Contrasting Languages

Trends in Linguistics
Studies and Monographs 51

Editor
Werner Winter

Mouton de Gruyter
Berlin · New York
Contrasting Languages
The Scope of Contrastive Linguistics

by
Tomasz P. Krzeszowski

Mouton de Gruyter
Berlin · New York 1990
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter & Co., Berlin.

© Printed on acid-free paper which falls within the guidelines of the ANSI to ensure
permanence and durability.

Library of Congress Cataloging in Publication Data

Contrasting languages the scope of contrastive linguistics /

edited by Tomasz P. Krzeszowski.
p. cm. (Trends in linguistics. Studies and mono-
graphs 51)
Includes bibliographical references and index.
ISBN 0-89925-590-6 (cloth acid-free paper)
1. Contrastive linguistics. I. Krzeszowski, Tomasz P.
II. Series.
P134.C58 1990
410 —dc20 90-45502
CIP

Deutsche Bibliothek Cataloging in Publication Data

Krzeszowski, Tomasz P.:

Contrasting languages the scope of contrastive linguistics /
by Tomasz Krzeszowski. Berlin New York Mouton de
Gruyter, 1991
(Trends in linguistics Studies and monographs ; 51)
ISBN 3-11-012133-6
NE: Trends in linguistics / Studies and monographs

© Copyright 1990 by Walter de Gruyter & Co. D-1000 Berlin 30

All rights reserved, including those of translation into foreign languages. No part of this
book may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopy, recording, or any information storage and retrieval
system, without permission in writing from the publisher.
Typesetting: Arthur Collignon GmbH, Berlin — Printing: Gerike GmbH, Berlin — Binding:
Lüderitz & Bauer, Berlin — Printed in Germany
Acknowledgements

I am indebted to all my friends and colleagues who took pains to read

the draft of this book in its entirety or in parts and suggested improve-
ments. The complete list would have to include many students of mine,
the first patient experiencers of the ideas contained in this volume. Facing
the impossible task of mentioning all the numerous persons whose com-
ments influenced my thinking, let me single out the following scholars:
Prof. Dr. Edmund Gussmann, Prof. Dr. Ruta Nagucka, Doc. Dr. Elzbieta
Muskat-Tabakowska, Dr. Tadeusz Danilewicz, Doc. Dr. Barbara Kryk,
Dr. Wojciech Kubinski, Dr. Krzysztof Kwasniewicz, Ms. Janina Ozga,
and Dr. Ewa Willim.
The book owes its existence to the tradition of contrastive studies in
Poland, inspired, promoted, and supported by their spiritus movens Prof.
Dr. Jacek Fisiak, to whom I hereby express my special thanks.
I am also very grateful to Marc Weinstein, who carefully read my
manuscript and suggested numerous stylistic improvements.
I also owe a debt of gratitude to Ms. Olga Sokolowska for her
assistance in compiling the indexes.
If the book suffers from inaccuracies, misinterpretations, and other
flaws, it is partly because its author did not pay sufficient heed to the
illuminating comments of its critics. Assuming total responsibility, the
author apologizes to both: the critics, for not listening carefully enough,
and to the readers, for obliging them to endure the remaining inadequa-
cies.
Gdahsk, March 1988 Tomasz P. Krzeszowski
Contents

Introduction 1

Chapter I

What is contrastive linguistics? 9

Chapter II

Tertium comparationis 15

Chapter III

Towards a classification of contrastive studies 23

Chapter IV

Three steps in "classical" contrastive studies 35

Chapter V

Contrastive studies at various levels of linguistic analysis 47

Chapter VI

Linguistic models and contrastive studies 107

Chapter VII

Towards a theory of semanto-syntactic equivalence 147

Chapter VIII

Contrastive Generative Grammar 169

Quantitative
Error
Chapteranalysis,
XIX contrastive
interlanguage,
studies
and Contrastive Generative Grammar 203
189
VIII Contents

Chapter XI

Cognitive linguistics and contrastive studies 213

Chapter XII
"Theoretical" and "applied" contrastive studies 233

Notes 245
References 251
Index of subjects 267
Index of authors 283
Introduction

Contrastive studies do not enjoy much respect among linguists. Although

many scholars of repute occasionally or systematically do practise what,
loosely speaking, falls within the domain of contrastive studies by evoking
cross-language evidence to support some theoretical claims (for example,
R. Lakoff 1972, Comrie 1976, or Siewierska 1984), they only do so
compelled by general linguistic motivation rather than by the require-
ments imposed by the rules of an altogether different game called con-
trastive linguistics.
Those who, by their own admission, undertake contrastive studies only
involve themselves in that part which we shall presently refer to as
contrastive analysis proper, paying insufficient heed to matters of prin-
ciple, which motivate the analyses and provide them with methodological
tools. In any case, the emphasis falls on actual practice and applications
to the detriment of the theory and methodology of contrastive studies.
Whatever issues arise in connection with these latter two aspects of
contrastive studies, they are treated only marginally, as it were, in passing,
and without sufficient attention paid to matters of finer detail. Conse-
quently, the number of works explicitly and exclusively devoted to the
theory and methodology of contrastive studies is negligible (but see Di
Pietro 1971 and James 1980).
For some years, I have been engaged in providing contrastive studies
with a more rigorous format. Several other linguists have advanced
illuminating comments and criticisms, notably Bouton (1976), Chester-
man (1980), James (1980), and, most extensively, Van Buren (1976). Thus,
although only embryonically, something that might be called a method-
ology of contrastive studies has begun to emerge.
The present book attempts to bring under one cover some discussions
and controversies connected with contrastive studies, to suggest possible
answers to some critical comments, and to provide a synthetic outlook
on the state of the art of contrastive studies with some modest suggestions
of improvement.
Contrastive studies have a very long history. As early as ca. 1000 A. D.
Aelfric wrote his Grammatica, a grammar of Latin and English, based
2 Introduction

on the implicit assumption that the knowledge of grammar of one lan-

guage may facilitate the learning of another language.
Among later grammarians, John Hewes, in the 17th century, was the
first to explicitly express the view that the knowledge of the native
grammar can not only facilitate learning a foreign language but also
interfere with it. In his A perfect survey of the English tongue taken
according to the use and analogie of the Latine, published in 1624, he
devoted a long introductory section to presenting fundamentals of English
in order to provide the learner with a "right knowledge or censure of
their owne Mother tongue, in regard it holdeth a great difference in it
selfe from the dialect of the Latine" Having provided some intricate
contrastive analyses of Latin and English, Hewes provides numerous
translational exercises to counteract what appear to be the effects of
negative transfer, focusing his attention on those phenomena which are
different in Latin and English.
Many other grammarians, like Howel(l) (1662), Coles (1675), and
Lewis (1670?) applied the idea of facilitation (positive transfer in modern
terms) by adjusting their grammars of English or of Latin to the needs
of speakers of various native languages. It is very interesting to note that
those early contrastive studies were motivated in almost the same way as
modern contrastive studies in the United States were motivated. As early
as 1670, Mark Lewis wrote the following words:
The most facil (sic!) way of introducing any in a Tongue unknown is
to show what Grammar it hath beyond, or short of his Mother tongue;
following that Maxime, to proceed a noto ad ignotum, making what
we know, a step to what we are to lean (sic!).
One wonders whether nearly three centuries later Charles C. Fries was
aware of these words, when he wrote the following:
The most efficient materials are those that are based upon a scientific
description of the language to be learned, carefully compared with a
parallel description of the native language of the learner (Fries 1945:
9)·
So, although the word "contrast" with reference to different phe-
nomena across languages had not appeared until the end of the 18th
century, when James Pickbourne (1789: 18) first used it ("I thought it
would be useful to contrast [italics supplied] the English verb with the
verb in other languages"), comparisons of languages for pedagogical
purposes probably go to the very beginnings of foreign language teach-
 Introduction 3

ing, while systematic written records of such procedures go back to at

least the 15th century (cf. Krzeszowski 1985, 1986 a, and in prepara-
tion; Meech 1935).
The early contrastive analysts did not concern themselves with meth-
odological problems, although they did work out a method of comparison
known as the "sign theory", the first method in contrastive studies
(Krzeszowski 1985). For many years contrastive studies were practised
and applied in the classroom in a more or less intuitive way, toutes
proportions gardees, like folk medicine, without much theory and without
much explanation.
But modern linguistic theories that began to flourish in the 20th century
could not fail to affect the state of affairs in contrastive studies. Interest
in methodology and theory of contrastive studies began to grow. Lin-
guistic explorations into the nature of language, its complex, multilayer,
and hierarchical structure, its systematic but changing nature, its function
in communication, and its relation to the human mind became a subject
of very close scrutiny of modern linguistics. This is not to say that many
of these problems never arose before. For centuries, people wrote gram-
mars (including contrastive grammars) and for centuries they were inter-
ested in how languages reflect human thoughts. But modern linguistic
theories have given new dimensions to old problems and have created
new problems (cf. Fisiak 1975 b; Fisiak — Lipinska-Grzegorek — Zabrocki
1978). Obvious things ceased to be obvious, and completely new ap-
proaches to language were proposed (e. g., generative grammar). Modern
linguistic theories have made contrastive analysts sensitive to methodo-
logical and theoretical problems in their own field. Contrastive studies
began to aspire to the status of a rigorous scientific discipline. What was
once a relatively simple, intuition-based procedure, began to assume the
format of an algorithm, which culminated in the attempt to construct
Contrastive Generative Grammar (see Chapter VIII). The rigorous, math-
ematical approach to contrastive studies has revealed a number of thorny
problems which at best make such efforts extremely difficult and at worst
make them hopeless. Ironically, the attempt to give contrastive studies a
rigorous, mathematical format could turn out to be self-defeating: it
could reveal the impossibility of conducting formalized contrastive studies
(cf. Van Buren 1976: 315).
The theoretical problems which bedevil contrastive studies can be
summarized as three paradoxes: the grammatical paradox, the semantic
paradox, and the pedagogical paradox.
4 Introduction

The grammatical paradox

Any grammatical contrastive studies performed without reference (at

least implicit) to meaning are doomed to failure. Yet, witness another
paradox, the revival of contrastive studies in the 20th century was pro-
moted by rather extreme versions of American structuralism, in which
the semantic aspect of language was removed from the mainstream of
linguistic analysis.
What makes a purely structural approach to contrastive studies the-
oretically impossible is the fact that any comparison presupposes simi-
larity as tertium comparationis (TC) against which differences can be
stated. Therefore, languages are structurally comparable in the extent to
which they are structurally similar. It must be remembered that early
structuralists were not yet familiar with the notion of underlying structure
nor with the notion of semantic representation from which, according to
later theories, surface structures in various languages are derived. These
concepts were acknowledged and appreciated in contrastive studies as
soon as they had made their way in the linguistic world. Originally,
however, contrastive studies concerned only those structures which later
became known as surface structures.
If one considers a taxonomy of surface structures in various languages,
it is possible to envisage a sort of a cline of similarities for every pair of
languages within the taxonomy. Theoretically, at one extreme, there will
be a pair of languages so radically different that no common set of
grammatical categories can be established for them, and at the other
extreme, there will be two languages identical in all respects. In-between,
there will be a spectrum of pairs of languages with varying degrees of
similarities. In such a hypothetical cline, the languages situated at the
two extremes would be completely incomparable in purely structural
terms. The two languages that would have nothing in common could not
be compared since there would be nothing that could serve as tertium
comparationis. The two structurally identical languages at the other ex-
treme, being identical, could not be compared either since there would
be nothing different to consider.
Graphically the cline can be represented as in Fig. 1, where L t and L 2
have no features in common while L n and L n + , are identical in all respects.
In fact, Fig. 1 represents a fragment of a more complex situation since,
in principle, the number of totally different languages and totally identical
languages may be more than two, and the respective numbers of such
 Introduction 5

Figure 1. Languages in a cline ranging from those with nothing in c o m m o n to

those that are identical.

languages cannot be determined a priori. Therefore, more accurately, our

cline should be represented by the following formulas: Let U stand for
the set of all possible human languages

U = [L„ L 2 Ln]

and let σ stand for the similarity ratio, i.e.,

σ = s/d

where s stands for the number of similar elements and where d stands
for the number of different elements, 1 and where both s and d > 0 , in
order for the L's to be comparable.
Let us assume, further, that reliable descriptions of all the languages
constituting U are available, and that these descriptions are all made
within the same theoretical framework or a set of mutually compatible
frameworks such that equivalence can be established for these frame-
6 Introduction

works. Provided these conditions are met, all languages in U can be

arranged pairwise on the scale of σ from η to m, where η > 0 and m < oo.
The grammatical paradox consists in the resulting situation, viz. that
languages are comparable only to the extent to which they are similar;
thus, the lower the σ for a given pair of languages, the less comparable
they are. It follows that the most voluminous contrastive analyses should
be performed on those languages for which σ would be the highest, while
contrastive analysis becomes increasingly impossible with the diminishing
σ. In either direction, purely structural contrastive studies reach an
impossible impasse, which cannot be overcome, even in principle. The
additional paradox, as we pointed out earlier, is that contrastive studies
in the 20th century began to flourish within the structural framework.
The paradox underlying such studies led to the situation in which,
thriving as they were, they were at the same time completely atheoretical.

The semantic paradox

Even before the difficulties connected with comparability within the

structural framework had been formulated, practising contrastive analysts
abandoned the strictly structural positions in favour of employing se-
mantic criteria in establishing comparability. At first, semantics was
employed implicitly as translational equivalence, the term borrowed from
the theory of translation as formulated by Catford (1965: 27 — 34) and
used in the context of contrastive studies by Halliday et al. (1964:
111 — 134). Translation equivalence was employed as the main criterion
in deciding what constructions and sentences in various languages are
comparable. The following principle was formulated:
If the items are not at least sometimes equivalent in translation, they
are not worth comparing (Halliday et al. 1964: 115).
This clearly implied the presence of semantics on the contrastive scene,
since translational equivalence presupposes at least some degree of se-
mantic resemblance: two sentences are mutually translatable if they share
at least some semantic properties.
Later, a more radical hypothesis was formulated. It claimed that
equivalent sentences across languages have identical semantic represen-
tations. This hypothesis was connected with the so-called universal base
hypothesis (Bach 1968: 91; Fillmore 1968: 51 ff.), which claimed that all
 Introduction 7

sentences in all languages are derived from a universal semantic base, for
which various metalanguages of representations were proposed, such as
a modified predicate calculus (Fillmore 1968), a system of roles (Fillmore
1968), or a system of labelled graphs (Krzeszowski 1974). All these
proposals had one thing in common: they assumed the existence of some
universal, underlying semantic representation, free of language-specific
syntactic categories, from which all sentences in all languages are derived
through language-specific categorical and syntactic rules. These rules can
be compared and contrasted in various languages, which provides a new
dimension for contrastive studies (see also Lipinska 1975: 50 ff.).
The hypothesis concerning the identity of semantic representations of
equivalent sentences leads to the semantic paradox, which is based on
the fact that what is identical is not subject to comparison, and what is
different is not comparable. Since equivalent sentences across languages
have identical semantic representations, it follows that differences occur
at less abstract levels, due to the operation of different, language-specific
grammatical rules. Therefore, languages differ at more superficial levels,
while in their deeper structure they are increasingly similar, and at the
level of semantic representation they are presumably universally identical.
Thus, the most interesting linguistic insights and generalizations provide
the least promise for contrastive studies since with the increasing univer-
sality of the grammar (whether expressed in terms of the universal base
hypothesis or in quite different terms, such as more recent versions of
the standard theory, as expounded by Chomsky 1975, 1982, 1984), there
is less and less for contrastive studies to deal with. Here is how Preston
formulates this paradox:
That the drive for universality should deny comparative detail between
even related languages should seem paradoxical is understandable at
a superficial level, but if we recall Chomsky's assertion that universal
grammar will eventually leave only idiosyncratic odds and ends and
irregularity behind in particular grammars, the conclusion that better
grammars provide less and less detail for contrastive analysis is self-
evident (Preston 1975: 69).
Therefore, what is left for comparison are essentially incomparable
idiosyncracies, such as suppletion, pronominal irregularities, morpholog-
ical curiosities, and specific selection features, as the only items of any
interest in the grammar of a particular language. According to Preston,
even if contrastivists do trouble themselves with all these idiosyncratic
details across languages, they will "make no real contribution to linguis-
tics" (Preston 1975: 65).
8 Introduction

The pedagogical paradox

Accepting the inevitable conclusion that what is universal must be familiar

to all foreign-language learners from their native languages and consti-
tutes the main focus of interest of theoretical linguists, we must also face
the conclusion that what remains to be learned are "non-comparable bits
of peculiar and idiosyncratic information generally ignored in contrastive
studies" (Preston 1975: 65).
So the pedagogical paradox is that what theoretical linguists consider
as trivial and uninteresting may be of utmost pedagogical importance.
Language teachers all over the world are particularly sensitive to errors
in the realm of morphology, transgressions against rare, exceptional forms
or unique pronunciations. All these phenomena, mainly connected with
surface structures, traditionally constitute the learner's hell and, if exces-
sively focused on, occasion many a failure in learning foreign languages
in classroom situations. Languages abounding in such phenomena have
the reputation of being "difficult" in contrast to those languages which
are relatively free of such surface phenomena as inflections.
Neither the semantic nor the pedagogical paradox undermine the
validity of contrastive studies since there is no way of telling what is
universal and what is idiosyncratic without conducting thorough con-
trastive studies of as many languages as possible. One cannot accept the
view that
a thorough investigation of [!] large number of languages is really
unnecessary for the universal features necessary to explanatory
adequacy are natural by-products of close, descriptively adequate work
on the structure of a particular language (Preston 1975: 67).
T. Zabrocki presents some convincing evidence demonstrating that
certain theoretical linguistic hypotheses cannot be verified without ref-
erence to cross-language data. In fact, Zabrocki is of the opinion that in
order to test any claim assigning a universal value to particular syntactic
rules or categories one should ideally test these claims "on a number of,
possibly all, languages, within general theoretical linguistic studies" (Za-
brocki 1976: 101).
The monograph presented here attempts to resolve these paradoxes
by suggesting such methods in contrastive studies which would be theo-
retically plausible and practically useful. In this way problems and con-
troversies beclouding contrastive studies may lose some of their edge,
which will give contrastive studies a chance to survive as an art, if not
as an algorithm.
Chapter I

What is contrastive linguistics?

When two or more languages are compared, it is possible to focus either

on similarities or on differences. 1 When a learner learns a new language,
he usually focuses attention on differences and remains largely unaware
of similarities. If he discovers some similarities, he is amused and surprised
since he ordinarily does not expect to find them. Grammarians, on the
other hand, quite early became interested in discovering what various
languages have in common, in the belief that making such similarities
explicit for the learner may facilitate the process of foreign language
learning. Early contrastive studies were motivated precisely by this as-
sumption (Krzeszowski 1985: 485).
There are several approaches to linguistic comparisons. The 19th
century witnessed the development of historical linguistics and the related
field of comparative historical studies, which aimed at finding the com-
mon genetic background for whole groups of languages. Another field
of linguistic comparisons emerged when languages were grouped on the
basis of various characteristics which they share. The area of linguistics
concerned with such comparisons is called typological linguistics. Lan-
guages grouped together in the same typological group need not be
genetically (historically) related. For example, English and Chinese, which
are not genetically related, share a large number of grammatical prop-
erties, such as relatively fixed and grammatically constrained word order,
paucity of inflections, and prominence of function words. These shared
features place the two languages quite close in the typological groupings,
in spite of the genetic distance separating them.
Contrastive linguistics is connected with yet another kind of compar-
ison: noting and describing similarities and differences in languages rather
than grouping them genetically or typologically. Ultimately, of course,
all kinds of comparisons may yield results which are relevant to linguistic
theory in general, as in the search for linguistic universals. Therefore,
differences between typological and contrastive linguistics are largely a
matter of focus: typological linguistics focuses on clusters of languages
10 Chapter I

united by some common feature or features, while contrastive linguistics

focuses on pairs of languages and explores similarities as well as differ-
ences between them.
Contrastive linguistics, like descriptive and historical linguistics, is
dependent on theoretical linguistics since no exact and reliable exploration
of facts can be conducted without a theoretical background, providing
concepts, hypotheses, and theories which enable the investigator to de-
scribe the relevant facts and to account for them in terms of significant
generalizations. But contrastive linguistics is also dependent on descriptive
linguistics since no comparison of languages is possible without their
prior description. In brief, then, contrastive linguistics is an area of
linguistics in which a linguistic theory is applied to a comparative de-
scription of two or more languages, which need not be genetically or
typologically related. The success of these comparisons is strictly depend-
ent on the theory applied. As will be seen later, in extreme cases, the
linguistic framework itself may preclude comparison. Therefore, contras-
tive linguistics imposes certain demands on the form and nature of the
linguistic theory which is to be "applied" in such comparisons. In many
less extreme situations the results of comparisons are strictly dependent
on the theoretical framework adopted in the comparisons (see Chapter
VI).
Originally, all contrastive studies were pedagogically motivated and
oriented. In recent years, however, distinctions have been drawn between
"theoretical" and "applied" contrastive studies (see Fisiak 1981: 2 — 9).
According to Fisiak
Theoretical CS give an exhaustive account of the differences and
similarities between two or more languages, provide an adequate model
for their comparison, determine how and which elements are compa-
rable, thus defining such notions as congruence, equivalence, corre-
spondence, etc. Applied CS are part of applied linguistics. Drawing
on the findings of theoretical contrastive studies they provide a frame-
work for the comparison of languages, selecting whatever information
is necessary for a specific purpose, e.g. teaching, bilingual analysis,
translating, etc. (Fisiak 1981: 9).
In Chapter XII we shall investigate to what extent this is a valid
distinction. Assuming, however, that "applied contrastive studies" are
sufficiently distinct from "theoretical contrastive studies", the former, as
part of applied linguistics, especially when related to teaching, must
necessarily depend not only on theoretical, descriptive, and comparative
 What is contrastive linguistics? 11

linguistics but also on other disciplines relevant to teaching; among them

are psycholinguistics, sociolinguistics, didactics, psychology of learning
and teaching, and possibly other areas which may be important in ways
difficult to evaluate at the present moment.
Finally, some comments are needed about terminology. Although the
word "contrastive" is used most frequently with reference to cross-
language comparisons of the sort described above, various authors have
been trying to replace it with other terms, such as "cross-linguistic
studies", "confrontative studies", and some even more esoteric terms, for
example, "diaglossic grammar" (Dingwall 1964 a), which enjoyed but a
brief existence. The word "contrastive" is likely to outlive all the com-
peting terms since it appears in titles of monographs and collections of
papers on the subject (cf. James 1980; Fisiak 1980, 1981, 1984).
Terminological complications manifest themselves in a rich array of
nouns which have been attested in collocation with the adjective "con-
trastive" In the current literature we find such terms as "contrastive
linguistics", "contrastive studies", "contrastive analysis", and "contrastive
grammar"
Although consistency is certainly wanting, there is an observable
tendency to select a particular collocation to refer to particular domains
of cross-language comparisons. And so the term "contrastive studies"
appears to be the least marked, as it fits all contexts in which other
collocations with "contrastive" are also appropriate. The term "contras-
tive linguistics" is also often used with reference to the whole field of
cross-language comparisons, with a slight tendency to focus on those
instances when theory or methodology of comparisons come into play.
The collocation "contrastive analysis" is often used interchangeably with
the above two terms, but there is a tendency to restrict its scope of
reference to comparison proper. In that restricted sense "contrastive
analysis" would refer to the third of the three steps in classical contrastive
studies, viz. description, juxtaposition, comparison proper, respectively
(see Chapter IV).
Finally, the collocation "contrastive grammar" is often used to refer
to the product of contrastive studies, as a bilingual grammar highlighting
differences across languages. In this sense, "contrastive grammars" con-
stitute an outcome of "contrastive linguistics"
Whether these observations are accurate or not, it would be desirable
to aspire towards some consistency in the use of these terms along the
lines suggested in the above generalizations.
12 Chapter I

Unfortunately, these observations do not exhaust all the contexts in

which the adjective "contrastive" appears. Many other collocations can
be encountered in an astonishingly varied assortment. Each such col-
location consists of the adjective "contrastive" followed by all manner
of nouns; so we get "contrastive pragmalinguistics" "contrastive soci-
olinguistics", "applied contrastive studies", "contrastive discourse anal-
ysis" "contrastive pragmatics", "contrastive syntax", "pragmatic con-
trastive analysis" "contrastive generative grammar", "contrastive phon-
ostylistics" "contrastive language studies", "contrastive lexicon", "con-
trastive considerations", "transformational contrastive studies", "theo-
retical contrastive studies", "classical contrastive studies", "deeper and
deeper contrastive analysis", "contrastive description", and quite a few
more.
The existing confusion is typical of a field where insufficient atten-
tion has been paid to various matters of principle. Preoccupied with
analytic details, investigators often lose sight of general distinctions
and aims of their research, and they do not question certain funda-
mental assumptions, which are often taken for granted. In contrastive
studies, the situation is further aggravated by the deeply-nourished
conviction of many practitioners in the field that theoretical problems
of relevance arise and can be solved only in the domain of pure and
descriptive linguistics. In the view of these linguists, contrastive studies
are merely a set of procedures involving mechanical application of
various findings in theoretical and descriptive linguistics. Implicit in
this position is the belief that contrastive studies do not require any
special theoretical framework to be fully effective and to bring forth
whatever results are expected of such analyses.
There is a handful of linguists concerned with contrastive analyses,
who occasionally feel guilty of the "sin of omission", and they stop to
ponder over certain vagaries of comparative procedures, trying to see
them in more general terms than those available from actual practice
(e.g., Van Buren 1974; Schwarze 1978; Kühlwein 1983). These few in-
vestigators are aware that contrastive linguistics requires its own theo-
retical framework in addition to specific linguistic models employed in
the description of the compared languages.
One of the purposes of this book is to remedy this situation by
highlighting and discussing some crucial issues which bedevil contrastive
linguistics. We are going to address notorious problems such as tertium
comparationis, equivalence, the relation between linguistic theory and
 What is contrastive linguistics? 13

contrastive linguistics as well as between contrastive linguistics and for-

eign language teaching. With respect to the existing terminological jungle,
referred to earlier on, we are going to suggest a taxonomy of contrastive
studies, which will provide motivation for some terms and render other
terms superfluous.
Chapter II

Tertium comparationis1

To compare them would be tantamount to

putting ten-ton lorries and banana skins
in the same class on the grounds that nei-
ther ought to be left on footpaths!
Carl James

One of the reasons why contrastive studies continue to perform the role
of the Cindarella of linguistics is the fact that its most fundamental
concept, tertium comparationis, remains as hazy as ever. The existing
contrastive analyses involve various platforms of interlinguistic reference,
determined by specific linguistic models which they employ and specific
levels of analysis which they embrace. Thus different tertia comparationis
are used for comparisons in lexicology, in phonology, and in syntax. In
few of these studies is explicit mention of any tertium comparationis made
or any justification for a specific choice presented. 2
All comparisons involve the basic assumption that the objects to be
compared share something in common, against which differences can be
stated. This common platform of reference is called tertium comparationis.
Moreover, any two or more objects can be compared with respect to
various features and, as a result, the compared objects may turn out to
be similar in some respects but different in others. Thus, a square and a
rectangle are similar in that both consist of four sides at right angles.
But they are also different, since in a square, but not in a rectangle, the
four sides are of equal length. 3 If we compare squares and rectangles
with respect to the angles, we ascertain that the two types of figures are
identical. If, on the other hand, we consider the length of their sides, we
find them to be different. Depending on the platform of reference (or
tertium comparationis) which we adopt, the same objects turn out to be
either similar or different.
In cross-language comparisons, the choice of tertium comparationis
will also constitute the determining factor in establishing similarities and
differences between the phenomena compared (cf. Lipinska 1975: 48;
Fisiak et al. 1978: 15). Since language is a complex hierarchical structure,
operating at various levels of organization, and since it manifests itself
16 Chapter II

as texts produced by its users, every aspect of language at every level of

organization, as well as every text and its constituents, can undergo
comparison with equivalent elements in another language. Therefore,
various kinds of contrastive studies can be distinguished, depending on
the tertium comparationis adopted and the kind of equivalence involved.
Before we attempt to classify contrastive studies, let us take a closer look
at various possible tertia comparationis.
Theoretical discussions tend to be limited to only two types: formal
correspondence and semantic equivalence (e.g., Lado 1957: 52 — 53; Spa-
latin 1969; Ivir 1969, 1970). Even a cursory glance at the wealth of the
existing contrastive studies suffices to notice that these two types of
tertium comparationis are not the only ones that are used in practice.
Formal correspondence and semantic equivalence can serve as tertia
comparationis for certain types of contrastive studies, such as syntactic
and lexical. Other types of contrastive studies, for example phonological,
pragmatic, or quantitative, must be based on other tertia comparationis.
Moreover, neither contrastive studies based on formal correspondence
nor those based on semantic equivalence are free from difficulties. For
example, it has been pointed out that formal likeness alone cannot serve
as a tertium comparationis without support from semantic equivalence
(Liston 1970: 44; Lipinska-Grzegorek 1977: 1 — 10). At best a comparison
based on formal criteria alone is incomplete, at worst it cannot be
performed at all, and in many cases it is misleading (see also Spalatin
1969: 3 1 - 3 4 ) .
If, for example, one compares Polish and English personal pronouns, a
formal analysis will ascertain the equivalence between the English you and
the Polish ty/wy and will be accurate as far as it goes. But such an analysis
is incomplete as it leaves out such forms as Pan/Pani and other possible
equivalents of you. These equivalents can only be established if other than
formal criteria are employed (see Chapter IV). English articles cannot be
compared to anything in those languages in which there are no articles, if
only formal criteria are considered. Finally, in the case of such phenomena
as the present perfect tense in English and passe compos0 in French, a formal
analysis is misleading since the formal similarity is not matched, at least in
this case, by semantic similarity, which creates a kind of situation which
often causes considerable learning problems (see Politzer 1968). Therefore,
it is generally recognized that a contrastive analysis based on purely formal
criteria falls short of both theoretical and practical expectations. We shall
return to this problem in Chapter VI.
 Tertium comparationis 17

Somewhat less obviously, a contrastive analysis based on semantic

similarity alone can also be inadequate and misleading. In the contrastive
practice, semantic equivalence is often erroneously identified with trans-
lation equivalence:

To establish that these [systems of deictics] are comparable, we first

need to show their contextual equivalence; this can be done most
simply by reference to translation (Halliday et al. 1964: 115).
Chapter VII will discuss differences between semantic equivalence and
translation. It will be shown that translation equivalents are often se-
mantically non-equivalent. At this point, it must only be noted that
semantic equivalence must be constrained formally, while translation
equivalence may, but does not have to, be thus constrained. When one
translates, one departs from semantic equivalence due to three types of
reasons: (1) errors in translation; (2) formal properties of respective
languages; and (3) what is loosely called "stylistic" reasons. These three
types of reasons lead to situations in which actual translation practice,
with the exception of that concerned with legal texts, seldom involves
semantic equivalents in the sense defined below in Chapter VII. This
means that only some translations can be used as data for systematic
contrastive studies (cf. Ivir 1969), while translation as a method of
contrasting must be regarded with caution:

Translation must be viewed amorphously as the rendition of a text

from one language to another. This is translation from the standpoint
of la parole: the text, the act of speech or writing, is the thing. Or it
may be viewed as a systematic comparison of two languages: this is
translation from the standpoint of la langue (Bolinger 1966: 130).
In fact, the use of translation in systematic contrastive studies is highly
limited:
Translation equivalence serves merely to help us isolate items of struc-
ture with shared meanings in the two languages (Ivir 1970: 15).
Even if we do distinguish translation equivalence from semantic equiv-
alence and base contrastive studies on the latter, we still face problems.
As has been stated earlier, semantic equivalence involves "formal" con-
straints. Thus, semantic equivalence is inherently connected with at least
some degree of formal correspondence. But the meaning of both concepts
is richer than is commonly recognized in contrastive studies. For instance,
"formal" can be extended to cover the entire plane of expression (cf.
18 Chapter II

Hjelmslev 1961: 59), whereas in most American studies the word "formal"
is restricted to word order, function words, inflections, affixation, and
suprasegmentale. In a broader perspective, "formal" would also embrace
such aspects of expression as aliteration, rhymes, and rhythm. Many of
these "formal" properties would find their place in the study of function-
ally (pragmatically) equivalent texts (see Chapter XI).
The notion "semantic" is also often extended to cover matters of
pragmatics, especially by those authors who identify semantic equivalence
with translation equivalence:
Our experience is that languages can be effectively contrasted only on
semantic basis, specifically, on the basis of translation equivalence
(Spalatin 1969: 34).
In reality many authors have shown that semantic equivalence is not
a necessary prerequisite of a good translation (cf. Rülker 1973: 29 — 35;
Krzeszowski 1974: 13, 1981 a; Kopczynski 1980: 4 1 - 4 2 ) . What is ex-
pected of a good translation is pragmatic or functional equivalence (see
Chapter V). It cannot be denied that pragmatic equivalence can serve as
tertium comparationis for contrastive analyses of such matters as the
structure of discourse, stylistic properties, and quantitative aspects of
texts. But syntactic contrastive studies, the primary concern of earlier
contrastive studies, must be conducted within the limits of the semantic
component of the language, or more specifically that part of the semantic
component which can be systematically and predictably correlated with
the grammatical structure of sentences. This restricted sense of "semantic"
still embraces some aspects of meaning which are traditionally relegated
to "pragmatics" or "interpersonal function" of sentences (Halliday 1970:
143). According to Halliday, the systems of mood and modality are
precisely those systems which relate sentences to their interpersonal func-
tions. It seems obvious that the notion "sentence semantics" should cover
those elements of "pragmatics" which can be correlated with the structure
of sentences, even if consistency in this area is definitely out of the
question; declarative, interrogative, and imperative sentences do not nec-
essarily perform the functions of statements, questions, and commands,
respectively. In so far as some correlation between form and function
does exist, those "functional" aspects of sentence structure constitute the
border area between pragmatics and semantics and should be included
in any semanto-syntactic contrastive studies (for details see Krzeszowski
1974). Therefore, James (1980) suggests that for the purposes of contras-
tive analyses translation equivalents should be limited to those which are
 Tertium comparationis 19

both semantically and pragmatically equivalent. However, this proposal

also raises doubts.
Presumably, what James means by "semantic" refers to Halliday's
"ideational" function of sentences, while Halliday's "interpersonal" and
possibly "textual" functions fall under "pragmatic" Under James' pro-
posal many sentences across languages would exhibit both "ideational"
(semantic) and "interpersonal" (pragmatic) equivalence; yet one would
hardly wish to use them as data for syntactic contrastive studies. For
example:
(1) Did he kill that dog?
(2) Zabil tego psa?
(literally: 'killed-he that dog?')
are equivalent, both ideationally (agent, transitive verb, patient) and
interpersonally (general question); yet, they falsely suggest a relationship
between syntactic types represented by (1) and (2) in English and Polish.
A systematic syntactic equivalence would have to be ascertained between
(1) and (3) rather than between (1) and (2):
(3) Czy on zabil tego psa?
(literally: 'whether he killed that dog?')
since (3) typically represents interrogative sentences in Polish, just as (1)
is a typical interrogative sentence in English.
In an earlier work (Krzeszowski 1981 b: 123), I suggested that syntactic
contrastive studies should be performed on data restricted in the following
way: a contrastive grammar will take as its primary data (to be assigned
the status of semanto-syntactic equivalence) the closest approximations
to grammatical word-for-word translations and their synonymous para-
phrases, if such forms exist. Such a constraining of primary data as the
basis for syntactic contrastive studies bypasses the inherent difficulties of
the proposals suggesting the use of unrestricted semantic equivalence as
the basis for comparison. Accepting any translation as a possible basis
for syntactic contrastive studies leads to two mutually exclusive and
undesirable consequences. Either (1) no comparative generalizations be-
come possible, as the number of well-formed translations of a particular
sentence into another language cannot be predicted a priori; or (2) purely
arbitrary decisions concerning formal correspondences in unconstrained
translations must be made. Any non-arbitrary decision involves circular-
ity: the investigator has to assume formal correspondences on the basis
of syntactic and/or morphological features which the compared texts
20 Chapter II

share. This circularity is even reflected in the use of the word "compa-
rable" in certain contrastive grammars. For example, Stockwell et al.
thus write about determiners in English and Spanish:
Both English and Spanish have two sets of determiners, commonly
referred to as definite and indefinite articles. In many respects they
are comparable [emphasis is my own]; in others they are different
(Stockwell et al. 1965: 65).
The circularity consists in the following: we compare in order to see
what is similar and what is different in the compared materials; we can
only compare items which are in some respect similar, but we cannot use
similarity as an independent criterion in deciding how to match items for
comparison since similarity (or difference) is to result from the compar-
ison and not to motivate it.
To avoid this undesirable circularity, in deciding about formal corre-
spondences, one needs a common tertium comparationis outside the for-
mal properties. The underlying meaning of the closest approximations to
well-formed word-for-word translations provides such a tertium compar-
ationis. Sentences and constructions sharing identical semantic represen-
tations at the level of sentence semantics (but necessarily exhibiting certain
idiosyncratic differences at the level of word-semantics) are semanto-
syntactically equivalent and constitute a constrained set of data for
syntactic contrastive studies. The approach through constrained trans-
lations does not require the initial recognition of shared syntactic cate-
gories as tertium comparationis for syntactic contrastive studies. Such a
recognition would illegitimately anticipate the results of contrastive stud-
ies. A detailed proposal along these lines will be presented in Chapter
VIII. (See also Krzeszowski 1974 and 1979).
Summarizing, let us say that formal properties alone do not provide
an adequate tertium comparationis for syntactic contrastive studies, while
a semantic tertium comparationis must be constrained through restricting
the scope of translation equivalents as primary linguistic data for syntactic
contrastive studies.
Such constrained but rigorous contrastive studies have a very limited
pedagogical relevance. Any extension of the scope of contrastive studies
to make them pedagogically more useful increases the likelihood of their
becoming less rigorous and hence less respectable as a "scientific" pro-
cedure. One has to look for ways of extending the scope of contrastive
studies without losing any of the rigour characterizing syntactic contras-
tive studies. Formal and semantic tertia comparationis, discussed so far,
 Tertium comparationis 21

will not suffice as bases for extended contrastive studies. For example,
phonetic and phonological contrastive studies cannot rely on semantic
equivalence as tertium comparationis.
The crucial notion in identifying various kinds of tertia comparationis
and determining their character is the concept of equivalence or the
relation which provides justifications for why things are chosen for
comparison, keeping in mind that only equivalent items across languages
are comparable. The various principles motivating equivalence and, eo
ipso, contrastive studies will provide grounds for dividing tertia compar-
ationis and, consequently, contrastive studies into various categories, each
being connected with a specific kind of equivalence which motivates the
comparisons (see Chapter III). In other words, equivalence is the principle
whereby tertium comparationis is established inasmuch as only such ele-
ments are equivalent for which some tertium comparationis can be found,
and the extent to which a tertium comparationis can be found for a
particular pair of items across languages determines the extent to which
these elements are equivalent. Thus, equivalence and tertium compara-
tionis are two sides of the same coin.