Verbal cues of organizational information in message decoding: An integrative approach to linguistic structure [Reprint 2019 ed.] 9783111350929, 9783110996388

Citation preview

JANUA LINGUARUM STUDIA MEMORIAE NICOLAI VAN WIJK DEDICATA edenda

curai

C. H. V A N SCHOONEVELD Indiana

University

Series Minor,

127

VERBAL CUES OF ORGANIZATIONAL INFORMATION IN MESSAGE DECODING An Integrative Approach to Linguistic Structure

by

MAMUD M. OKBY State

University

of New York at

1972

MOUTON THE HAGUE • PARIS

Binghamton

© Copyright 1972 in The Netherlands. Mouton & Co. N.V., Publishers, The Hague. No part of this book may be translated or reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publishers.

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 75-189706

Printed in Belgium

INTRODUCTION

A discipline ordinarily develops around a method and a theory or two. The method becomes a canon of methods and the theory a canon of theories, as the discipline grows. It develops its own instruments, its own terminology; often its own mathematics, its own logic, its own rhetoric. It begins to have traditions, customs, and a history; a philosophy and an ethic emerge, and perhaps a code of law. It finds enemies; forms alliances; firms up its boundaries; establishes appropriate sanctions in self-protection. It finds it necessary to inculcate its values in its recruits. It becomes an establishment. Several disciplines may impinge on the same area of knowledge. They may set up their separate camps there, and may compete for domination of the territory, or attempt to co-exist in relative harmony. Without reaching the stage of positive cooperation, they may combine to drive out others who encroach. When disciplines which impinge on the same area of knowledge do cooperate positively, an interdisciplinary field has been established. As task forces representing the cooperating disciplines work harmoniously to advance knowledge, they may develop a method, and a theory or two, and then some instruments, and terminology, for the interdisciplinary field — and it is well on its way to becoming a new discipline, and ultimately a new establishment. But must this always be the pattern? We know that there have always been scientists and scholars who have been something more than disciplinary, and more than representatives of their

6

INTRODUCTION

disciplines in an interdisciplinary field. They have been multidisciplinary. They have mastered the methods and theories of first one and then another and another discipline, willingly spending the thousands of hours required for each. They are then no longer "discipline-bound"; they may become citizens of the world of disciplines; they each represent an "interdisciplinary task force" of one, advancing knowledge without jealousy, or prejudice, or territorial imperatives. From Herodotus to Leonardo to Otto Neurath and Hans Reichenbach there have been such scientists and scholars. One such multi-disciplinary scientist and scholar whose contribution to this book was perhaps more significant than any other is Dr. Floyd Henry Allport. In his Theories of Perception and the Concept of Structure (as well as his subsequent papers developing his theory of enestruence) we find two basic challenges that Dr. Okby has responded to in his multi-disciplinary approach to language: 1) We need to solve the unsolved problem of meaning. 2) " We need to work toward a theory of structure that will deal with the organization of includedness as well as of breadth, that will explain not only organization but the organization of organizations, that will apply both within and between the levels, — a theory, moreover, that will hold not only for perception but for all behavior."1 Another such multi-disciplinary scientist and scholar is Dr. Okby himself. A study that is multi-disciplinary may bring together theories and concepts from several disciplines to throw light on one. Or, it may explore "gray areas" peripheral to but not subsumed by existing disciplines in order to fill in gaps in human knowledge. Or, it may achieve supra-disciplinary or meta-disciplinary meaning and utility, throwing light on all the disciplines that are relevant. Dr. Okby's book does all these things, with quiet urbanity and sound common sense. His gentleness and humility nearly make us forget the magnitude of the forward strides he is making. Cowpet Bay, St. Thomas, September, 1971 John Ball 1

Theories of Perception and the Concept of Structure, p. 563.

ACKNOWLEDGEMENTS

In extending an expression of appreciation to those who have been instrumental in shaping the kind of thinking of which this study is exemplary, I would like to dwell on the powerful influences of a number of scholars, with whom I have studied and/or done research at both the University of Michigan and Michigan State University between 1957 and 1962. Needless to say, though the 'integrative', 'eclectic' approach to general linguistics is most topical today, it was not so, when I wrote my Ph. D. thesis under the same title of this volume in 1962. At the time, my academic advisors in structural linguistics and communication theory were confident that in no less than one decade, I could hope to arouse sufficient interest in the significance of such a thesis. Structuralists were still working within the relative confinement of the triple categories of 'phonology', 'morphology', 'syntax'; and the Generative-Transformationalists had just called attention to the need for explicit accounts of relationships between 'surface' and 'deep' structures in language. However, I was fortunate enough to receive instruction and get advice from a group of eminent scholars, noted for their evenhanded treatment of linguistic issues, and for the breadth of context against which they view the exact study of language. In my opinion, these scholars may be considered some of the most powerful proponents of research contributing to the evolution of a unified theory of language. Professor John Ball of M.S.U., with his deep insights into the twin notions of 'form' and 'structure', gave me the chance to explore the side of 'meaning' as an integral part of the

8

ACKNOWLEDGEMENTS

study of language. He was able to remove all inhibitions against research contributions from philosophy, psychology, sociology, and cultural anthropology. Professor Donald Lloyd of Wayne helped clarify many of the misconceptions propagated by the narrow 'atomists' in linguistics, those responsible for all the dryness, all the piece-meal treatment that has sickened students of language for ages. He gave me the benefit of participation in his Modern Language Audio-Visual Research Project in 1961. Professors David Berlo, Erving Bettinghaus, Hidya Kumata, and Malcolm MacLean, Jr. of M.S.U. have all contributed richly to my understanding of communication theory and process, which I have applied to the study of language in its socio-psychological and cultural setting. It was under their influence, and while participating in the Code Behavior Project, that my interest in psycholinguistics was aroused. Professor John Winburne's specific contribution to the present study must be acknowledged with gratitude. His research in structural semantics was generously made available to me, while I was working on my thesis. Professor Robert Lado of the U.O.M. at the time gave me a lot of encouragement to pursue the study of meaning across cultures. His research in applied linguistics opened up new vistas of investigation connected with the perception and production of language in the teaching-learning situation. Professor Albert Markwardt demonstrated his ability to reconcile the two sides of the dichotomy of 'synchrony' versus 'diachrony'; his approach to historical linguistics offered practical demonstrations. Professor Kenneth Pike also taught what he preached with respect to the location of language within a unified theory of human behavior; and I had the benefit of his teaching. Professor Thomas Sebeok of Indiana was visiting at the U.O.M., when his Introduction to Linguistics was offered. Thanks to his efforts, I was strongly impressed with the meaningfulness of a communication approach to linguistics. Subsequent teaching and further research, discussions with students and colleagues have constantly provided additional confirmation of the need for 'integrative' research. Obviously, what would normally sound 'subversive' to the 'atomist' of all times is

ACKNOWLEDGEMENTS

9

now spelled out as most reasonable. It is indeed very gratifying to reflect on the current enthusiasm for 'eclecticism' in linguistics; to think back on the circumstances attending the original motivation and stimulation on my part; to recollect the challenge, the venture, and the competent guidance and generous cooperation, which made it possible for me to go one step beyond the established limits or limitations. In memory of all this, it is the author's pleasure to acknowledge with gratitude the most helpful influences of the following scholars: Professors John Ball, David Berlo, Erwin Bettinghaus, Hidya Kumata, Robert Lado, Donald Lloyd, Malcolm MacLean, Jr., Albert Markwardt, Kenneth Pike, Thomas Sebeok, and John Winburne. Binghamton, N.Y. 1971

M. Okby

PREFATORY

In view of the controversies raging over recent developments in modern linguistic theory, and the current expressions of uncertainty as to what has been accomplished which could be identified as truly worthwhile and sufficiently reliable, one is made increasingly aware of the existence of certain challenges that could no longer be overlooked. 1 These challenges may be briefly summed up in the finding of ways and means whereby: to bring about a new 'unity' and order into the study; to make linguistics truly the interdisciplinary science we 'talk' about; and to communicate more 'effectively' with the outside world, peopled by those who do not speak our jargon. A recognition of these felt needs and some positive action towards their fulfillment are basic to any worthwhile contribution to scholarship in the exact science of language. The first challenge is actually centred on the question of 'orientation'. Modern linguistics is assumed to have outgrown the deficiency it developed under the mechanistic orientation, whereby the reality of language was attached almost exclusively to its physical aspects. Modern linguists are assumed to have overcome the tendency to quibble over 'structure' VERSUS 'function', 'synchrony' VERSUS 'diachrony', and the rest of the muddled issues, innocently aroused by some valid distinctions put into the form of dichotomies, that were never meant to distort one's 1

Noam Chomsky, "Linguistic Theory" in Mark Lester's Readings in Applied Transformational Grammar, p. 53: "What seemed to be well-established doctrine a few years ago may now be the subject of extensive debate. Although it would be difficult to document this generalization, it seems to me that there has been a significant decline, over the past ten or fifteen years, in the degree of confidence in the scope and security of foundations in both psychology and linguistics."

12

PREFATORY

thinking or threaten to disintegrate the totality of general linguistic theory. However, such a danger will remain immanent, and these monotonous issues will keep recurring in any serious discussion of current trends or future developments, certainly with no avail, unless we were to be positive enough to adopt a 'basic orientation' for the study of linguistics, that would be conducive to consorted action, and productive of truly 'integrative' research. 2 The observable lack of integration within the discipline is also attributable to insufficiency of interest in 'synthetic' research. Generally, analytic research is conducted with little or no concern for the total context of operation, within which the minute findings would assume significance. Obviously, we must analyze data at all levels, if we are to pursue our investigations inductively but the fact is that we should analyze with the ultimate aim of synthesizing, and some of us should certainly keep busy charting and recharting models of correlation, and/or designing research of the 'molar' variety to utilize the findings of 'molecular' investigations. Undoubtedly, the common failure to discover the operation of general 'laws' or principles to account for verbal behavior arises fundamentally from a common failure to formulate sufficiently meaningful hypotheses. And this, in turn, is attributable to a lack of correlated concepts in the field. The second challenge has to do with the actualization of a notion that is generally endorsed by many linguists, without sufficient effort being made towards its implementation, namely that linguistics is an 'interdisciplinary' science. The fundamental reason for this lag or inconsistency is most probably rooted in the referred to lack of cohesion within the discipline of linguistics as hitherto conceived. Naturally, one cannot establish adequate contacts with 2

It may be worthwhile to quote the observation reiterated in the early Fifties concerning a 'framework of operation' for the study of language in its totality. See H.A. Gleason, An Introduction to Descriptive Linguistics, p. 373: "The analysis of a language must proceed by separating out the various parts, but a full understanding of language cannot be gotten if they are left as detached details unrelated to one another. The various elements are of significance and interest primarily because they fit together into one integrated system which people use in communication. This function of language provides a framework within which language can be looked at more or less as a whole."

PREFATORY

13

one's neighbors, if one's own house is internally divided. Rationally, however, it should be feasible to contemplate an inner form of coalition on the basis of mutual interest in a shifting focus clarifying a diversity of issues labelled 'descriptive', 'generative-transformational', 'synchronic', 'diachronic', etc. — all of which are worthy of note and thorough investigation. Meanwhile, further succour towards the same coalition would be expected to come from outward contacts with the social and behavioral sciences, which are assumed to provide information on the total context of verbal communication. Such information should prove most helpful in sustaining the desired integration. Another reason for the very limited measure of mutual interest and cooperation, effected so far along the interdisciplinary dimension, may be the natural inclination on the part of scholars from either side to keep within the confines of familiar paths over a relatively well explored territory. It is observable, however, that the literature on verbal behavior presented by psychologists interested in language is not only voluminous, but in many ways suggestive of profitable insights that could be utilized by some linguists. A few psychologists have made an effort to keep abreast of recent developments in linguistics, and their attempts show that they can address themselves intelligently to psycholinguistic phenomena, without necessarily being also 'linguists'. A similar standpoint should be taken by more linguists with regard to psychology and other behavioral sciences. It is no longer possible for linguistics to be treated as a 'self-sufficient' science. It is and has to be treated as 'interdisciplinary'. Those of us who could not help showing their skepticism of this integrative view of language, when isolationist tendencies were in full sway, would not be expected, under present conditions, to set unnecessary limits on more fruitful contacts across neighboring fields. The third challenge relates to communicability in the metalanguage of linguistic science. It is generally felt that the diversity characteristic of 'language' at all levels of usage should not be equally characteristic of the meta-language of a 'science' of language. However, because linguistics, as a science, has had to

14

PREFATORY

cater to several evolutionary developments, the load of technical terminology is not only excessive, but also confusing at times, especially to those who may be seriously interested in the study of language without being 'linguists'. Unfortunately, some of us, who believe in 'calling a spade a spade', and who can distinguish between simplicity of 'presentation' and simplicity of 'issues', may not make the effort to give sufficient clarity to whatever they offer for publication, for fear of being accused of 'popularizing' the science of linguistics. Assuredly, many will agree that, particularly in an age suffering from inadequate communication, simplicity of presentation is most advantageous. It is bound to give to the science of linguistics more credit, not less credit. The dignity of any science lies in the strength of its combined modesty and clarity. If we do recognize the 'uniqueness' of the study of language, in that it could arouse the interest of users at all levels, then in addressing all those concerned, we should not limit ourselves to fellow linguists, who happen to belong to our school, and who speak our jargon. What if we opened out for a more generous sharing of ideas ? In the light of the diverse reactions, we could then perhaps determine more accurately the relative significance of the heterogeneous little pieces of research, usually undertaken in the absence of a comprehensive view of needs arising in academia and elsewhere. We could at least do our bit towards finding a solution to the problem of 'relevance' or rather 'irrelevance' in research. And we could probably help mitigate the detrimental effects of the ivory tower isolation, and the simultaneous alienation of those who really care for the fruit of academic labor in all fields of linguistic investigation.3 3

See Report by Albert H. Marckwardt, LSA Delegate to the American Council of Learned Societies, LSA Bulletin, April 1970, p. 29: "In general, the papers and discussions centred about such questions as ... the problem of relevance in research as well as in teaching, and with reference to the learned society as well as to the individual scholar. ... There was on the part of the conferees a serious concern over the unhappiness and alienation so frequently felt and expressed by students, both graduate and undergraduate, and younger members of the learned professions. This situation was generally recognized as arising from a whole complex of causes, calling not for a single, simple solution but rather a broad reorientation of the entire academic community; a process involving changes in values and long-held attitudes which may require years to achieve."

CONTENTS

Introduction

5

Acknowledgements

7

Prefatory

11

PART i : HYPOTHETICAL

1. In Brief

19

2. Operational Definitions

26

2.1. Integrative Approach

26

2.2. Organizational Information

28

2.3. Encoding-Decoding Process

31

3. Two Basic Assumptions

35

3.1. Communication Orientation

35

3.2. Psycholinguistic View of Structure

42

4. Cues and Organizational Operations : A Proposed Model

48

4.1. Correlation at the Projection Level

48

4.2. Correlation at the Integration Level

51

PART II : ANALYTIC AND SYNTHETIC

5. Procedure

59

6. Contouring

64

7. Accentuation

77

8. Transition

88

16

9. Combination

CONTENTS

107

10. Integration

121

APPENDIX I

135

Conclusion

137

APPENDIX II

145

Corpus Material

147

Bibliography

159

Index

167

PART I

HYPOTHETICAL

1

IN BRIEF

This is an exploratory psycholinguistic study of correlations between verbal and conceptual structures in language, consisting in an investigation of the verbal cues of organizational information, and the corresponding operations of sequential structure, utilized in the encoding-decoding process.1 The study is based on two fundamental assumptions. First, that within the general framework of verbal communication, language could be profitably examined with respect to its characteristics as a unified form of behavior. Second, that a psycholinguistic interpretation of the notion of 'grammar' would be needed to account for transformation between the structures of content and of expression; it would present one basic variety of categories and relationships, assumed to be cued or signalled differently at the various levels of verbal structure. It is in this sense that grammar is operationally defined in this work as 'organizational information'. 1 Although 'production' and 'perception' as two modes of language processing are not equated, they are viewed as belonging to one and the same process. However, the focus of attention in this study is on 'perception'. See Pribram, K. H., ed., Brain and Behavior, p. 105. Article by A.M. Liberman, F.S. Cooper, D.P. Shankweiler and M. Studdert- Kennedy, "Perception of the Speech Code": "We may suppose that most of the encoding occurs in the conversion from muscle contraction to shape of the articulatory tract. If so, the commands to the muscles, which precede that conversion, should bear a simple relation to the phonemic (and subphonemic) structure of the language. We assume therefore that the speech decoder works by referring the incoming speech to the commands that would be appropriate for its production. On this view, the decoding and encoding of speech are two aspects of the same process."

20

HYPOTHETICAL

Accordingly, an attempt is made to elucidate some of the parameters of correlation that make up such a grammar, to categorize the referred to type of information, and to observe the hypothetical operation of the corresponding sets of cues at different levels of the verbal perceptual process. The central hypothesis is that grammar, in this broad sense, is psychologically accountable; that it constitutes a certain type of information, which is basic to the perception of verbal and conceptual structures at large; that, as such, it is the grammar of transformation in verbal behavior. In terms of this hypothesis, and pending rigorous experimental evidence, certain correlations are suggested between grammatical operations, or more specifically, operations of sequential organization in perception, and sets of verbal items normally employed to signal the transformation between conceptual and verbal structures. The proposed synthesis is the outcome of some analytic exploration in message decoding, centred on the examination of a few interrelated assumptions, pertaining to the referred to correspondence between the content and the expression structures in language. The following model of correlation underlies both analysis and synthesis. First Premise: 'Contouring' and 'Accentuation' cues are presumed to operate at the 'Projection' level of verbal response. Second Premise: 'Transition', 'Combination', and 'Integration' cues are treated as pertaining to the 'Integration' level. Third Premise: These cues are two main types under each category, or in correspondence with each sequential operation: 'explicitly sequential', and 'implicitly combinatorial'; here to be referred to simply as 'syntactic' and 'contextual'. Fourth Premise: These two main types of cues would naturally operate 'simultaneously' at both levels of verbal structuring, namely 'projection' and 'integration'. The study has been suggested in the main by recommendations of the 1953 Summer Seminar on Psycholinguistics,2 and by applications to the process of verbal communication of pertinent insights 2

See C. Osgood and T. Sebeok, eds., Psycholinguistics,

1954.

IN BRIEF

21

from Information theory, Transformation theory, Contextual theory, and Tagmemic theory. Meanwhile, the psychology of verbal perception drawn upon in the study is not sensitive to the distinctions established between Association theory and Field theory, simply because both are needed to account for the phenomena under observation. Perception of the verbal message may, in part, be accounted for in terms of the succession of verbal items in a chain of psycho-linguistic discrete reflex units, i.e. by word association, which is semantically controlled by the context of communication, and structurally constrained by established patterns. Meanwhile, such perception may occur as a result of some adjustment to the verbal context in its totality, which is based on a grasp of configurations, rather than discrete units. Obviously, the two perceptual procedures are not mutually exclusive. As far as the postulated model of correlation is concerned, however, the psychological principles recognized as basic should present no problems. They are well established in the field, and cannot be subject to controversy or dispute. They relate to such phenomena as the significance of relationships among objects and events; selectivity of sensations in the process of perception; organization of these in the perceptual field; and the relative constancy of such organization. This will explain why the Gestalt psychology of visual perceptual organization is heavily drawn upon in the study. With respect to the fundamentals of visual perception, Gestalt points of emphasis have not been outdated, and they turn out to be very suggestive in the study of verbal behavior, particularly with respect to the structure of stimuli in message decoding, and the unit characteristics of such a structure.3 As to why 'visual' perception is resorted to in an 3

See Osgood and Sebeok, eds., p. 51: "On the input side of the equation, Gestalt psychologists have been most active in concern about units ... Units are segregated as self-integrated aspects of the environment, which stand out as figures on a more or less homogeneous ground. Figures are characterized by shaped boundedness (contour), dynamic properties (e.g. obedience to gestalt laws), and constancy. Many of the accepted empirical laws of perceptual organization have resulted from observations under the Gestalt impetus."

22

HYPOTHETICAL

investigation of verbal behavior, rather than 'auditory' perception, another quote from the same Report provides an answer. 4 Now that the substance of this research has been roughly sketched, and its broad dimensions outlined, it may be worthwhile to dwell on two points: the link with psychology at large in research termed 'psycholinguistic', and the reference to Transformation theory. Concerning the former, the author is fully aware of the fact that psychology is no area of competence for him. And yet he is convinced that if one is prepared to explore, one would have to go in 'where angels fear to tread'. Or else, how could linguistics ever be the 'inter-disciplinary' science we all 'talk' about. The same would apply to scholars in psychology, who might wish to explore the linguistic 'territory'. We have the best of intentions, and all we ask for is some indulgence, if we do not sound deep enough, or sufficiently sophisticated. Perhaps sophistication is not what is called for at this point. There has been a bit of understandable confusion resulting from using interchangeably the terms 'psycholinguistics', 'psychology of language', 'verbal behavior', and 'verbal learning'. The confusion comes from the natural overlapping in these areas, but it is still possible to make some useful distinctions. Psycholinguistics is 'linguistics' psychologically oriented. The psychology of language is 'psychology' that focuses on problems of language development, language competency, etc. Psycholinguistics and the study of 'verbal behavior' are more or less the same thing in terms of content. However, a researcher who considers himself a student of behavioral science, first and foremost, might prefer to use the latter term more often than otherwise. As for 'verbal learning', it may be set aside in view of the fact that it has to do with the 'learning process' as such. 4

See Osgood and Sebeok, eds., p. 51: "It is unfortunate for our present purposes that so much of the work on perception has been concerned with vision; it would probably be correct to say that over 90 per cent of the research here has dealt with one aspect or another of vision. Most of the work done in audition has dealt with sensory rather than perceptual processes. This being the case, we shall have to assume a general analogy between visual and auditory modalities."

IN BRIEF

23

Next comes the reference to Transformation theory in this study. Transformation, as a 'general' concept accommodated in many branches of science, answers the need to account for 'change', by adopting a 'process' view of the phenomena under observation. The Chomskian theory is a version of Transformation theory, applied to linguistics. It is a valuable contribution, with which we will presently reveal points of linkage. Work on transformation in linguistics has been going on for quite a while, probably without being recognized as such. Semanticists have been busy considering transformation from the substance of the content unit to the form of the content unit, from things and events to words via their perceived significates. Dealing with problems of reference, they have been constantly emphasizing the fact that the relationships between words and things are not direct, but 'mediated' through the existence of perceived 'referents' which are not absolute entities. On the other hand, structural linguists have concerned themselves with transformation from the substance of expression to the form of the expression unit; and that explains their enthusiasm for the revelation of the so-called 'allo-eme' relationship, which they have applied to linguistic structure at phonological, morphological, and syntactic levels, with a possible extension to the semantic 'component'. And yet the link between the 'structure of content' and the 'structure of expression' in the process of language 'production' is the big challenge facing the Chomskian G-Transformationalists at present. They have established their model of correlation between levels of verbalization. They assume the existence of basic kernel strings, as a starting point, and they work their way 'up' from there, considering how 'new' and probably more complex strings get generated, and how the transformations could be accounted for in terms of rules. Some interested psychologists have been working hard on transformational problems of this nature. They are convinced that the G-Transformational grammar model is most suited to psychological investigations, because, as we have noted earlier, the notion of transformation seeks to accommodate change.

24

HYPOTHETICAL

Now, where do we stand in this context? We are also concerned with the same phenomenon of transformation between conceptual and verbal structures. And yet, our model of correlation is different in two respects. First, it underlies transformation, not from hypothetical kernel sentences to actual utterances, or from a given 'string' to a new 'string'; but rather from the actual cues of sequential structure in verbalization to the broad categories of organizational information internalized through certain operations of perceptual organization. In other words, our point of focus is 'decoding', or the process of language perception, rather than that of language production. Second, the presented model is more fundamental in the sense that it underlies most of the GenerativeTransformational assumptions. To put it differently, if the grammatical categories in the G-T model are 'permutations', 'deletions', 'additions', etc., the underlying operations, from the point of view of a decoder, are presented as 'contouring', 'accentuation', 'transition', 'combination', and 'integration'. These operations would account for the existence of permutations, additions and so on. They would constitute more direct links with the psychology of perception, and as such, they would naturally strengthen the foundations of language universals.5 Finally, a word about the order of presentation may not be amiss. The main thesis that has just been outlined in Chapter 1 is gradually developed through the rest of Part I. In the interests of clarity, the key concepts are briefly and operationally defined in the order of their occurrence: 'integrative approach', 'organizational information', and 'encoding-decoding'. This makes up the content of Chapter 2. Then, by way of sketching the hypothetical content with the addition of more detail, Chapter 3 treats the orientation of integrative research in linguistics, and the unifying concept of 'organization', as applicable to both sides of the dicho5

Noam Chomsky, "Linguistic Theory", in Mark Lester, ed., Readings in Applied Transformational Grammar, p. 59: "... It seems to me likely that questions of this sort will dominate research in the coming years, and to hazard a further guess, that this research will show that certain highly abstract structures and highly specific principles of organization are characteristic of all human languages, etc. etc."

IN BRIEF

25

tomy: the structures of content and of expression. Actually, it deals with the two basic assumptions already stated, namely: that the communication orientation is suited to the synthetic study of verbal behavior; and that the psycholinguistic view of structure encompasses the totality of message organization in the encodingdecoding process. Chapter 4 then sketches the suggested model of correlation between sets of verbal cues and certain operations of sequential organization. Part II starts with an account of the analytic procedure, dwelling on its nature as determined by aims and objectives, on its worth and its limitations under exploratory conditions. The data, which consist of excerpts from a recorded panel discussion, are briefly described, together with the method to be followed in processing, prior to the actual testing of assumptions. This covers the whole of Chapter 5. Chapters 6 through 10 are devoted to discussions of the five grammatical operations, focused on the assumed correspondence between cues and categories of organizational information. Interpretations are given to the tentative empirical evidence available, and comments are made on the possible amount of support that further investigations might provide. Accordingly, the thesis outlined in Part I receives its full amplification in Part II, with the inclusion of 'topical' or incidental references, suggested by the nature of the data or the tentative results of the analysis. An attempt is made to illustrate the fact that what may be contributed by an integrative study pertains both to the nature of a research hypothesis, and to the interpretation of research findings.

2 OPERATIONAL DEFINITIONS

2.1.

INTEGRATIVE APPROACH

Through the diversity of analytic research in linguistic structure largely aimed at the discovery of basic units, such notions as 'phoneme', 'morpheme', 'tagmeme', and subsequently 'sememe', have been established within a general framework, where language is approached as a 'code' and a 'system'. However, once the focus is shifted from 'anatomy' to 'physiology', from dissection to integration, things start to look hazy, and one encounters plenty of reference to how the different systematic aspects of language 'interlock' and how they 'interact', without enough empirical evidence of such interlocking or interaction. The reason, of course, is not a lack of empirical evidence, but insufficient awareness of the specific orientation of linguistic study, within which language could reflect these 'dynamics'. A notion like 'interaction' is simply inconceivable without reference to language as a form of 'behavior'. It is unaccountable out of the context of 'perceptual organization'. Accordingly, it is necessary to emphasize that, in verbal perception, the structures of content and of expression are inseparable: the former consisting of an organization of experiential data — auditory, visual, and other sense data; the latter comprising an organization of speech data along a time dimension, as auditory sense data. Clearly, the unifying phenomenon, whereby these two types of structure are correlated, is one of perceptual organization. Therefore, it is suggested that, from a psycholinguistic angle, it would be profitable to explore the parameters of correlation,

OPERATIONAL DEFINITIONS

27

which make it possible for language users to transform one type of organization into the other in the process of verbal communication. In a sense, this is viewed as an operation similar to the transformation of sound waves into electric waves in radio transmission, and then back into sound waves in reception. The two types of structure generated in the process are not identical, but there is enough of a correspondence between them to make them mutually convertible. With the turn of the 19th century and beginning of the 20th, the main stream of linguistic research was more or less limited to the structure of expression, which had to be maintained in its own rights, independently of the structure of content. Accordingly, analytic research in what came to be called 'descriptive' linguistics was confined to two major organizational categories: phonematization and grammatization, of course, in a sense limited to the nature of speech as a succession of discrete units. Now this type of approach had the virtue of a sharp focus, a lot of scrutinizing and attention to detail, meticulous classification, and rigorous manipulation of the data under observation. And yet, it has proved to be lop-sided, unrealistic, and incomplete. Consequently, a move in the direction of 'integration' is now by far the most formidable challenge in modern linguistics. Hopefully, we are better clarified now than we were then on what makes science scientific, on the unique characteristics of the social and behavioral sciences, which make them different from the physical and natural sciences. We are less skeptical of, probably more accommodating to, a possible combination of the mechanistic and mentalistic orientations in a fashion, as well as of the atomistic and wholistic view points. Accordingly, the 'integrative' approach in linguistics would be one that seeks correlation between content and expression structures, through comprehensive investigations into the common denominator, namely the structuring phenomenon, and its applicability at these two levels in the main. It is suggested that, whether we concern ourselves with transformation from the substance of content to the form of content; from the substance of expression to the form of expression; or whether we are interested in the corres-

28

HYPOTHETICAL

pondence between both content and expression structures in verbal communication, we might all benefit by research into what constitutes grammaticalness in the broadest sense of 'organization'. The underlying assumption is that a psychologically accountable grammar would provide a unifying concept, whereby the patterning of content and expression in language can be linked.

2.2.

ORGANIZATIONAL INFORMATION

This is the name given to the basic grammatical relationships utilized in verbal perception, and presented as simultaneously applicable to the structuring of message content and message form. 1 The sequential relationships with which we are mainly concerned seem to fall within two major categories: relationships of 'positional' organization, and relationships of 'substitutional' organization. The terms 'syntagmatic' and 'paradigmatic' have often been used to label such relationships with reference to the 'expression' structure. For instance, syntagmatic relationships would apply to sequences like the phonological 'consonantvowel-consonant', the morphological 'root-stem', or the syntactic 'noun-verb-noun'. What is particularly worthy of attention, however, is the fact that at the content level of structuring, the same relationships are equally operative. In evidence, consider such sequences as 'subject-predicate', or 'modifier-modified'. These sequences have been persistent enough to appear in almost all grammatical descriptions, even under austere mechanist conditions, forbidding any reference to the content side of language. Obviously, when we talk about 'subject' and 'predicate', we are simply using other names for 'operator' and 'operation', or 'actor' and 'action', in describing the organization of 'content' units. Similarly, the difference between the 'modifier1

In dealing with the configurational aspects of perception, the terms 'form' and 'structure' are sometimes used synonymously, and sometimes with some discrimination as to whether the notion of 'constituents' is implied, in which case, 'structure' would be preferable.

OPERATIONAL DEFINITIONS

29

modified' relationship and, say, the 'adjective-noun', or 'nounnoun' relationship, is one between a description of the content structure and a description of the expression structure. On the other hand, 'paradigmatic' relationships would 'normally' apply to the patterns of inflection for such form classes as nouns, verbs, etc., with possible extensions to include all instances of variation at the phonological or syntactic levels. In other words, it is possible to think of 'allophones' in distribution as constituting the paradigm or the 'inflectional pattern' of the phoneme. Similarly, it is possible to consider 'allotagmas' in distribution as making up the paradigm or the 'inflectional' pattern of the tagmeme. By analogy, further extension of the same concept of the paradigm is presented in this study as applicable to the 'structure of content', 2 just as well. It proves serviceable in demonstrating the basic unity of language, through the necessary incorporation of the semantic component, as represented by lexical items. An additional value of such an extension is the disclosure of the role of the 'semantic paradigm' in sequential organization. In brief, both dimensions of 'positional' and 'substitutional' organization are seen operative in the criss-cross of verbalization and conceptualization at various levels.3 As for the notion of 'information' used in this study, it is simply an extension of the denotation in the direction of its normal usage, which is, of course, different from the meaning of the same term in Information theory. The 'information' we are interested in is 'wanted' information. It is conversely related to randomness and haziness of contour, negatively related to entropy and uncertainty. In other words, the more structured the message, the more 'informative' it becomes; moreover, the referred to organizational 2 The 'structure of content' is in actuality an 'organization' of content units at the message level, to be distinguished from the so-called 'semantic structure', which consists in a summation of the organizations of content units in messages produced by members of a speech community. See Charles F. Hockett A Course in Modem Linguistics, p. 142. 3 This extension is quite in accord with the principle of linguistic duality eloquently expounded by Gustav Herdan. See: Gustav Herdan, Language as Choice and Chance, Part. IV.

30

HYPOTHETICAL

information is presumed to influence different aspects of the structuring phenomenon, not just transition, or knowledge of subsequent events on the basis of antecedent events. Subsequent discussion of the presented model will show the entire scope of operation. The proposed organizational operations, namely 'contouring', 'accentuation', 'transition', 'combination', and 'integration' are assumed to provide the 'parameters of correlation' between the two major types of structure employed in verbal communication. The reason underlying the assumption is simply their applicability to both. A language user, whether encoding or decoding, would need to contour the content of a verbal message by stages, depending on how complex or involved that message may be; he would want to accentuate certain segments, make transitions, combinations, insertions, etc. Meanwhile, there are sets of 'verbal cues', which he could manipulate in responding to, or signalling, those organizational aspects of the content structure. Tentatively, let us determine what we mean by 'verbal cues'. The term 'verbal' is used to mean two complementary things: first, as pertaining to the overt structures of speech, while correlated with the covert structures of conceptualization; second, in the sense of oral, rather than written. In this latter connection, the prosodic features of pitch, stress, and juncture are sometimes referred to as 'non-verbal' or supra-segmental. However, these features are here considered part and parcel of the 'segmental' units, whether these units are phonemes, morphemes, words, or even word combinations; 4 what makes speech a continuum of events are those hierarchies of configurational features, which give it its recognizable constitution. Accordingly, the term 'non-verbal' would be used with reference to gestural and other signalling phenomena in the face-to-face communication situation. Obviously, an exchange of hints on the gestural-visual band, for instance, could be utilized in reinforcing those cues delivered on the vocal-auditory band for the transmission of organizational information. Still, the present 4

Reference is here made to psycholinguistic units.

OPERATIONAL DEFINITIONS

31

investigation has to have limits, which encompass only those sequential relationships between what is externalized in 'speech' and what is internalized in perception. As for the term 'cue', it is also employed in a dual capacity: first, to mean stimulus, anything that excites to action in response; second, to mean carrier, here information carrier, or transmitter, to intimate and suggest, rather than actually 'transport' patterning information. Probably what needs to be spelled out in this connection is the notion of 'mediation' between a cue as stimulus and the potential or calculated response it is supposed to elicit. A verbal cue is only a hint that ripples into further hints, which finally trigger such a response. Should the same item occur twice in a message, it would act as two different cues, reinforcing each other in establishing certain relationships at one or more levels of perceptual organization.

2.3.

ENCODING-DECODING PROCESS

Before we get into a tentative explanation of what is involved in this process, let us first decide what it is that a language user puts into a code, or converts back from a code; then what major types of information are made available by a code. In terms of communication theory, the object of encoding for speech production and decoding for speech comprehension is a 'message' that has content and that has form; the content tells what it is about, and the form shows how it is put together both verbally and conceptually. Now a verbal message is normally defined as that part of the total speech output of a source, which, under specific circumstances both personal and situational, has been attended to, comprehended, interpreted, and acted5 upon by a receiver. Pertinent to our research at this point are three message characteristics. First, that both as content and as expression, a message must be structured. Second, that whether as output with a source or as input with a receiver, a message must be structured. Third, that the information transmitted by a message, as derived from the code in use, is both 5

This is 'mental' rather than 'practical' action.

32

HYPOTHETICAL

representational and organizational. It just so happens that we are here interested in the latter type of information. Through the activity of the central nervous system, it is possible for humans to engage in a process whereby representational and organizational patterns of stimulation can be converted into auditory speech patterns, and back from these into representational and organizational patterns in response. This is the encodingdecoding process. To put it differently, whether with one and the same person, with two or more, it is the process of transformation of an organized input of psychological constructs into an organized output of speech sound waves; and back from the latter to the former. What makes this progression of events a 'process' is not just circularity of occurrence, but circularity on more than one plane. On the one hand, we would need to remember that the perceptual organization of 'stored information' suggests a 'selection' of items from the total sensory input, and a 'patterning' of such items for future reference; which reference will determine subsequent 'selections' and subsequent 'patterning', and so on and so forth ad infinitum. On the other hand, when it comes to verbal communication of data from the information in stock, another form of circularity takes place. It is commonplace to mention that every language encoder is simultaneously a decoder, and that is fundamentally how he makes his 'appropriate' choices in the course of message production. Conversely, every language decoder is simultaneously a potential encoder, and that is how he can afford to be 'selective' in his turn, in the course of message reception. Now what is particularly germain to our study is the concept of transformation in this context, and how it applies to organizational information. The thesis maintained here is that in order for patterning to be transformational, it has to be based on psychological principles of organization; in which case, the same notions would operate alternately between a structure of content and a structure of expression, in message production as well as in message comprehension.6 6

B. Malmberg, Structural Linguistics and Hitman Communication, p. 25: "The encoding that takes place in the brain of the sender when the extra-

OPERATIONAL DEFINITIONS

33

At this point, it is appropriate to consider the general levels of organization of linguistic response. Charles Osgood distinguishes three: a Skill level; an Integrational level; and a Representational level.7 In correspondence with these, are three levels of organization of sensory input: 8 1. Projection level, having to do with the 'summation of points of maximal stimulation, and suppression of other activity'.9 2. Integrational level, having to do with the 'ordering of semantic units, concrete-relational classification of these, concordai agreement, and certain other relational phenomena'. 10 3. Representational level, with which this enquiry is not directly concerned. Information as to what the content of a message is, though related to, is still different from information as to the organization of such a content, and how this latter type of information is verbally cued. Accordingly, the first two levels are elaborated upon in the present investigation. Projection and Integration are assumed to cover the notion of grammaticalness in the broadest sense: the grammar of experiential data, and the grammar of verbal data. Whether we 'encode' or 'decode' messages, we are assumed to project structures and to integrate structures; and in so doing we unconsciously apply psychological principles, which could work for 'conceptualization' as well as for 'verbalization'. There are different interpretations of 'how' we could perform these complex operations, as we attempt to comprehend and produce verbal messages. If we are interested in one such interpretation, here is what the communication engineer is presently offering. A theory of Template Matching. According to a 'feature selection procedure', a human being would perhaps be able to project a configuration, which linguistic phenomena are linguistically structured, as well as the decoding which precedes the interpretation of the message in the brain of the receiver (as illustrated in the scheme below), are particular instances of a more general procedure called conceptualization, or concept formation." 7 See Osgood and Sebeok, eds., PsychoHnguistics, p. 95. 8 See Osgood and Sebeok, eds., PsychoHnguistics, p. 54. 9 See Osgood and Sebeok, eds., PsychoHnguistics, p. 54. 10 See Osgood and Sebeok, eds., PsychoHnguistics, p. 95.

34

HYPOTHETICAL

through subsequent integrative activity, would create a 'simulation', whose function could be that of the 'template' in template matching. We do not have to accept this theory for humans, if we do not like it; but it might be interesting. At least, we are generally agreed on this: that human beings perform the task of pattern projection in both production and recognition of verbal stimulae; and that the integrative quality of input or output is always determined by preconceived data in storage. It is no news today that machines designed for pattern recognition are those programmed to read printed characters, classify electrocardiograms, reorganize some spoken words, sort photographs, etc. The key to all this is the concept of organization. Of course, mechanical operations have their limitations; and a human being is still a different thing. But some psychological principles seem to work in application to programming; and that alone should provide some confirmation.

3 TWO BASIC ASSUMPTIONS

3.1.

COMMUNICATION ORIENTATION

"That within the general framework of verbal communication, language could be profitably examined with respect to its characteristics as a unified form of behavior."

It will be recollected that this is one of the two basic assumptions stated in Chapter 1, and adduced as pertinent to the thesis under investigation. Through our preliminary discussion of the integrative approach in Chapter 2, the essential unity of language has been very briefly touched upon. We have noted how in the course of language production or language comprehension, the two structures of content and expression do interact; how it is therefore necessary to attempt correlations of these interacting structures in 'synthetic' research; and how such attempts would eventually help maintain the integrity of linguistic science, and make it more meaningful to all 'students of language'. It may now be appropriate to dwell on the salient features of the communication orientation, to find out how much help we could get from it, in terms of the unified perspective on language. By far the most powerful notion that keeps cropping up in almost any serious discourse about language, without necessarily being fully exploited or utilized, as indeed it should be, is that of 'communication'. Undoubtedly, there is a general awareness of its importance, and yet, because this awareness seems to be rather vague or inarticulate, the student of linguistics is not normally taken far beyond the usual reference to language as a 'means of interaction' among humans. Socially or culturally oriented research

36

HYPOTHETICAL

would naturally capitalize on the 'instrumentality' of language as a communication system, and concentrate on its role or function in social contacts as well as in the creation and perpetuation of culture. Psycholinguistic research, however, would focus on the 'dynamics' of communication in language as a form of behavior, on the processes and operations underlying both phenomena of speech perception and speech production. Without some understanding of these phenomena, we should not really expect any more explicit accounts of the full impact of the concept of 'communication' on that of 'language'. We will either keep dwelling on the instrumentality of language in general terms or else go into detailed accounts of 'mechanical' communication, suggesting analogical inferences for the 'human' process at different levels of social contact, and offering nothing beyond a 'footnote' on language, that does not seem to justify the pains taken to present such intricate accounts. Therefore, there is no need here to go through the whole exposition of an SMCR model, mathematical or otherwise.1 We will simply hint at the referred to 'dynamics' of communication in the behavioral phenomena we call 'language', and discuss some of the implications of this view for the treatment of language as a unified whole. As already noted, at the 'inter-personal' level of communication, language is often described as a means of 'interaction' between a source and a receiver. And yet the full implication of this description is rarely reckoned in the actual treatment of linguistic phenomena under observation. In a 'static' stimulus-response model, representing only one way of this circular process of 'interaction', the (r-s) sequence is the verbal act between (S) and (R) in a practical situation, where (r) stands for the production of utterances by a speaker, while (s) stands for the perception of these utterances.2 1

For this and similar models, see David K. Berlo, The Process of Communication; C. Shannon and W. Weaver, The Mathematical Theory of Communication; and B. Malmberg, Structural Linguistics and Human Communication. 2 C.C. Fries, The Structure of English, pp. 32-34. Fries's reference to his S-r-s-R 'formula' seeks to steer away from any psychological implication, and it is difficult to tell what he means by 'strictly schematic' in: "This formula must not be taken as representing any psychological assertions or assumptions

TWO BASIC ASSUMPTIONS

37

In actuality, however, language consists in a mediation process between 'alternating' sides, not 'static' sides. Without due consideration given to the concept of 'alternation', it would be meaningless to talk of interaction in verbal behavior. In order to study language from this perspective, it is therefore necessary that the full verbal intercourse should be re-discovered in research. When a sufficiently broad context in dialogue form is available, affording ample chance to observe the verbal message exchange, it then becomes feasible to identify and account for the two major types of verbal cues we are concerned with at this point: those that may be correlated with how language is produced, and those that may be correlated with how it is perceived. Then, and only then, could the 'process' view be maintained in screening the interaction between source and receiver engaged in an act of verbal communication. Dialogues abound in cues suggestive of this 'living' side of language. And yet they have not been warranted sufficient attention, because linguistic research is still anchored in code-level abstractions; because it is still bound, cribbed and confined within sentence limits, to say nothing of the 'simple, affirmative, declarative' sentence that has been granted fundamental status as a basic unit at this level of structure; and because there is a lot of talk about 'interaction' out of the context of interaction.3 In order to account for the animated aspect of verbal behavior, in order to correlate language perception with language production, those cues would have to be identified and adequately described in the appropriate terminology, of the whatever. It is strictly schematic and limited to the particular point dealt with here — the representation of certain broad aspects of the functioning of language." 3 It seems that the narrow interpretation of de Saussure's 'object of linguistics' is still in sway. See F. de Saussure, Course in General Linguistics, p. 232: "From the incursions we have made into the borderlands of our science, one lesson stands out. It is wholly negative, but is all the more interesting because it agrees with the fundamental idea of this course: the true and unique objects of linguistics is language studied in and for itself." The necessary addition would be that language "in and for itself" is a matter of 'focus' that could not possibly delete 'connections', without which the 'thing' we call language ceases to 'be' what it is.

38

HYPOTHETICAL

psychology of perception and of individual 'adjustment' to the speech situation in its entirety. As for research on 'sentences' as such, or on separate linguistic items under one grammatical category or another, it is only fair to state that it has nothing to do with the dynamics of verbal communication, nothing to do with language alive. This is not to underestimate the significance of such research, but only to emphasize that it is not conducive to the integration between language as produced and language as perceived; that it does not subscribe to the totality of language as 'behavior' between source and destination. Obviously, work currently done on transformation is enough evidence that, at the 'intra-personal' level of communication, the dynamics of adjustment and interaction have been recognized in a fashion. 4 How 'sharp' the focus is on these dynamics, and how 'realistic' the available accounts may be — these are significant questions, for sure. What is relevant to our present investigation, however, is the increasing awareness of the dynamics of interaction within the individual communication system, and the attempts made to identify cues in the 'surface' structure that may be correlated with the different phases or 'appearances' of such interaction at 'deeper' levels of structure. Hence, the communication orientation would not limit one's view to the observation of those manifestations of interaction between one source and a different receiver; it would basically emcompass the 'primary' mediation process that goes on between stimulation and response in verbal behavior within each individual communicator. 5 One implication in this connection is that research in transformation should include both encoding and decoding. So far, the focus is on 'encoding', on production, or rather on the process of intrapersonal communication whereby conceptual structures get con4

The notion of 'adjustment' is here limited to whatever modifications are deemed necessary to transform verbal behavior at the point of initiation or reception; subsequently, however, reference will be made to the tension-release or cathartic function, which is the net result of successfully accomplishing the desired modifications. 5 See Charles Osgood and J.J. Jenkins, "A Psycholinguistic Analysis of Decoding and Encoding", Suppl. J. Abnorm. Soc. Psychol.

TWO BASIC ASSUMPTIONS

39

verted into verbal structures; but not on the reverse process. Even though encoding and decoding cannot be separated, the fact is that they are different operations. Therefore, a more comprehensive treatment of transformation would have to include 'decoding', and establish adequate links with the psychology of perception and the psychology of adjustment. Needless to say, recognition of this process of intra-personal interaction for speech production and speech perception is basic to the understanding of verbal inter-personal communication. Consider, for instance, one's decision as to whether a certain meaning is adequately expressed at one point in the course of conversation with an audience. If there is a conflict between what has just been said and what was initially intended, one says something like, "That is not what I meant", and injects another expression or a series of expressions to elucidate, or even totally substitute the original statement. This internal censorship occurs on the basis of and in response to feedback information, without which the communicator could be thrown out of the balance of adjustment that keeps him in effective performance. Similarly, the receiver of such communication would be expected to make his own selections, organizations, and alterations in the course of perception. Sometimes this is indicated to the source in the form of 'interruptions' or 'interjections', signalling his preference for a certain item to substitute one provided by the source, or even completing an utterance for him the way he expects to perceive it. The problem, however, is that once 'perception' is verbally signalled, it turns into 'production'. The circularity of the process is such that one's indices to the nature of 'decoding' are themselves manifestations of 'encoding'. 6 As for the homeostatic function served by verbalization as a tool 6

J.P.B. Allen and Paul Van Buren, eds., Chomsky: Selected Readings, p. 132: "I have never seen a precise characterization of a grammar for the encoder or a grammar for the decoder that was not convertible, by a notational change, into the other. Furtheimore, this is not surprising. The grammars that linguists construct are, in fact, quite neutral as between speaker and hearer. The problems of constructing models of performance, for the speaker and the hearer, incorporating these grammars, are rather similar."

40

HYPOTHETICAL

of individual adjustment, it would be inconceivable to study language as verbal behavior, without reference to its effect on the state of balance or imbalance for each user.7 Again, modern linguistic research seems to shy away from investigating verbal expressions along this dimension, most probably on the assumption that matters related to individual mental states are subjective, elusive phenomena that cannot be rigorously controlled for objective observation. Therefore, they are often set aside for the psychologist to pick up, if he wants. One would insist, however, that once the 'intra-personal' dimension in verbal communication is bypassed or underestimated, some problems of language description — even if we are only after language description — are bound to arise; and ways of tackling these problems will of necessity be inadequate. For one example, consider the class of utterance known as 'exclamations'. These utterances certainly communicate meaning. However, because the communicated meaning inside one and the same nervous system cannot be identified as 'eliciting regular, predictable responses' in others, this whole class has been simply dismissed with the name 'non-communicative utterances'. 8 Nothing is farther from actuality. The communicated meaning is simply intra-personal rather than inter-personal, and the communicative function is sometimes probably related more to adjustment for inner balance than to adjustment for mutual interaction with others, though the two are often hard to separate. Finally, within the framework of communication, the dynamics of both interpersonal and intrapersonal contacts are interrelated through the 'process' mobility of those operations that constitute the totality of verbal behavior. Once its impact is sufficiently felt, the very notion of 'process' should prove helpful in solving some of the problems pinned on the artificial antitheses responsible for distorted views on 'language as X' VERSUS 'language as Y', or on what is 'linguistic' VERSUS what is 'exolinguistic'. Reference is here made to some of the cherished dichotomies that have been 7 For further information on dissonance reduction, see Leon Festinger, A Theory of Cognitive Dissonance. 8 C.C. Fries, p. 52.

TWO BASIC ASSUMPTIONS

41

'literally' adhered to long enough to condition one's thinking and limit one's span of vision, even today. In the absence of a basic unifying orientation for the exact study of language, it was natural that the purport of these distinctions should be slanted to provide room for 'scholastic' division and segregation. Now that we have had more than enough of those varieties of linguistic research largely dictated by the demands of such dichotomies, it is assumed that the process view would contribute greatly to the integration of language and linguistics, through the formulation of new premises showing the sides of each of the referred to dichotomies in their true perspective, not in terms of polarity, but of circularity. Even the most recent distinction made between language as 'performance' and language as 'competence' could be thus abused, unless the circular relationship between the two sides of the dichotomy is made manifest. Clearly, the distinction is quite valid; it is meant to underscore the difference between what we do when we verbalize on the spur of actual communication situations, and what we do throughout our lives, building up our individual 'idiolects' through interaction with other members of the speech community, as well as with ourselves. One level of operation consists in a summation of occurrences on the other level; both levels of operation consist in an amalgam of intra-personal and interpersonal communication. However, as a result of textbook oversimplifications in attempting to explain the occurrence of error in speech as related to 'performance', and not to 'competence', one is sometimes put under the impression that competence is probably a little daemon-like entity, that is not subject to the oscillations of performance, sometimes manifested by error. 9 A moment's reflection, however, will show that if verbal communication is essentially the encoding-decoding process, characterized by the 9

Ronald W. Langacker, Language and its Structure, p. 35. Langacker seems to equate 'competence' with the 'structure of a language', and argues that "the structure of a language is not affected when its speakers make mistakes in talking any more than a symphony is affected when it is poorly performed". What is said of the 'structure of a language' is definitely true, but 'one's competence' is still something else; it is 'personal'; it consists in one's internalized knowledge of the structure of a language.

42

HYPOTHETICAL

previously alluded to circularity of operation simultaneously within the individual and outside of him, then what is called 'competence' simply shares the characteristic feature of all 'abstractions', namely 'relative constancy' amidst the flux of variant manifestations. It is built up the same way, through the dynamics of selection and organization; and consequently, it keeps on affecting and being affected by the individual's subsequent selections and organizations at the 'performance' level. This final implication of the process view of verbal events actually sums up all the rest. From this vantage point, the verbal message can be treated as the embodiment of a variety of 'cues' relevant to the study of language in its entire complexity: an ongoing, fluctuating phenomenon, simultaneously suggestive of the state of individual adjustment and group interaction. The verbal message can thus provide a point of focus, where different lines of linguistic investigation may be seen to converge, provided they are all pursued with equal emphasis on the concept of language as behavior and on the dynamics of communication as fundamentally characteristic of such behavior. 3.2.

PSYCHOLINGUISTIC VIEW OF STRUCTURE

"Second, that a psycholinguistic interpretation of the notion of 'grammar' would be needed to account for transformation between the structures of content and expression; it would present one basic variety of categories and relationships, to be cued or signalled differently at the various levels of structure." 1 0

This assumption is meant to furnish some guidelines for the adopted view of grammar as 'organizational information', and to show how the breadth of this view is consonant with its application to the totality of discourse, as structured content in structured form, or expression. In the early Fifties, when linguistic science was 10

Courtesy of the Northeast Modern Language Association Newsletter (Vol. II, No. 3, Oct., 1970), where the same assumption is discussed in an article of mine under the title: "Psycholinguistic Insights into the Notion of Structure".

TWO BASIC ASSUMPTIONS

43

identified with 'structural linguistics', and when the latter was almost exclusively limited to the study of the phonogrammatic constituents of what the Generative-Transformationalists now call 'surface structure', enthusiastic proponents of 'Structuralism' were not interested in correlating 'surface' structure with anything but itself. They had their misgivings about any reference to 'meaning' as entailing a subjectivity of approach, which could not be followed at a risk of losing the rigor of scientific investigation, as was characteristic of their research. 11 However, it would be unfair to claim, as indeed some people do, that the Structuralists, as a rule, underestimated the 'semantic' structure in the study of language, or that they were unaware of the role of syntactic variation in the expression of meaning. I think all we can say in essence about their rather sharp focus on the expression structure, is that it left out a substantial area for subsequent investigation, namely that occupied by relationships between sound and meaning. And yet, the Structuralists did point towards a more comprehensive theory of language, simply through upholding the concept of 'structure', which is now more fully utilized, upon the realization that it provides the basic parameter for all integrative research in linguistics. 12 Today, when 'Structuralism' is assumed to be fading out of existence, it is the author's contention that it is not; that it is simply undergoing a process of development in a new key, covering correlation between two major level organizations. The present focus on Generative-Transformational grammar, on Parametric theory, and on Contextual theory at large in the study of human behavior marks a firmer grasp of the full implications of the notion of 'structure', based on a discovery of its psychological foundations, 11

C.L. Ebeling, Linguistic Units, p. 83: "It is certainly possible to establish phonemic systems without having recourse to meaning at all. In fact, I have tried to do so in the present study." 12 Noam Chomsky, Language and Mind, p. 23: "The person who has acquired knowledge of a language has internalized a system of rules that relate sound and meaning in a particular way. The linguist constructing a grammar of a language is in effect proposing a hypothesis concerning this internalized system."

44

HYPOTHETICAL

and aimed at the necessary unification of the theory of language as form inseparable from meaning. Accordingly, what is 'Grammar' after all ? Let us not persuade ourselves that grammar stands for the package containing knowledge or description of the two linguistic systems known as 'morphology' and 'syntax'. If we did, we would simply be following a conventional attempt at the classification and naming of data, that leaves out 'phonology', even though it has organizational manifestations, to say nothing of the conceptual structure with which these manifestations are correlated. Meanwhile, let us not think of grammar as that entity which has been labeled 'Traditional', 'Structural', and 'Generative-Transformational' in some sort of chronological order; we are not here concerned with an evaluation of the salient features characteristic of these 'approaches' to the study of grammar. For our purpose, it would be more worthwhile to dwell on the 'nature' of grammar and give it such a definition as would cut across categories of the meaning-form composite and inter-relate them within the total framework of verbal communication. Some such definition would seek to articulate an 'organizational variety of information' underlying all manifestations of patterning in verbal behavior; it would stress the correlation of elements in whatever form they may be conceived, as sensory input or output. To put it differently, in as much as it pertains to human perception, grammar would stand for those organizational phenomena that help one structure the elements of a perceptual field. As such, the concept of grammar would simultaneously explain the patterning of message content and its verbal manifestation, account for both representation, or sign-significate relationships, and organization, or sign-sign relationships. From a communication point of view, a message, whether transmitted or received, must be structured both semantically and verbally. The verbal constituents of the expression structure are usually recognized as 'lexical' items in 'syntactic' combinations, which are aurally perceived through 'phonological' manifestation. The adopted view of grammar will reflect the interaction and overlapping among all three constituents, or rather aspects of the

TWO BASIC ASSUMPTIONS

45

expression structure. Accordingly, lexical items may fulfill representational 'and', 'or' organizational functions. It so happens that relational items like 'conjunctions', 'prepositions', etc. are sometimes referred to as 'pseudo-words', simply because their function is syntactic; they are not considered 'lexical' items proper, as it were. Conversely, lexical items of the variety referred to as 'content' words are not normally conceived of as fulfilling any relational or organizational function in a verbal context. We will see how they do. At this point, however, it is pertinent to reiterate that the distinction between 'representation' and 'organization' is not meant to place lexical items in water-tight compartments. A content word in a verbal context must of necessity serve a structural function; conversely, the same applies to a structure word, which must have a referential value. In either case, a message decoder has to synthesize a 'construct' of elements and project a configuration based on how the particular linguistic item operates in this dual capacity, once it is combined with other items in a verbal context. Obviously, there can be no representation without organization. This is the circularity of functions which assumes special significance in an integrative approach. The utterance 'husband the' does not reflect a meaningful organization in English speech. Reverse the order of these words, and at once 'things' become more meaningful. Now, here is an interesting question. Is meaningfulness attached to syntactic grouping as such, or to the semantic outcome of this syntactic grouping, or to both ? In this connection, it might be refreshing to reflect on the superficial controversy between extremists; those who believe in the supersession of form, and those who believe in the supersession of meaning; those inclined to strip syntax of all meaning, and those who think of syntax as the outcome of a semantic 'melange'. Obviously, the fact that the elements of syntax are formative and relational features does not nullify the fact that these features are correlated with representations of semantic constructs. Meanwhile, the fact that syntactic grouping is strictly coded, thereby commanding automatic response from force of habit, does not cancel out the simultaneous choices made at the level of

46

HYPOTHETICAL

semantic representation. Therefore, even though syntactic grouping may have meaning 'per se' at one level, as may be demonstrated through the use of nonsense words in grammatical utterances, yet it undoubtedly becomes more meaningful, when coupled with semantic grouping. If syntactic grouping has meaning for us and the meaning is structural, semantic grouping also has meaning, and the meaning is structural. So the case could not be one of syntax without meaning versus syntax with meaning. This is sheer contradiction in terms. Semantic and syntactic structures would have to interlock, because they are actually two aspects of 'one' structure. Finally, let us take a look at the structure of language as a 'phonological' product. It has been claimed by extremists again that, at this level of analysis, language could be 'profitably studied' without reference to meaning. However, in terms of the proposed integrative approach to language as behavior, as long as the phonological aspect of language is strictly coded, as long as it is 'psychologically' determined, it would be worth our while, not only to describe constituent elements and note their distribution, but also to study the dynamics of combination in the light of the psychological factors that make sound correlations 'intelligible', or meaningful. Whether segmental or sequential, the distinctive features of speech are recognized and produced according to structural principles of correlation. As we will find out in more detail, one's awareness of these features may be accounted for in terms of a developed sensitivity to occurrences involving the operation of such organizational principles as 'similarity', 'contiguity', 'contrast', etc. Accordingly, in English the phonemes /s/, /z/, and /§/, /z/, for instance, would get correlated in one's perception as voiceless 'versus' voiced counterparts, sharing the 'same' points of articulation. Here is a phonological meaningful relationship at the segmental level, in which both 'similarity' and 'contrast' are utilized. In the same way, an intonation contour is recognized as rising or falling, only in comparison with preceding points of that same contour. We will observe the significance of such phenomena, as we examine other prosodic features in their assumed power of correlation with certain categories of organizational information.

TWO BASIC ASSUMPTIONS

47

In brief, we have observed how the lexical components of our utterances mean little or nothing as individual constructs; they must be combined with other constructs, or rather put into some syntactic relationship to reduce their haziness in perception and sharpen their contours. Meanwhile, syntactic features mean little or nothing as empty correlatives; they must 'carry' the semantic load, for which messages are encoded and decoded in communication. As for the phonological components of the same structure, they too could be utterly meaningless, without the necessary correlations that bind 'tokens' in their multiplicity under each 'type' of construct, without the patterns underlying their controlled distribution at the different levels of concatenation. This quick glance at the different aspects of the expression structure from a decoder's point of view would probably suggest at least two basic considerations. First, that linguistic structure as perceived at the level of performance does not simply correlate the two sides of the dichotomy of 'expression' and 'content', but more importantly, that it does so through a considerable amount of overlapping among its lexical, syntactic, and phonological aspects. Second, that the observable overlapping and circularity point to the common denominators we are after, namely psychological principles of organization.

4

CUES AND ORGANIZATIONAL OPERATIONS: A PROPOSED MODEL

In terms of code-level abstractions, the structure of message content has been referred to in linguistic research as 'conceptual' structure, 'deep' structure, and 'semantic' structure. Setting aside irrelevant distinctions at this point, all we need to remember is that 'whatever' we choose to communicate to others must be 'organized' at both ends of transmission and reception. Our major hypothesis in this study has to do with the information provided by certain organizational operations, and how this information is transformationally cued 'up on the surface' of the verbalized product for subsequent recognition and interpretation in decoding. In this chapter, we will have a three-fold task, consisting of: a brief exposition of what each grammatical operation means in the broadest terms; a hint at its accountability in the psychology of perceptual organization; and some reference to its hypothetical 'correlation' with certain features of the 'expression' structure.

4.1.

CORRELATION AT THE PROJECTION LEVEL

This is the level in the verbal perceptual process, previously defined as having to do with the 'summation of points of maximal stimulation, and suppression of other activity'.1 A) Contouring. A contour is generally defined as 'the outline of a figure', or 'the line or combination of lines that define or bound anything'. In the context of any structure characterized by various 1

Charles Osgood & T. Sebeok, eds., Psycholinguistics, p. 54.

CUES AND ORGANIZATIONAL OPERATIONS

49

levels or degrees of elevation, however, a contour may be used to mean 'one or more lines joining points of equal elevation on a surface'. Accordingly, an initial step in message decoding may consist in responding to certain verbal cues, which help the receiver envisage a contour bounding certain content units, as meaningful combinations at one level or another. In this sense, contouring, as providing a category of organizational information, would perhaps underlie the attempt on the part of a message decoder to recognize the hierarchical structure of every utterance, as the bundles of distinctive and configurational features are simultaneously unreeled. He would want to distinguish the gross segments or portions of the content structure, or rather the minimum thematic sequences that make a difference in response. He would want to know, for instance, from the contour cues of a specific utterance, whether he should plug in an actor-action frame, a modifier-modified frame; whether there is a positioning of opposites, or a drawing together of similarities, no matter what these are in detail at this stage. In the psychology of visual perceptual organization, contouring is recognized as the outcome of intensity differences, or rather greater contrast between figure and ground. A triangle of light, projected very dimly on a screen is nothing more than a blur of light. But as the intensity of illumination is increased, a contour will emerge. Increased awareness of the divisions of an utterance, that segregate content units as belonging to one level of structure or another, is probably based on similar intensity differences. Assuming that Immediate Constituents are signalled by juncture patterns, among other configurational features, one might expect that the more contrastive and differentiated the junctures are, the more intensity difference is provided, and the more sharply defined is the contour. As a matter of fact, all the prosodic features of speech operate synchronously to give the decoder the contouring cues in question. However, it is here suggested that juncture, in particular, might be singled out for observation, as a verbal cue signalling the simultaneous occurrence of grouping and separation, the two complementary aspects involved in contouring. The other supra-

50

HYPOTHETICAL

segmental phonemic features, namely pitch and stress, will be considered redundant features. B) Accentuation. This operation has to do with the degree of articulation or differentiation, that the decoder perceives as maintained in connection with certain points of the contour. Accent placement on any unit of some conceptual pattern is usually associated with prominence in comparison with other units of the same pattern. While contouring is supposed to help the decoder 'figure out' the gross form of content, accentuation is assumed to enable him to distinguish the main features of the outline. Thereby he can tell which units are meant to carry more weight than others; which are stressed on account of their presumed significance to the source; which are calculated to create more awareness with the receiver. In Gestalt terminology, accentuation is probably based on all three interrelated notions of 'contrast', 'fixation', and 'figure on ground'. Any stimulus which contrasts particularly strongly with its background will cause a temporary change or cessation in the perceptual process, immediately followed by increased attention and awareness of the point of focus. Thus it will be readily and particularly attended to. Accordingly, those elements or units of the content structure that need to be maximally accentuated in order to stand out as more significant than others in the background must be tied up on the expression structure with one or more of the so-called 'emphatic' devices. One's normal exposure to the verbal techniques of emphasis in everyday speech will show that accentuation in discourse is not limited to the linguistic feature known as 'sentence stress', as different from 'word stress' or 'accent' or to 'emphatic stress' as different from normal sentence stress. Accentuation could be effected and raised to various 'degrees of elevation' through a multiplicity of ways, including utterance truncation into a one-word form, inversion or transposition of the accentuated item, repetition, etc. However, just for the sake of consistency, we will again limit ourselves to one more prosodic feature, namely 'sentence stress', and we will treat other synchronous features as redundant. It is commonplace to observe that the placement of a particularly strong

CUES A N D ORGANIZATIONAL OPERATIONS

51

accent or stress on one or more segments of the speech continuum is inevitably accompanied by a rise in pitch; and it is just as obvious that the employment of more than one technique of accentuation simultaneously is characteristic of the more 'expressive' form of messages.

4.2.

CORRELATION AT THE INTEGRATION LEVEL

At this level in the verbal perceptual process, it will be recollected that the focus of attention is on those 'relational' phenomena of 'semantic ordering' and 'concordal agreement'. 2 A) Transition. One of the basic phenomena to be investigated at this level is transition, which, in message decoding, would suggest a guided progressive motion from one content unit to another, or a guided passage from one stage of thematic development to the next. In analyzing the message to observe correlations between the structures of content and of expression, transition as a sequential operation is not viewed with reference to the 'probability' of occurrence of microlinguistic units, but in terms of the 'possibility' of occurrence of macrolinguistic or psycholinguistic units, as represented by words and word combinations. Gustav Herdan's concept of language as 'choice and chance' assumes great significance in this respect. Here one's interest in the organization value of the signal or cue at the point of transition does not center on the statistics related to the chance occurrence of 'X' items as succeeding units to 'Y' items, but rather on the 'possibility' of occurrence of such items as may be suggested by the semantic inflectional pattern or 'paradigm', previously alluded to within the framework of a particular content structure. In other words, our examination of transitional dependencies must be 'context-sensitive'. This is one primary consideration. We cannot afford to be entirely statistical, the way Information theorists are, simply because we are not interested in the organization value of verbal signals as 'chance' 2

Charles Osgood and T. Sebeok, eds., Psycholinguistics, p. 95.

52

HYPOTHETICAL

events. As the decoder makes his own selections by association, the element of 'choice' is obviously a privilege of his. "The statistical procedure usually ignores, however, a matter which is basic to linguistics — the distinguishing of levels of structure." 3 Now, this is the second important consideration with respect to our study of transition among other sequential phenomena. The hierarchical structure of verbal expression is assumed to provide helpful cues of transitional significance. In contemporary psycholinguistic research, these two tendencies, namely response to the level of structure and response to the context of communication, are generally accepted. They may be referred to respectively as 'syntactic' and 'associative' habits, 'syntagmatic' and 'paradigmatic' dependencies of a sort. The rationale is simple enough. If an utterance were to be broken up into single words, or into sections of equal length, it would lose its sequential value in the process of decoding. One reason is that the 'bracketing' or 'contouring' procedure, normally represented by (IC) diagrams, whereby meaningful word groupings are made, would in that case be impossible to accomplish. Another is that the amount of guesswork by word association in the course of transition, instead of being limited and restricted by the context of communication, would in that case be utterly randomized. Accordingly, the decoder would normally need two major types of transition cues: one to signal the Immediate Constituent or phrase structure level; and one to furnish the thematic environment of the transmitted message. The former would facilitate embeddedness or structural integration; the latter would channel association and set limits on its randomness, thereby permitting transition. In the early days of 'descriptive' linguistics, it was natural that the transition phenomenon should receive only partial treatment. Only syntactic or 'level' cues seem to be identifiable between sentences. According to C.C. Fries, the links between what he calls a'situation' sentence and a 'sequence' sentence must be structural units, or rather, grammatical in a strict 'linguistic' sense. Association is for 3

Charles Osgood and T. Sebeok, eds., Psycholinguistics, p. 94.

CUES AND ORGANIZATIONAL OPERATIONS

53

psychologists. Hence, he has to limit his investigation to the 'so-called' definitive article, the 'so-called' adverbs, and the demonstratives. However, his subsequent inclusion of such items as 'else' and 'other', implying 'context-sensitivity', 4 now leaves the door open for the warranted addition of the whole category of contextual cues within the framework of sequential psycholinguistics. Apart from those two main types of cues, the prosodic features also play an important part in signalling transition from one content unit to the next. B) Combination. In the course of progression from one unit of a message content to the next, there occurs an accumulation of similar parts, an addition of segments, that somehow are felt to belong together. Of course this recognition of belongingness through perception of similarity or analogy on the part of the decoder is based on his awareness of the information content of the message being delivered. And the stacking of similar or analogical constituents is what is meant by combination here. As to what is involved in the combination of elements, it is briefly 'redundancy'. While we recognize the natural overlapping among the complex operations of verbal behavior, whether organizational or representational, we would expect redundancy, as a fact of combination, to be chiefly paired with transition; forthcoming events keep getting correlated with 'similar' or 'contrastive' preceding events, without necessarily getting bound by anything like a Markov-process of strict progressive linearity from 'X' to 'Z' via 'Y' on the expression structure. The variables to account for this measure of 'freedom' from a Markov process in combination are chiefly two: the strength of association particularly for similarity within the framework of 'semantic inflection'; and the hierarchical structure of language, with its effect on the bracketing or contouring procedure. Consider an utterance like THIS. The referent of the word this would not be combined with that of the word like, as the immediately preceeding item in strict linearity, but most probably, it would be linked up with the word utterance. The fact that this sentence would re-write 4

See treatment of sequence signals in C.C. Fries, The Structure of English, pp. 241-255.

54

HYPOTHETICAL

as Consider an utterance like this utterance may be considered a tentative proof of the assumption. The decoder might or might not be fully aware of the repetition, reiteration or overlapping of items, but the end result for him is cumulative. Combination enriches context, and this in turn affects prediction of subsequent events.5 In terms of Information theory, prediction would be made on the basis of both 'discrete', and 'cumulative' probabilities, or rather, 'possibilities', as far as we are concerned. This sums up the effect of redundancy, as conducive to what is called 'stimulus patterning'. The stimulus presented in one form after another, thereby accumulating additional information for the decoder, becomes more specific, and hence more precisely associated with a given response than with other responses. This is in essence the psychological function of combination in verbal behavior. However, the principles of perceptual organization underlying combination, other than 'similarity', might be 'contiguity', and even 'contrast'. Here again, it is necessary to recognize the two chief categories of 'level' or 'syntactic' cues, and of 'contextual' cues. Prosodic features may be considered redundant. C) Integration. Perception is essentially an integrative process, one whereby several impulses, that have passed through auditory, visual, olfactory, and/or other sense modalities, get interpreted and correlated, thereby forming a joint 'whole' characterized by internal unity, and constituting the best possible 'fit' into the framework of previous experiences. Speech perception is no exception. It is assumed that in verbal behavior, a similar process involves an attempt on the part of the decoder to respond to those verbal cues, that would suggest a scheme of interrelations or a 'logic' for communicability in language. Conjunctions provide ready examples of this type of cues, to say nothing at this point about the 'contextual' 5 Charles Osgood and T. Sebeok, eds., Psycholinguistics, p. 97: "Given only knowledge of the immediately antecedent event at any one point of these levels, uncertainty as to the subsequent event is maximal (within limits imposed by the structure of the hierarchy). As we increase our knowledge by taking into account more and more of the sequence of the antecedent events, as well as subsequent events in the case of decoding, uncertainty as to the subsequent event decreases."

CUES AND ORGANIZATIONAL OPERATIONS

55

variety that should also be included. Needless to say, verbal cues operative at this level of organization would again be expected to have a dual capacity: 1) as markers of the basic formulae of reasoning and persuasion in the human context; and 2) as signals for thematic development in the course of communication. To put it differently, what is involved in integration seems to be the outcome of all the foregoing operations; not that these operations follow a consecutive linear order, of course, but that they all contribute to this phenomenon simultaneously. When contour and accent are discriminated, when the elements of content are processed in succession, as well as in combination, the end result is the integration of a 'new' structure of relationships, that must still 'fit' into the established framework of communicable experience. The concept of integration is accountable in terms of the principle of fusion or summation in perceptual organization. "Whether the comprehended thought becomes our own, whether we retain it, and whether we can use it, does not depend on comprehension as such. The process of 'fitting in' only defines more closely the meaning." 6 As for cues to this type of organizational information, they happen to be identified by the 'descriptive' linguist in terms of 'function words', which according to our model, would qualify as 'level' cues or just syntactic cues. However, within the presented framework of operation, whereby the verbal message is treated as structuredcontent in structured-form, it would be necessary to recognize the other major category of 'content words', those lexical items to be referred to as 'contextual cues'.

6

David Rapaport, Organization and Pathology of Thought, p. 45.

PART II

ANALYTIC AND SYNTHETIC

5 PROCEDURE

Within the theoretic framework sketched in Part I of this study, the corpus material in Appendix II is analyzed with the aim of testing a few assumptions which have to do with the proposed correlations between certain categories of verbal cues and the organizational operations hypothetically presented as underlying speech perception. The dialogue is treated in its totality as structured-content in structured-expression, and the focus of attention is placed on certain relationships of sequential organization. Under the operation named 'Contouring', an investigation is made into those relationships, assumed to be operative in decoding, between juncture and intonation, as two of the sequential phonological features, on the one hand, and the 'gross structuring' of content on the other. Under 'Accentuation', the study is concerned with the relationships between stress and juncture phenomena, and the 'prominence', or maximal stimulation of specific content units. Analysis under these two categories is related to the 'Projection' level of organization. At the 'Integration' level, represented by the three operations of 'Transition', 'Combination', and 'Integration', analysis is again applied to the totality of the verbal message, as both overt and covert structures. Transition is viewed as 'progression' from one content unit or group of units to the next. Combination is assumed to consist in the cumulative 'clustering' of such units on a differential basis. Integration, as the final stage in perception, is interpreted in terms of a 'cohesion' of the constituent units against a background of related conceptual structures. Under all three operations, identifiable verbal 'cues' are found to fall into

60

ANALYTIC AND SYNTHETIC

two major categories of organizational information: explicitly sequential; and implicitly combinatorial. These two categories of cues which reflect the interaction between the syntactic and semantic components will simply be referred to as 'syntactic', and 'contextual'. The analysis is auditory, broadly differential, and molar in character. It is meant to chart lines of 'integration' in linguistic research, based on a demonstration of the correlation between form and content in language. Accordingly, it is realized that in order to project a model of correlation in broad outline, the referred to relationships and operations must first be postulated, then provisionally tested on a purely intuitive basis. Eventually, however, confirmations should come from experimental evidence, and other types of analyses would then be conducted. As for the author analyst, who is a decoder of the messages he listens to, he could not possibly qualify as a 'casual' listener under the same conditions attending the particular communication being analyzed. While losing all the nonverbal cues previously alluded to, as helpful redundant features, he still has the advantage over the casual listener decoder in other respects. For instance, in the process of verbal perception, a decoder would normally utilize in particular a two-dimensional view of things; one prospective and anticipatory of forthcoming events, and the other retrospective and recapitulatory of antecedent events. The analyst has a better chance of commanding such a view of events both ways; prior training and availability of playback would naturally be of assistance to him. However, we would again need to remember the two-fold aim of this analysis as being exploratory and integrative; otherwise, it could not have been planned to cover so much territory. It is supposed to demonstrate: 1) what cues to look for under each of the postulated organizational operations in message decoding; 2) at what level of verbal perceptual organization they are supposed to function; 3) how to identify them in a corpus of linguistic material; and 4) how to go about testing a few assumptions with reference to their correlative value, drawing on pertinent psychological principles.

PROCEDURE

61

The data under observation consist of excerpts from a symposium on 'Form VERSUS Content in Art'. The various contributions, referred to as 'entries', which are made by members of the panel in discussing this issue, are given serial numbers in the order of their occurrence. And because the brief statements made by both moderator and announcer provide links in between the exchanged messages, the contributions of these two individuals are also included in the general count of entries to facilitate reference. At the Projection level of organization of linguistic response, hypothetically represented by 'contouring' and 'accentuation', any utterance composed of a group of words may be analyzed, however brief such an utterance may be. While data for such analysis could be chosen from any part of the corpus material, it will be convenient at this level to treat only the shorter entries. At the Integration level of organization, however, discourse units must be strung long enough to permit observation of such phenomena as 'transition', 'combination', and 'integration'. The longer entries will therefore provide data for analysis in terms of those operations. The procedure of analysis is two-fold: general and specific. The general consists of three steps. First, transcription of recorded material, special attention being given to the prosodic features as well as to pauses of hesitation. Second, selection of entries to be charted for observation of the phonological sequential features of pitch, stress, and juncture. Third, tabulation of lexical items for the rest of the entries, to help identify both 'syntactic' and 'contextual' cues. 1 The specific procedure starts, when the data have been reduced to these charts and tables. It consists of three steps. First, Formation of assumptions under each category, suggested by the charted data. Generally, Assumption I would be about the assumed capacity of a particular category of cues to elicit response in terms of a certain organizational operation. Assumption II would be concerned directly or indirectly with the 'possibility' of occurrence of a given 1

The basis of tabulation for 'content' items is briefly what has already been introduced as 'semantic inflection'. It will be elaborated on in dealing with certain assumptions.

62

ANALYTIC AND SYNTHETIC

cue, or type of cues, as may be determined through some preliminary investigation of a hypothetical correlation between what has occurred and what is pending occurrence. In other words, it would have to do with the operation of principles affecting predictability in verbal decoding. Assumption III would be concerned with the dynamics of verbal interpersonal communication, which might lead to utterance modifications, amplifications, etc. Should certain terms occur in any of the assumptions, which need further clarification of the intended meaning, such clarification is done, as a rule, in the course of interpretation. Second, Identification of cues. One or two entries are usually chosen for observation. If the specified cues are phonological sequential features, the patterns of these features are consistently charted according to one system of phonemic notation, such as the Pike system, adherence to which is meant to permit collation for points of similarity or contrast. On the other hand, if the specified cues are 'syntactic', a consistent classification of such cues is also observed, and is usually accompanied by a brief note explaining the criteria on the basis of which the classification has been performed. As for the identification of 'contextual' cues, categorical tables are compiled as follows. On the basis of the significant content unit known as the 'sememe', or 'senseme', according to Prof. Winburne, all the lexical items occurring in any verbal message of 'moderate' length (excluding those that are explicitly relational) are found to fall into two basic categories: one 'major' and one 'minor', depending on the density of clustering under each. At the head of each category, a representative lexical item is listed as a 'primary' cue. The items following in the same column are called 'reinforcement' cues, because they are perceived as redundant one way or another. They may be identical repetitions, analogous occurrences, items suggestive of referential overlapping, etc. Third, Interpretation. In a process of differential analysis, mere identification and reproducibility are all the empirical evidence needed to support a given assumption. In other words, verification is inherent in the demonstrable observation. However, because interpretation under circumstances dictated by the exploratory

PROCEDURE

63

nature of this research would still have the limitations previously alluded to, the presented verification of assumptions should be considered tentative.

6 CONTOURING

The notion of contouring may be summed up in a few words. It is a presumed tendency on the part of a decoder to respond to certain verbal cues, which signal two main structural features of simultaneous occurrence: grouping and separation of items. Whenever grouping of elements occurs, separation of these from other elements will concurrently take place. Verbal cues, which are supposed to signal the two structural relationships in question, may be more than the ones selected here for observation. Only two of the prosodic features of speech will be considered in this analysis, namely, juncture and intonation. It is assumed that juncture in particular is probably the chief variable affecting word grouping and segmentation, as far as field properties are concerned. Although intonation is recognized as a concomitant factor of perceptual organization at this level, it is not the 'only' other factor. With a view to the purpose of this analysis, it may be helpful to distinguish between two types of juncture, a) Tentative juncture /// may be defined as a pause of structural significance, occurring in the course of an utterance. It signals a 'sustain' in the encodingdecoding process. Therefore, it can occur at the end of a 'sentence', if such an end is not relatively 'conclusive'. It is often accompanied by a -4-3 or a -2-3 intonation contour, according to the Pike system of phonemic notation. 1 b) Terminal juncture / # / is a pause invariably occurring at the end of an utterance, the given signal 1

See Kenneth Pike, Phonemics.

CONTOURING

65

being, relatively 'conclusive'. It is more often than otherwise accompanied by the -2-4 intonation contour. 1. Assumption

I

Junctures and intonation contours are supposed to operate 'jointly' as cues for the 'simultaneous' separation and grouping of content elements, as represented by word concatenation in an utterance. 2. Identification of Cues {Entry 10: Ciardi)

But can you go on to the human question/ -2-3 until you know what stone will do/ -4-3 and what reinforced concrete will do/ -4-3 and what steel will do/ -4-3 and what glass will do/ -4-3 and what plastic facings will do # -2-4 Uh until your knowledge of these materials is such/ -2-3 that you ... you can put them in second place # -2-4 {Entry 13: Ciardi)

I'd like to ask William Butler Yeats that question/ uh and his answer is/ Oh body swayed to music Oh bright'ning glance/ how shall I tell the dancer from the dance # -2-4 You can't separate these th... these two things/ There is no dancer/

66

ANALYTIC AND SYNTHETIC

unless... the... the dance is nowhere until someone dances it # -2-4 And he's not a dancer until he's dancing # -2-4 3. Interpretation Entry 10 is divided into two sections by means of the terminal juncture after will do. This juncture may be viewed as performing two organizational functions simultaneously. While it separates the preceding frame from the one following it, it actually groups or unifies all the preceding segments, marked off by tentative junctures. A possible interpretation of this phenomenon may be given in terms of the Gestalt principles of proximity and similarity, on the one hand, and of contrast on the other. Positionally, what stone will do, what reinforced concrete will do, etc. are contiguous segments, marked off as 'segments' by means of successive pauses. And in duration, these consecutive pauses, or tentative junctures are very much alike. However, they become the more similar, when the terminal juncture after will do occurs, presenting a significant contrast in the form of a markedly dissimilar pause. In the particular utterance being examined, other aspects of similarity may be noted; for instance, the intervals between tentative junctures are almost equal in length, varying between 5 and 6 words each. Also the segments between tentative junctures are almost identical in their syntactic structure. Although these aspects may be viewed as contributing to contouring relationships of grouping and separation, yet they do not pertain to juncture cues as such. A look at the juncture pattern for Entry 10 may help clarify this and further illustrations. A #

1 9ws/

2 6ws/

3 6ws/

4 5ws/

5 B 5ws/ 6 w s #

1 9ws/

C 9ws#

It is suggested that the occurrence of B, while separating the A-B frame from the forthcoming B-C frame, actually gives more unity and cohesion to the segments A-Bl, 2, 3, etc. By the same token, when C occurs, it would potentially set the B-C frame apart, while

67

CONTOURING

furthering the unity, not only of B-Cl and B-C2, but also that of the whole B-C frame with the A-B frame. Similarity of duration between B and C as terminal junctures, coupled with their relative proximity could be some of the underlying factors. A #

1 9ws/

2 5ws/

3 B 8ws/ 9ws#

1 8ws/

2 C 4ws/ 10ws#

D 8ws#

The above terminal juncture pattern of Entry 13 would probably give a better idea of the possible effect on contouring of the distance between these junctures. Judging by the same criterion of contiguity, it would appear that B-C and C-D have a strong tendency to unite. Regardless of the content of these segments, the sheer structural phenomenon is borne out by experiments with metronome beats distributed over longer and shorter distances. It may be noted in addition that, in terms of progression, the distance between terminal junctures tends to get shorter, as the speaker moves on to conclude his point. It is not clear as yet, if this is more or less generally the case, or if it is characteristic of the style of individual speakers in argument. In any event, the observation is true of both patterns here. With regard to the operation of tentative junctures in their assumed capacity to signal both separation and grouping relationships, one would probably find an exact replica of what happens in the case of terminal junctures, only on a smaller scale. The same principle of contrast may account for the separation of frames or segments between points of tentative juncture, while the same notions of contiguity and similarity may explain word grouping within these segments. For example, in the A-B frame of Entry 13, the components of the A-Bl segment, already sharing a rough similarity of spacing, together with proximity in sequence, would, upon the occurrence of juncture 1, become the more similar and the more contiguous, while getting separated from the forthcoming components of segment A-B2. Considering intonation in the presumed scheme ofjoint operation with juncture, the general current of speech is supposed to flow at pitch level 3, according to the Pike system of notation. From there,

68

ANALYTIC AND SYNTHETIC

glides take place, up to levels 2 and 1, and down to level 4. Presumably, any departure from level 3, whether up or down, is in a sense, a signal of separation, unless a glide back ensues. Similarly, any return or glide back to level 3 from up or down is, in a sense, a signal of grouping or unification. Meanwhile, recurrence of the same contour is in itself a signal of grouping from similarity. The data being examined shows three types of intonation contours: /-2-3, -4-3, and -2-4/. The /-2-3/ and/-4-3/ contours glide both ways up and down, but they do return to level 3. As such, they seem to indicate temporary separation from the main current, followed by a junction. These are the contours observed to accompany the tentative junctures. On the other hand, the /-2-4/ contour observed to accompany terminal junctures in both entries, seems to represent departure down to level 4, with no 'apparent' return to level 3, or rather, a stronger type of separation. But as the stream of speech flows on, the decoder is made aware that such a separation is only momentary, though for a comparatively longer duration than in the case of the other two contours. The second assumption to be examined in this chapter has to do with the predictability of junctures, as verbal cues of organizational information. It may be worthwhile to find out whether a decoder's expectations of the occurrence of junctures tally with the normal syntactic divisions of the utterances he receives. If the Gestalt principles of perceptual organization may be accepted as adequate 'grammatical' descriptions of how junctureand-intonation cues help arouse in a decoder the response of lumping and dividing speech items upon their reception, these psychological principles would simultaneously be expected to account for the syntactic divisions that correspond to such verbal cues. The question now arises, whether the occurrence of juncture, or rather, its presumed operation in the encoding-decoding process is grammatical also in the limited syntactic sense, that may be discussed in terms of subject-predicate relationships, modifier-modified relationships, etc. It is reasonable to suppose that the relationships

CONTOURING

69

aroused by the verbal cues in question are grammatical, both in the broad configurational sense, encompassing such notions as similarity, contiguity, contrast, etc., and in the strict sense of syntactic structure, based on the notion of immediate constituents. The presented examination of assumptions I and II deals with juncture cues in this double frame of reference respectively. 1. Assumption II In order for junctures to operate as organizational cues of contouring, they must fall at syntactic points of structure. 2. Identification of Cues: Juncture and Corresponding IC Pattern (Entry 1: Announcer) Early in the discussion/Ll John Ciardi had made the statement/L2 that a rhythm is a way of knowing something # Asked to elaborate on what he meant/Ll Mr. Ciardi said/L2 Example: (L2) LI Asked to elaborateI /on what he meant/ ¡Mr. Ciardi said/ L2 L2 L2 LI LI (Entry 19: Neutra) Well, technology/L4 ? to-day which envelops us/L3 ? in such a terrible fashion/Ll has produced/L3 unbearable situations/L4 biologically unbearable situations in our community/L2 in our cities # We have been so in love/L2 and the technicians/L5

L2 /... L2

70

ANALYTIC AND

SYNTHETIC

and the technologists/L4 and also the architects with materials/Ll ? that it was overlooked that man the consumer/L3 is the greatest subject # To understand his responses and reactions I say/Ll is the thing # (Entry 29:

Ciardi)

And... and... its ... it's ... it's an experience # It says/Ll in making this/L2 I live/L3 better/L4 more richly/L4 more truly/L4 than I do/L5 by not making this # Whether it means anything or not/Ll I'm more alive in the process # As Robert Frost said/L2 a while back/Ll a poem/L2 is a momentary stay/L3 against confusion # You can't get clarified to stay so # Uh you musn't think that # You have to do it all over again # But for a minute/Ll the poem clarifies a thing # It uh it's an experience of life/Ll caught # I think that is the subject of all art # 3.

Interpretation

In order to find out how far the given pauses tally with syntactic

71

CONTOURING

points of structure, the corresponding IC pattern is sketched in every case, as shown by the example in Table 1. TABLE 1 Corresponding

IC

Patterns

Entry

Patterns

1

# # # # # #

19

29

# # # #

1-2 # 1-2 # 4-3-1-3-4-2 # 2-5-4-1-3 # 1 # 1-2-3-4-4-4-5 # 1 # 2-1-2-3 # 1 # 1 #

The patterns for Entries 1 and 29 share one characteristic, which Entry 19 does not have. It is the regularity of transition between levels in the hierarchical structure of utterances. It would normally take a regular ascending order ( # 1-2-3-, etc. # ) , a regular descending order ( # 3-2-1 # ) , or both for right branching and left branching. It would appear that the obvious irregularity of patterns in Entry 19 may give a more or less accurate picture of a jerky way of speaking, where pauses are made at various levels of the syntactic structure, irrespective of the regularity of transition. So, although some pauses may fall at syntactic points of structure, yet, when the hierarchy of levels gets shuffled, as it does in Entry 19, the decoder is furnished with such combinations as well technology, to-day which envelops us, architects with materials, etc.

In the light of such evidence, it would seem reasonable to modify the original assumption through the additional specification of the feature of regularity pertaining to transition between levels of the hierarchy of linguistic structure. Without this particular characteristic, 'pausing' could not be identified with 'juncture'; it could

72

ANALYTIC AND SYNTHETIC

not possibly perform its function, as a cue of organizational information. Accordingly, the following modifying phrase may be annexed: "..., following a regular course of transition between levels of the hierarchy." One other possible modification of the assumption may have to do with the range of levels up to which a speaker would normally be able to manipulate juncture cues with efficiency. It would appear that manipulation of such cues above the third level of structure would require adequate support from emphatic stress and pitch variations, or else a speaker would run the risk of breaking up the unity and continuity of the structure of content in the process of perception. Entry 29 gives an excellent example of a successful manipulation of juncture cues up to the 4th and 5th levels in the pattern # 1-2-3-4-4-4-5 # . Above all, however, the question of 'equal' or 'semi-equal' efficiency in performance between encoder and decoder would assume importance here. The occurrence of juncture would be predictable, if it fulfilled the two conditions of the regularity of transition, and the adequate limit for a range of operation between levels of structure. On the whole, it would be the more predictable, if these two individuals shared approximately the same degree of efficiency in encoding and decoding in a particular speech situation. It has been demonstrated through the foregoing examination of Assumption II how encoders differ in their manipulation of juncture cues, and how the decoding process is affected by such a difference with respect to the organization of items at the projection level. Assuming that the difference touches the encoder in terms of manipulation in production and the decoder in terms of prediction for perception, it would be understandable to think of contouring as 'given' and of contouring as 'expected', not as a presentation of the same or similar patterns of organization in every instance, but of characteristically different patterns at times. In short, the notion of 'relativity' seems impelling in this connection. There is reason to believe that while the grammatical base for juncture is an actuality with language users, differences will keep occurring between source and receiver, depending on the personal and situational factors

CONTOURING

73

responsible for error in performance. If an encoder chose to place his 'junctures' at points a, b, and d, while the decoder expected them at points b, e, and f, there would be difference between 'given' and 'expected'. To the extent that difference existed, the organization of content units at this level with the help of such cues would be affected. 1. Assumption

III

The organization value of juncture cues depends on the amount of agreement between contouring as 'given' and contouring as 'expected'. 2. Identification of Cues

Coupled Juncture and IC patterns for both 'given' and 'expected'. {Entry 3: Announcer) Given

Conflict with John Ciardi's point of view/Ll was provided by Millard Sheets # Responding to the question/L2 Is the big idea/L3 more important than concern with the actual material medium of the painter/Ll Mr. Sheets said/L2 Expected

Conflict with John Ciardi's point of view/Ll was provided by Millard Sheets # Responding to the question/L2 Is the big idea/L3 more important/L4 than concern with the actual material medium of the painter/Ll Mr. Sheets said/L2

74 (Entry 21:

ANALYTIC AND SYNTHETIC

Announcer)

Given

A question which seemed for a moment to go unanswered/Ll provided one of the best expressions/L2 of the view point of Millard Sheets # The panel was asked what they meant by the word art/Ll and whether they are not in fact talking about several different things/L2 when they used the word art # Expected

A question/L2 which seemed for a moment to go unanswered/Ll provided one of the best expressions/L2 of the view point of Millard Sheets # The panel was asked what they meant by the word art/Ll and whether they were not in fact talking about several different things/L2 when they used the word art # (Entry 11: Given

Neutra)

Most assuredly I have to know first the human beings/L2 Before I know/L4 what a ... cold wet stone will do/L5 do to my ... skin where I have (through which I have) heat losses # The biological side of it/L3 and the empathic side of it/L2 understanding human beings/L3 and systematically understanding them/L4 as well as empathically understanding them/Ll is the premise # to know/Ll what to do about stone steel and any other thing #

75

CONTOURING

I have to understand physiological optics before I know how to use glass or mirrors and so forth # (Entry 11: Neutra) Expected Most assuredly/Ll I have to know first the human beings/L2 before I know what a cold wet stone will do to my skin/L3 where I have (through which I have) heat losses # The biological side of it/L3 and the empathic side of it/L2 understanding human beings and systematically understanding them as well as empathically understanding them/Ll is the premise/L2 to know what to do about stone steel and any other thing # I have to understand physiological optics/Ll before I know how to use glass or mirrors and so forth # 3. Interpretation Table 2 shows the result of a congruency test in which given and expected patterns are compared. In either case, every juncture point is tagged by the IC level at which it occurs. TABLE 2 Congruency Test Entry 3 21

Given Pattern

Expected Pattern

# 1 #

# 1 #

# 2-3-1-2 # if 1-2 #

# # # # if

# 11

1-2 #

if- 2-4-5 # •# 3-2-3-4-1 # 1 #

#

2-3-4-1-2 if 2-1-2 # 1-2 # 1-2-3 # 3-2-1-2 #

if 1 #

The comparison between given and expected patterns shows 1 discrepancy for Entry 3, 1 for Entry 21, and 7 for Entry 11. In terms of a simple numerical calculation, it is easy to see that for

76

ANALYTIC AND SYTHETIC

the decoder, the organization value of juncture cues employed in both Entries 3 and 21 must be higher than that of cues in Entry 11. Especially as the total number of words for the entries in question is approximately the same in each case, it would be tempting to compute a ratio of 'organization value' for this particular set of cues, and the results would look more or less satisfactory. However, it is not so much the number of discrepancies between given and expected patterns that really counts, as it is the type of such discrepancies. In the foregoing interpretation of observations connected with assumption II, it has been pointed out that variability with regard to the operation of these verbal cues may be viewed in terms of: a) the regularity of transition from level to level, b) the depth of the nodes at which junctures are located, and c) the degree of proximity in efficiency of performance between encoder and decoder. Accordingly, the one discrepancy noted in Entry 3 would not be expected to affect contouring in any drastic way, especially as both given and expected patterns are, otherwise, identical. The decoder would be better accommodated, if one more juncture point were added at level 4; in other words, if the range were a little wider to suit an utterance carrying so much 'content weight'. Almost the same thing may be said with regard to the one point of disagreement in Entry 21. The added juncture, however, is not out of the range already covered; it is of the nature of adding one more stroke to a finished contour. In contrast with these points of disagreement are the ones noted in comparing 'given' and 'expected' for Entry 11. These are, for the most part, connected with the observed irregularity of transition from level to level. However, an additional factor gets into the picture, namely the lack of balance between left branching (i.e., in the NP or subject position) and right branching (VP or predicate position). Hence, there is a lop-sided distribution of junctures, putting too much weight on the NP side, as may be exemplified by: The biological side of it, etc. This also accounts for part of the 'entropy' or uncertainty connected with the reception of such an utterance.

7 ACCENTUATION

While contouring is described as a perceptual operation connected with the establishment among content units, as represented by words, of relationships having to do with the simultaneous occurrence of grouping and separation, accentuation may be defined as a presumed tendency on the part of a decoder to respond to certain verbal cues, which signal two other structural features of simultaneous occurrence, namely, projection and suppression of items. It would be impossible to think of the relative prominence of certain items in a speech continuum, without the concurrent levelling of other items in the vicinity. There could be no projection without suppression. It is assumed that sentence stress, as distinguished from word stress or accent, is a basic verbal cue to this particular operation. Other prosodic features, such as intonation and juncture, do get into the picture. And so do the syntactic mechanisms of transformation known as 'transposition' or 'inversion', 'deletion' or 'truncation', etc. However, the present analysis is concerned only with sentence stress, as a chief variable affecting the prominence and suppression of items, and with juncture as a concomitant variable. For the purpose of this auditory analysis, sentence stress stands for two main types: a) normal /'/, which is the more common type of 'sentence' stress, and b) emphatic or peak stress /"/, which is the strongest type that may be conspicuously distinguished among all the rest. Meanwhile, by way of incorporating into our investigation at this level the once neglected dimension of 'semantic import', at least one of our three assumptions under 'accentuation' will be

78

ANALYTIC AND SYNTHETIC

directly concerned with the comparative 'weight' of such semantic import, as may be correlated with stress variation. For the purpose, content units will be classified under categories of 'major' and 'minor', and the results of this classification will be utilized in testing this particular assumption. 1. Assumption

I

Stress and juncture operate 'jointly' as cues for the simultaneous projection and suppression of items, as represented by the words of an utterance. 2. Identification (Entry 16.,

of Cues

Ciardi)

Really all you're asking is that/ uh when you say you want a poet to have ideas / or an architect to have ideas / or an artist to have ideas / all you're really asking is that he be a human being # uh and I don't think that's too much to grant the artist # uh I don't care who he is / the chances are he's a human being # If he's a small human being / he'll say small things about being human / he will have small experiences in his art # If he's a larger human being / he will have larger ones / but the only way he can have an experience in his art form is technical # But uh you see that's the dirty word # One thinks immediately / empty correctness # Uh that's the ... not it at all # It's the way of going # The language must be spoken #

79

ACCENTUATION T A B L E 3a Accentuated Stress only asking poet architect artist asking I too grant who chances small small small larger larger only can art form

Units

Stress and Juncture really ideas ideas ideas all human being artist is human being human art technical dirty word immediately empty correctness all going spoken language

3. Interpretation All stressed units are supposed to have one thing in common, though, of course, in varying degrees. They all represent variations in intensity, compared with other units in the vicinity. Whether pre-junctural, post-junctural, or medial, these stressed units represent points of prominence, against a more or less levelled background. Without these points, the whole configuration of content units would be vague, and sometimes ambiguous. In the utterance, all you're really asking is that he be a human being, the stressed units are respectively, a function word and two content units, one minor and one major. The major unit, receiving emphatic stress, happens to be a key unit, upon which the whole argument is based. It is the only unit repeated four times without change or

80

ANALYTIC AND SYNTHETIC TABLE 3b Content Units Major poet architect artist he artist he he technical dirty word empty correctness ? way of going language — spoken

Minor

ideas ideas ideas things experiences ones experiences

human being human being human being being human human being you you I I you one

asking say asking think thinks small small small larger larger

art art form

modification. The minor unit, within the larger context of the whole symposium, is not actually minor. There is a lot of interaction among members of the panel, signalled by the asking and answering. As for the function word given prominence in this utterance, namely all, it is the most significant in the whole frame. The speaker makes use of it to pull the strings together and give point to his argument. In order for a decoder to give due attention to such words, it is necessary that they should be distinguished in a fashion, from the other words in the background. In terms of the Gestalt principle of figure-on-ground, when part A of a pattern is presented with more intensity than parts B, C, and D, part A will be perceived as a figure. It is not difficult to find out empirically that stressed units in a speech continuum are the only units to be distinguished against a background of noise that could drown almost everything else. When stressed units happen to be pre- or post-junctural, their prominence from increased intensity, would, through the additional

ACCENTUATION

81

effect of gradually arrested movement, or of gradually started or resumed movement, be given a more clear-cut form against the 'levelled' background. This appears to be very much the case with words like ideas, human being, technical, all, etc. The stress responded to under the circumstances is characterized by a more extended rendition of the syllable carrying it. This would probably increase the attention value of such a cue, particularly if in terms of intensity, the stress happens to be of the peak type, with respect to the entire utterance. An example of this is: Uh / don't care who he is / the chances are he's a human being #. So much for the configurational aspect of the joint operation of stress and juncture, as may be viewed from an examination of this portion of data. Quantitatively, stressed units are 37 in all: 16 of these are pre-junctural; 3 are post-junctural; and 18 are of medial occurrence in their respective utterances. It is quite obvious that a substantial number of stressed words occur medially, irrespective of juncture points. Naturally, the aim of the original assumption is not to prove or disprove the joint occurrence of stress and juncture as such, but rather to find out how far the enquiry could be pursued on the basis of such a count, with reference to 'semantic import', i.e., in terms of major and minor content units. Stress and intonation could also be profitably investigated in terms of a joint partnership along these lines, whereby verbalization could be correlated with conceptualization. Judging by evidence from the analysis of Entry 16, the assumption being examined here could be tentatively qualified in the light of findings from Tables 3a, and 3b. The table of accentuated units shows which of these are differentiated by stress, in the main. It also shows which of these are additionally differentiated by juncture. The table of content units is prepared with a view to discovering which of the total number of 37 units belong to the major category, and which belong to the minor one. Comparing the information provided by these two tables, one finds that out of 19 units differentiated by means of both stress and juncture cues, 15 may be identified as major content units. Therefore, it is probable that one way of qualifying the original assumption is to say that "stress and

82

ANALYTIC AND SYNTHETIC

juncture operate jointly, when they are applied to words of major content items". This is open to further investigation. 1. Assumption II Content units differentiated through the combined operation of stress and juncture appear to be generally more 'significant' than those projected by stress in a medial position. 2. Identification of Cues {Entry 15: Stevenson) I think that Mr. Ciardi has the ... the right point of view # I can't conceive of the kind of music that would interest anyone / without absolute control of materials first # {Entry 18: Ciardi) But I'm saying what's wrong with having a technical passion # What's wrong with loving your medium # What's wrong with believing that the only way you can get expressed what you are/ is through your medium # TABLE 4a Accentuated

Units

Stress

Stress and Juncture

Ciardi right conceive absolute loving expressed

view anyone first technical-passion medium are medium

3. Interpretation The notion of 'significance' seems to suggest at least two main issues in this context:

83

ACCENTUATION TABLE 4b Content

Units

Major point of view conceive materials first technical medium medium getting expressed wrong wrong wrong

Minor think anyone Ciardi passion loving

saying

I you you

a) Whether such units are significantly 'located' in the utterance; whether by virtue of their positions, they are likely to call the decoder's attention more strongly; b) Whether such units belong to a major content category. Thesetwo issues are touched upon briefly in this study. A third issue, which needs a much more thorough investigation in experimental research, has to do with the possible correlation between these two features of significance in the study of individual style. With regard to the attention value of units located in the prejunctural position, the table of accentuated units seems to indicate that in 2 cases out of 7, these units seem to receive the strongest stress. Therefore, if they do call attention more than stressed units in the medial position, then intensity alone as a field property would not explain the phenomenon. It is probable that the change of movement at points of juncture, coupled with the accompanying variations of pitch level, might account in part for the assumed significance of this location. According to the Gestalt principle of 'common movement', units may be viewed in separation, if they move toward or away from each other. The notion of separation is involved in the operation of both juncture and intonation as concomitant variables.

84

ANALYTIC AND SYNTHETIC

In the case of stressed units before terminal junctures, it is true that the anticipated drop to level 4 — with words like 'view' and 'medium', would cause a stress to gather additional intensity. Meanwhile, it is quite probable that the down-glide, signalling a separation from pitch level 3, followed by a temporary arrest of movement, should contribute to the attention value of these units. The question of whether the significance of such units has to do with their belongingness to major content categories may be tentatively answered from Table 4b. There is some indication that the assumption is well worth investigation. The element of choice may have been gleaned through the initial quantitative enquiry into how much of the total of major content units receives accentuation or differentiation, from more than one set of cues. The implication of this enquiry is that the distinction between major and minor content units, supposedly operating with both encoder and decoder, seems to underscore the notion of relativity. While differentiation of cuing is relative to the encoder's awareness of the multiplicity and variety of available types of cues, the amount of attention given by the decoder is also relative, partly to this kind of awareness, and partly to the strength and adequacy of cues, which are actually utilized, to say nothing of motivational factors on either side. The picture is even much more complex: what is relevant to this investigation, however, is the notion of 'selection' for the decoder. Accordingly, it would be meaningful to consider two sides of a comparison between what is 'given' in terms of accentuated units, and what is 'expected', and to find out what the relationship between these two could mean with respect to this feature of organization. What is 'given' is available for empirical checks. What is 'expected', however, may be difficult to determine with unquestionable precision. The nearest approximation would probably be for individuals to study a particular message, silently, outside the communication situation, with the purpose of locating the central content units, or those that they consider to be essential for understanding the message in question. It is assumed that such units would be the expected loci of accentuation. Accordingly, the

ACCENTUATION

85

expected pattern would vary from one individual to another. This variability is the feature that could be utilized in determining the organization value of these cues (stress cues) at this level of language perception. 1. Assumption III The organizational value of 'normal' and 'emphatic' types of sentence stress depends on the amount of agreement between accentuation as 'given' and accentuation as 'expected'. 2. Identification of Cues (Entry 9 : Neutra) The (matérial) was put (ahéad) of everything élse # And it is my (considered opinion) / and so I can (éasily) say and beyond (suspicion) / that I haven't used new (matérials) in a (daring) way / I would say that all these (matérial) (considerations) are (véry) much in (sécond) place in my mind # I (do) think that the (best) (material) the architect ever gets under his (hands) / are the (human béings) / who are his (clients) # And uh to understand (human béings) / to be (fascinated) by understanding their (responsiveness) their (responses) their (reactions) / is the (gréat) (art) of structuring an (environment) and arranging (stimuli) for them # Given stress /'/ Expected Stress ( ) (.Entry 12: Houghton) It would (appéar) / (off hand) / that this (panel) is moving in that (age-óld) discussion of (form) versus (content) /

86

ANALYTIC AND SYNTHETIC

over on to the side of (form) # And uh let's not worry about whether we're saying (anything) at (all) # The (important) thing is the (word) the (rhythm) the (shape) the (line) # Uh and I would like to come (back) to the support of (Sheets)' point which was that / (form) and (rhythm) and (shape) and (line) in (themselves) are (not) enough # And that (somehow) / there must be the (balance) of this (form) and (shape) and (rhythm) and (line) / addressed to (something) that relates to (human beings) # 3. Interpretation Accentuation as 'given' is simply denoted by its occurrence in sheer physical form as may be identified in the recorded material. Accentuation as 'expected', however, must be the outcome of a multiplicity of factors, both representational and organizational. To the former category would belong such considerations as the weight of the general semantic import for each of the marked units, together with the special prominence attached to these units in a specific context on the basis of their 'content' relevance. To the latter would pertain considerations of positional organization and other redundant configurational phenomena. To determine the organization value of stress cues, a simple congruency count is made as follows. In comparing 'given' and 'expected' loci of accentuation, the total number of marked cues, i.e. those 'given' plus those 'expected', plus those simultaneously 'given' and 'expected' equals the number of units (NU). The number of congruent units, i.e. those simultaneously 'given' and 'expected' equals (NCU). Accordingly the organization value of accentuation cues is a fraction of NCU/NU. The count for Entry 9 shows a ratio of 23/39 as against 27/36 for Entry 12. The difference is interesting; it would be worthwhile to determine how significant this is.

ACCENTUATION

87

The underlying assumption is that what is 'given' represents 'performance' on the part of an individual encoder; what is 'expected' represents 'competence' on the part of an individual decoder (here the analyst). The discrepancy between the two may be considered an indication of 'error', only if the number of 'expected' were to be statistically determined in terms of an average for a representative sample from the same speech community. On the other hand, it would be interesting to note that the extent of agreement among members of such a sample would provide a confirmation of the presumed tendency for such language users to decode signals against a common background of preconceived data of organizational information. Obviously, whether it is for contouring, accentuation, or any other operation of the organizational process, response to the verbal cue in question is not limited to its potency as a sheer physical phenomenon.

8

TRANSITION

Analysis at the 'projection' level of message decoding has been conducted in terms of the two presumed operations of 'contouring' and 'accentuation'. At this level, the organization of message content with the help of the prosodic features of speech is viewed as a gross structuring of units in the aggregate. The only relationships under investigation have been those of grouping versus separation, and projection versus suppression. At the 'integration' level of message decoding, however, organization of verbal content is viewed in terms of its more diverse relationships, which may be subsumed under three main operations: 'transition', 'combination', and 'integration'. Naturally, the overlapping of such relationships is expected, both among themselves, and with other relationships at the 'representational' level. It would reflect the essential unity of language in the process of perception. It would demonstrate the fact that there is no transition without combination of items; no integration without meaningful cohesion among the combined elements within the established framework of prior experience. Hence, at the integration level, and in the course of transition and combination of content elements, one would expect the utilization of such cues as would permit 'substitution', 'modification', and 'amplification' all along the line, until the total structure of the perceived message emerges in the mind of the decoder and is identified by him as an acceptable 'fit' — in which case, communication will have taken place. While integration, as an end result, or a final operation has to do with cohesion of parts, transition pertains to the particular aspect of

TRANSITION

89

progression. There is a presumed tendency on the part of a decoder to reconstruct the sequence of incoming speech events, as the source proceeds from one point to another. Such a tendency would probably underlie the normal question one would want to ask at every point: "what next?" The concept of language as behavior suggests process 'mobility' as basic to all verbal phenomena, including the relational operations under investigation. Transition for one such operation is therefore recognized as the living aspect of 'syntax'. It is the operation whereby the succession of speech events is recognized as a dynamic 'sequence'. Linguistic research that has been confined to utterance or sentence limits, that has been guarded against the 'illusive' study of content, could not possibly give a more-or-less complete picture of organizational phenomena, that correlate 'form' and 'meaning' in language. A t best, one side of the picture could be given, namely the narrowly 'syntactic' or 'explicitly' combinatorial aspect, whereby 'subjects' lead up to 'objects', and 'modifiers' to 'modified', with or without 'function' words. In terms of our present investigation, this view of syntax within sentence limits is static and incomplete. It has to be extended to cover relationships within and between utterances; and it has to be energized through contact between the two major structures of content and expression. Accordingly, if the linguistic form performing the function of transition from one sentence or utterance to the next happens to be a 'content' unit, and not a 'function' word, it is considered plausible from a communication point of view, to treat such a form as a 'syntactic' cue, in a broad, dynamic sense, here to be referred to as 'contextual'. T o limit syntax to 'surface structure' organization, and discuss it only in terms of N P and V P is to deny the very function of syntax, whereby transition is accomplished from one content unit to the next. Henceforth, reference to cues as 'syntactic' and 'contextual' should be interpreted simply as syntactic in the narrower and broader senses respectively. 1. Assumption I Transition in message decoding is effected partly through the

90

ANALYTIC AND SYNTHETIC

operation of 'syntactic' cues, and partly through that of 'contextual' cues. 2. Identification of Cues {Entry 4: Sheets)

T A B L E 5a Syntactic

3.

Cues

Reference Cues

Shift Cues

Putting it another way But by the same token But

And I think It seems to me I believe I think I believe I don't think I think And I think

Interpretation With a shift of focus from intra-sentential to inter-sentential relationships, our treatment of transition shows it as a product of linkage at points of terminal juncture and 'elsewhere'. The former type of linkage, which is the more obvious, is furnished by means of syntactic cues. These cues seem to fall into two major divisions: one of them 'refers' back to something of previous occurrence in speech; the other 'shifts' on to what is forthcoming. Of course both types suggest the onward motion, and both provide the connection between what has gone and what is coming; but this distinction appears to be pertinent especially in a comparison between one style of speech and another. It may also be correlated with particular types of context in speech. This is a point that may be picked up for future research. While Table 5a for Entry 4 shows 3 reference cues as against 8 shift cues, Table 6a for Entry 6 shows 17 as against 16. Also, the

TRANSITION

91

3 cues from Entry 4 belong to one variety: they introduce amplifications either through similarity or through contrast. We've advanced ... in raising the level of understanding ... 1 Putting it another way', the organization of a painting etc. Also, I think we've made a tremendous progress ... 'By the same token', I don't think that the things we are producing ... will ever hold their place etc. Again, I think that the experimental age ... will contribute tremendously to the younger painters ... 'But' it's a pretty empty art... etc. As for the shift cues in the same Entry, they also belong to one variety without a single exception. That particular variety introduces a 'personal opinion', or a 'personal conviction'. Eight such cues say: I think, I believe. TABLE 5b Contextual

Cues

Differentiated Cue Clusters greatest art great arts great ideas great art ideas human values great ideas ideas

insight power force ideas

{Entry 6 : Fadiman)

have advanced aesthetic understanding organization painting ... a complete unit accords of color handling space texture values technical building progress quantities easier to produce experimental age new vocabulary new power empty art ideas will come back

92

ANALYTIC AND SYNTHETIC TABLE 6a Syntactic

Cues

Reference Cues

Shift Cues

That The They That But Most of 'thsm' They They But The first But The first The second The The The The

And Now You What...? You And Now Now Really And of course Suppose Here Here Now And Is...?

On the other hand, Table 6a for Entry 6 shows other varieties of both reference and shift cues. One class of reference cues is for 'definitives' like the, the first; another for 'demonstratives' like that, etc.; another for 'pronouns'; and another for 'contrastive reference', like but. Shift cues for the same Entry consist of these groups of items: emphatic items like really, and of course; intermission items like now, here; items inviting audience participation or involvement, like you, suppose, what... ? etc. Finally, the word and is of particular interest. While it operates here as a 'shift' cue, it is observed to act elsewhere as a 'combination' cue, serving the function of addition or annexation. However, this double frame of operation is understandable in terms of the inevitable overlapping of organizational phenomena already discussed. The only solution for the practical purpose of analysis would be to make this arbitrary decision: if and is located at the head of a sequence sentence, i.e., following a

TRANSITION

93

t e r m i n a l j u n c t u r e , it is a u t o m a t i c a l l y c o n s i d e r e d a ' s h i f t ' c u e ; i f it is l o c a t e d a n y w h e r e else, it is c o n s i d e r e d a ' c o m b i n a t i o n ' cue. I n i n t r o d u c i n g s y n t a c t i c cues, r e f e r e n c e h a s b e e n m a d e t o t h o s e speech items w h i c h provide linkage at points o f terminal juncture, a s d i s t i n g u i s h e d f r o m t h o s e t h a t establish l i n k a g e e l s e w h e r e . T h e l a t t e r are g i v e n the n a m e o f ' c o n t e x t u a l ' c u e s . T A B L E 6b Contextual

Cues

Differentiated Cue Clusters ideas ideas summed up list 'them' ideas ideas ideas ideas ideas

'words' poet's ideas prose writer's ideas in the words themselves ideas of Shakespeare notions about the world notions as the expresses them

'o/' young novelists of very bad ... novelists ideas ideas rhythms Two dates one date another date Eighty seven years ago four score and seven years ago same date the first the second no rhythm rhythm rhythm rhythm that first statement second statement fact same statement different statement four score etc. beginning of great idea idea idea less operative idea operative eighty seven years ago

94

ANALYTIC AND

SYNTHETIC

Context is supposed to be structured in the course of speech through a process of 'differentiation' and constant 'reference'. While differentiation reveals the speaker's awareness of the sides of an issue, constant reference indicates his progression. How far he proceeds, or how fast — these are questions pertaining to the rate of transition. The important point to note here is that progression is achieved through a form of redundancy, once the 'primary' cues are given, and the differentiation established. Table 5b for Entry 4 shows the differentiation established horizontally by means of two columns; meanwhile, transition is represented vertically down the columns through the occurrence of redundant cues which get attached to the corresponding 'primary' cue (here italicized), as the speaker goes on. In a comparative frame of reference, Sheets speaks of: 1) the greatest art of past epochs; 2) the fact that painters have advanced; 3) that advancement is recorded in terms of quantities; 4) that this art is empty art; and 5) that painters need more insight. Each of these segments of content is considered a primary cue, which is continually reinforced by redundant cues of reference. A speaker could handle two or more primary cues at a time, and have them simultaneously reinforced. That is how he is enabled to proceed with his argument. It may be observed that while some of these reinforcement cues are so close to the corresponding primary cue as to be identical repetitions, others are only indirectly related. In either case, however, the function being served is the concurrent development of the content units initially introduced by the primary cues. Transition from one point of the structure of content to the next is naturally accompanied by some development or modification of the primary cue, by which the particular point has been previously ushered in. Table 6b for Entry 6 shows the same principles in operation: differentiation horizontally, and transition accompanied by amplification and modification vertically. The obvious difference between the charted sequence here and in Table 5b for Entry 4 is an indication of two styles of treatment. Sheets differentiates between

TRANSITION

95

classical and modern art in terms of two primary cues, and moves on from past to present, where he remains to the end of the entry. That is why most of his content units cluster in the second column. Fadiman, on the other hand, holds the two sides of the comparison all the way through. He introduces his two primary cues of 'ideas' versus 'words' almost simultaneously, and keeps piling up reinforcements on both sides till he moves up to the climax with the amplification: young novelists, full of ideas, etc. Then he makes a big shift to the question of rhythm; but there, again, he maintains the same balance of cues with one date here, and another date there; with one having rhythm and the other no rhythm. This is one way of exploring a pattern for transition cues. While syntactic cues may be considered explicitly relational, and explicitly transitional, contextual cues are implicitly so. They further progression through the referred to clustering phenomenon, which is explainable in Gestalt terminology. In a sense, every reinforcement cue must be 'similar' to the primary cue to which it belongs. And the more 'contiguous' it is to the primary cue, or to the preceding reinforcement cue, the more operative it is likely to be in achieving unity with the rest of the cluster. Curiously enough although syntactic cues are the more obvious, they seem to be the less operative as tools of transition. A simple test would be to remove them completely from the message under study, and observe the results. Progression could be effected through the operation of contextual cues alone. But this is not the way people speak; people use, and will go on using 'reference' cues and 'shift' cues, and any serious displacement of these could affect comprehension of their messages. Experiments along these lines might be both interesting and profitable. 1. Assumption II The pattern of transition cues, as manipulated by individual communicators, seems to show two interlocking frames: a macrostructure and a micro-structure. 2. Identification of Cues {Entry 2: Ciardi)

96

ANALYTIC AND SYNTHETIC TABLE 7a Syntactic

Cues

Reference Cues

Shift Cues

It He It That Something That It It Whether It One of the It It That It That It

A while back Then Now How...? And You You You What There There In In May...? What...? What...? Because And

3. Interpretation The distinction already made between syntactic and contextual cues serving the function of transition, coupled with the tentative discovery that such a function appears to be largely fulfilled through the operation of the latter type of cues — have suggested further investigation into what basically affects a decoder's comprehension of the sequence of events. If contextual cues appear to be so operative in revealing sequential progression, where in particular do they exercise maximum power? One's immediate impression is that probably the notion of a 'macro-structure' might offer a solution. If the big land-marks are visible, if the various turns and points of junction are held in view, it is very likely that one should find one's way about, without hitting blind alleys. The problem would then be how to identify those big land-marks which make up the macro-structure.

TRANSITION TABLE 7b Contextual Cues

Differentiated Cue Clusters

R. Frost ... put it said said said had to say know ... have missed know have missed know aiming at know have missed know have hit know have missed know know know know know know

rhythm way of knowing something ... wave length something ... way it goes something ... primitive rhythm (English Language) way of going meter of the race iambic other languages, other ways king of beat rhythmic (sequence) visual series of lines in time series of sounds (structure) in rhythm structure way of forming bringing to shape basic to form rhythm (way of knowing) human behavior ... poem doing ... writes ... poem doing ... receives poem poem never idea ... always ... experience what idea feels like live in it experience experience (pace) rate (rhythm)...unfolding rhythm wave length (way ... testing) way one knows

98

ANALYTIC AND SYNTHETIC TABLE 7c Macro-Structure Transition Frames 1. 2. 3. 4. 5. 6. 7. 8.

I was interviewing Robert Frost ... He put it ... He said ... Then he said ... What he had to say next ... That's your way of knowing it. One of the questions I'd like to raise ... But I want to say very emphatically ...

One way would be to chart all differentiated cue clusters, and to mark primary cues, as is shown in Table 7b. Presumably when all primary cues are identified, there will be nothing else to look for. These will be the constituents of the macro-structure, without which the decoder would be at a loss, regarding the sequence of events. If for any reason, he were to become momentarily inattentive and miss these particular cues, the chances are that he would lose points of significance in terms of sequence. This is not to say, however, that all primary cues are of equal significance. According to Table 7b, rhythmic 'pace' is reinforced only once, but to 'know' whether one has hit or missed something, through one's awareness of rhythmic sequences, gets the heaviest cluster of 11 reinforcement cues. Should the decoder miss as many as 8 or even 9 of these, he would be able to follow the presented sequence nonetheless. But if he were to miss the name of Robert Frost or subsequent reference to his name, he would not only miss the sequence of presentation at this point, but also the whole idea of source credibility attached to the better known American poet. Another way of identifying constituents of the transition macrostructure may be to obtain collective judgement on the matter from various decoders, and study the results. Table 7c shows the investigator's own impression of the minimum number of transition frames, without which the decoder could miss the sequence at one

TRANSITION

99

point or another. Frames 1-5 are essential for identifying the source of quoted information. Frame 6 signals a start for Ciardi's own personal reactions aside from what Robert Frost said or thought. Frame 7 signals a shift to the next major point, namely what sort of human behavior a poem is. Frame 8 signals his conclusion that a poem is never about an idea, but always about the experience of an idea. Tables 7b and 7c would make it possible to compare the resultant macro-structures identified in these two different ways: the quantitative and the impressionistic. TABLE 7b R. Frost ... put it know ... have missed

rhythm human behavior ... poem ? idea ... experience

TABLE 7c

1. 2. 3. 4. 5. 6. 7. 8.

A while back I was ... He put it ... He said ... Then he said ... What he had to say next ... That's your way of knowing ... One of the questions ... But I want to say ...

The fact that in either case the decoder is the analyst would certainly be expected to affect the results. However, the assumption regarding the correspondence between constituents of the macro-structure and major content units seems to be justified. According to Table 7b, primary cues are no other than the major content units of the Entry. Therefore, it would be safer to employ the quantitative method of identifying the macro-structure, as represented by Table 7b. As for the micro-structure, which has no separate existence per se, probably two things may be said about it in brief. First, it would naturally be expected to merge into the macro-structure at every point, both syntactically and contextually. Second, the distinction between 'macro' and 'micro' is relative. In Gestalt terminology, one thinks of the inter-dependence of action of the 'parts' within the 'whole'. Therefore, it is possible to think of the interlocking

100

ANALYTIC AND SYNTHETIC

frames as a number of interjacent circles, ranging from the smallest to the largest. This would probably explain the position of the cues put between brackets in Table 7b, as distinguished from the italicized cues on the one hand, and from any of the reinforcement cues on the other hand. Finally, concerning the effect on comprehension of the presence or absence of a macro-structure, it may be tested through total or partial deletion of its constituents, followed by comparison of results for various decoders. The underlying assumption is that predictability of forthcoming items in a particular order is calculated in terms of macro-structure units. For instance, when Ciardi starts by quoting Robert Frost, the chances are that when he moves on to discuss rhythm, the motion is more or less expected; and when his discussion is concluded by the statement It's a way of knowing something, the conclusion is more or less foreseen, because it is made precisely in terms of the initial remark quoted from Robert Frost. TABLE 8a (Part 1) Syntactic Panel

Entry

Announcer Ciardi Announcer Sheets Adams Fadiman Adams Neutra Adams Stevenson Adams Ciardi Announcer Sheets Adams Ciardi Announcer Ciardi

1 2 3 4 5 6 8 9 14 15 17 18 21 22 26 27 32 33

Cues

Reference Cues

Shift Cues

A while back They I just want Well I think But I think I think No

101

TRANSITION

TABLE 8a (Part 2) Syntactic Cues Panel Fadiman Ciardi Neutra Ciardi Ciardi Neutra Neutra Houghton Houghton Ciardi Stevenson Ciardi Ciardi Neutra Neutra Ciardi Sheets Ciardi Ciardi Sheets Sheets Ciardi Ciardi Sheets Sheets Ciardi Ciardi Sheets Sheets Ciardi

Entry 6 7 9 10 10 11 11 12 12 13 15 16 18 19 19 20 22 23 23 24 24 25 27 28 28 29 29 30 30 31

Reference Cues

Shift Cues

That May...? Most assuredly It would appear I'd like ... Really Well Absolutely But that That I mean That It I think I think

In other words, once the decoder is let in o n the gradual unfolding of the 'big frame', his expectations of items in the 'follow-through' would be more or less accurate. 1. Assumption III In interpersonal communication of the group situation, the total frame of transition is revealed through an exchange of cues, occurring with the change of speakers in succession.

102

ANALYTIC AND SYNTHETIC

TABLE 8b Contextual Cues Entry 1 2 2 3 4 4 5 6 6 7 7 8 9 9 10 10 11 11 12 12 13 13 14 15 15 16 16 17 18 18 19 19

Differentiated Cues

Ciardi rhythm way of knowing something rhythm way of knowing something conflict with Ciardi Millard Sheets big idea material medium both important ideas will come back into art (Adams) anecdote idea operative what... meant, John part of it human feeling technical passion (Adams) material of architect once eternal issues question materials second place ? human beings first fault of P. Avenue renewal fault of P. Avenue renewal panel moving side of form content along with form question ... W. B. Yeats dancer dance (Adams) Ciardi right materials first have ideas being human language spoken (Adams) lightning bugs have passion technical passion wrong? getting expressed through medium technology envelops us understand human the thing

TRANSITION

103

TABLE 8b (Cont.) Contextual Entry

Differentiated Cues

20 20 21

saying same thing human within form

22 22

23 24 25 25 26 (Adams) 27 27 28 idea! 29 29 30 great idea! 31 32 write a poem convey idea ? 33 33

Cues

question art? several different things M. Sheets question pertinent poetry and music two different arts problem nonsense bad mean remark parable? may have something to say It says: I make this experience subject of all art greater experience

not an idea an experience

ideas ... textbooks

2. Identification of Cues 3. Interpretation While Assumption I treats transition cues in terms of categories, Assumptions II and III consider the frames within which they operate. Speaking of a 'macro-structure' with reference to Entries, or individual units of discourse, would naturally suggest an examination of the 'total frame' of transition in a group communication situation.

104

ANALYTIC AND SYNTHETIC

To a certain extent, the total frame of transition for the recorded excerpts which constitute the corpus material of this study, is in a way 'meddled with' by virtue of the nature of 'editing', even if it consisted of nothing more than putting excerpts together, providing the most essential links in between. In other words, the 'presented' sequence is slightly different from the 'actual' sequence here and there. However, the fact remains that it 'is' a sequence calculated to give the transition cues necessary for one's perception of the total frame. Such a frame is therefore examined with this understanding of the circumstances attending its projection. A general look at the transition cues occurring with the change of speakers shows the two major types of syntactic cues previously identified and discussed in the presented examination of Assumption I. What is interesting to note, however, is the almost systematic occurrence of only shift cues, when the turn is between Adams, the moderator, and any of the speakers (Table 8a, Part 1), and of both reference and shift cues, when the turn is between members of the panel engaged in the conduct of the discussion (Table 8a, Part 2). The obvious reason is that Adams says nothing beyond "Mr. So-and-so wants to speak'" thereby breaking the 'referential' sequence, and making it necessary for the next speaker to start with a shift cue. On one single occasion, however, he takes part in the discussion (Entry 17): "I think it can be rather arbitrary. As Mr. Neutra pointed out, even lightning bugs have passion", upon which Ciardi retorts: "But I'm saying what's wrong with having a technical passion?" It is only here that the turn from moderator to 'speaker' produces a reference cue of transition. With the announcer, however, the case is bound to be different. He is not an actual participant in the communication, but he is, in a sense, superimposed. Because he can control his own entrances and exits, these could not be looked upon as affecting subsequent cues. The opposite would rather be true. He is the one to take the cue from the speaker he introduces; he is not a 'cue giver' but a 'cue taker'. This being the case, he sounds twice like 'Adams' (Entries 1 and 21), and twice like a participant (Entries 3 and 32).

TRANSITION

105

Table 8b shows the contextual cues occurring in the first and last sentence of every Entry. It may be observed how these always bear some reference to the main content units treated throughout the discussion. While the first sentence is supposed to pick up these significant cues from the foregoing speaker, the last may be expected to hand these down to the following speaker, as modified, amplified, or perhaps substituted for by other cues. As previously mentioned, the effect of editing on the total frame of transition should be noted. In other words, whenever transition occurs between the announcer and any member of the panel, it should be remembered that this transition is not 'actual', but calculated to resemble actuality. These shifts, though valuable in these terms, should still be identified as preconceived and superimposed. Therefore, they may be studied separately as is indicated by Entry numbers. Therewith, it is suggested that the moderator's brief statements with which he ushers in one speaker after another should also be noted separately. In a sense, these statements are also preconceived. In spite of his presence with the group, he is not really involved in the communication situation, especially as he puts in very little or nothing in the way of participation. His sole contribution is to turn in speakers. This being the case, the name Adams is just mentioned in Table 8b, with nothing to follow. But when he does contribute something, it is treated as with the other members of the panel. The assumption, that the first and last sentences uttered by panel speakers as they follow one another would normally contain 'pickup' signals of reference to significant content items of previous occurrence, seems to be justified in quite a number of cases. The closest fit to this pattern of operation is probably demonstrated in shifts between the announcer and the next panel speaker. In the shift between Entries 1 and 2, the pickup items are rhythm and way of knowing something-, they are the very words with which Ciardi closes Entry 2. In the shift between Entries 21 and 22, again, the pickup items are identical. It is on these two occasions that the announcer cannot be mistaken for a panel speaker, or an actual participant. The 'fit' in transition is too close to be true.

106

ANALYTIC AND SYNTHETIC

In the 3-4 and 32-33 shifts, however, transition is more or less like that between the actual participants in the discussion. The pickup signals are there, but they are not exact replicas. They are in a way similar to those in the 6-7 shift; they sound more incidental. In fact, it is this incidental quality that may be considered characteristic of all other shifts between panel speakers. There are always pickup signals, which vary from the most to the least incidental, from the most to the least casual. If the speaker chose to start with some anecdote, or to introduce his point at some length, it might appear for a moment that he had failed to pick up his cue from his forerunner. And yet such a cue would sooner or later come up. Again, should a speaker be interrupted before he was able to formulate his conclusions, the pickup signals for his follower might be anywhere, not necessarily in the final utterance. In other words, the first and last sentences do not always contain the pickup items. Though syntactically, both reference and shift cues among speakers seem to have a more or less fixed location, yet contextually, the so-called pickup signals could occur anywhere. And it is through the operation of these two major types of cues that the total frame of transition may be envisaged.

9 COMBINATION

Our treatment of combination, as an organizational operation, comes between transition and integration for the simple reason that it is basic to both. It does not really stand for a 'different' category of information; it is 'the' operation to account for communication in symbols. With a separate chapter devoted to its treatment, it may appear as a separate entity. This is far from being the case with any of the operations postulated in this study. It is only for the sake of analysis that combination, like transition, is considered tentatively 'in its own rights'. Combination of items, as may be seen from the examination of assumptions in the preceding chapter, is fundamental to the identification of cues serving the function of transition. The whole phenomenon of cue clustering shown to account for progression is explainable in terms of combination. Therefore, with respect to the decoder, whose task is basically the identification of cues, it is reasonable to suppose his having a tendency to add or accumulate items repeated identically, analogously, or by association. It is assumed that combination cues fall into the same two classes of 'syntactic' and 'contextual'. Syntactic cues of combination, in the sense of 'annexation', are limited to the words and, or, and their equivalents, such as in addition, furthermore, etc.; apart from these, a host of composite syntactic cues may be used for combination, when selected for 'reiteration' in close proximity. In other words, when units like and, or or are used to combine content elements like A, B, C, etc., the combination is something like A + B + C. The Gestalt depends more on contiguity than on similarity. But

108

ANALYTIC AND SYNTHETIC

when function units like that is, not until, etc. are reiterated in succession, and in close proximity, they produce combinations which are like AA, BB, etc. The Gestalt depends on both similarity and contiguity. This is considered the basis for distinguishing between 'annexation' and 'reiteration' cues on the broad syntactic side. They are not two kinds; but they produce two varieties of combination. It appears that any syntactic cue, including and, and or can operate as a reiteration cue for combination, once it is selected for reiteration in close proximity. Contextually, combination cues are the same two classes of 'primary' and 'reinforcement' cues. Differentiation being established, the decoder would expect the same recurrence of items, noted before, producing combinations which have either 'identical', or 'analogous' constituents. Both of the Gestalt principles of similarity and contiguity seem to be in operation here. 1. Assumption I Combination of items in message decoding at the integration level is produced partly through 'annexation' and partly through 'reiteration'. 2. Identification of Cues {Entry 7: Ciardi)

TABLE 9a Syntactic Annexation Cues basic (and) true husband (or) citizen about notation (or) etc.

Cues Reiteration Cues

That is This is It is That means beautiful and intense That means beautiful and intense Till the passion beautiful and intense Until the passion

109

COMBINATION TABLE 9b Contextual Cues Differentiated Cue Clusters human feelings beauty good husband good citizen humanity souls ... beautiful intense beautiful souls intense souls delivering sermons beautiful sermons intense semons

it notion it anything technical art form technically concerned notation way hands behave business playing piano technical art arts technical

passion

kisses

human feeling

poems poetry both one

technical

blubber technical passion

(.Entry 9 : Neutra) TABLE 10a Syntactic Cues Annexation Cues

Reiteration Cues

complex (and) shifting thirty (or) forty years to the Philistines (and) structuring (and) arranging and and whether ... or or or or or and and and

we have we have what have you

110

ANALYTIC AND SYNTHETIC T A B L E 10b Contextual Cues Differentiated Cue Clusters material

of the architect complex shifting new materials stainless steel plastics reinforced concrete (old stuff) new materials material considerations

best material human beings clients human beings responsiveness responses reactions them eyes ears auditive reception thermal reception inner ear sense of position sense of acceleration human beings

3. Interpretation When Ciardi employs such combinations as basic and true, husband or citizen, about notation or about the way the hands behave... the function units and and or are supposed to do the business of combination in an explicit way. These units, by themselves, seem to signal combination, positively or negatively. And when some correlation is established in terms of content value between basic and true, or between husband and citizen, etc., as would normally be expected, combination may be considered to be implicitly signalled in addition. And should the whole frame basic and true be reiterated in some analogous form, combination would get still further implicit reinforcement. Whenever this is the case with annexation cues, they are placed in a medial position between the two columns of the 'a' tables. The frame beautiful and intense is both annexational and reiterative; its constituents are both explicitly and implicitly combined. In fact, they are more implicitly than explicitly combined. Conse-

COMBINATION

111

quently, in a cluster count, they would go with reiteration cues. In order to show how many primary cues are responded to, whether these are syntactic or contextual, italics are used to indicate these cues. Whatever is listed under them is supposed to belong to them. Table 9a for Entry 7 shows 3 clusters of reiteration cues, as against 3 individual annexation cues. While the former are supposed to combine 10 items, the latter combine 6; that is, if one were to equalize items as 'single' in every case. Actually, if the sentence starting: That is ... combined two or more content units, to be combined with some two or more constituents of its parallel This is, then the net result would be a combination of 'two', not of 'four', or more. This is done for the sake of equalization. Table 10a for Entry 9 shows 4 clusters of reiteration cues combining 13 constituents, as against 4 individual annexation cues combining 8 constituents. Therefore, as far as evidence goes, it appears that reiteration overweighs annexation for syntactic signals. Contextually, however, the picture of combined items is entirely redundant. At this point, it should be re-stated that redundancy must be conceived of as a form of repetition, sometimes identical, and 'more often' analogous, or purely associative. It must be broadly maintained, if it is to account for the progression and development of items in a message organization. As previously mentioned, combination is not performed for its own sake. A comparison between cue clusters in Table 9b for Entry 7 and those in the corresponding Table for Entry 9 shows different forms of redundancy utilized for combination. Probably the most striking difference lies in greater variety among cluster constituents in Entry 7. The grouping of such items as beauty, good husband, good citizen, and humanity under the primary cue human feelings of Entry 7 is certainly different from attaching responsiveness, responses, and reactions to human beings of Entry 9; or from attaching stainless steel, plastics, etc. to new materials of Entry 9. While the former type of clustering is accomplished through subtle, ingenious repetitions, the latter is done through blank, unqualified, or slightly qualified repetitions.

112

ANALYTIC AND SYNTHETIC

However, in either case of contextual cue clustering, the form of combination is redundant and the underlying principles of perceptual organization are similarity in various degrees as well as contiguity. As far as syntactic combination cues are concerned, it has been demonstrated how these tend to be more reiterative than annexational. Generally speaking, the results seem to argue strongly for redundancy in the 'implicit', rather than the 'explicit' forms, even though stylistic individual differences would be expected to exist. 1. Assumption II The principle of linguistic duality seems to underlie the operation of combination cues. 2. Identification of Cues {Entry 12: Houghton) - Table 11 3. Interpretation The notion of 'linguistic duality' has been borne out by linguistic research at various levels, but within the narrow framework of 'phonematization' and 'grammatization', no reference being made to the 'content' side of verbal expression. However, from the point of view of communication, one cannot conceive of a linguistic principle as separate in operation from the conceptual framework of experience, as may be represented by message content in structure. The presented procedure of identifying combination cues in the structure of content would not be possible if the basic scheme of language operation were not so amenable to analysis in terms of a system of duality which encompasses both the structure of content, and the structure of expression. Fundamentally, it is possible to think of combination cues as 'syntactic' and 'contextual'. Syntactic cues are those grammatical units that may be identified as 'joiners' of content units. The 'a' tables for Entries 12 and 22 comprise a number of such connecting units. Contextual cues are those content units, which carry the sum and substance of the message. Whatever may have been left out of

COMBINATION TABLE 11a Syntactic Cues Annexation Cues

Reiteration Cues

that form, etc. (and) that somehow ... and and and

that that

and and and

TABLE l i b Contextual Cues Differentiated Cue Clusters content

form

saying anything something related to human beings what use a point something to say

form word rhythm shape line

something close to his heart

Mr. Sheet's point balance form rhythm shape line form shape rhythm line formalism formal materials words for sake of words material means to end content along with form

114

ANALYTIC AND SYNTHETIC

Entry 22: Sheets) TABLE 12a Syntactic Cues Annexation Cues

Reiteration Cues

idea (or) content rhythm (and) all ... to go over ... (and) defend speaks (and) writes

It isn't literature It is music It is a special ... l o u ' r e not dealing You're dealing It's important It's terribly ... It's not the ... It isn't just ... that never say ... that doesn't say ... that is written ...

TABLE 12b Contextual Cues Differentiated Cue Clusters different music music it special area of human feeling tone rhythm music music no idea no content doesn't say anything work of art form manner form

kinds of art literature ideas clearly uniquely stated not mere rhythm not just rhythm ideas context basic ideas concepts feelings beliefs

lecturer speaks and writes beautifully fascinating to hear him speak doesn't say anything never say anything two different arts problem

COMBINATION

115

the b tables for these two Entries belongs to the category of minor content units, those that are subsidiary ramifications, which receive no further development in the course of the message. Should one be interested in giving a detailed picture of content units, such minor elements would certainly make their appearance in the table. It is then possible to think of syntactic cues serving the function of combination as 'annexational' and 'reiterative'. This does not mean that the two classes are mutually exclusive, but that they produce two different types of combination. All the annexation cues in Table 11a for Entry 12 are simultaneously reiterative cues, with only one single exception. On the other hand, none of the annexation cues in Table 12a for Entry 22 is selected for reiteration. This seems to be a feature of individual style. Meanwhile, contextual cues, serving the same combinatorial function may be divided into 'identically redundant', and 'analogously redundant'. While there are various forms of analogous repetition, none of them can be confused with identical forms. It has been stated earlier how both classes of redundant units are essential for communication. However, so long as the aim is to achieve progression with the concurrent development of content units, there would normally be more items of the analogously redundant type than of the identically redundant. Under content in Table 1 l b for Entry 12, there is not a single item identically repeated. Under form, there is one in this particular cluster. Considering each cue cluster individually, the internal structure has two components: a 'primary' cue, followed by 'reinforcement' cues. Sometimes, the primary cue does not appear to be 'central' in the cluster following it. How far organizational primacy would coincide with conceptual centrality appears to be relative to the point of view of the encoder on one side, and the decoder on the other. For example, considering the cluster under music in Table 12b for Entry 22, should the item special area of human feeling occur first, it would be arbitrarily called a 'primary' cue for this particular cluster, according to the presented scheme of analysis. And should a decoder think it proper to call music still the primary cue, the difference between the 'given' order and the 'expected'

116

ANALYTIC AND SYNTHETIC

order would not really matter in this connection, as long as there 'is' a primary cue with other reiterative cues in the vicinity. In short, manifestations of the principle of linguistic duality seem to be copiously in existence, and would call for further tappings. 1. Assumption III In interpersonal communication of the group situation, an encoder's reaction to opposition in argument seems to be reflected in increased density of combination cues. 2. Identification of Cues Entries 13, 16, 18, and 20: Ciardi 3. Interpretation In debating this controversial issue, Ciardi appears to be the TABLE 13a Syntactic

Entry

Annexation Cues

13

None

16

18

None

20

None

Cues

Reiteration Cues

until someone ... until he ... poet or an architect or an artist / / h e ' s a small ... If he's a larger ... That's the dirty word. That's not it ... It's the way What's wrong with ... What's wrong with ... What's wrong with ... If your're going ... If you violate ... If you violate ... No one objects ... N o one objects ...

COMBINATION

117

TABLE 13b Contextual Cues

Entry 13: Ciardi

Differentiated Cue Clusters W. B. Yeats question answer

body dancer dancer someone dancer

music dance dance it dancing

Entry 16: Ciardi

ideas ideas ideas things ones

Differentiated Cue Clusters

human being human being human being being human human being

Entry 18: Ciardi passion loving getting expressed

poet architect artist artist he he he he he experiences experience art art form technical dirty word empty correctness ? way of going language spoken

Differentiated Cue Clusters wrong wrong wrong

technical medium medium

118

ANALYTIC AND SYNTHETIC T A B L E 13b (Continued) Contextual

Cues

Entry 20: Ciardi same thing at it differently music structures premises technical premises

poem not something said something happening happens

pianist what fingers doing what fingers doing

poet way thing flows form empty technically? human good human

TABLE 13c Comparative

Distribution

Entry

Total Words

Total Cues

Density per Cluster

Density per Cue

10 13 7 16 18

76 60 207 146 35

10 12 25 28 9

3 3 6 6 3

3* 4 4* 4Va* 3

20

174

20

6

3 x /4*

NOTE: The figures for Entries 10 and 7 stand for available samples of 'normal' distribution, as far as Ciardi is concerned. These are calculated to permit a tentative comparison within understandable limits. »Approximately.

predominant figure, fighting for form and technical passion. As long as his opposition can be kept under control, he is naturally relaxed: he speaks at some length, interjects incidental remarks, and pro-

COMBINATION

119

ceeds at a moderate pace. On two occasions, however, he seems to be gradually shaken by his opposition: once, following Houghton's Entry 12 and once, following Sheets's Entry 22. On the latter occasion, he retorts violently with the phrase: That's nonsense, Mr. Sheets, and says little under the circumstances, that could be studied in terms of the effect of opposition on his manipulation of combination cues. Houghton's opposition being milder, Ciardi goes on at a faster tempo, apparently saying more in less words. This is the general impression a decoder would probably get, as he listens to Entries 13, 16, 18, and 20. Table 13a shows no annexation cues in 4 Entries. All syntactic cues are reiterative: 2 clusters for Entry 13; 2 for 16; 1 for 18; and 2 for 20. This distribution of syntactic cues seems to be more or less normal, as far as Ciardi's style is concerned. See Entry 7, for example. As for contextual cues, the general picture shows density with brevity. Entry 13 shows 3 clusters, with a total of 12 units. Practically nothing is said, which is not redundant in a measure. The density ratio seems to rise even higher in Entry 16, as may be seen in Table 13c of Comparative Distribution. What is noticeable in addition is that the speaker seems to sacrifice motion and variety for the sake of greater emphasis; his reinforcement cues are, generally speaking, more or less identical repetitions of his primary cues: the only variation for human being is being human: experience has no reiterative form but experience', technical is made synonymous with some dirty word, to maintain the diversity to which he is accustomed. Entry 18 is probably the most representative of increased density of combination cues coupled with brevity. It is not only that the three main content units are trebly reiterated, but the whole syntactic frame of the question is constantly repeated. Nothing else is said beyond this. Entry 20 shows a return to normal, according to Table 13c. Comparing Entry 20 with Entry 16, one observes how the total number of words rises, while density per cue falls. On the other hand, the samples of 'normal' distribution for Entries 10

120

ANALYTIC AND SYNTHETIC

and 7 in Table 13c demonstrate, in a very 'tentative' fashion, how, under the effect of opposition in argument, density per cue rises in Entries 13 and 16, while the number of words falls. In other words, the speaker seems to reinforce his primary cues more than he usually does, and simultaneously, to reduce the total number of words.

10 INTEGRATION

Combination of the elements of a message content has been introduced as basic to transition on the one hand, and to integration on the other. It has also been stated that while transition is synonymous with progression from one end of a message to the other, integration is presumed to mean cohesion among parts within the framework of existing experience. The assumptions to be examined in this chapter have to do partly with what cohesion of parts is, and partly with how it is presumably effected, as far as the decoder is concerned. Preliminary to the analytic procedure at this stage however, one question should be raised and taken account of. In an investigation into what constitutes organizational information in verbal communication, are we interested in the 'logic' of such communication? And if so, what is our standpoint, as far as the 'quality' of reasoning is concerned? In answer to the first part of the question, one would say that this investigation is concerned at this point with the linguistic features pertaining to organization for reasoning, not with a value judgement of the reasoning process as such. In other words, when people reason with one another, they are supposed to use certain verbal techniques. They might 'substitute' some items, 'modify' other items, 'amplify' other items, etc. These are the organizational linguistic devices, which are here postulated as basic to any form of reasoning. Naturally, we are interested neither in the 'mode' of reasoning, nor in the 'quality' of reasoning. These are questions of more interest to the logician, and they are naturally outside the scope of this study.

122

ANALYTIC AND SYNTHETIC

Therefore, whether the communicator is reasoning or 'rationalizing', all we propose to do is to note what type of organizational cues he manipulates, where, how often, and with what effect on the decoder. Our aim is to discover a basic 'pattern', which would be expected to recur with other speakers of the same language. When it is stated or implied that 'substitution' cues, for instance, help a decoder accommodate a suggested relationship better, and therefore perceive it better, it should be understood that one is not concerned at all with whether such a decoder is convinced, whether he will accept or reject the presented relationship. Integration cues are, again, either syntactic or contextual. Syntactic cues are those grammatical items recognized as markers of speech segments, which 'substitute' content units, 'modify', or 'amplify' them. Contextual cues are, according to the basic scheme of combination, either 'primary' or 'reinforcement' cues. The same principle of redundancy adduced to account for combination is ultimately utilized to achieve the integration or cohesion of parts. Contextual cues would fit into the same functional framework of syntactic cues, which holds 'substitution', 'modification', and 'amplification', as organizational devices characterized by the same natural overlapping that we have witnessed elsewhere. 1. Assumption I Integration of content units is accomplished 'jointly' by syntactic and contextual cues. 2. Identification of Cues (Entry 27: Ciardi) 3. Interpretation One's view is here limited to inter-sentential syntactic cues signalling the three structural features specified as basic to the process of integration among decoded content units. In Table 14a of Entry 27, some units are listed as 'substitution' cues. What is meant in the first case, for instance, is that when the listener hears Ciardi say I want to make it into a parable of..., he excepts the

123

INTEGRATION TABLE 14a Syntactic Cues Modification

Substitution

of (what) goes (when) spring discovered (that) (while) he was sweeping this is (what) (while) he was cleaning It is (that) anything ... idea (that)

Amplification parable: (There) juggling: (He) specialists: (One) Desecration: (This) moral: (I)

TABLE 14b Contextual Cues Differentiated Cue Clusters

monastery island of peace monastery dedicated to praise

short story parable of artistic process juggler juggling juggler sweeping sweeping

monks specialists wrote hymns raised flowers illuminated manuscripts did decorations Desecration What Luce says Desecration

Sheets says poetry about ideas

juggled sessions of juggling cleaning juggler juggling sweat Virgin come down wipe sweat moral real one gist anything intensely done anything positively done any good shape praise statement experience prayer I say experience of ideas one idea praise

124

ANALYTIC AND SYNTHETIC

(Entry 29 : Ciardi) TABLE 15a Syntactic

Cues

Substitution

Modification

Amplification

think (that) is ...

(whether ... or)

Experience: (It) (more ... than) (as) R. Frost

TABLE 15b Contextual

Cues

Differentiated Cue Clusters (idea)

Experience live better more richly more alive in process Robert Frost poem: momentary stay against confusion not stay as (clarified) for a minute clarifies thing experience caught of all art

occurrence of 'something', an object or action, which could be 'artistic thinking', 'creative thinking', 'the artistic process', or anything suitable in this context. However, as soon as the decoder receives the cue: what, his expectations switch over to an expanded statement of what was of probable occurrence, and so he gets ... what goes on in the artistic process. This expansion of a content unit pending expression is introduced by means of such cues as what, that, etc. And this expansion is presumably calculated to give better accommodation to integration between the idea of the parable and that of the artistic process, by way of introduction. This is

INTEGRATION

125

not to say that other factors may not be recognized as affecting a speaker's choice of the expanded frame rather than the short frame. And this is not to say that the expanded frame always provides a better contribution to integration, but that it offers another alternative to a decoder who might need it in order to perceive the presented relationships better. Moving on to the second column of Table 14a, one observes another set of cues used to signal another type of expanded frames, referred to as 'modification' frames. The chief difference between these and substitution frames is a difference of function. Modification frames are supposed to delimit actions already expressed or pending expression. In the case of When spring came around, he found himself recovered, the whole frame, when spring came around may be looked upon as 'delimiting' the forthcoming action of the juggler 'finding himself recovered'. In this case, modification is 'prospectively' applied to action 'pending expression'. On the other hand, They were just about to run in and throw him out, when they saw the Virgin, etc., is an illustration of how an expanded frame is signalled to modify 'retrospectively' some action already expressed. The third column comprises a composite number of cues, which signal the structural device given the name of 'amplification' in the sense of expansion of something of prior occurrence. In Entry 25, Ciardi requests three more minutes to give a parable, and his request is granted with the words Speak, master. Thereupon, he starts ... There's a lovely short story of Anatole France's, etc., etc. The cue shown between brackets in this column is supposed to signal an amplification of the unit parable. Similarly, he is related to juggling, one to specialists, this to Desecration, and / to moral. To say that there, he, one, this, and I, etc., are 'amplification' cues is not to contradict one's calling them previously by the name of 'transition' cues. They are both transition and amplification cues; they fulfil complementary functions in one and the same process of language perception. It is only that one is here interested in a distinction between the different functions fulfilled by these and other units of speech.

126

ANALYTIC AND SYNTHETIC

Along these lines, one may read the results shown in Table 15a of Entry 29. That is the subject of all arts is a substitution frame. Whether it means anything or not is a modification frame, prospectively applied to Ciardi being alive in the process. As for the list of amplification cues, it is sufficiently varied in scope, although it comprises only three units. The notion of experience in this context happens to be one of those nebulous concepts, that can hardly be communicated even to artists, without ambiguity. It therefore needs amplification in various ways. One way, according to Ciardi, is to denote what it means in the broadest possible terms: It says, etc., goes the first cue. Meanwhile, a comparison is needed to demonstrate the advantage of living through an experience better, more richly than I do etc., goes the second cue. That does not seem enough to convince, not only a panel of fellow artists and creative thinkers, but also a large audience of mixed thoughts and mixed feelings. So, out goes the name of Robert Frost again: As Robert Frost said a while back, etc. The cue as, followed by the particular name of Robert Frost, is calculated to usher in amplifying information characterized by source credibility. As a matter of fact, one could list more cues under this category than just three; by and large, the development of content items consists in amplification to permit clearer vision. However, the most conspicuous cues that could be listed systematically here may be collectively recognized as units introducing: 1) comparisons explicitly stated; and 2) implicit or purely suggested analogies. With respect to contextual cues, there is practically nothing to add to what has already been demonstrated in the previous chapter about differentiated cue clusters; how they are supposed to congregate, and how they are presumed to operate with the decoder in terms of primary and reinforcement units. What specifically pertains to integration, however, is the scheme of their operation in terms of the three structural devices of 'substitution', 'modification', and 'amplification'. A look at the contents of Table 14b for Entry 27 will show how the total structure of differentiated cue clusters can be accommodated within the three-dimensional frame of reference. Juggling, sweeping,

127

INTEGRATION

and the artistic process seem to be handled as interchangeable substitutes; and so are the monks, Luce, and probably Millard Sheets. The moral modifies everything said: anything intensely well done, anything positively well done, etc. is, according to Ciardi, synonymous with the artistic experience. The whole parable as has been previously alluded to, is an amplification of the notion of the artistic process. Similarly, Table 15b of Entry 29 shows substitution in the inter-changeable character of a poem as a momentary stay against confusion, and as an experience of life caught. It shows modification through specifying the volatile nature of organization in the artistic experience. And it shows amplification in terms of Robert Frost's contribution to John Ciardi's initial statement about the notion of experience. 1. Assumption II The basic scheme of operation for integration cues seems to reveal the binary principle in action. 2. Identification of Cues (Entry 33: Ciardi)

T A B L E 16a Syntactic

Cues

Substitution

Modification

Amplification

(What) I am conveying (What) interests

(When) y o u put behaves (when)

. . . i s (what) portion/ tell m e (how) tell you (how) in (what) it means in (what) it does the fact (that) told us (what)

around (which) irrationalities (that) things (that) experiences (that)

Idea in action: (In) philosophy Creative thinking: (He) said Basic things: (You) can't get ... Profound obscurity (As)

128

ANALYTIC AND SYNTHETIC TABLE 16b Contextual Cues Differentiated Cue Clusters Idea ideas ideas into monographs

idea concept thing in the abstract idea Mr. Neutra 'pure idea' lost 18th cent, cause

problem solving rational Dean of engineering five steps limit problem qualitative analysis quantitative mathematics check quant, mathematics mechanical implementation chess playing problem solving problems to solve defining qualitative analysis quantitative mathematics

Experience experience ideas into human being portion of a life enactment of idea experience idea in action in philosophy pragmatic consequences how human being behaves accepts idea ... motivation way of measuring ideas experience what idea means to human b. what it does inside him whole bush of dendrites idea with all glands on with all fears on with all anxieties on with all joys ready with all irrationalities basic irrationalities emotions creative thinking

creative thinking basic things of life get married get married beget child die decently cardinal points of life hinge points

INTEGRATION

129

TABLE 16b (Continued) Contextual

text-book ideas

Cues important points irrationalities profound obscurities B marvellous obscurities experiences ... stir us resonance ... vibration

3. Interpretation The main verbal devices utilized for the integration of message content in perception have so far been presented as belonging to a tri-dimensional pattern. On closer scrutiny, however, it appears that the pattern of operation for these devices is basically bi-dimensional. Whether we reason or rationalize, whether we follow an inductive or a deductive procedure, or both, the underlying constituents of such an organization are presumably these two in the main 'substitution', and 'modification'. A content unit or a combination of content units is either 'replaced', or 'altered' in the course of reasoning. It may be wholly or partially replaced; and in the process of alteration, it may be qualified, tempered, delimited, or else, expanded, broadened, developed, etc. This dual scheme of operation for integration cues does not contradict the one already discussed under the foregoing assumption; it only presents a possible reduction in binary terms. As a matter of fact, the operation of the binary principle in language seems to be so pervasive as almost to dictate a scheme of dualities all the way through. In the first place, Table 16a for Entry 33 presents one of two categories of integration cues, namely the syntactic. A look at column 1 will show two varieties of substitution cues: one occupying an 'initial' position in the sentence; the other occupying a 'terminal' position. It would be interesting to investigate the relative significance of these two positions from the point of view of the decoder. Moving on to column 2, one observes two varieties of modification cues: those in the upper part modify both prospectively and retrospectively, according to their

130

ANALYTIC AND SYNTHETIC

position in the sentence. This feature has already been discussed in the previous examination of Assumption I. Those in the lower part of the column must always follow whatever they modify. They never precede it. As it now stands, column 3 comprises two types of amplification cues: one is distinctively analogical and extensional; it often suggests a comparison of some sort. This type is here signalled by the unit as. It is as distinctively analogical and extensional as and is distinctively 'combinational', and as then is distinctively 'transitional'. The other type is a composite group of shift cues employed to introduce the amplifying combination of items. However, when column 3 is mashed into column 2, the result will be a compound of two chief elements: one characterized by introducing further development and extension (now amplification); and the other comprising different features of alteration (now modification). On the other hand, Table 16b, which stands for the other chief category of integration cues, namely the contextual, shows, in turn, a dual distribution of these cues, as represented by the two columns of differentiated cue clusters: one for the notion of 'idea', and the other for the notion of 'experience'. It may be argued that the observable duality of distribution may very probably reflect the two-sided issue under discussion: 'form' versus 'content'. It is also possible that this basic duality is a characteristic of Ciardi's personal style. All this may be true to a certain extent. Observation, however, seems to support the assumption that 'duality' is inherent in the very process of differentiation, without which there could be no distribution of cues, no communication. A content analysis of any of the entries treated in this study would invariably reveal the binary structure of content in every case. In Entry 4, for instance, the major content units cluster around 1) the great art of the past; and 2) the empty art of the present. Entry 9 centres on two points: 1) technical concern with raw materials; and 2) technical concern with human response. Within this dual frame, the same devices of substitution and modification are observable about the distribution of contextual

INTEGRATION

131

cues in Table 16b for Entry 33. In a sense, all clusters under 'experience' are offered as substitutes for those under 'idea'. The whole bush of dendrites would replace the pure idea; creative thinking would replace problem solving, etc. Meanwhile, the notion of pragmatic consequences in philosophy is adduced to amplify the idea in action. The notion of the basic irrationalities is modified by the very mention of the item emotion, and by the thought of dying decently, etc. There is ample evidence here of the operation of the binary principle. 1. Assumption III For the total integration of points at conflict, a decoder's reaction suggests a summation of the main contentions and oppositions within the bi-dimensional framework of 'substitution' and 'alteration'. 2. Identification of Cues TABLE 17a Cue Exchange Contention

Reaction

(Entry 2: Ciardi) rhythm way of knowing basic to form poem not about idea about experience

(Entry 6: Fadiman) Lincoln's four score and... idea 'operative' Ideas 'of Shakespeare' as he expresses them cannot list them

(Entry 9: Neutia) best materials human beings (Entry 13: Ciardi) human question second (Entry 18: Ciardi) technical passion loving medium

(Entry 10: Ciardi) human question second (Entry 15: Stevenson) Mr. Ciardi right materials first (Entry 19: Neutra) technology envelops us man, greatest subject

132

ANALYTIC AND SYNTHETIC TABLE 17b

Cue Exchange (Entry 33 : Ciardi) (Entry 22: Sheets) four score ... out of context nonsense two different arts problem (Entry 28: Sheets) (Entry 27: Ciardi) poetry ... experience an idea one idea ... praise (Entry 29: Ciardi) (Entry 30: Sheets) an experience praise ... great idea (Entry 31 : Ciardi) greater experience

3. Interpretation It is assumed that, in the decoder's perception, organizational information in general would, under circumstances favorable to reception, promote the establishment of such interrelationships among the elements of message content, as could ultimately lead to a balance of cognition, accommodating new structures with a minimum of deviation or 'dissonance'. It is in this sense that integration has been associated with a 'cohesion' of parts, constituting the best possible 'fit' of in-coming information into the individual's repertory of conceptual data. Accordingly, in a communication situation of the type under investigation, where members of the panel form two opposing sides, the analyst decoder as an external observer, is provided with clues from the two interacting sides, which might shed some light on the intra-personal encodingdecoding process, characterized by similar interaction. In other words, the verbal techniques whereby members of the panel can accommodate their differences should, in a sense, suggest similar reactions on the part of a message decoder, who is also interested in achieving a similar equilibrium through the resolution of cognitive dissonance. Accordingly, from the point of view of a message decoder, the total integration of contributions from both sides of the panel, is correlated with the operation of cues signalling approval on the one hand, and disapproval on the other. Cases of doubt or indeci-

INTEGRATION

133

sion may be tentatively set aside. Now, in terms of the basic bi-dimensional pattern postulated for integration cues, signals of agreement between two or more members of the panel would be associated with the 'substitution' of something 'similar', or with 'amplification' of the 'same'. Conversely, signals of strong disagreement would be associated with the 'substitution' of something completely different, or with 'rejection', followed by a pause of suspense. Halfway between, are signals of mild disagreement, which might suggest some form of 'modification', or 'alteration'. In dealing with Assumption III, the distinction between syntactic and contextual cues is not considered pertinent to a discussion of the concept of 'summation' for the accommodation of differences and maintenance of cognitive assonance. Therefore, Table 17a shows a listing of the most salient cue or cues taken in the aggregate for one side of the controversy; while Table 17b is for those cues eliciting information on the other side. The aim is to reveal the most significant verbalizations at points of contact in the course of the discussion. As for the question of saliency or significance of these cues, it is determined partly by their frequency of occurrence, as previously shown in tables of differentiated cue clusters, and partly by the decoder's spontaneous reactions in his attempt to correlate such cues with the 'points at issue'. Comparing Entries 2 and 6, between Ciardi and Fadiman, one observes an attempt on the part of the latter to resolve a minor difference with the former. This is signalled by ideas as operative with the artist, then exemplified by the ideas of Shakespeare. Hoping that this 'amplification' will satisfy Ciardi, Fadiman asks: "Is that all you meant, J o h n ? " To which John answers: "That's very much a part of it." In Entries 9 and 10 between Neutra and Ciardi, followed by 13 and 15 between Ciardi and Stevenson, almost the same identical cues are exchanged. The technique is obviously 'substitution' of similar cues, and the anticipated outcome is agreement, as explicitly stated. In Entries 18 and 19 between Ciardi and Neutra, the same cue, technical passion, gets slightly 'modified' by the latter, to which the former signals his agreement: "Absolutely, we are saying the same thing, but coming at it differently."

134

ANALYTIC AND SYNTHETIC

On the other hand, a look at Table 17b will show how the 'substitution' of a different cue that could have been resorted to by Ciardi in Entry 23 is out of the question at this point. His opposition is too strong to accommodate substitution of any sort. Nothing less then 'rejection' would satisfy him, and the 'nonsense' signal rings out loud. However, as he cools off a bit, Ciardi gets into the proper mood for a more balanced interaction with Sheets in Entries 27 to the end. The technique manipulated on either side is the 'substitution' of different cues all the way through, even though basically in this type of argument some sort of resolution of the differences is often indirectly reached.

APPENDIX I

CONCLUSION

From the point of view of a message decoder, and in the light of some empirical evidence, an attempt has been made to discuss the foregoing assumptions, as categorically listed, within the basic theoretic framework postulated to sketch a model of sequential organization in verbal behavior. The three main challenges outlined in the Preface have at least determined the nature of this study, as an 'eclectic' approach to linguistics, communication oriented, both in terms of the dimensions of investigation, and the mode of presentation to the reader. As for the aim of this research, it is specifically to offer some contribution to the current trend towards maintaining the unity of linguistic theory amidst a flux of advances from neighboring fields, such as psychology, sociology, etc., that could no longer be ignored or treated as irrelevant. Besides, no matter what inclination one happens to favor in linguistics, whether it is Structural-Descriptive, Generative-Transformational, or Historical-Comparative, it would not be difficult to realize the value of integrative research, emphasizing the unity of linguistic structure at all levels. It is not enough to study 'grammar' in terms of structural signals that somehow operate with language users; or in terms of transformational rules, which though obviously artificial, are calculated to have the virtue of accounting for the correlation between 'surface' and 'deep' structures. Therefore, at least two important complementary additions have prompted the present research, which is also aimed at integration. The first is to keep in touch with the 'actual' and the 'real' in verbal

138

APPENDIX I

communication, through linguistic analysis and synthesis at the 'performance' level. The second is to hypothesize and postulate in terms of psychological principles, rather than logical or mathematical formulations. A move in the direction of 'plausible explanation' is here considered a necessary supplement. Instead of vaguely concluding that the grammar of verbal expression is a body of constraints on the freedom of occurrence of linguistic items, it is considered worthwhile to examine the psychological bases of these constraints, in an effort to account for them. Obviously, however arbitrary these constraints may be, they are there to fulfill the communicative function of a symbolic system, a function that has to do with perceptual organization for establishing correlations between message content and message form in verbal interaction. Between a theory to chart out such correlations and 'tentative' empirical evidence to support a few fundamental assumptions, the integration of language as structured content in structured verbal expression sounds sufficiently accessible to further research. In the course of this exploration, we have released ourselves from the limitations of the S - D tripartite scheme, and we have stepped out of the sentence bounds observed by the G - T school. W e have exploited the 'reality' of an 'actual' discourse, and we have attempted to sound the psychological basis of grammar in terms of human perception. Empirically, language is the type of on-going behavior we have tried to anchor between 'cues' and 'operations'. A n y postulates about language in the process of existence between source and destination must therefore be inferential. Inferential statements about language, however, must take into consideration two factors: 1) its 'process mobility', and 2) its 'integrated actuality'. The postulated correspondence between sets of cues and certain organizational operations is based on the former notion. Meanwhile, empirical demonstrations of the actual interlocking of semantic representation and verbal organization seem to provide ample evidence in support of the latter, namely the notion of 'integrated actuality'. The isolationist tendencies characteristic of early twentieth century linguistics had no 'real' support on 'scientific' grounds, as

CONCLUSION

139

some would allege. The discovery of such units as 'phoneme', 'morpheme', and subsequently 'tagmeme' was not based on a 'verifiable' assumption to the effect that each of these units of structure enjoys a degree of 'independence'. The reverse is true. There is no getting away from the 'demonstrable' 'interdependence' among all three and between all three and the semantic component. However, once the discovery of linguistic units at the code level has been made, the focus should shift on to the interdependence of constituent manifestations, in order to account for the essence of 'life' in verbal behavior. Accordingly, it is assumed that the presented work points to that stage in linguistic analysis, which is concerned with interrelations among the successive phases of verbal behavior from intention to comprehension through innervation and production, audition and perception. Hence, the term 'psycholinguistic' as applied to this type of investigation. In both parts of our presentation, the theoretic and the analytic, while the focus of attention has been deliberately kept on speech perception, incidental reference has been constantly made to other phases of the process, namely production on the one hand, and comprehension on the other. The features of circularity and process mobility as characteristic of language-in-performance have been sufficiently emphasized. Meanwhile, the selection of 'sequential' phenomena for observation has been suggested by the nature of the enquiry as 'integrative'. The aim, as previously stated, is to demonstrate 'interlocking' and 'interaction', and to go one step, beyond classification and increasingly rigorous description, into the area of plausible explanation. Due to the referred to isolationist tendencies, very little has been attempted by modern linguists within such a span. In particular, very little exists, that is sufficiently articulated to demonstrate the role played by semantic connections in the grammar of a language. This will explain the systematic presentation of 'contextual' cues as differentiated 'clusters', each of which is ear-marked by a 'primary cue', thereby demonstrating the same phenomenon of 'paradigmatic' inflectional patterning at a semantic level. Just as well, one could even conceive of 'embeddedness' at a semantic

140

APPENDIX I

level along the 'syntagmatic' dimension. So, hopefully, the student of linguistics is thus made aware of a more comprehensive view of sequential grammatical relationships, encompassing both phonological and semantic structures, and correlating the organization of message content with that of message form, or its verbal manifestation. It is in accordance with this view that our discussion of 'transformation' has been introduced. In a sense, the presented grammar of contouring, accentuation, transition, combination, and integration is the grammar that makes transformation possible in the encoding-decoding process. Briefly, the categories of organizational information elicited through the foregoing perceptual operations are the parameters of correlation between the structures of content and of expression. In the verbal communication process, message content needs to be contoured; certain segments of it need accentuation; the various units must be combined in such a way as to permit transition and total integration. These 'very' categories of organizational information underlie the use of those verbal signals or cues we have attempted to identify in our sample of a verbal discourse. They constitute the perceptual organizational functions served by all forms of communication in symbols, whether verbal or non-verbal. It is interesting to note that in other communication media, when the overt structures are manifested in lines, shapes, or colors, or even in the combined segments of a musical score, the same correspondence is presumed to hold between such manifestations and the underlying conceptual structures. The same parameters of correlation exist; only the surface cues are of a different nature. Hence, the power of this view of grammar with respect to the process of conversion named 'transformation'. Because there is nothing presumptuous about this presentation, the reader is invited to share some of the problems encountered by the author. One of these pertains to the appropriate naming of concepts outside the established water-tight distinctions in the field of grammar. The difficulty was to determine how a broader view of grammar could be discussed in terms of a narrower view. The other alternative was to risk intelligibility. The two main types of cues are

CONCLUSION

141

grammatical in the broader sense; the so-called 'contextual' cues, however, are simultaneously lexical items serving semantic representational functions. So, with the understanding that the term 'contextual' does not rule out the grammatical relational function, it has been adopted for the desired inclusion of that aspect of linguistic structure which reflects semantic connections. Accordingly similar instances of extension have had to be accommodated in description or explanation. Students and colleagues who have obliged me with their criticisms have raised a number of questions that must be dealt with at this point. Some have expressed their doubt about the order of the steps taken in performing IC cuts, in which case the levels given serial numbers up the hierarchy should not be taken seriously. Most likely, this kind of order would not work as a metric for quantifying manifestations of the contouring phenomenon, because there is much room for individual choice as to when a step will be taken, and therefore one's decisions in any case are bound to be arbitrary. The whole point is to suggest the possibility to rank order transitions in terms of the level at which they represent a constituent division. Some have enquired why the contextual cues are sometimes listed as single words, and sometimes as phrases. In answer, it has been indicated that language is assumed to be encoded and decoded variously, in narrower as well as in wider spans, but in nothing less than word stretch. In encoding, the 'span' is signalled by all the prosodic features of speech, which provide contouring cues for decoding. Some have been curious to know, judging by the absence of 'trees' and letter 'formulae', why the so-called 'phrase-structure grammar' has been resorted to in the analysis, rather than 'transformational grammar'. In answer, it should be remembered that neither 'trees' nor 'formulae' make grammar transformational. It is what the transformationalists call 'context-sensitivity' that makes grammar transformational. It is assumed that in this investigation, we have attempted to explore the correspondence between those cues that make up a verbal context and the types of organizational information that make up a grammar. Other minor

142

APPENDIX I

questions have actually been answered on page 60 under 'interpretation'. It is most likely; in fact it is expected that other individuals listening to the same panel discussion might react differently, and suggest some changes on the charting of intonation contours, junctures, etc. Reference to the discrepancies between 'given' and 'expected' under 'contouring' and 'accentuation' will show that such different reactions have actually been capitalized upon, because they support some of our main assumptions. It may now be appropriate to dwell on those features of interest and significance that have been disclosed or emphasized in the course of this study. The most outstanding of these, which still bears a little more reiteration is the role played by semantic representation in grammatization, and specifically in aiding 'transition', as the 'dynamic' aspect of 'syntax'. It has always been vaguely assumed that in verbal behavior, the two symbolic functions known as 'representation', and 'organization' are inseparable. We have had a chance to verify such an assumption at more than one level of the verbal perceptual process. The various tables of lexical cue clusters demonstrate how 'semantic inflection' may be viewed as a form of 'substitutional' organization on the one hand, and of 'positional' organization on the other. Hence, the feasibility of screening sequential relationships in discourse, guided by those lexical landmarks, and with little or no reference to those 'function' words, which are normally considered 'the' grammatical cues or signals of transition. Consonant with the expanded view of syntax to cover the operation of both verbal and semantic structures, has been the shift of focus from intra-sentential to inter-sentential relationships. In order to realize the value of that shift, it would be enough to remember that the 'surface' structure of any utterance could not possibly be correlated with a 'deep' structure in any 'realistic' sense in the absence of an adequate verbal context — to say nothing of a non-verbal context — against which such an utterance may be examined. Hence, it is hopelessly erroneous to do research on heterogeneous phrases or sentences, with the aim of disclosing the correspondence between surface and 'deep' structures. Under the

CONCLUSION

143

circumstance, the so-called 'deep' structure would not be deep enough, to say nothing of laboring on the false assumption that the semantic content of such heterogeneous phrases or sentences can be held 'constant' in changing contextual environments. This brings us to a consideration of our use of the full discourse for a corpus of verbal data. Obviously, this is meant to provide an opportunity to study language at the performance level, that is, within the framework of communication. The aim is to screen the dynamics of interaction, which are never reflected in any statement of rules, or code level generalization. In dealing with contouring and accentuation, the suggested computation of a ratio of organization value has been based on a comparison between 'given' and 'expected' cues. What is given reflects performance; what is expected reflects competence, and the discrepancies are manifestations of error. Analysis along these lines is assumed to show the actual behavior of a message decoder, who needs to get adjusted to a communication situation, to get organized, and to make predictions relative to such organization in the course of the verbal exchange. The soundness of linguistic analysis in these terms derives from the fact that it consists in extensions of, and elaborations on, the psycholinguistic responses of language users in actual speech situations. The question of rigor in such investigations then becomes relative to the amount of control provided by the researcher. Finally, the design of this exploratory study is meant to mirror the normal scientific procedure that goes from hypothesis, through data processing to verification and interpretation of findings. Although quite a few of the 'assumptions' made under each of the organizational operations may sound too familiar to warrant further investigation, the new significance attached to them in this study comes from their placement in a 'new' structure of relationships, incorporating both semantic and grammatical considerations, both verbalized output and perceived input. For instance, an austere Structuralist would vouch that the suprasegmental features of speech operate synchronously, but he would not want to shift the focus on to the function of such operation, and correlate outward form with the inner communicative aspects of verbal behavior —

144

APPENDIX I

which we have attempted to do. He would not want to 'mix' lexical items with 'grammatical' sequential features, on the assumption that these are two separate 'domains'. We are here reacting positively to this artificiality; we are here attempting to counteract the effect of lop-sided, fragmentary research in linguistics. Moreover, many students of general linguistics have expressed their anxiety to know how we normally reach the conclusions, which are often vested in the garb of 'facts' and listed under categorical headings. This legitimate source of anxiety is here taken into consideration. The reader is let in on the total structure of operation for 'comprehension' of what goes on in a sample of linguistic analysis, treating the totality of discourse, and conducted within the basic framework of one central hypothesis. Rather than provide him with heterogeneous examples from one or more languages to illustrate the various categorical aspects of some superimposed theory, or theories, and leave him with a confused image of what we do in linguistics, it is considered essential in this type of enquiry to lead the reader systematically, and go step by step, until the so-called 'conclusions' are faced in their unfinished state as 'tentative' findings. This way, we can aiford to attract attention, sustain interest and communicate more effectively, as we seek to emphasize the unity of language, and the integrity of linguistic science.

APPENDIX II

CORPUS MATERIAL

THE CREATIVE ARTIST

A Symposium held at Miami University, Oxford, Ohio, on March 23, 1959 Panel J. Ciardi, poet and Poetry Editor of the Saturday Review C. Fadiman, author, lecturer and critic N. Houghton, Managing Director of the Phoenix Theater, New York R. Neutra, Los Angeles architect M. Sheets, artist designer and Director of Los Angeles Art Institute H. Stevenson, composer musician, chairman of the Composition Department, University of Southern California P. Adams, Director of the Cincinnati Art Museum Excerpts from Recordings on form versus content in art Entry 1 : Announcer. — Early in the discussions, John Ciardi had made the statement that a rhythm is a way of knowing something. Asked to elaborate on what he meant, Mr. Ciardi said: Entry 2 : Ciardi. — A while back I was interviewing Robert Frost uh in Florida, and he put it uh as well as I uh I've ever heard it said. Uh it's something that's been buzzing around the edge of my my head for a long time, and he he helped locate it. He said when you

148

APPENDIX II

start a poem, you don't know where it's going. It's on its way uh and you don't know where it's coming out, but you do know when you've missed it. Uh then he said, now there's a question. How do you know when you've missed what you didn't know you were aiming at? Uh and he doesn't know. You do know. Y o u do know when you've missed it. Y o u know, you hope you know when you've hit it, but you can be positive you know when you've missed it. Uh the what he what he had to say next on it was that the rhythm is in it. That's that's your way of knowing it. Something about the wave length of the way it goes. There's something as primitive as the beat of Mother Goose. That's that's the rhythm of the English language. It's the meter of the race. Uh there's there's a kind of way of going. In English it's basically iambic. In other languages it uh it goes other ways. It's a kind of beat. Rhythmic sequences. Whether they're uh uh visual as in a series of lines, or in time as in a series of sounds, all structure's involved in rhythm. Structure is your way of forming a poem: it's your way of forming anything. Bringing it to shape. May I have to insist that a rhythm is a way of knowing something? That it's very basic to form? It's a way of knowing something about ourselves. One of the questions I'd like to raise is what sort of human behavior is a poem. What is a human being doing when he writes a poem ? What is a human being doing when he receives a poem? Because I'd I don't know how the others would feel about this, but I want to say very emphatically a poem is never about an idea. It's always and only about the experience of an idea. And that's a very different thing. It's what what an idea feels like when you live in it. That is, it is an experience and it has to come to you at a pace. Experience has to happen at a certain rate. And the rhythm and and the wave length of the language are the basic ways uh poetry has of testing the validity of the esthetic experience. It's the way one knows. Entry 3: Announcer. — Conflict with John Ciardi's point of view was provided by Millard Sheets. Responding to the question, is the big idea more important than concern with the actual material medium of the painter, Mr. Sheets said:

CORPUS MATERIAL

149

Entry 4: Sheets. — They're both terribly important. And I think the greatest art, uh as I look back upon the great arts of many epochs, contains both of these qualities in fair equality. It seems to me that if we do look our own age comparatively, we must recognize that we've advanced, 11 believe very honestly, tremendously in the last twenty or twenty-five years in America and throughout the world, in raising the level of esthetic understanding and feeling about painting. Putting it another way, the organization of a painting as a complete unit, the accords of color, the new concepts of handling space within space, texture values and all the things that are part of the the technical building of of a of a painting. I think we've made a tremendous progress over the perfectly banal, cheap kind of literal painting that that degenerated in the late 19th and early 20th century, where where realism was an end in itself. But by the same token, I don't think that the things that we are producing today in these large quantities, which are so much easier to produce than dealing with great ideas, will ever hold their place in terms of of comparison with great art of of many epochs of the past. I believe that ideas and deep human values are permanent. I don't think that the twentieth century has ruled these out for a moment. I think that the experimental age that we've gone through will contribute tremendously to the younger painters coming on in the next few generations, who will inherit with one stride new vocabulary, new power. But it's a pretty empty art if you compare it to great ideas. And I think that the ideas will come back into art, if we as a society have the the insight and the power and the force to develop these ideas. Entry 5 : Adams. — Mr. Fadiman wants to speak of that too, Mr. Sheets. May he? Entry 6 : Fadiman. — I just want to add a little anecdote which may be familiar to all of you involving the great French uh painter Derain and the great French poet Mallarmé. Derain was a very intelligent man who was a friend of Mallarmé who once said to him, "Mallarmé," he said, "I cannot understand why I I cannot write poetry as you do. I have such remarkable ideas." And

150

APPENDIX II

Mallarmé said, "That's your trouble; poetry is made with words, not ideas." Now that may seem paradoxical. You might say this is art for art's sake. Not at all. The poet's ideas and the prosewriter's ideas are in the words themselves. They are not something which can be summed up. What are the ideas of Shakespeare? No one knows. No one knows. You cannot list them. And if your English professor here tells you to list Shakespeare's ideas, you must see that he is fired. Now that doesn't mean Shakespeare has no ideas. Not at all. Not at all. But his notions about the world, and you may call them ideas if you will, these notions are only perfect and clear when they are expressed precisely as he expresses them and in no other way. Now a great many, 11 in the course of my job as an editor, I used to meet a great many young novelists. Most of them very bad young novelists, but they had a lot of ideas. They're full of ideas. Really, they had many more ideas than Dickens ever had in all of his life, I assure you. But the ideas were not of the sort that could be married to this particular medium, narrative prose. The first question asked of Mr. Ciardi was in connection with what he meant by his statement about rhythms. And of course he said he couldn't answer you exactly and he's quite right. But I think we can make it a little clearer this way. We can say, suppose I gave you two dates. Here's one date. Eighty-seven years ago. Here's another date. Four score and seven years ago. Now both the same date, aren't they ? The first date has no rhythm. The second date has rhythm. Mr. Ciardi could tell you just what rhythm that is and why it is that Lincoln chose to say that rather than eighty-seven. Now the second statement is a different statement than the first statement, curiously enough. The fact is the same. The statement is different. The statement four score and seven years ago is the beginning of a great idea, and you know what the idea was. You've read the Gettysburg Address. And we remember that idea is operative. The idea would be less operative if Lincoln had said eighty-seven years ago. Does that make, is that all what you meant, John ? Entry 7 : Ciardi. — That's very much a part of it. This this is such

CORPUS MATERIAL

151

a basic and true notion, and yet, it's one the audience resists so much. As soon as ... I've had this experience uh uh ever since I began with the Saturday Review... If you try to say anything technical about the art form, that means you have no human feelings. Uh that that means you hate beauty uh and you're not really a very good husband or citizen. It's it's it's as if a musician destroyed his humanity by being technically concerned about notation or about the way the hands behave in the business in the business of uh playing the piano. People think consistently that so long as their souls are beautiful and intense, they're poets. Now I think there are many beautiful and intense souls delivering sermons, for example, delivering perhaps beautiful and intense sermons. But they're not poems. Poetry is a technical art. All arts are technical. And until the passion is technical, it doesn't go. I made an aphorism for that once. Till passion is technical, kisses are blubber. Uh the that's that's you've you've got to be both ... the human feeling and the technical passion have to be one. Entry 8: Adams. — Mr. Neutra would like to speak to that too. Entry 9: Neutra. — Well, uh perhaps the material of the architect is a little bit more complex and shifting than the material of the poet, and that's why I would uh I could make a much clearer statement about it from my point of view. About thirty, thirty-five, or forty years ago, modern architecture started to be explained to the Philistines and to all the good people who didn't know why we need a new architecture. You know uh we have all these new materials. We have the stainless steel and the plastics and the reinforced concrete, what have you, and we can't go on with this old stuff. The material was put ahead of everything else, and it is my considered opinion and so I can easily say and beyond suspicion, that I haven't used new materials in a daring way. I would say that all these material considerations are very much in second place in my in my mind. I do think that the best material the architect ever gets under his hands are the human beings who are his clients. And uh to understand human beings, to be fascinated by understanding their responsiveness, their responses, their reactions is the

152

APPENDIX II

great art of structuring an environment and arranging stimuli for them. Whether it's for their eyes or for their ears or for their auditive reception or for their thermal reception or for their inner ear sense of position or acceleration and so forth. He has to understand human beings, and this is his material. I suppose that every issue of the trade magazines bring new materials and new engineering and new installations and they are always played up as the decisive factor in architecture. But once upon a time architecture was concerned with eternal issues. Entry 10: Ciardi. — May I May I ask one question of that, Mr. Neutra? Uh I'm interested in your emphasis, but can you can you go on to the human question until you know what stone will do, and what reinforced concrete will do, and what steel will do and what glass will do, and what plastic facings will do ? Uh until until your knowledge of these materials is such that you you can put them in second place? Entry 11: Neutra. — Most assuredly, I have to know first the human beings before I know what a cold wet stone will do to my skin where I have heat losses. The biological side of it and the empathic side of it, understanding human beings and systematically understanding them is the premise to know what to do about stainless steel and any other thing. I have to understand physiological optics before I can know how to use glass mirrors and so forth. I don't start with mirrors and glass. And this is the great fault of Park Avenue re uh renewal. Entry 12: Houghton. — It would appear off hand that this panel is moving in that age-old discussion of form versus content over on to the side of form. And uh let's not worry about whether we're saying anything at all. Uh the important thing is the word, the rhythm, the shape, the line uh, and I would like to come back to the support of Mr. Sheets' point which was that form and rhythm and shape and line in themselves are not enough and that somehow there must be the balance of this form and shape and rhythm and line addressed to something that relates to human beings. I don't think anybody would really dispute me, but the uh maybe they

CORPUS MATERIAL

153

would, this panel I think would. But we have heard so much emphasis on the formalism, on the formal materials as opposed to what use these were to be put to, that I would like to to try to put my weight back on the side of of suggesting that certainly four score and seven years ago is better than eighty-seven years ago and one does understand the meaning of the difference, but Lincoln was utilizing the phrase in order to make a point and he had something to say. And he wasn't just playing with words for the sake of the words in themselves. They were the material which was a means to an end, and the end was something that he had very much close to his heart to say to other people, and the content is important along with the form in which it's said. Entry 13: Ciardi. — I'd like to ask William Butler Yeats that question, uh and his answer is "O, body swayed to music, O, bright'ning glance how shall I tell the dancer from the dance." You can't separate these things, these two things. There's no dancer unless ... the the dance is nowhere until someone dances it, and he's not a dancer until he's dancing. Entry 14: Adams. — I'd like to ask a musician who practices the most purely artificial and abstract of all art forms. Mr. Stevenson, how do you speak to this point? Entry 15: Stevenson. — I think that Mr. Ciardi has the the right point of view. I can't conceive of the kind of music that would interest anyone without absolute control of materials first. Entry 16: Ciardi. — Really what you're asking is that when you say you want a poet to have ideas, or an architect to have ideas or an artist to have ideas, all you're really asking is that he be a human being. Uh and I don't think that's too much to to grant the the artist. Uh I don't care who he is, the chances are he's a human being. If he's a small human being he'll say small things about being human. He will have have small experiences in his art. If he's a larger human being, he will have larger ones, but the only way he can have an experience in his art form is technical. But uh you see, that's that's the dirty word. One thinks immediately empty

154

APPENDIX II

correctness. Uh that's the not it at all. It's the way of going. The language must be spoken. Entry 17: Adams. — I think it can be rather arbitrary. As Mr. Neutra pointed out, even lightning bugs have passions. Entry 18: Ciardi. — But I'm saying what's wrong with having a technical passion? What's wrong with loving your medium? What's wrong with believing that the only way you can get expressed what you are is through your medium? Entry 19: Neutra. — Well, technology today, which envelops us in such a terrible fashion, has produced unbearable situations, biologically unbearable situations in our community, in our cities. We have been so in love, and the technicians and the technologists and also the architects, with materials that it was overlooked that man the consumer is the greatest subject. To understand his responses and reaction, I say, is the thing. Entry 20: Ciardi. — Absolutely. We're saying the same thing but coming at it differently. Uh maybe across the hedge of of uh our different art forms we see this differently. But a poem is not something said. It's a something happening. And how does it happen unless one engages it in the terms of what it's already started? Uh if you're going to write music, you have to, you have to watch your uh structures. If you violate your premises, you're going to be in trouble. If you violate your technical premises, you're going to be in trouble. No one objects to a pianist being worried about what his fingers are doing. Uh he couldn't be a good human pianist without worrying about what his fingers are doing. Uh no one objects to this in a dancer. Uh yet, if a poet says that he is in love with the way one thing flows into another in his poem, he's supposed to be empty technically. This is the only way he can be human within his form. Entry 21: Announcer. — A question which seemed for a moment to go unanswered provided one of the best expressions of the view point of Millard Sheets. The panel was asked what they meant by

CORPUS MATERIAL

155

the word 'art' and whether they were not, in fact, talking about several different things when they used the word 'art'. Entry 22: Sheets. — I think this this question is one of the most pertinent questions that have been asked. In other words, we are talking about several different kinds of art. What's true in music, it doesn't have to represent an idea or a content. It isn't literature; it's music. It's a special area of human feeling that best expresses itself through tone, rhythm and all the other definitions that you might give for music. But in the word, in in the art of literature, you have another problem. You're not dealing with mere rhythm alone, you're dealing with ideas. The four score and seven years ago is out of context. It's important. It's terribly important, the definition that uh, the comparison you've given, Mr. Fadiman. But it's not the basic reason that we respect that particular address. It isn't just the rhythm. There are some basic ideas, concepts, feelings, beliefs that were very clearly, uniquely stated. Now I feel that the the problem that we are talking about here is is one that we can be we can leave everyone confused about by merely stressing the importance, because we are nearly always as artists, as critics on the defensive about what constitutes an organic work of art, the form, the manner in which it's done. And therefore, we are a little prone to go over on that side continuously and defend the importance of the form. But I went to a lecture one time which someone summed up like this. He speaks and writes beautifully. Of course, he doesn't say anything, but it's just fascinating to hear him speak. I don't think we have time in a world like this to listen to things that are just spoken well, that never say anything. I would love to hear music that doesn't say anything, that is written beautifully; but they are talking about two different arts, and I think we've got to face the fact that this is our problem. Entry 23: Ciardi. — But that's that's nonsense, Mr. Sheets. I mean, that's the kind of thing one says. Entry 24: Sheets. — That's bad! Entry 25: Ciardi. — I mean this remark that you quote is a piece

156

APPENDIX II

of nonsense. May I take about three minutes to give a a parable. I uh I Entry 26: Adams. — Speak, master. Entry 27: Ciardi. — Uh I think it may have something to say. Uh there's a lovely short story of Anatole France's, and I want to make it into a parable of what goes on in the artistic process. You recall the juggler of Notre Dame uh went through France for a long time and made himself a good living by juggling. He would get down on his back and toss balls up in the air and juggle them with his hands and his feet and his nose and all went well and he made a good living out of it. Then some years later one winter, ill and broke and homeless, he found himself wandering down a back road of France, and he came to a monastery. The brothers took him into the monastery and they kept him through the winter and tended him. When spring came around, he found himself recovered. He was well. But he decided he didn't want to leave this monastery. He was going to stay there; he had found a little island of peace. But he also discovered that this monastery was dedicated to the praise of the Virgin, and he liked that. But all of the monks were specialists of some sort. One wrote hymns in praise of the Virgin and one raised flowers for her altar and another illuminated manuscripts and another did decorations of a sort and only the juggler had nothing to give her. So he took to sweeping out the chapel. But one day while he was sweeping out, he was taken by an urge to to do something real for the Virgin, and he took out his mat and juggled for her. There there thereafter he uh he began to sneak in little sessions of juggling before the statue while he was cleaning out the chapel. And one day one of the monks looked in and saw this happening and called all the others and said, "Desecration of the temple." This is what the what Luce says about the novel every once in a while, 'desecration of the temple'. And the monks looked in through the window and sure enough, there was the juggler juggling in front of the statue. They were just about to run in and throw him out, when they saw the Virgin come down from the pedestal and wipe the sweat from the juggler's brow. And I uh

CORPUS MATERIAL

157

I'll leave you to pick out the moral. I uh I think it's a real one, but I think the gist of it is that anything intensely well done, any anything positively done, any good shape is a praise, is a statement, is an experience. It's a prayer. I don't know how much more meaning there is than that. Mr. Sheets says poetry is about ideas. I have to go back to my original phrasing. It's not about ideas. It's about the experience of ideas and basically the one idea that a poem has to give is praise. It says I make this. Entry 28: Sheets. — That's an idea. Entry 29: Ciardi. — And and it's it's it's an experience. It says, in making this, I live better, more richly, more truly than I do by not making this. Whether it means anything or not, I'm more alive in the process. As Robert Frost said a while back, a poem is a momentary stay against confusion. You can't get clarified to stay so. Uh you mustn't think that. You have to do it all over again. But for a minute, the poem clarifies a thing. It uh it's an experience of life caught. I think that is the subject of all art. Entry 30: Sheets. — I think praise is a great idea. Entry 31: Ciardi. — I think it's a greater experience. Entry 32: Announcer. — John Ciardi was asked if he was not, when he writes a poem, in fact trying to convey an idea. Entry 33: Ciardi. — Uh no, it is not an idea I want to convey to the reader. Let him find his own ideas. I'm trying to give him an experience. Uh yes, ideas exist. I leave those to academicians to put into monographs. What I'm conveying, what interests me is when you put this idea into a human being, what portion of a life follows in the enactment of this idea. That's an experience. It's the idea in action. Uh if you wa in philosophy, they speak of pragmatic consequences. Sometimes it's hard to evaluate an idea, a concept. The o the only real reason turns out to be all right. I don't know how to define this thing in the abstract. But tell me how a human being behaves when he accepts this idea as a motivation, and then I'll tell you how I feel about it. That's a way of measuring ideas,

158

APPENDIX II

you see. But it's the experience. All all I'm interested in for any idea is what it means to a human being, what it does inside him. Uh Mr. Neutra was speaking earlier of the fact that there's no such thing as pure idea. That was a lost 18th century cause. You've got this whole bush of dendrites turned on, your glands on, with all your fears on, with all your anxieties on, with all your joys ready, with all your irrationalities. And I refuse to be rational about this process. It's not a matter of problem solving. Uh it's it's a matter of those basic irrationalities we call the emotions and that do the most important things of our lives. Uh I found myself on a panel with the Dean of Engineering. And this panel was called a Panel on Creative Writing, on creative thinking. And he told us what creative thinking was. He said it consists of five steps: you define the limits of your problem, uh you perform a qualitative analysis, you then perform the quantitative mathematics, check the quantitative mathematics and find the mechanical implementation of your solution. And that's not creative thinking, that's chess playing. That's problem solving. I have a great deal of respect for it. We all have problems to solve. But I defy you to do any of the basic things of a life in these terms. Y o u can't get married this way, be defining the girl and performing quantitative mathematics. You can't get married, you can't beget a child, you can't die decently this way. And all the cardinal points of our life, all the hinge points, all all the things that make our lives important to us are irrationalities. They're they're full of ringing marvellous obscurities because they involve these experiences that stir us. Resonance is the word here. There's a vibration. That is nothing to do with textbook ideas as stated.

BIBLIOGRAPHY

Adrian, E.D., The Physical Background of Perception, The Waynflete Lectures (Oxford: Clarendon Press, 1946). , Sensory Integration (= The Sherrington Lectures I) (Liverpool, Eng.: University Press of Liverpool, 1949). Allen, J.P.B., and Paul Van Buren, eds., Chomsky: Selected Readings (New York: Oxford University Press, 1971). Allport, Floyd Henry, Theories of Perception and the Concept of Structure (New York: John Wiley & Sons, Inc., 1955). Anderson, Richard, and David Ausubel, eds., Readings in the Psychology of Cognition (New York: Holt, Rinehart & Winston, Inc., 1965). Ball, John, and Cecil B. Williams, Report Writing (New York: The Ronald Press Co., 1955). Bartlett, Sir Frederic, Thinking: An Experimental and Social Study (London: George Allen & Unwin, 1956). Bending, A., "Twenty Questions: An Information Analysis", Journal of Experimental Psychology 1953), 46: 346-348. Bennis, Warren G., et al., Interpersonal Dynamics (Homewood, 111.: Dorsey Press), 1964. Berlo, David K., The Process of Communication (New York: Holt, Rinehart, & Winston, 1960). Bloch, Bernard, and G.L. Trager, Outline of Linguistic Analysis (Baltimore, Md.,: Waverly Press, 1942). Special Publications of the Linguistic Society of America. Bloomfield, L., Language (New York: Holt, Rinehart, and Winston, 1933). Boulding, Kenneth, The Image (Ann Arbor, Mich.: University of Michigan Press, 1956). Broadbent, D.E., Perception and Communication (New York: Pergamon Press, 1958). Brown, R., Words and Things (New York: The Free Press, 1958). Brown, Roger, and D. Dulaney, "A Stimulus-Response Analysis of Language and Meaning", in P. Henle, ed., Language, Thought and Culture (Ann Arbor, Mich.: University of Michigan Press, 1958), 49-95. Brown, Roger, and E. Lenneberg, "A Study in Language and Cognition", Journal of Abnormal and Social Psychology (1954) 49 : 454-462.

160

BIBLIOGRAPHY

Bruner, J.S., J.J. Goodnow, and G.A. Austin, A Study of Thinking (New York: John Wiley & Sons, 1956). Carnap, R., Introduction to Semantics (Cambridge, Mass. : Harvard University Press, 1946). , The Logical Syntax of Language (London: Routledge and Kegan Paul Ltd., 1951). Carroll, John B., "The Analysis of Verbal Behavior", Psychological Review (1944) 51: 102-119. , The Study of Language (Cambridge Mass. : Harvard University Press, 1953). , "Communication Theory, Linguistics, and Psycholinguistics", Review of Educational Research 28: 79-88. , "The Interdisciplinary Summer Seminar on Linguistics and Psychology", Social Sciences Research Council, Item 5, 40-42. Carroll, Lewis, The Game of Logic (New York: Macmillan and Co., 1887). Chase, Stuart, Power of Words (New York: Harcourt, Brace and Co., 1954). Cherry, Colin, On Human Communication (New York : Massachusetts Institute of Technology, 1957). Chomsky, Noam, Aspects of the Theory of Syntax (Cambridge, Mass. : M.I.T. Press, 1965). , Cartesian Linguistics (New York: Harper & Row, 1966). ——, Language and Mind (New York: Harcourt, Brace and Co., 1968). Cofer, Charles N., "An Experimental Analysis of the Role of Context in Verbal Behavior", Trans, of the New York Academy of Science (1960) Ser. II, 22: 341-347. Condon, John C., Jr., Semantics and Communication (New York: The Macmillan Company, 1966). Cooper, F.S., A.M. Liberman, and J.M. Borst, "Interconversion of Audible and Visible Patterns as a Basis for Research in the Perception of Speech", Proceedings of the National Academy of Science, (1951) 37: 318. Current Trends in Psychology and the Behavioral Sciences (Pittsburgh: Dept. of Psychology, College of the University of Pittsburgh, 1954). Current Trends in Information Theory (Pittsburgh: Dept. of Psychology, College of the University of Pittsburgh, 1951). Dember, William N., The Psychology of Perception (New York: Holt, Rinehart & Winston, 1960). De Saussure, Ferdinand, Course in General Linguistics, Charles Bally, Albert Sechehay, eds., in collaboration with Albert Redilinger, Baskin Wade trans. (New York: Philosophical Library, 1959). Ebeling, C.L., Linguistic Units (The Hague: Mouton, 1960). Eisenson, John, J. Jeffrey Auer, and John Irwin, The Psychology of Communication (New York: Appleton-Century-Crofts, 1963). Ellis, W.D., A Source Book of Gestalt Psychology (New York: Harcourt, Brace and Co., 1938). Fant, G., Acoustic Theory of Speech Production (The Hague: Mouton, 1960). Fantz, Robert L., "The Origin of Form Perception", Scientific American, (May 1961) 204: 66-72.

BIBLIOGRAPHY

161

Festinger, Leon, A Theory of Cognitive Dissonance (Stanford, California: Stanford University Press, 1957). Flanagan, J.L., Speech Analysis, Synthesis and Perception (Berlin: SpringerVerlag, 1965). Foder, Jerry A., and Jerrold J. Katz, eds., The Structure of Language: Readings in the Philosophy of Language (Englewood Cliffs, N.J.: Prentice Hall, 1964). Francis, W.N., The Structure of American English (New York: Ronald, 1958). Fries, C.C., The Structure of English (New York: Harcourt, Brace and Co., 1952). Gleason, H.A., An Introduction to Descriptive Linguistics (New York: Holt, Rinehart and Winston, 1961). Gonzales, Richard C., and Charles N. Cofer, The Role of Adjectives as Context in the Organization of Meaningful Materials (College Park: University of Maryland, 1957a). Department of Psychology Technical Report No. 15. , The Effects of Modifiers on the Organization of Verbal Material as Shown in Clustering in Free Recall (College Park: University of Maryland, 1957b). Department of Psychology Technical Report No. 6. , A Study of Clustering in Verbs, Adverbs and Four Word Sequences (College Park: University of Maryland, 1957c). Department of Psychology Technical Report No. 16. Greenberg, Joseph H., Essays in Linguistics (Chicago, 111.: University of Chicago Press, 1958). ——, "Current Trends in Linguistics", Science, (Oct. 1959) 130, No. 3383. Grinker, R.R., Toward a Unified Theory of Communication (New York: Basic Books, 1956). Guttman, N., "Laws of Behavior and Facts of Perception", S. Koch, ed., Psychology: A Study of Science, Vol. 5 (New York: McGraw-Hill), (1963), 114-178. Halle, M., "The Strategy of Phonemics", Word (Aug.-Dec. 1954) 10, 197-209. Hayakawa, S.I., Language in Thought and Action (New York: Harcourt, Brace & Co., 1949). Hebb, D.O., The Organization of Behavior (New York: John Wiley and Sons, 1949). Herdan, Gustav, Language as Choice and Chance (Groningen: P. Noordhoff, 1956). Hill, A.A., Introduction to Linguistic Structures (New York: Harcourt, Brace & Co., 1958). Hockett, C.F., "Problems in Morphemic Analysis", Language (1947) 23: 321343. , A Course in Modern Linguistics (New York: The Macmillan Co., 1958). ."Review of Shannon and Weaver: The Mathematical Theory of Communication", Language (1953) 29: 69-92. , "Grammar for the Hearer", R. Jakobson, ed., Structure of Language and Its Mathematical Aspects, Proceedings of the 12th Symposium in Applied Mathematics (Providence, R.I.: American Mathematics Society, 1961), 220-236.

162

BIBLIOGRAPHY

Hoijer, Harry, ed., Language in Culture (Chicago, 111. : University of Chicago Press, 1954). Hovland, Carl I., The Order of Presentation in Persuasion (New Haven: Yale University Press, 1957). Howes, D.H., "The Definition and Measurement of Word Probability". Unpublished Ph. D. Thesis, Harvard University (1950-1951). , "The Intelligibility of Spoken Messages", American Journal of Psychology (1952) 65: 460-465. , "On the Interpretation of Word Frequency as a Variable Affecting Speed of Recognition", Journal of Experimental Psychology (1954) 48: 106-112. , "On the Relation between the Intelligibility and Frequency of Occurence of English Words", Journal of the Acoustical Society of America (1957) 29: 296-305. Howes, D.H., and C.E. Osgood, "On the Combination of Associative Probabilities in Linguistic Contexts", American Journal of Psychology (1954) 67, 241-258. Hull, C.L., Principles of Behavior (New York: 1945). Huppé, Bernard F., and Jack Kaminsky, Logic and Language (New York: Alfred A. Knopf, 1956). Jackson, Willis, id., Information Theory (London : Ministry of Supply, 1950). , Proceedings of a Symposium on Applications of Communication Theory (London: Butterworth Scientific Publications, 1952). Jacobson, R., C.G. Fant, and M. Halle, Preliminaries to Speech Analysis: The Distinctive Features and Their Correlates (Cambridge, Mass.: Acoustic Laboratory, Massachusetts Institute of Technology, 1952). Technical Report No. 13. Jakobson, R., "The Cardinal Dichotomy in Language", R.N. Anshen, ed., Language: An Enquiry into Its Meaning and Function (New York: Harper and Brothers, 1957), 155-178. Jakobson, R., and M. Halle, Fundamentals of Language (The Hague: Mouton, 1956). Jenkins, J.J., "Transitional Organization: Association Techniques", C.E. Osgood and T.A. Sebeok, eds., "Psycholinguistics", Supplement to Journal of Abnormal Social Psychology (1954) 52: 112-118. , "A Study of Mediated Association, Report No. 2. Studies in Verbal Behavior", N.S.F. Grant, University of Minnesota (1959). , "A Mediational Account of Grammatical Phenomena", Journal of Communication (1964). Karwoski, T.F., H.S. Odbert, and Charles Osgood, "Studies in Synthetic Thinking: II The Role of Form in Visual Response to Music", Journal of General Psychology (1942) 26: 199-222. Katona, G., Organization and Memorizing (New York: Columbia University Press, 1940). Katz, David, Gestalt Psychology. Translated by Robert Tyson (New York: The Ronald Press Co., 1950). Katz, Jerrold J., The Philosophy of Language (New York: Harper and Row, 1966).

BIBLIOGRAPHY

163

Koffka, M., Principles of Gestalt Psychology (New York: Harcourt, Brace and Co., 1935). Kohler, Wolfgang, Dynamics in Psychology (New York: Grove Press, 1960). Korner, Stephen, Conceptual Thinking (Cambridge, Eng.: Cambridge University Press, 1955). Lado, Robert, Linguistics Across Cultures (Ann Arbor, Mich.: University of Michigan Press, 1957). Langacker, Ronald W., Language and Its Structure (New Yoik: Harcourt, Brace, 1968). Lenneberg, Eric, "A Note on Cassirer's Philosophy of Language", Philosophy and Phenomenological Research (1955) 15: 512-522. Lenneberg, Eric, and R. Brown, "A Study in Language and Cognition", Journal of Abnormal and Social Psychology (1954) 49: 454-462. Lester, Mark, ed., Readings in Applied Transformational Grammar (New York: Holt, Rinehart and Winston, 1970). Licklider, J.C.R., "Basic Correlates of the Auditory Stimulus", in S.S. Stevens, ed., Handbook of Experimental Psychology (New York: John Wiley and Sons, 1951). Linsky, Leonard, ed., Semantics (Urbana, 111.: University of Illinois, 1952). Lloyd, Donald J., and Harry R. Warfel, American English in its Cultural Setting (New York: Alfred A. Knopf, 1956). L.S.A. Bulletin, No. 43 (April, 1970). Maclay, Howard and Charles E. Osgood, "Hesitation Phenomena in Spontaneous English Speech", Word (April 1959) 15, No. 1. McCawley, James D., "The Role of Semantics in a Grammar", in Bach and Harms, eds., Universals in Linguistic Theory (New York: Holt, Rinehart and Winston, 1968). Malmberg, B., Structural Linguistics and Human Communication (New York: Academic Press, 1963). Miller, George A., Language and Communication (New York: McGraw-Hill, 1951a). , "Speech and Language", in S.S. Stevens, ed., Handbook of Experimental Psychology (New York: John Wiley & Sons, 1951b). , "Information Theory and the Study of Speech", in R.A. Patton, ed., Current Trends in Information Theory (Pittsburgh: University of Pittsburgh Press, 1954a). , "Psycholinguistics", in Gardner Lindsey, ed., Handbook of Social Psychology, Vol. II. (Cambridge, Mass.: Addison-Wesley Press, 1954b). , "Speech and Communication", The Journal of the Acoustical Society of America (1958) 30: 397-398. , "Some Psychological Studies of Grammar", American Psychologist (1962) 17: 748-762. Miller, George A., and J.C.R. Licklider, "The Intelligibility of Interrupted Speech", The Journal of the Acoustical Society of America (1950) 22: 167-173. Morris, Charles W., "Foundations of the Theory of Signs", International Encyclopedia of Unified Science, Vol. I, No. 2. (Chicago: University of

164

BIBLIOGRAPHY

Chicago Press, 1938), 77-137. Reprinted in Writings on the General Theory of Signs (The Hague: Mouton, 1971). ,Signs, Language and Behavior (New York: Prentice-Hall, Inc., 1946). Reprinted in Writings on the General Theory of Signs (The Hague: Mouton, 1971). Morton, J., "A Preliminary Functional Model for Language Behavior", International Audiology (1964) 3: 216-225. Mowrer, O.H., "The Psychologist Looks at Language", American Psychologist (1954) 9: 660-694. Nagel, Ernest, "Principles of the Theory of Probability", International Encyclopedia of Unified Sciences, Vol. I, No. 6. (Chicago: University of Chicago Press, 1939), 343-422. Newman, S.E., "Effects of Contiguity and Similarity on the Learning of Concepts", Journal of Experimental Psychology (1956) 52: 349-353. Nunnally, J.C., and R.L. Flaugher, "Correlates of Semantic Habits", Journal of Personality (1963a) 31: 192-202. , "Psychological Implications of Word Usage", Science (1963b) 40: 775-781. Ogden, C.K., and I. A. Richards, The Meaning of Meaning (New York: Harcourt, Brace, and Co., 1948). Osgood, Charles E., Method and Theory in Experimental Psychology (New York: Oxford University Press, 1953). , "Factor Analysis of Meaning", Journal of Experimental Psychology (1955) 50: 325-338. , "Psycholinguistics", in S. Koch, ed., Psychology: A Study of a Science, Vol. 6. (New York: McGraw-Hill, 1963a). - , "On Understanding and Creating Sentences", American Psychologist (1963b) 18: 735-751. Osgood, Charles E., G.J. Suci, and P.H. Tannenbaum, The Measurement of Meaning (Urbana, 111.: University of Illinois Press, 1957). Osgood, Charles E., and J.J. Jenkins, "A Psycholinguistic Analysis of Decoding and Encoding", Supplement to Journal of Abnormal and Social Psychology (1954)49: 126-135. Osgood, Charles E. and Thomas A. Scbeok, eds., "Psycholinguistics: A Survey of Theory and Research", Journal of Abnormal and Social Psychology (1954) 49, supplement to No. 4. Penfield, Wilder, and Lamar Roberts, Speech and Brain-Mechanisms (Princeton, N.J.: Princeton University Press, 1959). Pike, K.L., Phonemics (Ann Arbor, Mich.: University of Michigan Press), 1947). Publications in Linguistics. , The Intonation of American English (Ann Arbor, Mich.: University of Michigan Press, 1956). Pribram, K.H., ed., Brain and Behavior (Baltimore: Penguin Books, 1969). Raffel, G., "Two Determinants of the Effect of Primacy", American Journal of Psychology (1936) 48: 654-657. Rapaport, David, trans., Organizaticn and Pathology of Thought (New York: Columbia University Press, 1951). Rosenblith, Walter A., ed., Sensory Communication (Cambridge, Mass.: M.I.T. Press, 1961).

BIBLIOGRAPHY

165

Rubenstein, H., and M. Aborn, "Psycholinguistics", Annual Review of Psychology (1960) 11: 291-322. Ruesch, Jurgen, and Gregory Bateson, Communication (New York: W.W. Norton and Co., 1951). Sapir, Edward, Language (New York: Harcourt, Brace & Co., 1921). , Selected Writings. Edited by David Mandelbaum (Berkeley, Calif.: University of California Press, 1949). Saporta, Sol, Psycholinguistics: A Book of Readings (New York: Holt, Rinehart and Winston, 1961). Saporta, Sol, and Thomas Sebeok, "Linguistics and Content Analysis", Trends in Content Analysis (Urbana, 111.: University of Illinois Press, 1959). Schatz, C.D., "The Role of Context in the Perception of Stops", Language (Jan.-March 1954) 30, No. 1, Part 1, 47. Shannon, C. and W. Weaver, The Mathematical Theory of Communication (Urbana, Illinois: University of Illinois Press, 1949). Skinner, B.F., Verbal Behavior (New York: Appleton-Century-Crofts, 1957). Sondel, Bess, The Humanity of Words (New York: The World Publishing Company, 1958). Spear, N.E., B.R. Ekstrand, and B.J. Underwood, "Association by Contiguity", Journal of Experimental Psychology (1964), 67: 151-161. Straus, O., "Relation of Phonetics and Linguistics to Communication Theory", Journal of the Acoustical Society of America (Nov. 1950) 22, No. 6, 709711. Voegelin, Charles, "Distinctive Features and Meaning Equivalence", Language (1948) 24: 132-135. , "Linguistics Without Meaning and Culture Without Words", Proceedings of the Ind. Academy of Science (1949) 58: 47-49. Abstract. Warfel, Harry, Language: A Science of Human Behavior (Cleveland, Ohio: Howard Allen, 1962). Werner, H., "A Psychological Analysis of Expressive Language", in Heinz Werner, ed., On Expressive Language: Papers Presented at the Clark University Conference on Expressive Language Behavior (Worcester: Clark University Press, 1955). Werner, H., and E. Kaplan, "Development of Word Meaning Through Verbal Context: An Experimental Study", Journal of Psychology (1950) 29: 251-257. Werner, H., and S. Wapner, "Toward a General Theory of Perception", Psychological Review (1952) 59: 324-338. Whatmough, J., Language (New York, 1957). Whorf, Benjamin Lee, Language, Thought and Reality, John B. Carroll, ed. (New York: John Wiley and Sons, 1956). Zipf, George, The Psycho-Biology of Language (Cambridge, Mass.: M.I.T. Press, 1965).

INDEX

Abstractions, code-level, 37, 48. Accentuation, 50-51, 77-87; given and expected, 85-86. Alternation, between source and receiver, 37. Analysis, 60-63, 143. Association theory, 21.

Design, of study, 143. Dialogue, 37. Dichotomies, 11, 40. Discourse, full in analysis, 143. Discrepancies, between 'given' and 'expected', 142. Duality, linguistic, 112, 129-130.

Choice, in IC cuts, 141. Cognitive assonance-dissonance, 132-133. Collective judgement, 98. Combination, 53-54, 107-120. Communication, intra-personal, inter-personal, 36-39. Competence, 87. Content units, major and minor, 84. Context, verbal, 94, 142. Context-sensitivity, 53-54, 141. Contextual theory, 43. Contouring, 48-50, 64-76; given and expected, 72-73. Cues, annexation and reiteration, 107; attention value, 81; contextual and syntactic, 62, 89, 141; density of combination, 116; potency, 88; primary and reinforcement, 62, 94; reference and shift, 90; saliency, 133.

Encoding-decoding, 31-33; circularity, 39; convergence, 39. Error, 87. Extension, instances of, 141.

Deep structures, 142-143.

Field theory, 21. Form and meaning, 45. Frame, actor-action, 49; modifier-modified, 49. Generative-Transformational, 43. Gestalt principle, of common movement, 83. Gestalt psychology, 21. Grammar as organizational information, 19. Grammar, psycholinguistic interpretation of, 42-45. Grouping versus separation, 64. Immediate Constituents, 49. Independence, of units of structure, 139. Information, representational and organizational, 32. Information theory, 51.

168

INDEX

Integration, 54-55, 121-134, 139. Integrative approach, 26. Interaction, verbal, 26; dynamics of, 36, 143; lack of, 37. Interdisciplinary approach, 12. Interlocking, of structures, 139. Irrelevance, in research, 14. Isolationist tendencies, 12-13, 139. Language, homeostatic function of, 39; instrumentality of, 36; as competence, 41-42; as performance, 41-42, 87, 143; as unified form of behavior, 19. Linguistics, Descriptive, 27, 52; self-sufficient discipline, 22.

Parameters, of correlation, 20, 140. Patterning, 44. Premises, 20. Presentation, order of, 24. Process, view of language, 37, 138. Projection versus suppression, 77. Prosodie features, 54. Psycholinguistics, 22; 1953 Summer Seminar, 20. Psychology, of adjustment, 38-39; of language, 22; of perception, 38. Reasoning process, 121. Redundancy, 111. Relationships, intra-sentential and inter-sentential, 52-53, 142. Representation and organization, overlapping of, 45.

Markov process, 53. Meaning, reference to, 43. Meaningfulness, of syntactic grouping, 45. Metalanguage, of linguistic science, 13-14. Model, hypothetical of correlation, 24.

Semantic connections, 139. Semantic import, 77. Semanticists, 23. Sequential phenomena, 139-140. Stress, 50. Structure, 43; content and expression, 27; levels, 52. Supra-segmental features, 49.

Organization, 28; levels, 33; paradigmatic, 28-29; positional, 28; substitutional, 28; syntagmatic, 28-29; value, 76, 86. Organizational information categories and operations, 29, 30, 60. Organizational principles, 46. Orientation, 11, 35-42.

Template matching, 33-34. Transformation, 20, 21, 38, 140. Transition, 51-53, 88-106; macro- and micro-structures, 95. Transitional dependencies, 51. Type-token relationships, 47. Verbal, behaviour, 22; cues, 30-31; learning, 22; message, 42; perception, 48, 51.