ADVANCES IN THE STUDY OF BEHAVIOR V 12

Visit the link below to download the full version of this book:

https://medipdf.com/product/advances-in-the-study-of-behavior-v-12/

Click Download Now

 Contents

List of Contributors ........................................................ vii

Preface ................................................................... ix
Erratum .................................................................. xi

Pavlovian Conditioning of Signal-Centered Action Patterns

and Autonomic Behavior: A Biological Analysis of Function
KAREN L . HOLLIS

I. Introduction ......................................... 1
I1. The Prefiguring Hypothesis ........................... 3
I11. Pavlovian Conditioning and Foraging Behavior .......... 6
IV . Pavlovian Conditioning of Defensive Behavior .......... 28
V. Pavlovian Conditioning of Reproductive Behavior ....... 43
VI . Ecological Implications of Prefiguring .................. 48
References .......................................... 53

Selective Costs and Benefits in the Evolution of Learning

TIMOTHY D. JOHNSTON

I . Introduction ......................................... 65
I1 . Learning and Evolution-Historical Background and
Current Concerns .................................... 66
I11. An Ecological Conception of Learning ................. 69
IV . Cost-Benefit Analysis and the Evolution of Adaptations . . 70
V . The Selective Benefits of Learning ..................... 73
VI . The Selective Costs of Learning ....................... 79
VII . Learning and the Adaptive Complex .................... 92
VIII . Implications for the Study of Learning ................. 96
References .......................................... 98

V
vi CONTENTS

Visceral- Somatic Integration in Behavior. Cognition.

and “Psychosomatic” Disease
BARRY R. KOMISARUK

I . Introduction ......................................... 107

I1. Visceral-Somatic Relationships ....................... 109
I11. Higher Order Integration of Visceral and
Somatic Activity ..................................... 112
IV . Visceral Activity and Ideational Imagery ............... 123
V . Toward a Concept of Psychogenic Organic
(“Psychosomatic”) Disease ........................... 129
VI . Conclusion .......................................... 134
References .......................................... 135

Language in the Great Apes: A Critical Review

CAROLYN A . RISTAU AND DONALD ROBBINS

I. Introduction ......................................... 142

I1. Brief History of the Ape Language Projects ............. 143
111. Theoretical Issues .................................... 145
IV . The Signing Apes .................................... 155
V. Artificial Lexicons ................................... 178
VI . Investigations into Meaning ........................... 188
VII . Investigations into Mental States ....................... 207
VIII . Relation to Animal Cognition and Natural Animal
Communication ...................................... 219
IX . Implications for Human Language and Cognitive
Development ........................................ 230
X. Problems Raised by the Language Research and
Suggestions for Future Research ....................... 237
XI . Concluding Statements ............................... 245
References .......................................... 247

Index ..................................................................... 257

Contents of Previous Volumes ............................................... 261
 List of Contributors
Numbers in parentheses indicate the pages on which the authors’ contributions begin.

KAREN L. HOLLIS,* Animal Behaviour Research Group, Department

of Zoology, University of Oxford, Oxford, England ( 1 )

TIMOTHY D. JOHNSTON, Research Branch, North Carolina Division

of Mental Health, Raleigh, North Carolina 2761 1 (65)

BARRY R. KOMISARUK, Institute of Animal Behavior, Rutgers Uni-

versity, Newark, New Jersey 07102 (107)

CAROLYN A. RISTAU, The Rockefeller University, New York, New

York 10021 (141)

DONALD ROBBINS, Division of Social Sciences, Fordham University

at Lincoln Center, New York, New York 10023 (141)

*Present address: Department of Psychology, University of Toronto, Toronto, Ontario,

Canada MSS 1Al.

vii
This Page Intentionally Left Blank
 Preface

With the publication of the twelfth volume of Advances in the Study of

Behavior, we wish to restate in more contemporary terms the aims stated
in the original preface, namely, to serve “. . . as a contribution to the
development of cooperation and communication among scientists in our
field.” Since that preface was written in 1%5, an increasing number of
scientists from disciplines as widely separated as behavioral ecology and
the biochemistry of behavior have become engaged in the study of animal
behavior, employing the specialized techniques and concepts of their dis-
ciplines. Even then, the boundaries of ethology and comparative psychol-
ogy were no longer distinct: now they have been merged with broader
syntheses of social and individual functioning and have together provided
the bases for studies of the neural and biochemical mechanisms of behav-
ior. New vigor has been given to traditional fields of animal behavior by
their coalescence with closely related fields and by the closer relationship
that now exists between those studying animal and human subjects. Sci-
entists engaged in studying animal behavior now range from ecologists
through evolutionary biologists, geneticists, endocrinologists, etholo-
gists, and comparative and developmental psychologists, to neurophysi-
ologists and neuropharmacologists. The task of developing cooperation
and communication among scientists whose skills and concepts necessar-
ily differ in accordance with the diversity of the phenomena they study
has become more difficult than it was at the inception of this publication.
Yet the need to do so has become even greater as it has become more
difficult. The Editors and publisher ofAdvances in the Study ofBehavior
will continue to provide the means by publishing critical reviews of re-
search in our field, by inviting extended presentations of signifcant re-
search programs, by encouraging the writing of theoretical syntheses and
reformulations of persistent problems, and by highlighting especially pen-
etrating research that introduces important new concepts.

ix
This Page Intentionally Left Blank
 Erratum
Advances in the Study of Behavior
Volume 11
Page 34 Figure 28 should appear as follows:
20
----- spermotoqeness
- MAS

- promo testosieron
- - - - - -- - - - - - wl?nwlOmO/
devzkwd
cpdIdymlS
I5 - - rpcrmolldr/
wrmomoo

- IpmurldS

- P'mory
s9ermotocy~

- wmwrlwpn~
\
'.- fully rqmrad

FIG.28. Hormonal control of male reproductive behavior in Anolis carolincnsis. Top:

Relationship between plasma testosterone levels and behavioral and physiological events
in the annual reproductive cycle of the male lizard Anolis carolinensis. Modified
from Crews (1975). The number of animals assayed in each monthly sample are as
follows: N = 6; D = 10; J = 14; F = 10; A = 10; M = 17; J = 8; S = 12. Tokarz
and Crews (unpublished). Bottom: Effect of castration and androgen (testosterone) re-
placement therapy on the sexual and aggressive behavior of the male lizard, Anolis
corolinrnsis. In these experiments. sexually active males were castrated after baseline
levels of sexual and aggressive behavior were determined. Two weeks following castra-
tion. males were given Silastic implants (0.06 cm i.d. x 0.12 cm 0.d.) containing testos-
terone subcutaneously. Males were tested daily first with a male intruder and then with a
female intruder for 15 min each. Mean and SEM are shown; n = 12. From Crews (1979a)
with permission of the Society for the Study of Reproduction.

xi
This Page Intentionally Left Blank
 Advances in
THE STUDY OF BEHAVIOR
VOLUME 12
This Page Intentionally Left Blank
 ADVANCES IN THE STUDY OF BEHAVIOR. VOL I2

Pavlovian Conditioning of Signal-Centered Action

Patterns and Autonomic Behavior:
A Biological Analysis of Function

KARENL. HOLLIS*
ANIMAL BEHAVIOUR RESEARCH GROUP
DEPARTMENT OF ZOOLOGY
UNIVERSITY OF OXFORD
OXFORD, ENGLAND

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
11. The Prefiguring Hypothesis . .
111. Pavlovian Conditioning and F
A. Locomotory Search and A
B. Consummatory and Food-Procuring Behavior ...................... 8
C. Pavlovian Processes in the Development of Food Recognition
D. Conditioning of Digestive
E. Pavlovian Processes in the Rejection-r Ingestion-f Toxins . . . . . . . . 19
IV. Pavlovian Conditioning of Defensive Behavior . . . . . . . . . . . . . . . . . . . . . . . . 28
A. Interspecific (Antipredator) Defense .............................. 28
B. Interspecific Defense and the Backw
C. Intraspecific Defense . . . . . . . . . . . . . . . . . .
V. Pavlovian Conditioning of Reproductive ...................... 43
A. Courtship ...................... ...... 44
B. Parental Behavior ....................................... 46
VI. Ecological Implication figuring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
A. Prefiguring and Non-Pavlovian Learning 49
B. Naturally Occurring Conditional Stimuli as Learned Releasers . . . . . . . . . 50
C. Concluding Comments . . . ...................... 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................. 53

I. INTRODUCTION

Consider the difficulties an animal faces if it must actively search for food,
water, or for a mate. Where and when can food be found and mates secured?
Likewise, where and when will predators most likely strike and rivals appear?
*Present address: Department of Psychology, University of Toronto, Toronto, Ontario, Canada,
M5S 1Al.

I
Copyright 0 1982 by Academic FYess. Inc.
All rights of reproduction in any form reserved.
ISBN 0-12-004512-5
2 KAREN L . HOLLIS

For many animals the strategic difficulties of searching-and remaining vigilant

to the search of others-would be insuperable obstacles to survival if the envi-
ronment were a capricious one. But given that an animal could predict where and
when food would be available, such information would provide considerable
savings in foraging time and energy reserves. And, if an animal could locate
predators based upon, say, some sort of warning signal, the danger of a surprise
attack would be negated altogether. Pavlovian conditioning is a mechanism for
such predictions.
But Pavlovian conditioning appears to involve more than mere “predictabil-
ity” or “information transfer,” as might be expected. The result of such condi-
tioning is the eventual elicitation of a response, the conditional response or CR,
which actually precedes the occurrence of the biologically important event and
which, if it is a skeletal behavior, is often directed at the conditionul stimulus
(CS).The CR would not only appear to be “jumping the gun” but would appear
to be somewhat misdirected. That is, in a typical Pavlovian conditioning experi-
ment (Mackintosh, 1974) the experimenter presents two stimuli to the animal, a
CS followed by an unconditional stimulus (US). The US is some biologically
relevant event, e.g., food, water, or noxious stimulation. The US was so-named
by Pavlov ( 1927) because it unconditionally elicits a stimulus-specific response,
the unconditional response (UR). Pecking is a UR to food in pigeons, for exam-
ple. The CS is sometimes described as a “neutral” stimulus because it does not
elicit a conditional response prior to training; that is, the CR is conditional upon
the CS-US pairing procedure. The animal need not perform some response to
obtain the US, as it must in an instrumental (Skinnerian) conditioning paradigm.
(These two paradigms, as well as experimental justification for distinguishing
between them, are discussed in a later section.) The procedural hallmark of
Pavlovian conditioning is that the CS and US are presented to the animal inde-
pendently of its behavior. Nonetheless, after sufficient training of this nature, the
CS elicits a conditional response which precedes US presentation.
This anticipatory conditional response is, more often than not, energetically
costly. For example, a territorial male Betra splendens will forcefully attack a
stimulus which in the past has preceded the appearance of a rival (Thompson and
Sturm, 1965). Similarly, if the illumination of a key light is reliably followed by
food, a pigeon will peck quite vigorously at the key, regardless of the fact that no
such response is required in this situation (Brown and Jenkins, 1968). Why does
conditioning result in an anticipatory response? What advantage does an animal
gain by responding to cues which reliably accompany US events when these USs
are, themselves, already effective in eliciting the appropriate response? One
might ask why the classically conditioned pigeon does not simply go to the food
magazine and wait, or why the classically conditioned Betta does not conserve
its energy for the real aggressive encounter. That is, while no one would question
whether the predictive function of Pavlovian conditioning is advantageous, why
 BIOLOGICAL FUNCTION OF PAVLOVIAN CONDITIONING 3

does the operation of this mechanism also result in a response whose true
referent has not yet appeared? In this article I will explore the biological function
of the anticipatory conditional response.
Psychologists’ views of learning phenomena, including Pavlovian condition-
ing, have changed radically in recent years. Among other things, learning is no
longer believed to provide an endless source of plasticity (Shettleworth, 1972;
Hinde and Stevenson-Hinde, 1973). Rather, an ability to learn merely permits
the animal to extend its utilization of the species-specific response repertoire to
new situations. From a biological standpoint, this ability is, in itself, the product
of its species evolutionary history. Ultimately, then, the explanation of all learn-
ing phenomena must be made consistent with ecological and ethological consid-
erations. An analysis of the function of Pavlovian conditional responses is but
one attempt to do so.

11. THEPREFIGURING
HYPOTHESIS

I suggest that the biological function of classically conditioned responding,

which I will call prefiguring, is to enable the animal to optimize interaction with
the forthcoming biologically important event (US). The performance of a CR,
although energetically costly, allows the animal better to deal with the US event,
and, as such, the CR is essentially preparatory.’
The animal is, of course, already physiologically equipped to respond to
biologically important events in the absence of their having been signaled.
Nonetheless, the prefiguring hypothesis maintains that anticipatory responses to
stimulus events (CSs), which in the past have reliably accompanied those USs,
provide a selective advantage over and above the ability to respond to the US
alone. In this sense the CR is nor an accidental “false start” of an otherwise
adequate conditioning process. The anticipatory conditional response is, itself,
the evolutionary ruison d’ttre of Pavlovian conditioning. Of course, the per-
formance of a CR is how psychologists have always measured Pavlovian condi-
tioning. Nonetheless, when the question of function has arisen, greater impor-
tance has been attached to the predictive power of the CS-US relationship (Mac-
kintosh, 1979) where “prediction” appears to take on the ordinary dictionary
meaning of “knowing in advance.” However, to predict that it will rain tomor-
row afternoon is one matter; to take my umbrella tomorrow morning is quite
another.
’The use of the word preparatory here is purely descriptive and refers to the funcrion of the CR, not
the mechanism whereby the CR is produced. That is, I do not intend that the term preparatory be
theoretically synonymous with its use in some animal learning contexts, e.g., the “preparatory-
response” hypothesis (Perkins, 1968; see also Prokasy, 1965). There the term is used to refer to a
non-Pavlovian mechanism of learning. This issue is further discussed in Section IV,A,2.
4 KAREN L . HOLLIS

It would seem that a conditional response to experimenter-selected CS events,

like illuminated discs and ticking metronomes, would offer little in the way of a
selective advantage, and, indeed, in some cases would seem to be maladaptive.
In the “long box,” for example (Hearst and Jenkins, 1974), pigeons could be trained
in such a way that they invariably missed a brief food presentation because they had
been pecking at a CS some distance away (although, of course, no such response
was required). However, the fact that we can produce “maladaptive” or “ineffi-
cient” behavior in the laboratory by distorting the temporal and/or spatial rela-
tionships between stimuli is irrelevant to the functional argument presented here.
Naturally occurring signals, with which evolution has dealt, would hardly be so
arbitrarily related to US events. The bird which approaches and pecks an illumi-
nated spot on the chamber wall because in the past this CS has been paired with
food presentation, is also the bird which, in the wild, is likely to approach and
peck certain small “spots” on the forest floor previously associated with food.
Here, however, these spots may be the tell-tale cues that buried insects un-
avoidably leave behind.
If a functional approach to Pavlovian conditioning, a phenomenon today al-
most exclusively concerned with causal mechanisms, somewhat muddles the
boundaries between biology and psychology, the idea is not wholly original.
Pavlov himself was a functionalist interested in “phenomena of adaptation ”

(1928, p. 83). Pavlov’s writings and those of his students clearly reflect those
functional considerations. However, E. A. Culler perhaps said it best:

[The] concept of a self-regulating mechanism has been amply documented by Cannon

(1932). Constancy of water content, of salt content, . . . maintenance of body temperature
are but special forms of a pervasive ‘homeostasis.’ Admirable as these autonomic stabiliz-
ers are, they do not approach in range and flexibility the adjustive mechanisms which
nature has provided in [Pavlovian]conditioning. . . . [If a] UR were his only recourse, the
animal would still be forced to wait in every case for the stimulus to arrive before
beginning to meet it. The veil of the future would hang just before his eyes. Nature began
long ago to push back the veil. Foresight proved to possess high survival value, and
conditioning is the means by which foresight was achieved.. . . The CR, in brief, is
nature’s way of getting ready for an important stimulus. The salivaty secretion prepares
the mouth for reception of food and gastric secretion for its proper ingestion. . . . Con-
ditioned lid-closure protects the cornea from a blow or jet of air. Conditioned iridic reflex
saves the retina from undue stimulation. Conditioned galvanic skin-response . , . prepares
for seizing and manipulating the stimulus, and so on (Culler, 1938. pp. 134-136).

And so on, indeed. In the remainder of this article I will attempt to show,
specifically, how “nature’s way of getting ready” is accomplished by Pavlovian
conditioning of the animal’s repertoire of adaptive responses. This applies not
only to the conditioning of autonomic responses, the domain of Pavlovian condi-
tioning known to an earlier era, but to the more recently discovered conditioning
of complex skeletal behaviors (autoshaping or signtracking) as well. Moreover,
 BIOLOGICAL FUNCTION OF PAVLOVIAN CONDITIONING 5

for any given US, I will stress the notion that the conditional response is not one
but actually a battery of responses, both autonomic and skeletal, some similar to
the UR and some opposite in direction, and together as one unit they function to
insure optimization of biologically relevant events.
A functional analysis of behavior involves determining the ultimate evolution-
ary advantage of indulging in a particular behavior which, in theory, would be
measured in terms of an animal’s inclusive fitness (Hinde, 1975). Thus, the
prefiguring hypothesis suggests that, by responding to CSs in anticipation of the
US event, an animal increases its own reproductive potential or that of its close
kin. Fitness is often difficult to measure, however, because many behaviors are
far removed in time from actual reproductive gains. Under such circumstances a
currency is chosen which, it is assumed, will ultimately be translated into repro-
ductive success. The behavior in question is then measured in terms of its net
contribution to this currency (McCleery, 1978). For example, a behavior which
enables an animal to obtain more food is recognized as adaptive because feeding
efficiency would eventually be translated into a reproductive advantage. Feeding
efficiency and other such currencies will be employed in this article as a means
with which to assess the contribution of prefiguring to fitness.
Nonetheless, in his treatise on adaptation, G. C. Williams (1966) warns us
against “unwarranted uses of the concept of adaptation” (p. 11). A benefit may
be the result of chance, not design, and we must be careful to distinguish
adaptations from mere fortuitous effects (see also Lewontin, 1979; Maynard
Smith, 1978). Thus, although one may be able to demonstrate successfully a
benefit which results from the performance of a CR, this in itself is insufficient.
One must also demonstrate that conditional responses are not merely accidental
by-products of the conditioning process. That is, an animal’s nervous system
might be such that it cannot help but anticipate the US, even though anticipation
is not the function of Pavlovian conditioning. This simplistic account of the
CR-that it is an accidental false start-seems unlikely: We will see that the CR
does not always involve the same response system as the UR (Wasserman, 1973;
Jenkins et al., 1978) and even where it does, it is not always similar to the UR
(Bykov, 1959). Indeed, in some cases the CR is in the opposite direction of the
UR (S. Siegel, 1979a). Moreover, the form of the CR changes over the course of
conditioning (Wasserman, 1973) and varies with both the intensity (Gray, 1965)
and type of the CS (Holland, 1977), and even the timing of CS and US events
(Holland, 1980; Rescorla, 1980b). Most importantly, however, the ability of the
prefiguring hypothesis to predict successfully such perturbations of the anticipa-
tory conditional response suggests that a functional analysis is correctly centered
on the CR.
Finally, functional explanations do not compete with causal explanations-
they involve orthogonal questions whose answers complement and support one
another ( N . Tinbergen, 1963; see Shettleworth, in press, for a discussion of this
6 KAREN L. HOLLIS

topic with regard to learning). A functional argument cannot explain how Pavlo-
vian conditioning works. On the other hand, as the remainder of this article will
reveal, the prefiguring hypothesis is able to make novel predictions about con-
ditioned behavior and to provide a systematic account of the many different
“response rules” governing the form, the appearance, and the timing of the
Pavlovian conditional response.

111. PAVLOVIAN
CONDITIONING
A N D FORAGING
BEHAVIOR

Obtaining food and water is a long and complicated process involving a

sequence of many different behavior patterns, both skeletal and autonomic. The
process begins with physiological changes responsible for hunger or thirst. These
are followed by locomotory sequences of behavior, sometimes called appetitive
behavior (Craig, 1918), which bring the animal in contact with food and water
sources. Later, when food or water is encountered, consummatory behaviors are
evoked. Concurrent with these activities, and continuing long past them,
physiological changes involving ingestion are taking place. Some of the more
familiar of these physiological responses include salivation (in some animals),
secretion of gastric and pancreatic fluids, and increased gastrointestinal motility.
Pavlovian signaling operations have been demonstrated in all of these various
behavioral contexts. Pavlov himself was primarily interested in digestion. But
not only was the prototypical Pavlovian conditioning experiment concerned with
ingestion; much of the autoshaping literature today is an investigation of con-
ditioned food-related behaviors. Although this section is not intended to provide
an exhaustive discussion of the literature, it will attempt to show through selected
animal learning and animal behavior experiments how an animal might incorpo-
rate a signaling operation in the various aspects of its search for food, including
locomotory search and approach behavior, consummatory and food-procuring
behavior, and autonomic behavior related to food ingestion. Finally, the impor-
tance of Pavlovian conditioning in the development of food selection and in the
avoidance-or consumption-of poisons will be discussed.

A. LOCOMOTORY
SEARCHA N D APPROACH
BEHAVIOR
Because Pavlovian conditioning has been studied within the confines of labora-
tory apparatus, locomotory search behaviors are necessarily restricted. Nonethe-
less, a large body of research (for reviews, see Hearst and Jenkins, 1974;
Schwartz and Gamzu, 1977) suggests that specific orienting, approaching, and
searching behaviors (which I will shorten to “approach”) are elicited by CSs
paired with food and water. In pigeons, for example, visually localizable signals
like the illumination of a key light which is followed by food presentation, evoke