS in modifying
illumination conditions or removing something intervening be“
tween the observer and the Obj ect. To observe more complex'things
necessitates instrumental aid, but certainly in the case of macro-
scopic fields no question arises as to the independent‘ existence of
the observed. AS Jordan remarks:
We know of the planet Pluto only because we possess astronomical
observatories; but we believe Pluto to have existed already in the
time of homo neandertalensis.15
Surely, no one with a “mind” undebauched by Berkeleyan theology
believes that Leeuwenhoeck created his animalcules just because he
was able to magnify them to 270 times their size. .It is a strange
circumstance of scientific thOught that writers seem to adopt cre—
ative epistemology as though they believed that Herschel and
Galle' are responsible for the existence of Uranus and Neptune, or
that Ritter and Herschel created ultraviolet and infrared radiation
respectively.
At the bottom of the modern physicist’s difficulties with quantum
measurements lies his sensationistic-psychological View. Observa-
13
Ibid., p. 54.
\

\
" Einstein, Reply, p. 669..
1’
Process, p. 271.
 METROLOGY AND METASCIENCE 309
tion as an operation he carries back to sensations. Accordingly, when
he can not assume the presence of sensations he questions the ex-
istence of events. The literature on quantum mechanics testifies to
the difliculties the psychistic view hasIn]ected into the problems of
cause, scientific Obj ectivity, and realityIn physics.
Modern objective psychology allows nowhere for the theory
that observation consumes the thing observed. No matter how in-
direct and remote the act of observation may be from the observed
object, the interbehavior is always the rock upon which all descrip—
tions and interpretations are founded. It may be helpful in illus-
trating this point to indicate a continuum of varying observational
interbehaviors.
In the case of macroscopic physics, the existential independence
of observed objects need never be questioned. But these objects
may be bound up with observation. Objects as perceived are inter-
dependent with the act of perceiving. The perceptive act, however,
does not create the perceived object. The psychologist is also con-
cerned with such independent objects as acts of persons—for ex-
ample, assertions which linguistically produce objects—and with
other no'less' independent objects once they are created, such as
the products of assertions—phlogiston, ether, and caloric. Though
such things exist only as human inventions they often constitute
objects lincultural situations as things talked about or believed in.
What is required is to distinguish (a) things whose existence is in—
dependent of human behavior but which come to be known, (b)
things in knowing relations, and (c) things existing only by human
construction.
Construction. Scientific construction consists basically of descrip-
tions or reports of what occurred during observational contacts with
events. To record or describe an event by constructing a protocol is
to create an index or representation of it. How simple or complex
the construction is depends upon the simplicity or complexity of the
events treated, the ease or difliculty of getting into contact with
things, and the type of interpretation needed. By interpretation is
meant, of course, the interrelation of present events with others Ob-
served at various previous or contemporaneous periods.
It follows that the measurement problems of modern physics de-
mand more complicated construction than the descriptions of classi-
cal meehanics, though the differences are Only difierences of de-
310 PSYCHOLOGY AND LOGIC

gree, not of kind. Many of the difl'iculties which physicists expe—

rience with relativity and quanta—measurement problems are trace—
able to the attempt to treat atomic and electronic events on the same
plan as macroscopic events.16 Obviously, measurements in modern
physics call for more spontaneous and original constructions. New
and bolder forms of mathematical equations are required. The
situation must be clearly faced that microscopic events are less eas—
ily approachable, less easily Observed. Still, the scientific work and
the systems built are directly continuous with each other.
Quantum measurements provide no occasion for the illicit epis-
temology and ontology with which physicists have recently been
struggling. Nothing in the history of atomic physics lends coun-
tenance to the confusion of authentic construction with any sort of
mysticism. The fact that knowledge concerning energy involves
intricate operations upon the type and quantity of energy provides
no 'warrant for assuming that the existence Of energy depends upon
anything the operator does. The operator’s most effective means of
controlling energy changes and directing their distribution consist
of his interbehavior with events.
In' the interest of clarity we must Sharply differentiate our view
of construction from that of psychistically-oriented writers like
Morgan, Pearson, and Margenau. A construct for Morgan and
Pearson consists of present and stored sense impressions (“sensa-
tions”). These constructs constitute external obj ects—that is,
things known as compared with knowing, mind, or consciousness.
Constructs or external objects are projected from consciousness to
make up the real world."
Margenau’s construct differs from the Morgan-Pearson type in
that it includes, besides sense material, also some factors like the
Kantian categories—for example, the property of permanence or
continuity of existence.18 Pearson’s constructs belong to Humian
phenomenological metaphysics, while Margenau’s 'stem from
Kant’s transcendental realism. Both of these constructs are worlds
apart from the descriptive and interpretative products of workers
who derive constructs from Observational, experimental, and men-
surative contacts with the things about
whichthey construct.
Loeb and Adams, Development, p. 627; Menzel and Layzer, Physical, p. 304.
17Pearson, Grammar, pp. 40-42, 58, 67f., et passim.
18Reality, p. 301; see also Bentley, Physicists.
 METROLOGY AND METASCIENCE 31 I
All constructs pertaining to nonscientific and scientific description
and interpretation are built up from such contacts. Thus, “sensa-
tions,” “sense impressions,” and “sense data” are constructed on the
basis of contacts with the primarily qualitative characteristics Of
things, whereas constructs like cause, permanence, and relation
arise from more direct contact with relational properties.
We repeatz‘The retention by writers on quantum measurement
of psychistic psychology leads them into an unnecessary and
troublesome entanglement with metaphysics. In the following sec-
tions we examine this issue with respect to causation, objectivity,
and reality.
MEASUREMENT AND CAUSE
Bohr writes:
Causality may be considered as a mode Of perception by which we
reduce our sense impressions to order.”
'
This Kantian view goes contrary to the observer’s actual operations
with stimulus objects. Cause is not ‘a mental or subjective principle
which unifies psychic qualities to form objects. We find no basis in
any fact for such a construction. What observed facts show are in-
dividuals Operating upon things directly or indirectly. Causal de-
scription results from the activity of analyzing events in order to
discover or construct a law which refers to the organization or in-
terrelation of the component factors in a field (p. 157f.).
In a similar manner Einstein,20 who is closer than Bohr21 to the
position that scientific work constitutes contacts with originally in-
dependent events, proceeds valiantly on the level of concrete in-
teraction with things, but in the end comes back to impressions.22
. And so we have the unusual paradox that measurement, which
is regarded as so powerful a criterion of precision and validity,
brings mysticism into science. At once we must differentiate
conclusions concerning free will, indeterminism, vitalism, and other
supernaturalistic matters from more legitimate questions concern-
ing the limits of observation and construction. Bohr23 has indicated-
\
1"
Atomic, p. I 1 6f.
”Einstein, Podolsky, Rosen, Can quantum-mechanical.
2’
Bohr,\Can quantum-mechanical.
22 See, Einstein and Infeld, Evolution, p. 310f ; Einstein, Reply, Physics, et passim.

Cau\sality.
3"
3I2 PSYCHOLOGY AND LOGIC

hisvrecoil from the former problems. But even the more modest in-
volvement with indeterminacies maintains a too-close'attachment of
science to historical philosophy. What is determinism? The'proper
answer is suggested by the entire scientific enterprise. The event
with which the scientist begins may consist of closely interbehaving
factors which he measures and modifies in order to reach descrip—
tions and laws.
No inconsiderable portion of a scientist’s activity must be de-
voted to the elimination of wrong views and badly constructed or
unverifiable hypotheses. Many initial problems center in alleged
events . To sum up: In quantum measurements the analogical pro-
cedures based uponIn classical mechanics turnout to be unsuitable.
As we concluded our study of causation, the creative doctrine
must be discarded1n favor of sets of factors5n specific fields. This
View makes useless any question of indeterminism (I 72f.).

MEASUREMENT AND OBJECTIVITY

In a Similar way the physicist’s psychistic premises have brought
the problem of objectivity into the quantum-measurement domain
by leading to the confusion in different measurement situations of
(a) interbehavior with things, (b) visual reactions to things, and
(c) the presence of “sensations.”
Quantum-mechanical situations obviously present great mensu-
rational difliculty. First, thereIs the inherent complexity of the
situation itself. It has been suggested that such difficulty stems at
least in part from an ineptness in adjusting from the technique of
classical or material mechanics to that of quantum or energy me—
chanics. As a consequence of this lag physicists have inferred that
by comparison with classical situations there is something irrational
and unreal in quantum mechanics. Because in classical mechanics
the measuring system could be set Up with a looser interrelation of
objects, of space—time coordinates, and of instruments, it appeared
that the tight interrelation Of factors in an energy system made the
thing measured inseparable from the measuring apparatus and even
from the measurer.
The question has thus been raised whether the data are objective
or exist onlyIn the process. Physicists haveindicated their belief
 M ETROLOGY AND M ETASCIENCE 3 I3
that a tree or a wave length is a construct or a mental creation.24 A
flower in Berkeleyan fashion is blue only when it is perceived, and
in general there is no difference between the world and the knowl-
edge of it.25 Clearly, on this basis one may either aflirm or deny the
existence of subatomic events genuinely or allegedly not amenable
to accurate estimation.
It is only the psychistic tradition which stands in the way of de—
scribing quantum-measuring conditions as they are actually encoun-
tered.” Once metaphysical constructions are rejected such histori-
cal questions as subject-obj ect and reality-illusion will not be
dragged into eXperimental science. Foremost in the situation will
be the facts of constructing hypotheses and theories, mathematical
or nonmathematical, as the result of interbehavior with things. A
measuring problem on this basis has nothing to do with an exist-
ence or objectivity problem. The scientist merely has to exert him-
self to the utmost to discover a means of Obtaining the increased
precision of knowledge concerning, for example, an electron’s
simultaneous momentum and position. Such technical difliculties
require no new theory of knowledge, no new cognitive principles.
Why should we introduce into any measuring situation, even the
most difficult one of quantum mechanics, any occult and spurious
factor such as an ego? Despite the futility of dabbling in psychics,
atomic'and quantum physicists indulge heavily in such theoretical
excesses as confounding a scientific observer with (I) an ego ob—
serving its brain or retina, (2) a measuring instrument, or (3) the
thing with which the ego is interacting.27

MEASUREMENT AND REALITY

Scientists accustomed to thinking of themselves as safely pro—
tected from the vagaries of metaphysics by the impenetrable cur—
tain of measurement found their complacency Shattered by the evo-
lution of quantum mechanics. After long entertaining the dogmas
that science is measurement and that physics, because built upon
mensurational Operations, is the primary science, they have had to
face the two following propositions
Margenau, Reality, p. 294.
−
’5
Ibid., p. 288; see also Lenzen, Concepts.
2'
See Menzel-Layzer, Physical; Frank, Foundations.
vdn\ Neumann, Mathematische, p. 223f.
'7
314. PSYCHOLOGY AND LOGIC

P(Measurement can not be regarded as operations performed upon

isolated things possessing unique autonomous properties).
P(To measure is to create the thing measured).
The most ironical thing about the injection of metaphysics into
physics is that there is no basis for it whatsoever. That it occurred
is evidence enough that physicists have lacked an objective psycho~
logical vieWpoint. Consider for a moment what precipitated the
difficulty. In terms of measurement or other direct description we
can not interpret a photon as an individual particle moving, say,
from the sun to the earth. It is impossible to state the precise po—
sition and momentum of isolated light particles, as in a classical me-
chanical system. As Frank says:
A particle by itself without the description of the whole eXperi_
mental setup is not a physical reality.”3 °
Now from an objective psychological standpoint there is no
occasion for physicists to be disturbed by the interdependence of
things measured and the measuring apparatus, and by the alleged
mutual interference of the two. This alleged interference is only
relatiVe. In a genuine sense every measurement disturbs the system
being measured.29 In other words, a measuring Situation constitutes
a different System from the nonmeasuring situation. In the former
the Obj ect measured does not remain independent of the measuring
instrument and the conditions required for the mensurational op—
erations. In the nonmeasuring system the object in question does
remainiindependent of the same list of factors. All technical or
precise observation requires the isolation or organization of a sys—
tem. If one describes the observer’s manipulations as interferences,
measurement certainly intensifies them. Other interferences are
also inevitable. Scale readings must be averaged and calculated;
this fact involves considerable transformation.
When we deal with an observer’s interbehavior with things, no
reality question obtrudes itself into any measuring situation. ‘If it is
nOt possible, as in classical mechanics, to set up a geometric map
lotating the position and motion of a-particle, one can invoke the
prediction criterion as proposed by Einsféin, Podolsky, and Rosen:

Foundations, p.48. i"

’3

2f'-MenzelLeyzeI-, Physical, p. 323; Grimsehl, Textbook, vol. V, p. 275.

 METROLOGY AND METASCIENCE SI 5
If, without in any way disturbing a system, we can predict with
certainty (i.e. with probability equal to unity) the value of a physical
quantity, then there exists an element of physical reality correspond-
ing to this physical quantity.30
The predict-ion criterion admittedly involves considerable con-
struction, but even if it is not formulated in terms of certainty it
emphasizes the continuity of the measurer’s contact with events.
The measuring system may require the employment of various cor—
respondences; momentum and position may have to be separately
determined, but at no point does any metaphysical reality problem
come in.
Nor need the observer’s and measurer’s participation in the
measuring system be regarded as introducing a subjective factor.
Since objective psychology has liquidated all psychics, the distinc-
tion between subjective and objective involves no scientifically det-
rimental factors. To demonstrate the position or momentum of a
subatomic or a light particle it may be necessary for the physicist to
arrange a diaphragm either to mark the point passed by the light or
the momentum it imparts. All such operations are continuous with
those necessary to (a) produce events such as the presence or ab-
sence of spectra, electronic—protonic discharges, heat, or (b) accel-
erate events, as in supplying catalysts or operating instruments like
cyclotrOns, betatrons, etc.
Only a psychistic background makes physicists wary of the ob-
server’splace1n the occurrence of a measuring event. In no way
does the worker’s participation introduce a factor of unreality,
There is nothing occult about an observer or about anything he can
possibly do. Whether, and how far, events involve such participa-
tion are matters concerning specific Situations. Also, the reduction
of such participation whenever desirable is the problem and chal-
lenge of the scientific job.
A number of writers have correctly pointed out that what seems
to precipitate reality questions in quantum mechanics is the fact that
the visual or pictorial form of descriptive model is lacking.’1 Like
the other difliculties, this, too, is based on the traditional psycho-
logical, doctrine of internal sense qualities and sense data. His—
Can \quantum-mechanical, p. 777.
3°

∂ Causality; Kaiser, Consequences; Berenda, Note.

316 PSYCHOLOGY AND LOGIC

torically the question has centered around the relationship of an

equation with sensory quality. Because of Mach’s prominence in
such matters we refer to his views concerning the relation both of
mathematical functions and of atoms to “experience.”
As an adherent of the Berkeley—Hume—Mill sensation tradition
Mach clung to elements or sensations as the only realities. Mathe—
matical functions or equations, accordingly, he regarded as a sort
of language or auxiliary mental artifice for facilitating the
“mental” reproduction of facts.
Although we represent vibrations by the harmonic formula, the
phenomena of cooling by exponentials, falls by squares of times, etc.,
no one will fancy that vibrations in themselves have anything to do
with the circular functions, or the motion of falling bodies with
squares.32
It is this view probably that is reflected in Einstein’s remark about
the discrepancy between mathematical exactness and reality."3
That the sensationistic view prevents a proper attitude toward
the relation of construction and observation is apparent from
Mach’s attitude toward atoms and the atomic theory of his time.
Ruark and Urey recount the dramatic scene when Mach rose up
after Boltzmann’s lecture on molecules to say: “You do not know
that molecules exist.”"‘ Mach rejected the atomic theory because
atoms “cannot be percieived by the senses.”35 With all his well-
intentioned effort to banish metaphysics from physics he accom-
plished the opposite. Though Mach retreated from his anti—atomic
position, he could not be expected in his time to realize that
“sensations,” “experience,” and “intuitive” items are constructs.
Moreover, as we have repeatedly pointed out, they are not helpful
in any science, but simply reflect historical inventions in nonscien-
tific situations."3 _
The view held by both physicists and psychologists that “sen-
sations” and “immediate experiences” are basic processes runs
back to spiritistic metaphysics which interpreted perception as a
process by which a soul or ego creates things, or somehow intuits
" Science, p. 492. v-
’3
Chap. 13, p. 6. ∙
Atoms, p. 1.
'5
∙
Sc1ence, p. 492. ∙
See chap. 6.
 METROLOGY AND METASCIENCE 317
or absorbs them. Current Obj ective psychology has left far behind
all questions related to subject and object, or how mind or thought
can be connected with extension and matter. Perception we treat
as complex interbehavior between organisms and things. Both
organisms and things become behaviorally modified through co-
ordinate conditions surrounding their successive contacts.37 The
history of these contacts Viewed as an evolution in a set of Situations
is the basic feature of objective perception theory. Instead of on a
theological basis, the objective View is grounded on the observa-
tion of the cultural evolution of the organism’s behavior.
At once we obviate all metaphysical questions in quantum me~
chanics, such as the nature of unobserved Obj ects. This problem
iS formulated as: “How do things look when we do not look at
them,’ ”33
and 1s consideredIn the context of the disturbance of an
object while it is being measured. Writers who do not regard such
problems as “nonsense,” on the ground that perceived Obj ects are
not created by the perceiving act or Observation,39 demonstrate
the insidious domination of cultural presuppositions over the most
competent philosophers of science. Such writers may declare that
quantum mechanics, like all other parts of physics, deals with
nothing but relations between physical things, and further that
all its statements can be made without reference to an observer;40
still they can not move away from the cultural institution concern-
ing “the existence of an eXternal world?”1
Such views bypass the fact that all descriptions and interpreta-
tions originate in contacts of organisms and objects. From such
contacts we derive our information concerning the properties of
things and our changing responses to them. These contacts consti-
tute an enormous history, which begins at a time when organisms
and objects are autonomous and independent; it continues through
a series of reciprocal behavior changes which are constantly being
conditioned by circumstances in the interbehavioral field.
The demand that reality in quantum mechanics be established
on the basis of experiments and measurements requires no justifi-

" Kantor, Principles, chap. 9, Survey, chap. Io.

I"Reichenbach, Philosophic, p. 18.
3°
Ibid.
Ibid.
Reichenbach, Experience.
318 PSYCHOLOGY AND LOGIC

cation.- Experiment and measurement constitute contacts with

things; every theory designed to describe those things and predict
how they will act must correspond to them.
Decisive for the solution of the reality problem in quantum
mechanics is the fact that we are studying the scientist’s interbe—
havior with radiant electromagnetic or electronic energy. Hence
all energy values as the products of measuring operations are de—
rived from the events investigated. The prominence of construc-
tional activity in atomic measuring situations should not allow us
to lose sight of the events which originally promote the work and
which naturally must be represented in the product.

METROLOGY AND LOGIC

Students of the logic and methodology oft, science appreciate in
some degree that measuring enterprises constitute system build-
ing. Also, scientists interested in measurement theory have expertly
discussed such system components as units, scales, standards, opera—
tional techniques, and homogeneity of things measured, in addi-
tion to the relation of all these items to their environing frame—
work.‘ Nevertheless, their adherence to formal logic prevents
logicians from integrating measurement systems and metrological
science with systemology as the general science of lOgic—that is,
system-building theory. We demonstrate this point by considering
the reflections Of two writers who are sympathetic to metrological
and systemological integration, but who do not differentiate be~
tween (1) systems of operations upon things and (2) systems of
statements or descriptions of those Operations.
The first of the two attempts to connect measurement with
formal logic is based upon an analysis of measurement into three
parts: (a) formal statements, in the sense of numbers in serial or
other relations, (b) the measurable experience '(or content), and
(c) a dictionary of interpretative statements in which the items of
(b) are so defined as to exhibit the logical properties of the group
in (a). By means of this dictionary the numbers of (a) are assigned
to the contents of (b)."'2
But now it turns out that since the dictionary consists of sentences
analyzable into (a) logical forms, (b) referring contents, and (c)
-the assignment of thesecontents to the variables of the logical
". Ballard, Paradox.
\
 METROLOGY AND METASCIENCE 319
forms, a second dictionary is required. With this type of analysis
the regress continues endlessly. Measurement is thus inflicted with
a serious paradox.”
Comment: The paradox is inherent only in the logical theory
according to which logic is made into an extraneous system con—
structed at best to serve a describing or symbolizing function. Con-
sider the enormous contrast with the interbehavioral View accord-
ing to which logic constitutes an enterprise of constructing systems
for ascertaining the values of things and properties. The description
of measurement on the formal basis does not even allow for such
concrete Operations as are required for number-assigning. A meas—
urement is defined as a proposition stating an assignment of num-
bers to objects, or classes of objects, discriminated in experience.“
Here is the typical conversion of interbehavior into words.
Our second example has its source in the quantum—mechanics
field. In consonance with the common belief that quantum me—
chanical events present altogether different Observational and men-
surational problems from the classical mechanical type, it is
assumed that quantum mechaniCs must be less concerned with
the structure Of the physical world than with the structure of the
language in which that world is described.45 Thus questions con—
cerning the existence of physical entities are transformed into-
questions concerning the meaning of propositions (language).
Because of the limited measurability of quantum events, and
their alleged anomalous character, they can only be described by
means of' uniquely descriptive languages. The corpuscle language
and the wave language each is restricted to some phase of quantum
action, and thus is deficient. The neutral language describes what
are called interphenomena consisting of energy occurrences when
there are no coincidental collisions with matter resulting in a record
by a Geiger counter, a phOtographic film, 'or a cloud-chamber
track. An example is a photon travelling between its source and
point of collision.46
N ow, it is argued that Since quantum mechanics can not escape
the indeterminacy principle, its language can not be patterned on a
’3
Ibid.
Ibid., p: 134.
’5
Reiéhenbach, Philosophic, p. I 36H.
’" Ibid.,\\p. 21.
320 PSYCHOLOGY AND LOGIC

two-valued logic which comports with propositions built on a true-

false dichotomy. Instead resort must be had to a probability logic
with at least three values, including the indeterminacy value.
Comment: Aside from the issue whether such differences exist
between the macroscopic and microscopic fields as to necessitate
different types of observation and measurement, some general
logical questions come to the surface. In the first place, why at-
tempt to confine scientific Situations in the straightjacket of a
formalistic system? Not even a three- or n-valued logic can do
justice to all mensurational problems. Again, why assume that
logic consists of setting up a language? To be correct, we ought
to say. we set up a large number of descriptive systems to deal with
the many features of the quantum situation. And what about the
motivation for quantum-mechanics system building? Is the logician
working with a view to a possible cooperation with the physicist
in solving some problem, or is he indulging in airbitrary construc—
tionism? In the latter case one must admit at once the irrelevance
of his activity. One can not in either case escape the multiplicity
and complexity of the specific conditions involved.
 CHAPTER XXV

SYSTEM BUILDING AND METALOGIC

THE CONTINUITY HUMAN BEHAVIOR

OF

BOTH psychology and logic point to a direct line connecting

system building and logical theory. This continuity extends
even beyond these limits. Below system building are interbehaviors
of persons with stimulus obj ects involving only partial structuring
of things, as well as completely nonorganizing behavior. Above
systemology lies the metalogical domain—the source of the system
builder’s presuppositions and motivations.
Metalogic, notice, is not confined to the technical definitions
and postulates connected with formal system construction. 1 It18
concerned also with the background of systems, with logical appli-
cations and possible consequences of logical work. Briefly, meta-,
logic comprises the basic philosophy of the logical theorist. It can
be no better described than:
. an exploration of the periphery of logic, the relations of logic
to the rest of the universe, the philosophical presuppositions which
give logic its meaning, and the applications which give it importance.2
At once we ask: What kind of philosophy? What kind ofuni-
verse? And precisely howIS logic related to both?
We renounce, of course, the absolutistic and
philosophy revealed in the following: supramundane
Formal logic is the heart of philosophy precisely because the sub-
ject matter of logic is the formal aspect of all being, an aspect not
only of objects and events in time and Space, but equally of nonspatial
and nontemporal relations of objects.’
Our universe c0nsists of nothing but our natural habitat plus
civilizational artifacts. Logic, then, is inevitably set within the
spatio—temporal milieux of individuals and cultural groups. All
transcendent and mystical systems built for escapist and senti-
mental reasons exist only within these behavioral limits.
1
3333P ‘93-
‘
2Cohen, Preface, p. xi.
aOp. cit.,p. x.
3 22 PSYCH OLOGY AND LOGIC

Nor can we assume that logic is the heart of philosophy. Re-

vealed here is the presupposition that philoSophy delivers a com-
plete and fixed cosmic system and that logic is exclusively an
instrument for constructing the system. However, the study of
philosophizing behavior discloses no necessity for building such
a philosOphical system. Again, if we reject such cosmic philosophy,
logic takes on a different aspect. Throughout this treatise we have
emphasized that logic constitutes a concrete kind of work resulting
in specific system products. Logic belongs to every department of
human behavior. Moreover, though _system building is exceed-
ingly important for effective adaptation to‘life conditions, logical
behavior is still only one type of behavior, systems but one kind
of behavioral product.
The application of logic brings up another important issue. If
logic is not a single inclusive entity which is profitably exploited by
being applied to all types of Situations, it loses its key or dominant
character. Logic pervades all human situations precisely because it
constitutes system—building behavior complementary to other sorts.
Logicians who find our philosophical presuppositions too narrow
because restricted to human behavior and to the human Scene can
not escape the fact that their dissatisfaction, after~all, can only have
significance within such a boundary. What they object to are re-
strictions placed on their invention of transcendent systems. But,
surely, to overlook the continuity of interbehavior with one’s
natural and cultural environs when constructing a system of uni-
versals, when inventing nonspatial and nontemporal objects, and
when creating a transcendental universe or logos is to prevent logic
from becoming a datum for study. Interbehavioral continuity
alone makes logical study vital and urgent.

METALOGIC AND LOGICAL PRACTICE

The interplay of philosophical presuppositions and system-
building practice provides a treasury of source material for observ-
ing and evaluating system—buildingbehavior. Transcendent sys—
tems are based on non-naturalistic philosophical presuppositions.
Naturalistic systems are constructed from specific and concrete
materials. But are not all systems, both transcendent and natural-
istic, products developed by particular persOnS having particular
backgrounds and interests? Not the least important advantage of
 SYSTEM BUILDING AND METALOGIC 323
keeping logic exclusively within the interbehavioral range is the
restraint it places on irresponsible system construction. Thus in the
system builder’s life conditions we have a touchstonefor evaluating
his system, From the standpoint of logic as a tangible pervasive
discipline we consider the potentialities and limitations of logic on
three levels: (A) affairs of everyday life, (B) science, and (C)
philosophy. ∙
A. Logic and Living
System as order or structure in everyday situations makes for
perspective and convenience; it aids in appreciating relationships,
fitness, and cogency. Such factors are basic for orderly and effective
adaptation to a vast variety of things and Situations. Even the child.
very early in life begins to systemize his reactions with respect to
the objects and persons about him. Elementary as such structuring
of behavior is at this level it nonetheless is continuous with logical
system. Order and system are also-definite aids in all effective
thinking, judging, reasoning, and debating behavior. The fact that
systemizing things is a prerequisite for achieving proof, inference,
and prediction, and for. gaining self-assurance and securing con-
viction means that logic becomes an intimate part of our daily-life
situations.
Unfortunately our innumerable system-building operations are
frequently overshadowed by finished structural products handed
down as cultural heritages—for example, the syllogism. The basis
for the syllogism’s appeal as the model of reasoning lies in its
systemic properties. The perfect system is the circle without corners
or loose ends. What is fitted into it is uniquely localized and firmly
fixed. Of course, to achieve perfect systemization one has to resort
to an emptying and formalizing procedure. System, however,
Should not be confused with the empty form, as has traditionally
been done.
One of the most impOrtant functions of logic in everyday activi-
ties is that it enables us to separate (a) systems of things from (b)
systems which are sheerly verbal. Systems of things are of two
types: (1) those existing in rerum ndtura without human or other
animal agency and (2) those previously structured through an or-
ganism’S contrivance. Purely verbal systems include linguistic
descriptions of things howsoever fashioned. The most marked
3 24. PSYCHOLOGY AND LOGIC

contrast between (a) and (b) types of system arises when the
latter consists of word structures which do not correspond to exist-
ing Objects.
I

To be able to differentiate between (1) acts, (2) objects acted

upon, and (3) products of action is to achieve considerable orienta—
tional power. It likewise does away with the perennial problem of
hnower and known. Knowing is interbehavior with (a) natural
things, (b) contrived things, and (c) statements concerning things
actually or allegedly existing. Interbehavioral logic is- useful in
finding one’s way among these-various situations.
Similarly, the system-building view of logic is of use in clarify-
ing traditional fallacies. For the most part these fallacies concern
improper or incomplete systems. The parts do not hold together,
as in the fallacy of the consequent or non se uitur, or the system
is not rounded out—that is, it is left incomp etc, as in the classical
petitio principii illustration. °

B. Logic and Science

System-building finds its most significant scope in the scientific
domain. Even though science is not merely an enterprise of
systemization, classification and organization are important features
of scientific work. Here we recall Poincaré’s classic allusion to
science as essentially systemization.
Science is built up with facts, as a house is with stones. But a col-
lection of facts is no more a science than a heap of stones is a house.‘
The importance of system and order in science constitutes the
raison d’etre of the methodistic logiCs. Indeed, classical inductive
logics are mainly sets of critical suggestions concerning scientific
systematics—prescriptions how grand-scale systems should be
built.
Overemphasis of system, however, may be detrimental to
scientific work. The task of science, after all, is to resolve problems
concerning events. Can these events be reduced to generalized
happenings? Can single methods be- devised to solve scientific
issues? How Often the history of sCi‘ence reveals the fallacy of
reducing many things to one, a reduction made in order to achieve
universality. Has it helped, for instance, to reduce astronomy and
Foundations, p. 127.
 SYSTEM BUILDING AND METALOGIC
325
physics to mechanics, mechanics to geometry.P5 The neatness and
-order obtained by Single comprehensive systems hardly do justice
to the events for the investigation of which science exists.
The advancement of knowledge and the success of investigation
point effectively to the futility of totalitarian systems. Generalized
systems restrict research, repress the claim of events for the atten-
tion of scientific workers. Systemization must be instrumental to
the work of science. But this means limiting Systems, keeping
them down to Specific data and problems. Certainly the closer one
stays to actual events the more specific systems will be. Conversely,
the more comprehensive systems are, the more they are involved
with formulae and words, with linguistic or symbolic elements.6
‘The growth of geometry, for instance, reduces the self-contained
and universal Euclidean system to one of the systems. Geometries
are multiplied by a large factor. The development of knowledge
concerning heat, electricity, and optics increases the number of
scientific systems. The rise of biology, geology, and psychology
calls for far mOre systems than were allowed by those who stressed
the system maker’s constructive products to the exclusion of the
materials conditioning his system—building work.
An instructive example of an all-encompassing system is fur-
nished by an eminent physicist who wants to reach out to the living
cell and “life,” to weld together into a whole the sum—total of all
that is known. SchrOdinger writes:
We have inherited from our forefathers the keen longing for
unified, all-embracing knowledge. The very name given to the highest
institutions of learning reminds us, that from antiquity and through-
out many centuries the universal aspect has been the only one to be
given full credit.7
The outcome is that each individual is God, a result which uni-
versalists want and believe in, but have not the courage to pro-
claim. SchrOdinger himself shrinks from this final conclusion and
escapes by adopting the Indian Athman-Brahman doctrine—

" Cf. Nagel, Meaning.

Interesting here is Reichenbach’s proposal. “Instead of speaking of the structure
of the physical world, we may consider the structure of the languages in which this
world‘cambe described; such analysis expresses the structure of the world indirectly,
but in
axmore precise way.” Philosophic, p. 177.
7 What\is Life? p. vii.
326 PSYCHOLOGY AND LOGIC

namely, the personal self equals the omnipresent, all-compre-

hending eternal self.8
The product is no more instructive than the procedure leading
to it, since it is a scientist who turns the wheels. The classical regu-
larities of nature are statistical. But even the genes of biology,
which, on this basis, can not be regarded as stable because of_the
fewness of the atoms in their composition, can be reduced to deter-
ministic mechanisms by a quantum theory of molecular structure.
Admittedly this is an extreme example. But even if SchrOdinger’s
freedom of creation is not completely attained by other univer-
salists, even if comprehensiveness of achievement is not aspired to
by other scientists, the difierence is one of degree only. When
the urge to universalization is present, an improper emphasis is
somehow impressed upon investigation. At the least, the specificity
of the job, the influence of data, and the tentative and probable
characteristics Of the work done are minimized. The present
example therefore recommends itself as a model precisely because
it shows us the power of verbal and imaginative behavior to create
the appearance of exalted system.
Scientific. system building has definite boundaries. Astronomy,
which is potentially limitless in its scale, stops Short on the basis of
its observational and calculative materials. Though a figure like
3,500,000,000 light years as an estimate of the diameter of the
universe lacks familiarity, it is still well within the range of arith-
metical operations. Similarly, the science of thermodynamics Speci-
fies definite limits to the generalizations it establishes. The first
law reminds us of the boundaries placed upon complex and intense
processes; there are optimal probabilities to be observed. The
second law limits the temporal interval of the energy interchanges,
such that perpetual motion is ruled out. And if we allow N ernst’s
heat theorem9 to pose as the third law, it asserts the impossibility
Of attaining absolute zero by any process whatsoever.10 AS another
example of scientific limitation and checking we may mention the
Le Chatelier principle, as formulated by Braun.
Every process caused by an external influence or by a primary proc-
ess in a system takes place in such a direCtion that it tends to oppose
’ Ibid., p. 88.
‘ ' New Heat Theorem.
1°
See Grimsehl, Textbook, vol. II, p. I 50f.
 SYSTEM BUILDING AND METALOGIC 327
the alteration of the system produced by the external influence or
primary process.11
The state of equilibrium which the Le Chatelier principle indi-
cates is a warning supplied by natural events against arbitrary
extrapolations of observed results.
If scientific systems are to be significant, limitations must also
be placed upoiI their unit constructs. To be valid and workable,
hierarchies of constructs must plainly show their derivation from
events and from observation of the latter. We plainly see the
work of the individual abstractor and system builder throughout
the continuum of word descriptions, pointer records, averages and
other statistical constructs, and formulae and equations.
CONSTRUCTING AN EXTERNAL WORLD
An interesting system—building convention harbored by philo-
sophiCally inclined scientists concerns the creation of an external
world. How created and from what? From “sensations,” is the
traditional answeri
In the first instance the elements of the physical order are given
in sensation. In physical science one constructs the fundamental con—
cepts of the things, characters and processes revealed by sensation.
A physical theory is the systematic representation of certain char—
acters of the physical order. The goal of physical science is a unitary
system Of physical theory.12
What seems to be overlOoked is that “sensations” are con-
structs, as indeed are all descriptions and interpretations of
things. Strange that scientists wish to create objects of physics out
of products previously constructed by traditional theology and
psychology. Sensation constructs are in no Sense derived from any
investigation; they are therefore not scientific constructs. If
physics, psychology, or any other science is to produce constructs
with descriptive and predictive value, it must create them on the
basis of relevant contacts with things.
∙ The notion of constructing a world is utterly misleading.
Actually what is constructed is a description of events and of the
operation of things, such as the interrelations of masses—planets,
\

11Ibid., p. 152.
1’
Lenzen, Nature, p. I.
328 PSYCHOLOGY AND LQGICI

terrestrial objects, and so on. Such a large-scale system as our

objective world goes far beyond the scientist’s scope or competence.
The primary objection to such ungoverned and futile system
building is that the scientist neglects his proper work. What he is
qualified to do is to investigate events and the problems they sug—
gest. Whatever system he builds in his capacity of scientist can be
done only on the basis of such investigation. Going beyond this
work he assumes a role other than scientist.
Evidence is not lacking that all notions of the construction of
a world” are based upon a confusion_of constructs with the things
stimulating the construction. Descriptions are confounded with
events described, quite in the Berkeleyan obfuscation tradition.
That the descriptions are made in terms of words, sentences,
mathematical or logical symbols instead of in terms of sensations.
constitutes no palliation.
.3

C. Logic and Philosophy

The philosophic level of logical activity is essentially that of
systems. Philosophers consider their work as, above all, the build-
ing of knowledge or world structures. For this reason logic is
regarded as intrinsically a philosophical enterprise or as identical
with all philosophy."
What constitutes a philosophical system?” For the most part,
systems constructed, on a philosophical level are as eXpansiVe as
speculative imagination can make them. A considerable portion of
traditional philosophy consists of method systems to transcend the
actual world. That these systems turn out to be fragile collocations
of words is a pathetic fact of philosophical history.
Methodistically inclined builders of philosophic systems seek
to construct absolute techniques for achieving equally abs'plute
certainty; they aim to give laws to nature or to develop omnipjotent
reason. Thus they create powers to extend their own powers. By
comparison the ontologists attempt to achieve some trarisexperi-
ential realm, an elysium to which they can escape frOm the actual
world, and thereby transformIevery observable thing according to
a scale of ultimate values. In some form or other all cBntent

Russell, External World; Reichenbach, EXperience.

" Russell, Mysticism, p. I I If.
 SYSTEM BUILDING AND METALOGIC 329
philosophies aim to prove there is a god, that man is free and at
least partially immortal.
Even sceptical philosophers do not wholly relinquish commerce
with such supramundane inventions. Although they question the
work and achievement of their opponents, they fail to touch the
nerve of the difficulty—namely, to provide an adequate basis for
checking such extravagant performances.

TOWERS AND SYSTEMS

Traditional philosophy supports Wilde’s mot that man does not
live by bread alone but by catchwords. He might have added, too,
that man the philosopher lives also by building towers. Not only
towers of ivory to enable him to withdraw from the affairs of
everyday life, but also towers of babel to transcend the concrete
world. From this tower:
The free intellect will see as God might see, without a here and
now, without hopes and fears, without the trammels of customary
beliefs and traditional prejudices, calmly, dispassionately, in the sole
and exclusive desire of knowledge—knowledge as impersonal, as
purely contemplative, as it is possible for man to attain. Hence also
the free intellect will value more the abstract and universal knowl-
edge into which the accidents of private history do not enter, than
the knowledge brought by the senses, and dependent, as such knowl-
edge mu'st be, upon an exclusive and personal point of view and a
body whose sense-organs distort as much as they reveal.15
The value of philosophy for the tower-dweller is:
t.hrough the greatness of the universe which philosophy con-
templates, the mind also15 rendered great, and becomes capable of
that union with the universe which constitutes the highest good. 1"

Doubtless the higher the tower the more pristine the good
achieved from union with the universe; also the farther man
transcends the world of which he is a part. He neither profits from
Antaeus or the biblical tower builders, nor does he consider the
lesson implicit in concrete system building.
Tower construction as the unrestricted system building of. abso-
I

lutistic\Iph\ilosophers goes on in reckless disregard of the kind of

1:’° Russell,\Problems,
Ibid., p. ’250.
p. 248.
330 PSYCHOLOGY AND LOGIC

acts performed and the means employed. The egoism displayed

completely masks the fact that the builder lacks any transcendent
powers by which to get beyond himself and his world. Support for
the production of such mystic and supernatural systems the
absolutist finds in his alleged ideals. _Reason, to deserve the name,
for most philosophic system builders must be mystical.
'
I
TECHNIQUES OF UNRESTRICTED SYSTEM BUILDING
Prominent among the techniques of unrestrained system build-
ing are (I) obeisance to utterance and (2) projection and fixation
of verbal creations.
(.1) Obeisance to Utterance. More thanwhimsyIS implied1n
the theory that myths are diseases of utterance. Utterance as
reference to things and events constitutes; to begin with, reliable
description. But linguistic behavior may easily,become unruly even
in the presence of fact. References and descriptions are distorted;
extrapolations are made way beyond anything observed or legiti—
mately inferred. The end point is to create the utterness Of the
supernatural.
Submission to utterance involves accepting and preserving verbal
creations. Upon such behavior depend the power and evil of myth
making. Probably the most effective mechanism to preserve the
products of verbal creation is argumentation for and against their
acceptance. ∙
(2) Projection and Fixation of Utterance. By hypostatizing and
fixating utterances men establish what they wish or what they
blindly draw from the well of folkbelief. In this way the great
mysteries of philosophy and religion are created and maintained.
Systems are built up without limit in number, without bound in
rationality.

ABNEGATION MOTIVE IN PHILOSOPHY !

A primary condition for the self—expanding exercise Of‘ building
supramundane systems is renunciation of kinship with concrete
things and events. To harp upon areal world beyond thethin‘gs
observed in nature is behavior stemmiII‘g from primitive theological
sources. Discernible, too, in such statements as we have quoted
from RussellIS the theological View of man’s Sin and corruption,
the devaluation of knowledge “brought by\the senses’’because
 SYSTEM BUILDING AND METALOGIC 331
dependent upon a “body whose sense-organs distort as much as
they reveal.” On what ground can anyone favor knowledge
verbalized in abstraction from a purely rational source to the
original domain of things and events?
It is an eloquent comment upon the hold that ancient theological
traditions exert upon modern thought when a writer is compelled
to urge an appreciation of the “infinite beauty of our natural
world” in support of a naturalistic philosophy."
What is it but a folkway to regard the speculative construction
of hypotheses concerning the integration of knowledge as inevita—
bly concerned with supramundane things? Such construction is
(really centered in the activities of persons as they interbehave with
their surroundings.

THE PROBLEM OF THE INDUCTIVE SYSTEM

Ever since David Hume made popular the notion that from past
and present data or knowledge we can not deduce any future
datum, philosophers have been disturbed by this limitation. Thus
has arisen the futile attempt to justify inductive inference. In—
ductive theory has been dubbed the despair of philosophy;18 many
writers have agonized over it, argued about it, only to leave the
problem where it has always been.
The conviction that there is no a priori foundation for induction:
. . I. will slowly shatter civilized life and thought, to a degree which
will make the modernist’s loss of confidence in Christian super-
naturalism, so often cited as the ultimate in Spiritual cataclysms, seem
a minor vicissitude.19
So writes a philosopher with the anguish of one who is being
deprived of a cherished hope. He continues:
,To dispute the rational validity of induction,.however, is to deny
that reason and good-will have a purchase on reality, to deny mind’s
hope of acclimating itself to any world whatever, natural or super-
natural.20
1’
Lamont, Naturalism, p. 597. Notice Lamont’s comparison of the Christian with
Chinese and Indian attitudes toward nature. It is suggested that appreciation of nature
is Ia cultural trait.
Whifehead, Science, p. 3 5.
1'
Williams, Ground, p. 16.
"11.121.
332 PSYCHOLOGY AND LOGIC

Here is a demand for an all-embracing system capable of en—

chaining all events occurring and to occur. Is not this craving for
cosmic mastery entirely out of harmony with any specific or sig—
nificant human problem or enterprise?
The assimilation of induction to deduction_lS circle squaring in
reverse. Those who wish to subject concrete events to the absolute-
ness and imperiousness of a priori principles wish to transform the
squareness, hence the incompleteness, of the.inductive system into
the completely rounded—out circle of deduction. Inevitably those
concerned with the problem of justifying the inductive principle
are dominated by the quest for supernatural knowledge and abso-
lute reason. Though the inductive problem is said to be an issue
of science and derived from contact with concrete things, actually
those who deal with it do not build systems°for scientific purposes.
It is encouraging that even philosophers who,are not altogether
free from the quest for certainty, and who regard the systems they
build as inherent in the nature of things, suggest abandoning the
inductive problem.21 But something more than logical policy is
called for. We must take into account the specific system—building
enterprise. Whether or not the procedureIS inductive or deductive
depends upon the kind of materials used.

STATISTICAL PROCEDURES AND ABSOLUTE SYSTEMS

A most interesting metalogical item concerns the absolutistic
system builder’s assimilation of scientific techniques. So persistent
are aprioristic presuppositions and motivations that statistical
methods, which can hardly deviate from concrete events, have been
invoked to establish absolute systems. Thus the instruments con—
structed in order to facilitate adjustments to events by predicting
their future occurrences are made into tools for extrapolating be-
yond events and our reactions to them. We can not too strongly
object to the fallacious belief that somehow statistical and arith—
metical procedures can transform contingent events into
things.
absOlute
Philosophical system builders regard any means as justified if
only it helps to attain some absolutistic end. They further Ttheir
activities by using the intricacies Of statistical \arithmetic to obscure
Cf. for example, Ambrose, Problem, Will, Problem,
Jeflreys, Theory; Bergmann, Comments, Wang, Notes.
Future, Black, Language,
 SYSTEM BUILDING AND METALOGIC 333
plain and definite issues. Accordingly, when we find a logician
making use of a statistical or proportional syllogism we must in—
quire into his presuppositions and motivations, as well as into his
procedures. In this way we can understand perfectly the flat failure
to achieve knowledge-less knowledge, and also Observe the con-
fusion regarding the actual things about which knowledge is de-
sired. A priorists start with the valid principle that predictive
propositions are subject to probability calculation 5 they end with
the assertion that induction possesses jurisdiction Over all branches
of philosOphy, as well as over the natural sciences.22

LOGIC AND THEPLENUM

On the whole we have stressed the magnification of the system-
izer’s appetite and emphasized the checks that should be placed
upon his ambitions. There is another and reciprocal impropriety
to be avoided—namely, limiting systems to abstractions.
In order to magnify himself and his work the system builder
minimizes his world. He stresses simplicity, vacuousness, and
quantity. All this in order to reduce the universe to a manageable
system by means of signs and symbols.
Paradoxically the emphasis of specific things and Operations
allows for a more comprehensive system coverage. Obviously more
things can be dealt with if systems are made for things and things
not reduced to elements of systems.
For logic cosmos is system; it is the organization and ordering
of the systemizer’s field of interest. But systemizing means creative
operations. Systems imply selection, the application of criteria of
use or of completeness. Under no circumstances must we lose sight
of the operational field. There is always a plenum—a set of events,
things, and entities—which can never be exhausted by the struc-
turing operations.
” Williams, Ground, p. 202.
 BIBLIOGRAPHY
Aiken, H. Notes on the categories of naturalism. Journal of Philosophy,
I946.43.517-523-
Ambrose, A. The problem of justifying inductive inference. Journal of
Philosophy) 1947: 44: 253'272°
Ancillon, J. P. F. Doutes sur les bases du calcul des probabilités. In
Mémoire: de l’llcadémie Royale de: Science: et Belles-Lettres, Classe
de Philosophic Speculative, Berlin, Decker, 1794-1795, 3-32.
Aristotle, The Works of. Translated into English (J. A. Smith and W. D.
Ross eds.) vol. I, Oxford, Oxford Univ. Press, 1928.
Avey, A. E. The law of contradiction: its lOgical status. Journal of
Philosophy, 1929, 26, 519-526.
Bain, A. Logic: Deductive and Inductive. London, Longmans, I870.
Ballard, E. G. The paradox of measurement. Philosophy of Science, 1949,
I6, '134-I36.
Balmer, J. J. Notiz fiber die Spectrallinien des Wasserstoffs. flnnalen der
Physile und Chemie, 1885, 25, 80-87.
Bartlett, R. J. MeasurementIn psychology. The ddvancement of Science,
I939'194-0, 1:422'441-
Beck, L. W. Constructions and inferred entities. Philosophy of Sczence,
1950, 17, 74-−∂
Becker, 0. Mathematische Existenz; Untersuchung zur Logic und On-
tolog ie Mathematischen Phanomene. Halle, Niemeyer, 1927.
Bell, E.
gT. The Development of Mathematics. New York, McGraw-
Hill, 1940.
Benjamin, A. C. Logic of measurement. Journal of Philosophy, 1933,‘ 26,
701-710.
of Science, 1938, 5, 132-
Bentley, A. F. Physicists and fairies. Philosophy
165
Berenda, C. W. A note on quantum theory and metaphysics. Journal of
Philosophy, 1942, 39, 608-611.
Bergmann, G. Some comments on Carnap’s logic of induction.
of Science, 1946,13, 71-78.
Philosophy
Birkhoff, G. D. Aesthetic Measure. Cambridge, Harvard Univ. Press,
1933'
Intuition, reason, and faithIn science. Science, 1938, 88, 6'01-
6091'
Black, M. The Nature of Mathematics: A Critical Survey. New York,
Harcourt, Brace, 1935.
‘

Language and PhilosOphy. Ithaca,Cornell Univ. Press, 194.9.

Bogoslovsky, B. B. The Technique of Controversy: Principles of Dynamic
Logic. New York, Harcourt, Brace, 1928.
Bohr,
N. Atomic Theory and the Description of“-
Cambridge Univ. Press, 1934.
Nature.Cambridge,
 BIBLIOGRAPHY 335
Can quantum-mechanical description of physical reality be con-
sidered complete? Physical Review, 1935, 48, 6-96702.
Causality and complementarity. Philosophy of Science, 1937, 4,
289-298.
Boole, G. An Investigation of the Laws of Thought, on which are
Founded the Mathematical Theories of
London, Walton and Maberly, I8 54. Logicg and Probabilities.

Born, M. The Restless Universe. New York, Harper, 1936.

Brewster, D. Memoirs of the Life, Writings, and Discoveries of Sir Isaac
Newton. Edinburgh, Murray, 185 5,‘2 vols.
Bridgman, P. W. Logic of Modern Physics. New York, Macmillan, 1927.
Statistical mechanics and the second law of thermodynamics.
Bulletin, flmerican Mathematical Society, 1932, 38, 225-245.
The nature and limitations of cosmical inquiries. Scientific Month—
ly. I933. 37. 385-397-
The Nature of Physical Theory. Princeton, Princeton Univ.
Press, 1936.
Some general principles of Operational analysis. Psychological Re-
w‘ew. I945. 52. 246-249-
Broad, C. D. On the relation between induction and probability. Mind,‘
1918, 27, 389-404; 1920, 29, 11-45.
Burks, A. W. Empiricism and vagueness. Journal of Philosophy, 1946,
43. 477-486.
Burtt, E. A. Value and existence. Journal of Philosophy, 1947, 44, 169-
I79.

J.
Butler, The Analogy of Religion, Natural and Revealed, to the Consti-
:tution and Course of Nature. London, Bell, 1878.
Butterbaugh, G. I. A Bibliography of Statistical Quality Control. Seattle,
Univ. of Washington Press, 1946.
Campbell, L., and Garnett, W. Life of James Clerk Maxwell. London,
Macmillan, 1882.
Campbell, N. R. Physics: The Elements. Cambridge, Cambridge Univ.
Press, 1920.
An Account of the Principles of Measurement and Calculation.
London, Longmans, 1928.
Measurement and its importance for philosophy. Aristotelian So-
ciety, Supplementary vol. XVII, London, Harrison, 1938.
Quantitative estimates of sensory events (Final Report). The
fld‘vtmcement of Science, 1939- 1940, 1, 331-349.
Carmichael, R. C. The Logic of Discovery. Chicago, Open Court, 1930.
Carnap, R. Testability and meaning. Philosophy of Science, 1936, 3, 419-
4713The
1937: 431‘40-
two concepts of probability. Philosophy and Phenomenological
Research. I945. s. 513-532.
‘u‘ Remarks on induction and truth. Philosophy and Phenomenologi-
\cal Research 1945--,1946 5, 590—602.
336 PSYCHOLOGY AND LOGIC

Probability'as a guide to life. Journal of Philosophy, 1947, 44, I41-

I48.
Carus, P. The problem of causality. Open Court, 1888, 2, 1200-1204.
Cassirer, E. Substanzbegriff und Functionsbegrifi. Berlin, Cassirer, 1910.
Child, A. On the theory of the categories. Philosophy and Phenomenologi-
calResearch,1947, 7, 3--I6-335.
Churchman, C.
W. Probability theory, I, II, III. Philosophy of Science,
1945, 12, I47— I73.
Theory of Experimental Inference. New York, Macmillan, 1948.
A materialist theory of measurement. In Philosophy for the Fu-
ture (R. W. Sellars, V. J. McGill,
Macmillan, 1949. M. Farber eds.) New York,

Cohen, M. R.
Reasonand Nature. New York, Harcourt, Brace, 1931.
A Preface to Logic. New York, Holt, 1944.
Cohen, M. R.., and Nagel, E. An Introduction to Logic and Scientific
Method. New York, Harcourt, Brace, 1934.
Comrey, A. L. An Operational approach to some problemsIn psychological
measurement. Psychological Review, 1950, 57, 211-228.
Costello, H. T. The naturalism of Woodbridge. In Naturalism and the
Human Spirit '(Y. H. Krikorian ed.) New York, Columbia Univ:
Press, 1944.
Cournot, A. A. Exposition de la Théorie des Chances et des Probabilités.
Paris, Hachctte,1843.
Couturat, L. The principles of logic. In Encyclopaedia of the Philosophical
Sciences, vol. I, Logic, London, Macmillan, 1913. -
Crookshank, F. C. The importance of a theory of signs and a critique of
language in the study of medicine. In The Meaning of Meaning by
C. K. Ogden and I. A. Richards, Supplement II, London, Kegan
Paul, 1923.
Czuber, E. Warscheinlichkeitsrechnung und ihre Anwendung auf
Fehlerausgleichung, Statistik und Lebensversicherung. Vol. I, Leipzig,
Teubner, 1903.
Dadourian, H. M. Force in mechanics. Science, 1939, 87, 388-389.
Dantzig, T. Number: The Language of Science. New York, Macmillan,
I939C°
Darwin,C..G Logic and probability in physics. Philosophy
6
1939. 48- 64
of
Scienicle,
Davis, H. T, and Nelson, W. F. C. Elements of Statistics, with Applica—
tions to Economic Data. Bloomington, Principia Press, 1935.
Dennes, W. T. The categories of naturalism. In Naturalism and the Hu-
man Spirit (Y. H. Krikorian ed.) New York, Columbia Univ. Press,
I94-4- l‘
Descartes, R. See Haldane and Ross. -
Dewey, J. Logic. The Theory of Inquiry. New York, Holt, 1938.
Dewey, J., and Bentley, A. F. Definition. Journal of Philosophy, 1947,
44, 281-306.
 BIBLIOGRAPHY 337
Knowing and the Known. Boston, Beacon Press, 1949. ’

Diels, H. Die Fragmente der Vorsokratiker Griechisch und Deutsch. Berlin,

Weidmann, 1922, 3 vols.
Dingle, H. Science and Human Experience. New York, Macmillan, 1932.
Through Science to Philosophy. Oxford, Clarendon, 1937.
Dubislav, W. Die Definition. Leipzig, Meiner (3rd. ed.), 1931.
Dubs, H. H. Rational Induction. Chicago, Univ. of Chicago Press, 1930.
Definition and its problems. Philosophical Review, 1943, 52, 566-
577-
Ducasse, C. J. On the nature and the observability of the causal relation.
Journal of Philosophy, 1926 23,5 7-6-8.
A neglected meaning of probability. Proceedings, Sixth Inter-
national Congress of Philosophy, New York, Longmans, 1927, 343-
34-7-
Eaton, R. M. General Logic: An Introductory Survey. New York, Scrib-
ners, 1931.
Eddington, A. 8. Space, Time and Gravitation: an Outline of the General
Relativity Theory. Cambridge, Cambridge Univ. Press, 1920.
The Philosophy of Physical Science. New York, Macmillan, 1939.
Edgeworth, F. Y. Probability. In Encyclopaedia Britannica, vol. XXII,
New York, Encyclopaedia Britannica, Inc. (13th ed.), 1926.
Einstein, A. Geometric und Erfahrung, erweiterte Fassung des Festvortrages
gehalten an der Preussischen Akademie der Wissenschaften zu
Berlin am 27 Januar 1921. Berlin, Springer 1921.
The Meaning of Relativity. Princeton, Princeton Univ. Press,
1923.
Physics and reality. Journal of the Franklin Institute, 1936 221,
348Autobiographical
382
notes. In Albert Einstein: Philosopher-Scientist
(P. A. Schilpp ed.) Evanston, Library of Living Philosophers, 1949.
Reply to criticisms. In Albert Einstein: Philosopher-Scientist (P.
A. Schilpp ed.) Evanston, Library of Living Philosophers, 1949.
Einstein, A., and Infeld, L. The Evolution of Physics. New York, Simon
and Schuster, 1938.
Einstein, A., Podolsky, B., and Rosen, N. Can quantum-mechanical de-
scription of physical reality be considered complete? Physical Review,
I935.47.777-~780-
Ellis, R. L. On the foundations of the theory of probability. Transactions,’
Cambridge Philosophical Society, 1843, 8,1-6. -
Enriques, F. The Historic DeveIOpment of Logic. (J. Rosenthal trans.)
New York, Holt, 1929.
Eve, A. S. Rutherford; Being the Life and Letters of the Rt. Hon. Lord
Rutherford, O. M. New York, Macmillan, 1939.
Fairbanks, A. The First Philosophers of Greece. New York, Scribners,
I898.
Faraday, M. Experimental Researches'In Chemistry and Physics. London,
Taylor and Francis, 1859.
338 PSYCHOLOGY AND LOGIC

Feigl, H. Operationism and scientific method. Psychological Review, 1945,

52, 250-259.
Existential hypotheses. Philosophy of Science, 1950, 17, 3 5-62.
Ferguson, A., et al. Quantitative estimates of sensory events. Report of the
British Association for the Advancement bf Science, London, British
Association, 1938.
Fisher, R. A. Statistical Methods for Research Workers. London, Oliver
and Boyd, 1942.
Fourier, J. B. J. The Analytical Theory of Heat. (A. Freeman trans.)
Cambridge, Cambridge Univ. Press, 1878.
Frank, P. Foundations of Physics. In International Encyclopedia of Unified
Science, vol. I, No. 7, Chicago, Univ. of Chicago Press, 1946.
Frege, F. L. G. Die Grundlagen der .Arithmetik, begrifszchriftlich
abgeleitet. Vol. I, Jena, Poh‘le, 1893; vol. II,\1903.
Gauss,.C. F. Intensitas vis magneticas terrestris a'd mensuram absolutam
revocata. In Werhe, vol. V, Leipzig, Teubner, 1863- 1903.
George, W. H. The ScientistIn Action: Scientific, Study of His Methods.
London, Williams and Norgate, 1936.
Ginsburg, B. The finite universe and scientific extrapolation. Journal of
Philosophy, 1935, 32, 85-92.
Goldenweiser, A. History, Psychology, and Culture. New York, Knopf,
1933-
Gotschalk, D. W. Of the nature and definition of a
Review. I93I. 49. 469--477
cause.
Philosophical
Greenwood, T. Probability. In The Dictionary of Philosophy (D. D.
Runes ed. ) New York, PhIIOSOphical Library, 1942.
Grimsehl, E. A Textbook of Physics. (Translated from 7th German edi-
tion. L. A. Woodward, ed.) London and Glasgow, Blackie, 1932-
1935, 5 vols.
Haldane, E. S. and Ross, G. R. T. (eds.) The Philosophical Work of
Descartes. Vol. I, Cambridge, Cambridge Univ. Press, 191 I.
Harding, T. S. Science at the tower of Babel. Philosophy of Science, 1938,
5. 338-353-
Hardy, G. H. The theory of numbers. Science, 1922, 56, 402-405.
von Hartmann, E. Kategorienlehre. Leipzig, Haacke,1896. \
Hawkins, D. Existential and epistemic probability. Philosophy of Science,
I943. 10. 255--26I
Heath, T. L. The Thirteen Books of Euclid’s Elements. (Translated frpm
the Text of Heiberg with Introduction and Commentary.) Cam-
bridge, Cambridge Univ. Press (2nd ed.), 1926, 3 vols.
Hedrick, E. R. Tendencies in the logic of mathematics. Science, 1933,‘
77. 335-343-
Helmholtz, H. Ziihlen und Messen. In Philosophische Aufsa'tze, Edward
Zeller zu seinem fu'nftzigja'hrigen Doctor-Jubila'um, Leipzig, Fues’s
Verlag (Reisland), 1887. x
 BIBLIOGRAPHY 339
Hempel, C. G., and Oppenheim, P. A definition of “degree of confirma-
tion.” Philosophy of Science, 1945, 12, 98-1 15.
Studies in the logic of eXplanation. Philosophy of Science, 1948,
15. I35'175-
Herodotus. History. (G. Rawlinson trans.) New York, Dial Press, 1928.
Hobbes, T. See Molesworth.
Hoffman, J. J. I. Der Pythagorische Lehrsatz mit 32 teils bekannten, teils
neuen Beweisen versehen. Mainz, Kupferberg (2nd ed. ), 1821.
Hofstadter, A. Causality and necessity. Journal of Philosophy, 1949, 56,
257-270.
Horst, P. Measurement, relationship, and correlation. Journal of Philoso-
phy. I932. 29. 631-637
Houston, W. V. Philosophy of physics. Science, 1937, 45, 413-419.
Hume,
D. A Treatise on Human Nature. (L. A.
Clarendon, 1896.
Selby-Bigge ed.) Oxford,
James, W. Principles of Psychology. Vol. I, New York, Holt, 1890-1 905.
Jasche, G. B. Immanuel Kant’s Logik. Leipzig, Meiner (3rd ed.), 1920.
Jefireys, H. Scientific Inference. Cambridge, Cambridge Univ. Press, 1937.
Theory of Probability. Oxford, Clarendon (2nd ed.) , 1948.
Jevons, W. S. Elementary Lessons in Logic. New York, Macmillan, 1912.
Principles of Science. London, Macmillan, 1924.
Johnson, H. M. Pseudomathematics in the mental and social sciences.
American Journal of Psychology, 1936, 48, 342-351.
Johnson, W. E. Logic. Cambridge, Cambridge Univ. Press, 1921.
Jones, P. C. Kant, Euclid and the non-Euclidians. Philosophy of Science,
I946. I3. I37~I43
Jordan, P. On the process of measurement in quantum mechanics. Phi-
losophy of Science, 1949, I6, 269-278.
Joseph, H. W. B. An Introduction to Logic. Oxford, Clarendon (2nd ed.),
1916.
Kaiser, C. H. The consequences for metaphysics of quantum mechanics.
I940. 37. 337-348.
Kantor, J. R. Principles of Psychology. New York, Knopf, 1924-26,
2 vols.
A Survey of the Science of Psychology. Bloomington, Principia,
I933-
An Objective Psychology of Grammar. Blomington, Indiana
Univ. Publications, Seience Series, no. I, 1936.
The Operational principle in the physical and psychological sciences.
Psychological Record, 1938, 2, 1-32.
An interbehavioral analysis of propositions. Psychological Record,
1943. S) 309-339- ∙ ∙ ∙
Problems of PhySIologIcal Psychology. Bloomington, PrincipIa,
x
1947-
Kaplan, A. Definition and specification of meaning. Journal of Philosophy,
\‘3946, 43, 281-288.
340 PSYCHOLOGY AND LOGIC

Kemble, E. C. The probability concept. Philosophy of Science, 1941, 8,

204-232.
Keynes, J. M. A Treatise on Probability. London, Macmillan, 1921.
Keynes, J. N. Studies and Exercises in Formal Logic. London, Macmillan
(4th ed.), 1906.
Klein, F. Elementary Mathematics from an Advanced Standpoint. (E. R.
Hedrick and C. A. Noble trans.) vol. I, New York, Dover, 1945.
Kneale, W. Probability and Induction. Oxford, Clarendon, I949.
Lamont, C. Naturalism and the appreciation of nature. Journal of Phi-
losophy, I947? 44': 597'"608
Laplace, P. S. Théorie Analytique des Probabilités. Paris, Courcier (3rd
ed.,) 1820.
Larmor, J. Dimensions of units. In Encyclopaedia Britannica, v01. XXV-II,
New York, Encyclopaedia Brittannica, Inc. (13th ed. ), 1926.
Leibniz, G. W. Monadology. Chicago, Open Court, 1902.
Lenzen, V. F. The Nature of Physical Theory: Study in Theory of
Knowledge. New York, Wiley, 1931. o
Indeterminism and the concept of physical reality. Journal of
Philosophy, I933, 30, 281-288.
Concepts and realityIn quantum mechanics.- Philosophy of Science,
1949, 16, 279-286.
Lewis, C. I. 'Mind and the World-Order. New York, Scribners, 1929.
Alternative systems of logic. Monist, 1932, 42, 481-507\.
Some logical considerations concerning the mental. Journal of
Philosophy, I941, 38, 225-233.
Lillie, R. S. Transmission of activation in passive metals as a mOdeI of the
protoplasmic or nervous type of transmission. Science, 1918,48,
51-60.
Transmission of physiological influence in protoplasmic systems,
especially nerve. Physiological Reviews, 1922, 2, 1-37.
The physical nature of nervous action. American Journal of
Psychiatry, I929, 9, 46I-479.
Lindsay, R. B., and Margenau, H. Foundations of Physics. New York,
Wiley, 1936.
Locke, J. An Essay Concerning Human Understanding. London, Johnson,
1801.
Loeb, L. B., and Adams, A. S. The Development of Physical Thought,
A Survey Course of Modern Physics. New York, Wiley, 1933.
Lotka, H. J. Probability —increase in shuffling, and the asymmetry? of
time. Science, I924, 59, 532--536.
Lotze, H. Logik, Drei Biicher vom Denken, vorn Untersuchung und vom
Erkennen. Leipzig, Hirzel ( 2nd ed.), \1880.
McGill, V. J. Concerning the laws of contradiction and excluded middle.
Philosophy of Science, 1939, 6, 196-2.11
McGilvary, E. B. RelationsIn general and universalsIn particular. Journal
of Philosophy, 1939, 36, 5- I 5, 29-40.
 BIBLIOGRAPHY 341
McGregor, D. Scientific measurement and psychology. Psychological Re-
view, 1935,42, 246-266.
Mach, E. Space and Geometry'In the Light of Physiological, Psychological,
and Physical Inquiry. Chicago, Open Court, 1906.
The Science of Mechanics. Chicago, Open Court, 1907.
Manse], W. L. Prolegomena Logica: An Inquiry into the Psychological
Character of Logical Processes. London, Livingston (I851), I862.
Margenau, H. Critical points in modern physical theory. Philosophy of
Science. I937. 4. 337-370-
Reality in quantum mechanics. Philosophy of Science, 1949, 16,
287-302.
Maxwell, J. C. Matter and Motion. (J. Larmor ed.) New York, Mac-
millan, 1920.
Meldrum, W. B. and Daggett, A. F. A Textbook of Qualitative Analysis
(Using the Semimicro Methods). New York, American Book Co.,
1946.
Menzel, D. H. and Layzer, D. The physical principles of the quantum
theory. Philosophy of Science, 1949, 16, 303-324.
Mercier, C. A. On Causation, with a Chapter on Belief. London, George,
1 19.

Merz,9J. T. A History of European ThoughtIn the Nineteenth Century.

Edinburgh, Blackwood, 1903-1923, 4 vols.
Michelson, A. A. Light Waves and Their Uses. Chicago, University of
Chicago Press, 1903.
Mill, J. S. A System of Logic, Ratiocinative and Inductive. London,
Parker, 1862.
An Examination of Sir William Hamilton’s Philosophy. New
_'York, Holt, 1884, 2 vols.
von Mises, R. Probability, Statistics, and Truth. (J. Newman, D. Sholl,
and E. Rabinowitsch trans.) New York, Macmillan, 1939.
Comment on Donald Williams’ reply. Philosophy and Phenomeno-
logical Research, 1945-1946, 6, 61 1-613.
Molesworth, W. The English Works of Thomas Hobbes of Malmesbury.
London, Bohn, 1839-1845, 11 vols.
Moore, G. E. Principia Ethica. Cambridge, Cambridge Univ. Press, 1922.
Moore, J. S. Why a realism of universals? Journal of Philosophy, 1939, 36,
884-888.
Murray, G. Five Stages of Greek Religion. New York, Columbia Univ.
Press, 1925.
Nagel, E. A. Can logic be divorced from ontology? Journal of Philosophy,
1929, 26, 705-7I2.
On the Logic of Measurement. (Doctoral Dissertation, New
York, Columbia University, 1930.
A frequency theory of probability. Journal of Philosophy, 1933,
533'554'
342 PSYCHOLOGY AND LOGIC

The “formation of modern conceptions of formal logic in the

development of geometry. Osiris, 1939, 7, 142-224.
Principles of the Theory of Probability. In International Encyclo-
pedia of Unified Science, vol. I, no. 6, Chicago, Univ. of Chicago
Press, 1939.
Logic without ontology. In Naturalism and the Human Spirit
(Y. H. Krikorian ed.) New York, Columbia Univ. Press, 1944.
Some reflections on the use of language in the natural sciences.
Journal of Philosophy, 1945, 42, 61 7-630.
The meaning of reduction in the natural sciences. In Science and
Civilization (R. C. Staufi'er
I949-
ed.),_Madison, Univ. of Wisconsin Press,
Needham, J. A biologist’s view Of Whitehead’s philosophy. In The Phi-
losophy of Whitehead. (.P A. Schilpp ed.) E.,vanston Northwestern
University, 1941.
Nernst,W. The New Heat Theorem, its Foundations in Theory and
Experiment. London, Methuen, 1926.
von Neumann, J. Mathematische Grundlagen der ,Quantenmechanik.
Berlin, Springer, 1932.
Newton, I. Philosophiae Naturalis Principia Mathematica. (Cajori revision)
Berkeley, Univ. of California Press (1934), 1946.
Neyman, J., and Pearson, E. Contributions to the theory of .testing
Statistical Hypotheses. Statistical Research Memoirs, 1936,1, 1-37;
1937) 2: 2.'5"57°
Nicod, J. Foundations of Geometry and Induction. New York, -Harcourt,
Brace, 1930.
Nisbet, R. H. The foundations of probability. Mind, 1926, 35, 1-27.
Oliver, W. D. Logic and necessity. Journal of Philosophy, 1950, 47, 69-73.
Pap, A. On the meaning of necessity. Journal of Philosophy, 1943, 40,
449-458-
Pearson, K. The Grammar of Science. London, Dent (Everyman ed.),
I937-
Peirce, C. S. Collected Papers. (C. S. Hartshorne and P.
Cambridge, Harvard Univ. Press, 1931- 1935, 6 vols.
Weiss
eds.)

Pepper, S. C. How to look for causality. In Univ. of California

Publications
in Philosophy (G. P. Adams, J. Lowenberg, and S. C. Pepper eds.)
vol. XV, Berkeley, Univ. Of California Press,11932.
Definition. In Univ. of California Publications in Philosophy
(G. P. Adams, J. Lowenberg, and S. C. Pepper eds), IXXI,
Berkeley, Univ. of California Press, 1939.1
The descriptive definition. Journal of Philosophy, 1946,43, 29-
36.
What are categories for? Journal
556- oJ
Philosophy, 1947, 44,5
546-
'Planck, M. The meaning and limits of exact scienc\e. Science, 1949, I 10,
3I9-327
 BIBLIOGRAPHY 343
Poincaré, H. The Foundations of Science. New York, Science Press, 1921.
Polya, G. Heuristic reasoning and the theory of probability. American
Mathematical Monthly, 1941, 48, 450-465.
POpper, K. Logik der Forsch‘ung, zur Erkenntnistheorie der modernen
Naturwissenschaft. Vienna, Springer, 1935.
Ramsey, F. P. The Foundations of Mathematics, and other Logical Es-
says. New York, Harcourt, Brace, 1931.
Rankine, W. J. M. Miscellaneous Scientific Papers. London, Griflin, I880.
Reese, T. W. The application of the theory of physical measurement to
the measurement of psychological magnitudes, with three experi-
mental examples. Psychological Monographs, 1943, 55, no. 3.
Reichenbach, H. Wahrscheinlichkeitslehre. Leiden, Sijthofi, 1935.
Experience and Prediction. Chicago, Univ. of Chicago Press,
1938. '

Philosophic Foundations of Quantum Mechanics. Berkeley and

Los Angeles, Univ. of California Press, 1944.
Reid, T. Essays on the Intellectual Powers of Man. Edinburgh, Bell, I 785.
Rickert, H. Zur Lehre von der Definition. Tiibingen, Mohr (2nd ed.) ,
1915.
Ritchie, A. D. Scientific Method, an Inquiry into the Character and
Validity of Natural Laws. London, Kegan Paul, 1923.
Robinson, R. Definition. Oxford, Clarendon, 1950.
Robson, W. A. Civilization and the Growth of Law. New York, Mac-
millan, 1935.
Rosenblueth, A. and Wiener, N. The role of models in science. Philosophy
of Science, 1945, 12, 316-321.
Royce, J. The Principles of LOgic. In Encyclopaedia of the Philosophical
Sciences, vol. I (A. Ruge ed.) London, Macmillan, 1913.
Ruark, A. E.., and Urey, H. C. Atoms, Molecules, and Quanta. New
York, McGraw-Hill, 1930.
Runes, D. D. The Dictionary of Philosophy. New York, Philosophical
Library, 1942.
Runge, C. Maass und Messen. In Encyclopa’die der Mathematischen Wis-
senschaften mit Einschluss'ihrer A nwendungen, 1903-1921, 5, 3—24.
Russell, B. The Problems of Philosophy. New York, Holt, 1912.
On the notion of cause. Proceedings, Aristotelian Society, New
Series, vol. XIII, London, Williams and Norgate, 1913.
Our Knowledge of the External World as a Field for Scientific
Method in Philosophy. London, Allen and Unwin, 1914.
Logical Atomism. In Contemporary British Philosophy. (J. H‘
Muirhead ed.) London, Allen and Unwin, New York, Macmillan,
1924.
Introduction to Mathematical PhilosoPhy. London, Allen and
Unwin(1919),1924.
Analysis of Matter. London, Kegan Paul, 1927.
Mysticism and Logic. New York, Norton, 1929.
34-4- PSYCHOLOGY AND LOGIC

Principles of Mathematics. Cambridge, Cambridge Univ. Press,

1903; (Reissue) New York, Norton, 1938.
An Inquiry into Meaning and Truth. New York, Norton, 1940.
My Mental Development. In The Philosophy of Bertrand Russell
(P. A. Schilpp ed.) Evanston, Northwestern University, 1944.
Reply to Criticisms. In The Philosophy of Bertrand Russell (P. A.
Schilpp ed.) Evanston, Northwestern University, 1944.
Human Knowledge, Its Scope and Limits. New York, Simon
and Schuster, 1948. ‘

Ryle, G. Induction and hypothesis. Aristotelian _ Society, Supplementary

vol. XVI, London, Harrison, 1937.
Scates, D. E. How science measures. Journal of Experimental Education,
1937, 5. 296-312.
Schiller, F. C. S. Formal Logic. London, Macmillan, 1912.
Schlick, M. CausalityIn everyday life andIn recent science. In University
of California Publications in Philosophy, vol. XV, no. I 5 (G. P.
Adams, J. Lowenberg, and S. .C. Pepper eds ), Berkeley, Univ. of
California Press, 1932.
Schopenhauer, A. On the Fourfold Root of the Principle of Sufficient
Reason, and On the Will of Nature, London, Bell, I 891.
Schrodinger, E. What Is Life? The Physical Aspects of the Living Cell.
Cambridge, Univ. Press, 1944.
Sears, J. E. Measurements. In Encyclopaedia Britannica, vol. XXX, New
. York, Encyclopaedia Britannica, Inc. (13th ed. ), 1926.
Sellars, R. W. The Philosophy of Physical Realism. New York, Macmillan,
I 32.

Semat,
9H. Fundamentals of Physics. New York, Rinehart, 1945.
(Sheppard, W. F. Probability. In Encyclopaedia Britannica, vol. XXXI,
New York, Encyclopaedia Britannica, Inc., 1926.
Shewhart, W. A. Economic Control of Quality of Manufactured Product.
New York, Van Nostrand, 1931.
Statistical Method from the Standpoint of Quality Control. Wash-
ington, Graduate School, Dept. of Agriculture, 1939.
Sommerfeld, A. Atomic Structure and Spectral Lines. (H. L. Brase trans.)
vol. I, New York, Dutton (3rd ed.,) 1934.
Spaier, A. La Pensée et la Quantité: Essai sur la Signification et la Rea-lité
des Grandeurs. Paris, Alcan, 1928.
Spencer, L. J. Abrasives. In Encyclopaedia Britannica, vol. I, New York,
Encylopaedia Britannica, Inc. (14th ed. ), 1929.
Spinney, L. B. A Textbook of Physics. New York, Macmillan
1925.
(3rd ed. ),
Spinoza, B. Ethics. London, Dent (Everymamed. ), 1910.
Stebbing, L. S. A Modern Introduction to Logic. London, Methuenl(3rd
ed. ), 1942.
Stevens, S. S. A classification of scales ofmeasurement. Paper read before
The International Congress for the Unity of Science, September, 1941.
 BIBLIOGRAPHY 34 5
On the theory of scales of measurement. Science, 1946, 103,
677-680.
Stranathan,J. D. The “Particles” of Modern Physics. Philadelphia and
Toronto, Blakiston, 1942.
Strong, E. W. Procedures and Metaphysics: a Study in the PhiIOSOphy
of Mathematical-Physical Sciences in the 16th and 17th Centuries.
Berkeley, Univ. of California Press, 1936.'

Struik, D. J.oOn the foundations of the theory of probability. Philosophy

of Science, 1934, 1, 50-70.
Stumpf, C. Neben den BegrifI der Mathematischen Wahrscheinlichkeit.
Berlin Bayr, Academic (phil. class.), 1892, 37-120.
Taylor, L. W. Physics, the Pioneer Science. Boston, Houghton, Mifflin,
1941.
Thomson, W. (Lord Kelvin) Popular Lectures and Addresses. London,
Macmillan, 1889, 3 vols.
Todhunter, I. A History of the Mathematical Theory Of Probability from
the Time of Pascal to that of Laplace. Cambridge and London, Mac-
mfllan, 1865.
Toms, E. The law of excluded middle. Philosophy of Science, 1941, 8,
33-38.
Vaihinger, H. Die Philosophie des als ob: System der Theoretischen, Prak-
tischen und Religiosen Fiktionen der Menschheit auf Grund eines
Idealistischen Positivismus. Leipzig, Meiner (7th ed.), 1922.
Veblen, O. Problems and solutions. American Mathematical Monthly,
I922. 29. 357-358.
Veblen, O. and Young, J. ‘W. Projective Geometry. Vol. I, Boston, Ginn,
'1910.
Venn, J. The Logic Of Chance: an Essay on the Foundations and Province
of the Theory of Probability with Special Reference to its logical
Changes and its Application to Moral and Social Science, and to
Statistics. London, Macmillan, 1888.
Waismann, F. Are there alternative logics? Proceedings'Aristotelian So-
ciety. I945--I946 46. 77-104.
Inference.
Wald, A. On the Principle of Statistical South Bend, Notre
Dame Univ.,1942.
Wallace, W. The Logic of Hegel. (Translated from the Encyclopaedia of
the Philosophical Sciences) Oxford, Clarendon, 1892.
Wang, H. Notes on the justification of induction. Journal of Philosophy,
I947: 44, 701'710°
Webster, D. L., Farwell, H. W., and Drew, E. G. General Physics for
Colleges. New York, Century, I92 3.
Weitz, M. Philosophy and the abuse of language. Journal of Philosophy,
I947. 44. 533- -546
Whewell, W. History of Scientific Ideas. London, Parker, 1858, 2 vols.
Novum Organum Renovatum. London, Parker, 1858.
346 PSYCHOLOGY AND LOGIC

Whitehead, A. N. The Concept of Nature. Cambridge, Cambridge Univ.

Press, 1920.
Science and the Modern World. New York, Macmillan, 1925.
Whitehead, A. N., and Russell, B. Principia Mathematica. Cambridge,
Cambridge Univ. Press, 1910-1912, 3 vols.; (2nd ed.) 1925-1927.
Wiener, N. A new theory of measurement. Proceedings, London Mathe-
maticalSocz’et’y, 1921,19, 181--.205
Will, F. Is there a problem of induction? Journal of Philosophy, 1942, 39.
Will the future be like the past? Mind, 1947, 332-347.
Williams, D. C. The definition of yellow and of good. Journal of Philoso-
phy. 1930. 27. 515-527.
The Ground of Induction. Cambridge, Harvard Univ. Press,
1947-
Windelband, W. Vom System der Kategorien. In Philosophischen A bhand-
lungen (Christoph Sigwart zu seinem siehzigsten Geburtstage.)
Tiibingen, Mohr, 1900.
Winderlich, R. Das Ding: Eine EinfuhrungIn das°Substanzproblem; Teil
I, Die Dinge der Naturwissenschaft. Karlsruhe, Braun, 1924.
Winn, R. B. The nature of causation. Philosophy of Science, 1940, 7, 192-
204.
Wittgenstein, L. Tractatus Logico-Philosophicus. New York, Harcourt,
Brace, 1922.
Wolif,.K. H. The unique and the general: toward a philosophy .,of soci-
ology.Philosophy of Science, 1948,15, 192-2-10.
Wundt,W .Logik. Stuttgart, Enke, (3rd ed.) 1906— 1908, 3 vols.
Young, J. W. Lectures on Fundamental Concepts of Algebra and Geome-
try. New York, Macmillan, 191 1.
 SUBJECT INDEX
Absolutism (see Universalism), in ab- Behavior, largely abstractional, 10f.;
stracting, 15; in definition, 33f., 46; source of logical forms, 140; system-
in formal systems, 82; in philosophical ological, and logical laws, 202f.;
systems, 76, 82f., 332f.; in categoriza- measurement of judging, 273; conti-
tion, 128; in causation, 147, 152f., 154, nuity of, 321f.
159; in laws of thought, 175f., 181; by Belief, probability as degree of, 208f.,
means of language, 194; in mathe- 212, 218, 223f., 2385.
matics, 196f.; in probability theory, Biology, instrument construction in, 103;
206, 215f., 226f., in measurement 'and ontology, 103; categories in, 126f.
theory, 2935.; and statistical pro-
cedures, 332f. Calculating systems, 302.
Abstracting, in system building, 1, 7, 26f., Calculation, in probability theory, 2065.,
8o, 106f., 333; as interbehavioral 210, 239, 245f., 2615.; operational
operation, 1f., 7, 10f.; field character character of, 220; in measurement,
of, I; in mathematics, 15., 23f.; logi- 271, 291f., 302.
cal forms as products of, 25., 1385.; Calculus, probability (see Probability
and symbolic logic, 25.; constructional calculus), 210, 220, 243.
principles in, 45.; in Euclidean ge- Categories, necessary in logic, 108; mis-
ometry, 45.; individual and cultural interpretations of, 108; are system-
operations of, 75.; and quantum me- building instruments, 108f.; basic in-
chanics, 8, 12; and the existence prob- tellectual tools, 109; Aristotelian,
lem,8 , 9f.., 1885.; of numbers, 8f.,10, 109f., 129; Kantian, 109; Hegelian,
25, 26,306; interrelated with gen- 109; historical evolution of, 1095.,
eralizing, 9f.; and perception, 9; reflect cultural background, I 12f., 1 16,
means formalizing, 10; specific char- 122, 129; current attitudes toward,
acteristics of, 10f.; in measurement, 10, 112; different sources of, 113; and
306f.; symbols as products of, 10f.; knowledge, 113; as psychic entities,
divisive, 11f.; multiplicative, 12f.; and 114; related to classes and universals,
atomism, 12; in psychology, 12f.; role 114f.; connect logic with language,
of [language in, 185., 27; criteria for, 115, 128; and symbols, 115; technique
20f.-; specificity principle in, 20f.., 24; of creating, 1 15; interbehavioral origin
limits of, 21f.; illicit, 215.; extra- Of, 1155.; space and time as, 116;
polation in, 24; and universals, I32, abstract continuous with concrete,
138, 143; in science, 189. 116f.; specific functions of, 117f.;
Action sources, of logical forms, 141. range of,118f.; in mechanics, 119f.;
Acts, as system materials, 73; fixated, 98. in electricity, 120; in geometry, 120;
Actuality, a probability factor, 225. everyday, 121; religious, 121f.~; in di-
Aesthetics, categories in, 129; measure- mensional analysis, 1225.; value,
ment in, 304. 124f.; in thermodynamics, i25f.; in
Algorithms, as mathematical instruments, biology, 126f.; ethical, 127f.; logical,
101. 128; grammatical, 128f.; philosophi—
Ambiguity,’In definition, 465. cal, 129; and ontology, 129; aesthetic,
Analogizing, as generalizing inter- 129.
behavior, 15f.; in mathematics, 15; Causation, as a logical problem, 145f.,
and scientific models, 15f.; definition 167; as an ontological problem, 145f.,
by. 51-
Analytical logics, 144.
1495. ; influence of deductive mathe-'
matics on, 146f. ; in methodistic logic,
Analyzing, definition by, 50f. 146f.-; cultural background of,1475;,
Antigenetic fallacy, 2. 159, 167; naturalistic Greek concep-
Atomic physics, 266, 3055., 3095. tion of, \148; and mysticism, 148;
AtOmism, and abstracting, 12. classical theories of, 1495.; substituted
Axiological categories, 124f. for, by sentences, 150, 154, 162; as
343 PSYCHOLOGY AND LOGIC

a priori knowledge, 150, 159; thing, world, 206, 224, 266, 270, 313; in
150f.; relation, 151-, event, 152, ∂∙ ∙ relation to Ob-
1685.; as universal, 152f. ; specificity servation, 311, 316.
of,15.2f, 159, 167, 171f.; and logical Constructional measurement, 274.
necessity, 154f.~, interrelation of field Constructs, logical instruments as, 94; as
components, 156f., 166; as correlation, causal data, 1615.; and causal law,
1565., I62; and statistics, 156; opera- 173; derived from events, 327.
tional theory in, 158f.; based on Contingency, of logical laws, 182, 186;
atomistic psychology, 1595.; constructs in science, 257; of measurement, 272.
as data for, 1615., 166; events as data Contradiction, law of, 175, 18of., 191,
for, 1635.; investigative operations in, 194, zoof.
165f.; interpretation of, 166f.; erect- Convergence principle, and generaliza-
ing a system Of, 1675.; instrumental tion, 18; in probability theory, 240.
systems of, 171f.; in scientific research,
'Correlation, and causation, 1565., 162.
171f.; paradoxes of, and quantum me- Cosmic systems, 76, 103,322, 3295., 333.
chanics, 172f.; and indeterminism, Counting and‘ measuring, 2965.
1725.; and measurement, 174, 305,
Counting systems, 301f.
311f. . Culture, influences system building, 7, 83,
Cause, vs. condition, 166; and e5ect, 166,
168. 90, 92, 106; in abstracting and gen-
Certainty, as aim of formal systems, 77, eralizing, 8§.; logical instruments re-
82; in probability theory, 207, 222f., flect, 92, 106; and categorization,
242. 112f., 116, 122,~‘129; causation theory
Chance, and probability theory, 204, 216, attached to, 1475., 159, 167, 216;
234.
and logical laws, 1795., 202; pre-
Chemical analysis, as operational system, suppositions of, and quantum me-
84f. . chanics, 317.
Classes, are systems, 58f.; as system ma-
terials; 72f.; as mathematical instru- Data, truth depends on, ”192.
ments, 101; and categories, 114f.; not Deduction, and system building, 86; and
restricted to formal structures, 1 15; as science, 86; vs. induction, .86, 229,
logical forms, 140; in analytical logics, 332; and probability, 205, 206, 211,
144; stressed by conceptualistic lO- 226, 228f., 255, 260.
gicians, 183. Definition, confusion concerning, 28, 35;
Classification, universalism in, 28; and concerned exclusively with words, 29,
Operationism, _29; defining by, 49f.; 37; formalistic and methodistic treat-
and language, 58; as system building, ments of, 29f.; One and Many prob-
58f., 118; criteria for, 59; may pre- lem in, 29; as search for essences,
cede definition, 59; and categorIcal 295.; and sensationistic psychology,
function, 118. 31; nominal, 31f., 41f.; real, 32,45;
Concepts, di5erentiated from things, 94; conceptual, 32f.; and symbolic logic,
as fixated behavior, 97; and categories, 33f.., 53; symbols confused with things
I 14;
as mental states, I 14; as forms of in, 33; syntactical, 33; absolutism‘\in,
interbehavior, 114; and universals, 33f., 46; semantic, 34; from inter-
I3 5.
Conceptualism, Kantian, 8, 114; in defi-
behavioral standpoint, 345. ., 36,
product vs. performance in, 35, 37;
45:.;f
nition, 32f., 36, 42; and universals, wide range of, 35; criteria for 35f.;
I32f., 137; logical laws treated on analysis of factorsIn, 36f.; by Conven-
basis of, 183. tion, 37; ostensive, 37, 41, 56; com-
Conditions, as constructional terms for pared with propositions, 38, 41; lin-
causes, 166. guistic factors in, 385.; relatidnto
Consciousness, threshold of, as psycho-
logical model, 106.
systembuilding, 40; and description,
40f. ; and meaning, 42f. ; and the in-
Consistency, of thinking, 187; as a law definable, 445. ; ambiguity and vague-
∙ of thought, 187, 201.
Construction, and abstraction, 4; in-
ness in, 465.; ‘classes of, 495.; fiat type
of, 52f. ; ∂ in, 535.; '

fluences upon, 8; de novo, 107; in manipulative vs. linguistic, 55; genus-

causal theory, 166f.; of \an external et-diflerentiam rule in,57f.; noncircu-
 INDEX 349
larity rule for, 58; classification may External world, 206, 224, 266, 270, 313.
precede, 59. Extrapolation, generalizing by, 14f.,
Delimitation, definition by, 49. 17f., 21f.; and abstraction, 24; in
Descriptions, substituted for things de- logic and science, 1885.; in measure-
scribed, 259; vs. operations, confused ment, 291f.
in measuring situations, 307; originate
in contacts of organisms and objects, Factorial coordination, probability as,
317. 2535.
Determinism, and causation, 1725., in. Facts, vs. laws in causation, 173; fitted
∙ science, 1725“; 312; and measurement to a priori principles, 227.
theory, 312, 320. Field theory (see Interbehavior), and ab-
Dialectical logics, 144. stracting behavior, 1; and causation,
Dichotomy, and polytomy, 194. 156f., 158, 166,172.
Dimension systems (see Dimensional Form, vs. content, 195.
analysis), 119. Formal logic (see Formalism, in logic),
Dimensional analysis, categories in, stresses structure, 76f. ; a closed system;
1225.; and measurement theory, 123, 77; primary categories of, 128.
294. Formalism, in abstraction, 10, 1885.; in
Disjunction, means of establishing truth, definition, 29f.., 335.., 53; in mathe-
193£ matics,“ 82, 197; as logically typical,
Divisive abstractions, 1 If. 82f.; in categorization, 109, 111; in
Dualism, 8, 104f., 114, 137, 216, 266f. logic, 143f., 154f., 1815., 185, 187f.,
1905.; and causation, 149f., 154f.,
Electricity, development of categories in 170; in science, 189.
field of, 120. Forms, logical (see Logical forms).
Epistemology, and causation, I50; and Formulae, as logical instruments, 91; as
logical laws, 1905.; influences prob- mathematical instruments, 101.
ability theory, 207, 2565. Free will, among physicists, 311i.
Equating systems, 394. Frequency theory, of probability, 209,
Equipossible ratio, in probability theory, 213f., 218f., 235, 2415
207. 217. 23415. Functionality, and causation, 1565.
Essences, in definition, 295., 34; con-
nected with the universal, 31 ;' as Gaming problems, in probability evolu-
philosophical category, 129. tion, 216f., 234.
Estimating systems, 302f. Generalizing, in system building, I, 45.;
Ethics, categories in, 127f. as interbehavioral operation, If, 13f.,
Evaluating behavior, measurement of, 22; logical forms as products of, 25.,
273. . 141; and symbolic logic, 25.; con-
Evaluating systems, 273, 303f. structional principles in, 45.; in
Events, confused with symbols, 39, 174, Euclidean geometry, 45.; individual
324; as system materials, 71f.; as operations of, 75.; cultural, 8f.; inter-
sources of logical forms, 1395.; vs. related with abstraction, 9f.; specific
constructs, 17 3, 327; relations between, characteristics of, 135.; and the law
misconstrued, 173f.; probability sys- of excluded middle, 14; by extrapola-
tems centered in, 218f., 241; replaced tion,'14f., 17f., 21f.; limits of, 14f.,
by propositions, 259. 21f.; by analogizing, 15f.; by mag—
Excluded middle, law of, 14, 94, 175, nification, 16; by reiteration, 16f, 24;
I77. I84.
1911., I97If and mathematics, 16f., 20f.; and the
Existence, problem of (see Ontology), in infinite, 16f., 23; by interpolation, 17;
causal theory, I59f., 164, 173f.; and and the convergence principle, 1 8; spec-
numbers I74, 270; and logical laws, ificity principle in, 20f., 24; criteria
1885.; and probability, 211f.; in for, 20f.; illicit, 225.; and material
. measurement, 266, 3085., 3135.; and implication, 24; in psychology, 25;
system building, 327f. fallacies resulting from, 25f.; in sys-
Experience, and the a priori controversy, tem building, 26f.; and universals,
24f. 132, 138.
Experimentation, 189, 226. Geometry, as abstracting and generalizing
Explanatory tools, in logic, 99. behavior, 45.; categories in, 120;
350 PSYCHOLOGY AND LOGIC

freed from Euclidean absolutism, 196f. Interval scale, of measurement, 287.

Grammar, categories in, 128f. Intuition, in mathematics, 82; in prob-
ability theory, 247f., 258.
Hegelian dialectics, 94; categories as ma- Investigation, in causation, 165f.; as sup-
terials of, 109. porting the a priori, 226; in probabil-
Hierarchies of events, in science, 12, 117, ity systems, 256 ; scientific vs. mensura-
324i tional practice, 2995.
Human-science vs. natural-science values,
124£ Judging, as a technical category of for-
Humanistic events, in logical systems, 72. mal logic, 128; measurement of, 273,
Identity, law of, I75, I77f., 180f., 184f., 275.
2005.
Ideology (see Culture), trends of, in Knower vs. known, 324.
measurement theory, 266f. ”Knowledge, causal, 174; problems of, re-
Ignorance, equal distribution of, 235; lated to logic, 190f.; propositional,
236f., 257. 192; probability as, 2235., 2375.; as
interbehavidral event, 259, 324.
Immediate inference, as system product,
77.
Implicatory systems, 77f. Language (see' Symbols, Terms), place
Indefinable, the, 445., 109. of, in abstracting and generalizing,
Indeterminacy principle (see Determin- 185.; and definition, 33f., 39f., 465.;
ism), 227, 299; in measurement, 319f. in system building, 735.; and logic,
Indi5erence, Principle of (see Ignorance, 95, 154f., 183, 195; categories connect
Equal Distribution of). logic and, 115; and universals, 136f.;
Induction, and system building, 79, 331f., stressed by logical epistemologists, 1 83;
333; complete, 79; vs. deduction, 86, serves to absolutize, 194; establishes
332; and logic, 193, 324; and prob- the a priori, 1 94; logic exclusively con-
ability, 205, 211, 2285. cerned with, 195; as basis "of truth,
Inference, 250, 258. 195; can not unify science, 195f. ; in
Infinite, the, as product of generalization, description of quantum events, 319f.
16f., 23; a name for iterative action, Laws of things, probability as (see Fre-
23; and categorization, 13o. quency theory, of probability), 2415.
Infinitesimal, the (see Infinite, the), a Laws of thought and things (see Logical
mathematical category, 1 3o. laws).
Instrument construction (see Logical'In- Linguistic logic, 1545.
struments), in logic, 100f.; in mathe- Linguistic metaphysics, 2 59.
matics, 101f.; in physics, 102f.; in Linguistic systems, consist of referential
biology, 103f.; in psychology, 104f.; actions, 735.
techniques in, 106f.; iconography as, Linguistic things, as system materials, 7 3.
106; abstraction in, 106f.; de novo, Logic, symbolic, 25., 24, 3 3f., 128, 144;
107; in the social sciences, 107. and mathematics, 4, 82, I46f.,154,
Insuflicient Reason, Principle of (see Ig- 182, 184; non sequitur fallacy'In, 25,
norance, Equal Distribution of). 324-, formal systems as typical of, 82f.;
Interbehavior, basic to abstraction, 1f., no a priori limits to processes of, 88;

∙
10f.; mathematical equations derived not concerned with autonomous forms,
from 2; generalization a specific pro- 95; instrument construction in, 100f.;
cedure of, 13f., 22; and definition, categories inevitably involved in,
345.; categories originate out of, is relative, 125, primary categories.I of,
1155.; and universals, 1375.; causal,
158; and probability, 207f., 233f.; and
128; and universals, 131, types'
144; domain of law, 175; integrated of:
measurement, 272, 279; continuum of with events, 175, 200, 228; and .cul-
observational, 309f.; perception as, ture, 1795., 202; and ontology,
317; knowing as, 324. 183, 1885., 202; illicit extrapolation
Interbehavioral theory, and pragmatic in, 1885.; ‘verification in, 191f.; as
realism, 138.. sentential, 195,202; probability, guide
Interpolation, generalizing by, 17; in to, 204; and science, 228f., 3245.;
measurement, 291f. interbehavioral view'of, 319; and phi-
 \
INDEX 351.
losophy, 321f., 328f.; and metalogic, Logical multiplexity, vs. logical uni-
322f.; and statistics, 332f. versalization in probability theory,
Logic of science, vs. logic of truth, 225. 2 5 5f.
Logical forms (see Universals), as things, Logical necessity (see Necessity).
136f.; as terms, 136f.; as thought, Logical positivism, 95, 100.
136f.; are interbehavioral products, Logically possible, the, 155.
137, 142; specificity of, 138f.; not Logos doctrine, 181
limited to universals, 139; thing or
event sources of, 1395.; qualities as, Magnification, generalizing by, 16; un-
139; properties as, 139; relations as, controlled, 2 56.
139f.; classes as, 140; behavior as Magnitudes, variety and specificity of,
source of, 140; numbers as, 14of.; 28 If.; “psychic” vs. physical nature of,
product sources of, 141 ; as abstractions 295i
and generalizations, 141; as system Material implication, 24.
components, 142f.; propositions as Material systems, comprise things, 615.;
compounds of, 143; in various logics, events, 71f.; relations, 72; classes, 72f.;
144 .. acts, 73; linguistic things, 73.
Logical instruments (see Instrument con- Mathematical systems, 795.
struction), ch. 16; employed also in Mathematics (see Numbers, Symbolic
science, 91f.; models, schemata, and logic), interbehavioral nature of, If.;
formulae as, 915.; and culture, 92, field operations in, 2, 82; formalism
106; exclude a priori principles, 92; in, 3f., 82, 197; reduced to logic, 3f.;
borrowing of, 93; originality in crea- as system building, 4; mysticism in, 7;
tion of, 93; autistic construction of, analogizing in, 15; as postulation,
93; specificity of, 94f.; interbehavioral 81f., 215f., 235; intuition in, 82; a
view of, 94f.; as orientational tools, domain of relations, 101f.; classes,
9 5; scientific instruments compared to, symbols, and algorithms as instruments
95f.; vs. nonlogical, 95f.; as miniature in, 101f.; instruments of, in physics,
systems, 96f.; as system items, 97; ma- 103; categorization in, 116, 130; evo-
terials of, 97f.; fixated behavior as lution of, 196f.; emancipation of,
source of, 97; for intellectual opera- a5ects logic, 197; and probability, 210,
tions, 985.; the syllogism and universe 2155., 2345., 245f.; measurement as
Of discourse as, 100; the system maker’s applied, 2675.; power of reasoning
techniques in producing, Io6f. by, 269; as science of order, 269; num-
Logical languages, 100. ber as the basis of, 269.
Logical laws, improperly interpreted, Meaning, and definition, 42f.
17 5; based upon observed events, Measurement, abstraction in, 10, 306f.;
175f.‘; as interbehavior, 176f., 186, defining by, 52; and dimensional
200; classical formulations of, 1775.; analysis, 12 3, 294; as creating things
as cultural products, 1795.; and tau- measured, 1 74, 308f. ; and the existence
tology, 179, 184, 195 ; ontological view problem, 174, 266, 270, 3085., 3135.,
of, 179, 183, 1885., 194f.; Aristotle’s 319; and causality, 174, 305, 311f.;
commonsense view of, 1795.; historical and probability, 228, 243, 247f.,
origins of, 1795.; suflicient reason, 3o5f.; place of, in logic, 265; theory
182, 185; current treatment of, cor- vs. practice in, 265f.; is system build-
responds to nominalism, realism, con- ing, 265, 271, 2995., 318; unsatisfac-
ceptualism, 183; and syntacticism, 183; tory constructional trends in, 266f.;
mind and reason as basis for, 184, indeterminacy principle in, 266, 319f.;
186f.; contingency of, 186; as system- as applied mathematics, 2675.; and
building enterprises, 186, 1995.; in numbers, 2685.; 297, 306; specifiicity
reasoning and thinking, 1865:; con- of, ~2715., 281; and calculation, 271,
nected with data, 1905.; related to 291f., 302; from interbehavioral
knowledge problems by epistemologists, standpoint, 272, 297, 314f.; requires
‘_1905.; role of language in, 1945.; specific objects, operations, and situa-
iden‘tified with mathematical laws, tions, 2725.; practical level of, 275f.;
1965.; are constructions, 200f.; use of, in science, 276f., ch. 24; technological
not\\ limited to logic, 200; and sys- level of, 276f.; formal level of, 277f.;
temOlogical behavior, 202f. quality and quantity in, 2785.; spec-
352 PSYCHOLOGY AND LOGIC

ificity of magnitudes in, 281f., 295f.; treatment Of, 16of.; continuity in,
systems of, 2825., 3005.; metaphysical 189i
or mathematical systems no criteria for, Natural science, place of language in, 48.
282; postulates of, 283f.; units of, Necessity, logical, and factual cause,
284, 2935.; instruments and operations 1545.; in logical laws, 1835., 202;
of, 285f.; construction of scales in, its metaphysical ‘use' in probability
2855.; philosophical bias in scale mak- theory, 222f.
ing, 288f.; and psychological scales, Negative categorization, 121.
288; interpolation and extrapolation Nominal scale, in measurement, 286f.
in, 291f.; statistical operations in, Nominalism, and definition, 31f., 36,
292f., 305; counting, 2965.; and 41f., 43; and universals, 132, 135,
modern science, 2995.; in microscopic 136; and logical laws, 183‘; in prob-
physics, 3005., 305, 320; evolution of ability theory, 257, 259.
systems of, 3005.; and quantum me- _Non sequitur, fallacy of, 25, 324.
chanics, 300, 305, 307f., 3105., 3175.; Nonnumerical frequency, and probability
Operator as instrument of, 303; Opera- theory, 2 52f
tions in, confused with descriptions, Normative principles, in logic and
307; construction and observation in, science, 175f.
3075.; brings mysticism into science, Numbers (see Mathematics), are con-
311f. ; objectivity in, 312f.°, and real- structions, 2, 8f.; 25, 26; abstraction
ity, 3135., theory of (metrology), in- Of, 8f.., 10, 65, 26, 306; as explanatory
tegrated with systemology (science of system tools, 99; as logical forms,
logic), 3185. 14of.; in measurement, 174, 2685.,
Mechanics, changing categories in, 119, 283, 286, 2965., 306; used to illus-
12 5. trate principle of verifiability, 193; as
Mentalism (see Dualism), 104f. supreme universals, 269f.; and objec-
Metalogic, the logical theorist’s philosoj tive reality, 270; glorification of, 270;
phy, 321f.; related to logical practice, and mental states, 270; nature of,
322f.- 295f.; interdependent with things, 306.
Metalogical
systems, 76.
Metaphysics (see Existence problem), 10; Objective necessity, vs. logical_'necessity,
and the category problem, 112; vs. 15°) 154-
science, 1 73; linguistic or nee-nominal- Objective psychology, speculative systems
istic, 259. handled by, 99.
Method systems, 75f. Objectivity, in treatment of natural
Methodistic logic, 144, 146f., 324. events, 16of.; and reality, 308; and
Methodology of science, 145, 172, 182, measurement, 312f.
260, 267. Objects, of thought, vs. logical laws,
I 88; as perceived, vs. act of perceiving,
Metrology (see Measurement), and logic,
3185. 309.
Microscopic events, and the determinism- Observation, in causal situations, 165,
indeterminism issue, 173f.; require 174; as interbehavior, 308f.; relation
to construction, 316.
. uniquely descriptive languages, 319f.
Ontology (see Existence problem), in
Mind, sensations as atoms of, 12; as locus
of essences, 32; as a cognitive power, categorization, 129; and universals,
113; as category source, 113; and the
134f., 141; in dialectical logics, 144;
universals problem, 136; theories of, andlcause, 145f., 147, 1505.; and
logical laws, 183, 1885.; confused
influence causation doctrine, 159f.;
changing ideas of, 184; as a unifying connection of, with logic, 202‘; and
probability, 207, 2565.
construct, 258. Operational principles, vs. a priori
Models (see Logical instruments).
Multiplicative abstraction, 12f. norms, 175. .
Operational systems, 845.
Mysticism, in logic, 3; in mathematics, Operationism, and definition, 29, 53f.,
7; and causation, 148; in science, 311. 5 5, 57; and-classification, 29; in sys-
tem building, 845.; in causal theory,
Naming, definition by, 49, 53. I58f., 215f.; ini'mathematics, 197f.;
Natural events, objective vs: rationalistic in truth-frequency theory, 250.
 INDEX 353
Opposition, square of, as system product, tions, 2095., 2195., 2465., 2585.;
77f- Laplacian theory of, 21 15., 2345., 256,
Ordering systems, 302. 261; and plausibility, 212; many cri-
Organismic, instrument for social and po- teria for, 215; statistics in, 215, 221,
litical descriptiOn, 106. 227f., 236, 2405., 2605.; still bound
Ordinal scale, of measurement, 286f. by tradition, 216; origin of, in games
Ostensive definition, 56f. of chance, 2 16f., 234; response-centered
systems of, 217f.; event-centered sys-
Participation, doctrine of, and universals, tems of, 218f.; truth-frequency theory
134f. of, 220, 2485.; and necessity, 222f.;
Particulars, vs. universals, 132, 1335. and possibility, 224f.; potentiality and
Perception, subjected to abstracting and actuality in, 225; science inevitably
generalizing procedures, 9f.; causal occupied with, 2255.; and prediction,
theory of, 18; a priori basis of, 130; 23of.; interbehavioral design for analy-
as a creational process, 316f., consti- sis Of, 2325.; typical formulations of,
tutes interbehavior, 3 1 7. 2345.; as readiness to act, 24o; con-
Petitio principii, fallacy of, 324. vergence principle in, 240; as empirical
PhilosoPhy, definitional problems'In, 48; frequency, 2415., 244; as statistical fre-
‘categories of, 129; presuppositions of, quency, 2435.; as mathematical fre-
and probability theory, 206f., 257, quency, 245f.; as propositional rela-
259; specifications for an Objective tions, 2465.; intuition in, 247f., 258;
type of, 256; and logic, 321f., 328f.; and nonnumerical frequency, 2 52f.; as
presuppositions of, and system building, factorial coordination, 2535.; and log-
322, 332; abnegation motive in, 33of. ical multiplexity, 255f.; ontological
Physics, continuity of abstracted units in, problem in, 25 65.; as “rational recon-
12; instrument construction in, 102f.; struction” of scientific procedures, 260;
introduction of probability into, 227; systems for measuring, 3o4f.
metaphysics in, 314f. Probability calculus, 207, 210, 246, 2615.
Plausibility, and probability, 212. Propositions, and definitions, 33, 38, 41;
Polytomy, and dichotomy, 194. as constructional products, 143; and
Possibility, and probability, 224f. ’ deductive mathematics, 147; truth Of,
Postulation, defining by, 50; place of 191f., 198f.;,and probability, 2095.,
Constructor in, 805.; mathematics as, 2195., 2465.
81f.‘ Psychology, instrument construction in,
Potentiality, and probability, 225. 104f.; scale making in, 288.
Prediction (see Determinism), and the “Psychic magnitudes,” vs. physical mag—
problem of deduction, 86; and prob- nitudes, 288.
ability, 23of.
Prescription, definition by, 52f. Quality, as logical form, 139; measure-
Probability, unique place of, in logic, ment of, 2785.
204; involves system'building, 204f.; Quantity, measurement of, 2785., 281;
paradox of, 204; many situational as property of things, 28o; confusion
types of, 205; absolutism in theory of, concerning its referent, 295f.
206f.; calculation in, 2065., 210, 220, Quantum mechanics, 8, 12, 645., 96; and
245f., 2615.; phiIOSOphical presupposi- causation, 172, 305; and statistical
tions and problems in, 206f., 256f.; frequency, 243; and measurement, 274,
and certainty, 207f., 222f., 242, 244, 3075., 313f.; avoidance of metaphysi-
246; as equipossible ratio, 207, 217, cal questions in, theory of, 326.
2 345.; applies to all complex situations,
207f.; truth as a form of, 208, 223f.; Ratio scales, 2885.
as quantity of knowledge, 208, 2375.; Rationality, in causal theory, I59f., 172;
,of single events, 208, 210f., 242f., in probability theory, 220f., 2 34f.,
251f.,263; specificity of, 208f., 2125, 246_f. ‘

23.5, 255,263; as degree of belief, 208, Realism, 32, 1-32, 135f., 202.
209, 212, 218, 223f., 2385.; froni
interbehavioral standpoint, 208; fre-
Reality, sought in numbers, 270; and ob-
jectivity, 308; problem of, in quantum
∏ theory of, 209, 210, 213f,, mechanics, 3135.
218f., 221, 235, 2415.; and proposi- Reasoning, principles for regulating,
354- PSYCHOLOGY AND LOGIC

181; conventional vs. interbehavioral 202°

3 10 ical laws as overnin ) 202.
treatment of, 186f., 200; power of .Sentential logic, 196.
mathematical, 269. Series, in probability theory, 209, 211f.,
Reduction, in science, IIf., 117, 190, 240, 242, 263i.
324i Social sciences, instrument construction
Reductive abstraction, 1 1f. in, 106.
Reiteration, generalizing by, 16f., 24. Space and time, as abstractions, 11, 15;
Relations, as system materials, 4, 33, 72; continuity of, 116f.-; as mathematical
as evaluative categories, 124f.; as logi- categories, 1 16; categories in me-
cal forms, 138f.; in symbolic logic, chanics, 1 19; boisndary of science, 190.
144; causes as, 152; in interbehavioral Specificity, of system building, 60, 81, 87,
logic, 161; stressed by realistic lo- 325; principle of, recognized in logical
gicians, 18 3; transformed into sen- writings, 138.
tences, 2 59; mathematical, 2'83f. Statistics, and cause, 156; and the prin-
Relativity of logical systems, 125. ciple of identity, 201; and probability,
Relativity theory, and metrology, 299. 2435. 261f.; and measurement, 292f.,
Response-centered probability systems, 305; in science, 241, 326, 332; and
2175 absolute logic, 332f.
Rules, definition, remote from logical Subject-object, 317.
work) 57f- Subjective vs:I objective, in discussion of
values, 124f.°, in probability theory,
Scales, number and types of, 2855., 288. 2385., 251; in measuring systems, 315.
Schemata, as logical instruments, ch. 16. Suficient reason, principle of, 182, 185.
Scholasticism, and definition, 30; and Syllogism, the, as logical instrument, 100,
logical laws, 181f. 333; as system-building enterprise, 143,
Science, reduction in, 12, 117, 190, 323; as logical law, 184, 185.
324f.; as interbehavior, 14, 312, 314, Symbolic behavior, in generalizing, 185.
318;- and system building 615., 86, Symbolic logic, 25.; material implication
126, 1885., 324'; role of deduction in, in, 24; and syntactical and semantic
86; unification of, 112, 189; categori- definition, 3 3f, and descriptive opera-
zation in, 112, 117, 118f.; methodolo- tions, 41f., 5 3; truth function table of,
. gy of, 145, 182, 267; causation in, 78; categories of, 128; constructed of
149, 152f., 156, 162f., 172; and de- relations, 144; influence of, on metrol-
terminism, 1725., 312; and logic, 175,
177, 181f., 188f., 3245.; abstracting ogy, 299°
Symbols (see Language, Symbolic logic),
and extrapolation in, 1885.; formal
description vs. operational achievement confused with things, 33, 39, 46,174,
324; confused with acts, 3 5; as system
in, 189; universality or unity in, 189f.; materials, 97f.; structures of, as logical
specificity of, 189f.; hesitation to ac-- instruments, 100f.; as physical instru-
cept probability in, 225f.; measuring
ments, 103°, and categories, 1 15.
behavior in, 284, 295, 297; cultural
presuppositions of, 317; limits of, 326; Syntacticism, in definition, 33, and logical
and the existence problem, 327f. laws, 183. i

Scientific investigation, the field of causal Synthesizing, definition by, 50f. l

problems, 171 ; vs. mensurational prac- System builder, role of, in systemization,
tice, 2995. 6o, 81, 106.
Scientific method, 145, 172, 182, 260, System building, cultural basis of, 7,' 83,
267. 87, 90, 92, 106; as interbehavior, 59,
Scientist, the, seeks absolute knowledge, 60. 73. 93. 97. 106. Io7. 114,1field
225f.; as a measuring instrument, 303. hypothesis in, 60; vs. system product,
Semantic definition, 34. 6o,\8\7, 325f.; interrelation of Opera-
Sensations, events reduced to, 8; are ab- tions and materials in, 86f.; continuity
stractions, 12f.; in causal theory, 16of., of, 99f.; categories are instruments of,
164; are constructs, 311. 108f., 115f;;\theory of, and metrologi—
'Sentences (see Propositions), substituted cal science, 318f.-, specificity of, 32 5;
for cause, 154, 162; verifiability of, unrestricted, 99,‘ 330.
1905., material for system building, System products, vs. system building, 60,
196; 198f.; identified \with events, 81, 87, 32 5f ; investigative schema for,
 INDEX 355
6of.; types and characteristics of, 615., Universality in science, 1 89f.
885. Universals (see Logical forms), as formal
Systemology, integrated with metrology, elements, 94; similar to categories,
2995. 114f.; in interbehavioral logic, 131,
1375.; ambiguity of, in conventional
Tautology, in logical laws, 179, 184, logic, 131; products of system build-
188; in sentence systems, 190, 192; ing, 131; historical _ treatment of,
formal logic concerned with, 195.
Terms (see Language, Symbols), uni-
1315.; vs. particulars, 132, 1335.;
abstracting and generalizing products,
as
versals, as, 136f. 132, 138; current treatment of, 1335.;
Thermodynamics, categories in, 125f. spatiotemporality, as criterion of,
Things, confused with symbols, 39; as I33f.; character value of, 134; as
system materials, 61; universals as, words or terms, 135f.; and concepts,
136f.; as sources of logical forms, 135f.; involved with language, 136;
1395.; causes as, 15of.; existence of, as things, terms, or thoughts, 136f.;
vs. knowing, 173f., 324; not created interbehavioral vs. mentalistic-psycho-
by measurement, 174, 308f.; Aristotle’s logical description of, 137f.; are par-
objective organization of, 180; lan- ticulars, 138f.; numbers, as supreme,
guage substitutes for, 194; specificity 269£
of measured, 272f. Universe, the, an illicit construction, 24.
Thinking, as a generalized power, 1865.; Universe of discourse, as logical instru-
consistency of, 187; as interbehavior, ment, 100.
zoo.
Thoughts, logical forms as, 136f.; laws Vagueness, and definition, 465.
of, ch. 20. Value, category of, 124f.
Truth, as a form of probability, 208; vs. Verbal assertion, vs. operational achieve-
existence, 223. ment, 189.
Truth—function (matrix) systems, 78f. Verbal symbols, vs. actual things, 181.
Truth value, of sentences, 1915., 198f.; Verbalism (see Language, Symbols,
interpretation of, 198; determined on Terms), in definition, 3 1 f.
basis of specific conditions, 199. Verification (see Truth), vs. verifiability,
192f.; and truth, 198.
Universalism, in logic (system building), Vitalism, 311f.
175,378, 179, 181f., 1835., 187, 192,
I99. 255f-
 NAME INDEX
Adams, 62, 310n. Carnap, 211, 260.
Aiken, 112n. Carnot, 92', 126.
Ampere, 277. Carus, 1 5211.
Ancillon, 259. Cassirer, 112n.
Angstrom, 63. Cavalieri, 12o.
Aristotle, 14, 30, 57n., Ing., 132, 148, Cavendish, 303.
'
1795.,
193. _ Child, 114n.
Avey, 184n. Church, 37n., 198.
Churchman, ‘26on., 28In.
Bain, 176n. Clausius, 189.
Ballard,‘318n., 319n. Cohen; 25n., 46n., Son” 86nq 138)
Balmer, 625., 655. 145n., 176n., 178n., 134D” 188n4
Bartlett, 288m. 229D” 321B.
Bayes, 217. Compton, 689
Beck, 164n. Comrey, 281n.
Becker, 198. Condorcet, 217. a
Becquerel, 43. Costello, 47.
Bell, 1a., 3a., 8In., 93n., 197n. Coulomb, 64.
Benjamin, 279n., 280n. Cournot, 218, 241.
Bentley, 28n., 42n., 53, 310n. Couturat, 3711.
Berenda, 315n. Crookshank, 124n.
Bergmann, 260a. Czuber, 2 59. ∙
Berkeley, 174, 316.
Bernoulli, Jacob, 217, 236, 237. Dadourian, I In.
Bernoulli, Johann, 217, 237. Daggett, 84n.
Bertrand, 204. D’Alembert, 217, 260, 291.
Birkho5, 22n., 304n. Dantzig, 93n.
Black, 198n. Darwin, 225a.
Bode, 61f. Davis, 49B.
Bogoslovsky, 178n. De Finetti, 240.
Bohr, 15, 645., 305n., 307, 311, 315n. De Méré, 234.
Boltzmann, 316. De Moivre, 217.
Boole, 178n., 182, 187, 259. De Morgan, 219, 238.
Born, 93n. Dennes, 112n.
Bosanquet, 144. Descartes, 7 5f., 1 I 7.
Brackett, 65. Dewey, 28n., 42n., 53; 56n., 94, 1.141,
Bradley, 24, 144.
Brahe, 226.
143n., 146n., 147, 184n., 193.
Diels, 148n. ∙
'
Braun, 326. Dingle, 8n., 24n., 96. l

Brewster, 120. Donkin, 238.

Bridgman, 2n., 8n., 24n., 53f., 189. l
Drebbel, 276.
Broad, 229n. Drew, 37n.,
Brouwer, 179, 197f. Dubislav, 28n.
Burks, 37n., 47n., 4811. Dubs, \42n.
Butler, 206. Ducass€,\151n., 166n., 253D. '
Butterbaugh, 26on.
Durkheim, 113.
Campbell, L., 3o3n.
'Campbell, N. R., 50n., 123, 229n., 268, Eaton, 25n., 58in, 145n.
271, 284n., 287,295, 29711. Eddington, 8n.
Cardan, 216, 237. Edgeworth, 242n.
Carmichael, 8011. Ehrenfest, 16.
 INDEX 357
Einstein, 64, 269, 271n., 305n., 307n., ‘Infeld, 305n., 307n., 31 In.
3o8n., 311, 314, 316.
Ellis, 218, 241. James, 24.
Enriques, 5 5n., 182n. Jasche, 175n., 194n.
Endid) 5f°1 49) 55) 56) 81° Je5reys, 229n., 243n., 248, .267n., 306.
Eudoxus, 81. Jevons, 176n., 178n., 195, 222, 238f.,
Euler, 217, 257. 246, 257, 291.
Eve, 6811. Johnson, H. M., 303n.
Johnson, W. E., 44f., 5o, 58n., 135.
Fairbanks, 1221‘1. Jones, 46.
Faraday, 291. Jordan, 305n., 308.
Farwell,‘ 37n. Joseph, 59n., 17on., I83n.
Feigl, I64n., 173n., 259n.
Ferguson, 288n. Kaiser, 315n.
Fermat, 217, 234, 237. Kant, 8, 46n., I46, 150, 151, 159, 175,
Fisher, 293a. I94. 310.
Fourier, 123. Kantor, H. J., 7011.
Frank, 313n., 314. Kantor J. R., 4n., 19n., 25n., 74n.,
Frege, 2. 105n., 12911., 165n., 186n., 228m,
Galileo, 182, 217, 234, 276. 317n.
Galle, 308. Kaplan, 4411.
Kelvin (see Thomson).
Garnett, 303n.
\

Kemble, 211f.
Gauss 53. 277. 293. 294.
Kepler, 217, 226, 234.
George, 165n., 29on.
_Gibbs, 189. Keynes, J. M., 22013., 236f., 243n.,
Ginsburg, 18n. 2465., 258, 259.
Goethe, 135, 258. Keynes, J. N., 176n., 185.
Goldenweiser, 118. Klein, 101.
Gotshalk, 152n. Kneale, 146n., 23on.
Grassmann, 120, 197. von Kries, 236.
Greenwood, 206n.
Grimsehl, 52n., 276n., 314n., 326n. Lagrange, 217.
Lamont, 33m.
Haas, 66.‘ Langley, 277.
Haldane, 75n. Laplace, 172, 2165., 235, 237.
Hamilton, 185, 197. Larmor, 123.
Harding, 124n. Layzer, 310n., 313n., 314n.
Hardy, 21, 269. Le Chatelier, 326.
von Hartmann, 11on., IIIf. Leeuwenhoeck, 308.
Hawkins, 211. Leibniz, 182, 195, 217, 237, 246.
Heath, 6, 56. Lenzen, 8n., 174n., 296, 313n., 32711.
Hedrick, 179n., 198n. Leucippus 147.
Hegel, 76, 111, 144, 178. Leverrier, 62.
Heisenberg, 227, 299. Lévy-Bruhl, 113.
Helmholtz, 267, 295, 296. Lewis, 58n., 192m, 193n., 199.
Hempel, 98n., 156n., 26on. Lillie, 104n.
Herodotus, 298. Lindsay, 116n.
Herschel 61, 238, 277, 308. Linnaeus, 104.
Hilbert, 81. Locke, 31, 224.
Hobbes, 151. Loeb, 310n.
Ho5man, 10m. Lotka, 24n.
Hofstadter, I54, 155n. Lotze, 41.
Hopkins, 165. 'Lyman, .645.
Horst; -290n.
Houston 8n. ‘McGill, .91‘.
Hume, 8, 19, 15of., 153, 156, 159f., McGilvary, 134, 138.
I73. 224. 23I. 316. 33I- McGregor, 288n.
Huygens 117, 217, 237. Mach, 4f., 6, 8, 12, 316.
358 PSYCHOLOGY AND LOGIC

M'ansel, 177a. Podolsky, 314.

Margenau, 116n., 305n., 310, 313n. Poincaré, 5, 6, 8,6411, 204n., 324.
Maxwell, 92, 227, 241, 301,303. Poisson, 237.
Meldrum, 84n. Polya, 212.
Mendelee5 61, 68. Popper, 213a.
Menzel, 310n., 313n., 314n. Porphyry, 89.
Mercier, I52n. Pythagoras, 197.
Merz, 241n.
Meyer, 68. Ramsey, 133, 240.
~Michelson, 294.. Rankine, 96.
Mill, 13, 4o, 42, 56n., 75n., 82f., 146, Recorde, 20.
151, 159f., 169f., 177n., 185,316. Reese, 288n.
von Mises, 213f., 235n., 243n., 245, __Reichenbach, 193n., 210, 224n., 317n.,
260. '
31911., 325n., 328n.
Mitchell, 2on. Reid) 31) 421 45-
Mohs, 287. Rickert, 32.
Montmort, 218. Riemann, 120.
Moore, G. E., 45n. Ritchie, 289n.
Moore, J. 8., 133a. Ritter, 277, 308.
Morgan, 310. Ritz, 66. 6
Mosley, 69. Robinson, 28n. ‘
Murray, 45n. Robson, 176n. a
Rosen, 314.
Nagel, 9a., 46n., 48, 80n., 145n., 176n., Rosenblueth, 15n., 96, 1041!.
178n., 184n., 188n., 202n., 210f., Ross, 7511.
235n-. 249. 254m. 259m. 279m. 325". Royce, 5o.
Nagaoka, 64n. Ruark, 316.
Needham, 103n. Runes, 37n., 206n.
Nelson, 4911. Runge, 29311.
Nernst, 326. Russell, 1n., 3n., 18n., 20f., 25, 33f., 4.5,
von Neumann, 313n. 53, 90n., I31n., 135, 136, 162, 164,
Newlands, 68. 171n., 193n., 197, 198, 210, 229n.,
Newton, 76, 120, 123, I53. 268, 278n., 295f., 328n. 329n., 330i.
Neyman, 26on. Rutherford, 15, 64f., 68.
Nicod, 3n. Rydberg, 64, 66f.
Nisbet, 248n. Ryle, 212.
Nobili, 277.
Scates, 288n.
Oersted, 277. Schiller, I85.
01]III. 277. 303- Schlick, 153n.
Oliver, 15511. Schopenhauer, 76,152n.
Oppenheim, 98n., 156n., 26on.
Ostwald, 12.
SchrOdinger, 325,
Sears, 293n.
326.
Sellars, 137.
Pascal, 217, 234, 237. Semat, 37n.
Pap, 150n. Sheppard, 243, 244m.
Paschen, 645. Shewhart, 26on.
Pearson, E., 26on. . Socrates, 20. .
Pearson, K., 245, 246, 310. Sommerfeld, I6n., 66n., 68. ‘
Peirce, 219f., 248f. Spaier, 268, 279n.
Pepper, 41f., 53, 112, 162. Spencer,113, 287n.
Perrin, 64n. Spinney, 37n.
Petrie, 70f. Spinoza, 182.
Philo, 121. Stebbing, 3n.
Pieri, 81. Stevens, 281n., 285n., 2865.
Planck, 64f., 292. Stranathan, 303n.
∙
∙
Plato, 20, 30, 132. Strong, 5f., 56.
Plotinus, 121, 181. Struik, 253. ‘