Lecture Notes ni Computer Science detidE by .G Goos dna .J Hartmanis 22 [[ lamroF Aspects fo Cognitive Processes Interdisciplinary Conference Ann Arbor, March 1972 detidE yb Storer Thomas dna David Winter Springer-Verlag Berlin-Heidelberg • New York 1975 Editorial Board: .P Brinch Hansen. .D Gries C. Meter • G. • Seegm011er N. Wirth .rD Thomas Storer Dr. David Winter The University of Michigan Department of Mathematics Ann Arbor, MI 48104/USA Library of Congress Cataloging in Publication Data Interdisciplinary Conference in the Formal Aspects of Cognitive Processes~ Ann Arbor, Mich., 1972. Formal aspects of cognitive processes. (Lecture notes in cc~iputer science 22) ; Includes index. .i Cognition--Congresses. .I Storer~ Thomas~ .de ii. Wiinter~ David J., ed. !II. Title. IV. Series. BF311. !56 1973 001.53 74-32111 ISBN 0-387-07016-8 AMS Subject Classifications (1970): 68-02, 68A30, 68A35, 68A45, 68A50, 68A55 CR Subject Classifications (1974): 3.6, 3.7 ISBN 3-540-07016-8 Springer-Verlag Berlin • Heidelberg • New York ISBN 0-387-07016-8 Springer-Verlag New York. Heidelberg. Berlin This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photo- copying machine or similar means, and storage in data banks. Under § 54 of the German Copyright Law where copies are made for other than private use, a fee is payable to the publisher, the amount of the fee to be determined by agreement with the publisher. © by Springer-Verlag Berlin - Heidelberg 1975. Printed in Germany. Offsetdruck: Julius Beltz, Hemsbach/Bergstr. PRE FACE The academic disciplines which have produced models of cognitive processes vary from psychology to sociology, linguistics, psyeholinguistics, mathematical biology, neuro- physiology, communication sciences, mathematics and logic, each of which may embody several distinct and diverse subdiseiplines. Further, the types of models involved range from heuristic to empirical, pragmatic, programmatic, and (varying degrees of) formalistic with every combination of these also being represented. The "Interdisciplinary Conference in the Formal Aspects of Cognitive Processes", which was held in Ann Arbor, Michigan in March 1972, was primarily devoted to formal models of cognitive processes with particular emphasis on the explication of the assumptions and hypotheses underlying new and existing models and theories in the above areas. As was anticipated, the atmosphere and the talks presented at the conference stimulated a free flow of ideas among the participants which, it is hoped, has opened continuing channels of com- munication among investigators from the disciplines represented, thereby contributing to continuing studies in the area of theoretical models of cognitive processes. We are publish- ing here a representative selection of the papers presented at the conference in further sup- pert of these studies. We take this opportunity to warmly thank those who helped us with the conference and these conference proceedings. The conference was funded by the Institute of Science and Technology at The University of Michigan. Most of the organization work was done by Lee Zukowski. The typing of the proceedings was done by Catherine Rader. Finally, we wish to thank Springer Verlag for agreeing to publish the proceedings. Thomas Storer, Buffalo and Ann Arbor David Winter, Ann Arbor November 1973 TABLE OF CONTENTS ~LCO~ ABOARD - Thomas Storer ................................................................ 1 THEORIES OF THE BRAIN; 5F/~AVIOR, THE ,DIZ]M ROBOTS AND COGNITIVE PROCESSES, RESPECTIVELY - John T. Lamendella ........................................................... 4 TWO CLASSES OF HOLOGRAPHIC PROCESSES EEALIZABIZ IN THE NEUI~ REAI~ - J: Patrick Cavanagh .......................................................... 14 SEMANTIC MEMORY RETFJEVAL: SOME DATA AND A MODEL - Elizabeth F. Loftus .......................................................... 41 IMPLICATION AS AN ALT~TIVE TO SET-II~gLUSION AS THE SEMANTIC PP_IMiTIVE - ArnoldL ewis Glass ........................................................... 55 STRUCTUEED-S~ORAGE AFA (Abstract)- Armen Gabrielian, Seymour Ginsburg ........................................... 70 PREDICATE CALC~US FEATL~ GF_~TION - David Rothenberg .............................................................. 72 A MATHEMATICAL MODEL FOR PERCEPTION APPLIED TO THE PERCEPTION OF PITCH - David Rothenberg .............................................................. 126 MODELS OF SPEECH PRODUCTION - Chin-W. Kim ................................................................... 142 TOWARDS A THEORY OF LINGUISTIC MEMORY - Terrence J. Keeney ............................................................ 159 THE G R ~ OF RELATIVE ADJECTIVES AND COMPARISON - Renate Bartsch, ~eo Verm~nann ................................................ 168 A SIMPLE HIERARCHICAL MODEL OFN ATURAL - SEIJZCTION David J. Winter ............................................................... 186 ON THE NOTION OF A RULE - Thorns Olshewsky .............................................................. 189 EMPIRICAL RESTRICTIONS ON THE POWER OF TRANSFORMATIONAL GRAMMARS - Royal Skousen ................................................................. 204 WELCOME ABOARD Thomas Storer The University of Michigan Welcome to the "Interdisciplinary Conference in the Formal Aspects of Cognitive Processes". The word "interdisciplinary" was put in the announced title in the advance flyer, by the way, as a hopeful expression of my own predisposition towards the conference participants. Things worked out pretty well in that regard, I think; the disciplines repre- sented by our speakers include computer and communication sciences, linguistics, mathe- matics, neurophysiology, philosophy, psychology, social research, and systems and indus- trial engineering--and the subdisciplines of each of these here represented is equally varied. Let me say now that the responsibility for the choice of the particular area representatives for this conference is entirely mine--the common thread, or "bond", being only (as sug- gested) an interest in cognitive processes and a formalistic approach. Perhaps a few words in that regard are warranted at this time. Speakers were t.t__on chosen on the basis of repu- tation or area but, rather, on the basis of a disciplined approach as set forth in their pre- liminary abstracts. This is not to say that I necessarily agree with either the particular approach or the conclusions of any given speaker; simply that each has something important to be said within a framework which allows for some sort of formal statement and/or analysis. And this, I feel, could be profitable for the whole community. This particular conference came about in the following way: in the past 2-1/2 years I have traveled to several conferences in one or more of the areas here represented--some- times speaking, sometimes only attending--and I found that several interesting observations were seemingly to be made. Among these are numbered: )1 These conferences were staunchly anti-formalistte. Most speakers, granted, borrowed freely from the more formal disciplines, but often abused the language borrowed from the parent discipline with respect to that discipline. Personal interchanges at these meetings left me somewhat skeptical as to whether or not precise terms, used imprecisely, were being substituted for imprecise terms. 2) Unlike the mathematics and logic conferences (the only ones of my previous active participation) the talks were often exercises in defensiveness (particularly of reputation) and developed into emotional involvement, sometimes to the exclusion of proposed content. 3) Often, at the very university which housed the conference, there were people (who, it appeared to me, were very good, and not to be dismissed lightly) who "never went to such things" even though there were in the apparent mainstream of the discipline in question, had something important and relevant to say, and were, quite obviously, very interested in the topics under discussion. Those to whom I am referring felt that the very lack of a common formal groundwork was a barrier to productive discussions and results. 4) A disproportionately small segment of the respective communities involved seemed to monopolize the addressed. In view of these (admittedly subjective) observations (on an admittedly small sample space), I resolved to someday test these observations by organizing a conference which 1) was of a formalistic bent, 2) invited speakers without regard to reputation, 3) aggressively advertised to "those usually excluded from such proceedings", and 4) excluded most of the usual run of speakers via 1), 2) and 3) above, and the injunc- tion that the only thing a speaker may legitimately assume from his audience is some degree of formal sophistication. The University of Michigan, under the auspices of the Department of Mathematics, with the sponsorship of the Institute of Science and Technology has made my resolution a reality at this time. And this conference was viewed as an experiment--a "trial run", as it were--and the response so far has been extremely positive, almost overwhelming. I (immodestly) ascribe this response to the correctness of (at least some of) my observations concerning the "usual" conference of this type, and that these reactions are shared by some subgroup of the broad community. Now I, too (in conjunction with John Lamendella), have a formal approach to cognitive processes--i e., a model (which it is not my purpose to discuss here)--and have spoken about it several times to various conferences and "interested" groups. Unfortunately, it suffers from the defect of not properly belonging to any discipline, presupposing some neuro- physiology, a little set theory, logic, graph theory (damn little), and one trivial topological result (which can be omitted if you'll take my word that various structures which are intro- duced are well--defined)--all in all (with the e~ception of the neurophysiology and the one "lifting theorem" from topology) material covered in the first four weeks of our Freshman Honors Calculus course. The model arose in conjunction with an attempt to formally char- acterize certain linguistic processes, but, recognizing the generality of the theory, we couched it in the language of "generalized" cognitive processes--a special case of which is the original characterization process under discussion Linguistic journals say, "Too formal; besides this stuff is psychology. " Psychology journals devoted to cognitive proces- ses say, "Too formal; besides, this stuff is mathematics. " Mathematics journals say, "Too interdisciplinary; and besides, the mathematics involved is trivial. Write a book." Some of the frustration generated by the above also serves as partial motive for this conference. This conference is, finally, an attempt to obviate the above and similar frustrations (which I know for a fact are not experiences unique to John Lamendella and myself) through the presentation of an interdisciplinary collection of papers devoted to the formal aspects of cognitive processes as a viable, valuable, productive discipline in its own right. The ~peakers here assembled represent my choice for the spokesmen in support of this viewpoint, and they will begin presenting the hard evidence at 1 P.M. this afternoon. THEORIES OF THE BRAIN; BEHAVIOR, THE MIND, ROBOTS AND COGNITIVE PROCESSES, RESPECTIVELY John T. Lamendella San Jose State College Many transformational grammarians have become convinced that the field of linguistics is, in reality, a branch of cognitive psychology and that a transformational gram- mar should be viewed as a cognitive theory of language. This position has led at least some linguists to wonder how one might validate claims of cognitive relevance; how one could choose between theories which equally well accounted for the emperical data but with dif- ferent psychological implications. Having asked these questions, linguists join the ranks of psychologists, physiologists, psycho-physiologists, neuro-psychologists, philosophers, and others who also do not know what they mean when they talk about cognitive processes. In this paper I will aLtempt to outline a framework for discussing questions like "What is a cognitive process ?" "What would a formal theory of cognitive processes be like ?" "What is the relationship betweea theories of the brain, theories of behavior, theories of mental processes, theories of robots, and theories of cognitive processes?" The distinctions I make apply only to an idealized world of scientific inquiry inhabited by strawmen who construct theories purely of one type rather than the messy types of theories which have a way to turn- ing up in the real world. What I give here should be taken as a plea for the establishment of a conceptual framework in terms of which we can understand what a given real world theory is a theory of. The first class of theories I would like to distinguish involves an anatomical and physiological description of the structures and processes of the nervous system. The theo- retical component of such descriptions would for the most part be hypotheses filling in the gaps of empirical observation. Physical description theories would include definitions of morphological units such as the neuron, spinal cord, and the ~q cranial nerve. In addition, they would describe functional units such as the auditory pathways, the extra-pyramidal motor system, and the timbic system. The term "functional" as used in this context reflects a concern with the spatio-temporal organization of activity in the anatomical units defined. The auditory pathways are defined as a "functional" entity on the basis of sequenced patterns of neural activity traversing specified structures in the nervous system over time. For the neurophysiologist, structure and function are inseparable and it would make no sense for the physiologist to talk of auditory pathways apart from such structures as the cochlear nuclei, medial geniculate body, etco Hypothetical functional entities such as the sodium pump in- volved in the conduction of the nerve impulse are properly posited by the physiologist only when they are construed as having some reasonably direct physical realization rather than merely being a way of looking at what happens. The notion of functional component found in many information processing models of cognitive processes is quite different. Thus, we might discuss a functional entity mousetrap apart from any spatio-temporal physical manifestation. Something is a mousetrap provided it catches mice. What physical form it has is irrelevant. Nevertheless, it should be pointed out that while a physiologist's definition of function involves a physical manifestation, physi- ological descriptions don't reduce to anatomical structures in any simple manner. Thus, for example, the auditory pathways have no direct physical existence as a distinct entity. The question is, could a physiological description of neural processes count as an explanation of cognitive processes. Certainly some scholars believe it is only because we don't know enough about the brain that we cannot give adequate physiological accounts of cog- nitive processes. However, just as a physiological description reduces to an anatomical description only in an oblique, as yet formally undefined way, so a description of cognitive processes reduces to physiological description in some even more oblique, even more unde- fined way. The concerns of the idealized psychologist are quite different from those of the idealized physiologist. For example, while a physiological theory would be concerned with the biochemical and/or structural modifications which underlie memory storage in the brain, it would not be concerned with the fact that people store information about the world and use this information in particular ways for particular reasons. If reverberating circuits, facili- tated synaptie transmission, network structures, protein synthesis, and glial processes were all involved in the storage of information about the world, the physiologist, as physiologist, would not notice. There is no physiological basis for identifying these diverse physical structures as functionally equivalent in the sense in which they all function to store informa- tion. This is the p~ychologist~s notion of function. Even if a totally adequate description of physiological processes, existed, a description of cognitive processes would still have to be given by the field of psychology. A very popular approach to the study of cognitive processes within the real world field of psychology has been to ignore them entirely and describe instead the external behavior which is the result of cognitive processes. Theories within this approach can take several forms. Early behaviorism and much so-called neo-behaviorism exemplify an approach I will call behaviorial taxonomy. Within this framework, a description of a specified domain of human behavior involves a system of predictions of the associations among external events and states; i.e., an inventory of stimulus-response pairs. The two major sources of evi- dence for the construction of such taxonomies are observation of overt behavior in relation