INFORMATION TO USERS
This was produced from a copy of a document sent to us for microfilming. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the material submitted.
The following explanation of techniques is provided to help you understand markings or notations which may appear on this reproduction.
1. The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)” . I f it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you of complete continuity.
2. When an image on the film is obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because of movement during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that should not have been filmed, you will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc,, is part of the material being photo graphed the photographer has followed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. I f necessary, sectioning is continued again—beginning below the first row and continuing on until complete.
4. For any illustrations that cannot be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In aU cases we have filmed the best available copy.
liniversiV MicixSilms International 300 N. ZEEB ROAD, ANN ARBOR, Ml 48106 18 BEDFORD ROW, LONDON WC1 R 4EJ, ENGLAND
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8119551
H o f , Su s a n n a L a r o m
IMPLICATIONS OF RESEARCH IN LANGUAGE AND THOUGHT FOR THE DEVELOPMENT OF AN INTEGRATED THEORY OF COGNITIVE STRUCTURE
The American University Pîî.D. 1981
University Microfilms I n ter n eti O n 3.1 300 N. Zœb Road. Am Arbor. MI 48106
Copyright 1981 by Hof, Susanna Larom All Rights Reserved
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IMPLICATIONS OF RESEARCH IN LANGUAGE AND THOUGHT
FOR THE DEVELOPMENT OF AN INTEGRATED THEORY
OF COGNITIVE STRUCTURE
By
Susanna Larom Hof
submitted to the
Faculty of the College of Arts and Sciences
of The American University
in Partial Fulfillment of
The Requirements for the Degree
of
Doctor of Philosophy
in
Higher Education
Signatures of Committet
Chairman :
^ / 7 Dean of the College yy
( ■L'j-’yV L { Ù /Vk / Date. / 1981
The American University Washington, D.C. 20016
THE ME R I C AH UNIVERSITY LIBRARY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IMPLICATIONS OF RESEARCH IN LANGUAGE AND THOUGHT
FOR THE DEVELOPMENT OF AN INTEGRATED THEORY
OF COGNITIVE STRUCTURE
BY
SUSANNA LAROM HOF
ABSTRACT
Recent studies within the fields of linguistics,
philosophy, psychology and neurophysiology have suggested
that inferences can be drawn from the general form of
language to the structure of human thought. More specifical
ly, such studies project the possibility of expanding
current understandings of human cognitive processes through
the analysis of language. Support for the feasibility of
such a project in the writings of Chomsky, Fodor and Katz
warrants that it be given careful attention.
The purpose of this study is to examine arguments to
the effect that valid inferences regarding the logical
structure of thought can be made on the basis of an analysis
of the logical structure of language.
The attempt to examine whether there is an underlying
basic structure to human thought and to determine how input
is received, organized and processed has traditionally been
an area of significant philosophical inquiry. In order to
establish the precedent for this study as a legitimate area
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of concern within philosophy and to pinpoint its precise
place within the range of philosophical inquiry, Part I of
the study examines the epistemological theories of Aristotle
and Kant, viewing them as attempts to analyze the structure
01: reality from within the context of our understanding of
it. While the goals of these philosophers were quite dif
ferent the approach which they took and the insights gained
as a result of their explorations have much in common with
current work being done in the cognitive sciences and in
recent philosophical research in language and mind. It is
argued that any attempt to analyze the structure of thought
through an examination of the structure of language finds
itself firmly grounded within this tradition of philosophical
concern.
Part II examines the psycholinguistic theory of Noam
Chomsky, interpreting it as partof acontinuing attempt to
adequately define the origin, nature and limits of human
knowledge through an examination of the logical structure
of language, and explores the implications of their theory
in the semantically based theories of Fodor and Katz. Part
III seeks empirical support for their conclusions within the
field of cognitive psychology considering the conclusions
of empirical studies of knowledge acquisition and knowledge
structure in the research of Bruner and Piaget. It is
concluded that while addressing the question from the
perspective of technical linguistics, Chomsky's research
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. orovides the outlines for a philosophical theory of knowledge
which establishes categories of thought and provides answers
to philosophically significant questions regarding its
structure. The additional strength of such a theory as that
proposed over those of Aristotle and Kant is that it is
firmly grounded within an empirically defensible framework.
If, indeed, it is possible to gain a significant
insight into the nature of our cognitive processes through
the analysis of language, these findings will have important
implications for a theory of learning. If the logical
structure of language is a reflection of how we logically
structure our thoughts, develop concepts and process
information, then an understanding of the form of language
and consequently of thought serves as a guideline for how
we can best facilitate learning. Part IV explores the
implications of this approach and develops an argument for
the role of philosophy within the liberal arts curriculum.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS
PART I
THE DEVELOPMENT OP A THEORY OF COGNITIVE STRUCTURE AS A TRADITIONAL CONCERN WITHIN PHILOSOPHY Chapter
I. ARISTOTLE...... 2
II. K A N T ...... 33
PART II
ATTEMPTS WITHIN LINGUISTIC THEORY TO DEVELOP A THEORY OF COGNITIVE STRUCTURE
III. THE SYNTACTICALLY BASED THEORY OF NOAM CHOMSKY . 6?
IV. SEMANTICALLY BASED THEORIES ...... 119
PART III
COGNITIVE APPROACHES TO LEARNING THEORY
V. JEROME BRUNER ...... 153
VI. JEAN PIAGET ...... l89
PART IV
CONCLUSIONS
VII. IMPLICATIONS OF RESEARCH IN LANGUAGE AND THOUGHT FOR THE DEVELOPMENT OF AN INTEGRATED THEORY OF COGNITIVE STRUCTURE . . ..222
BIBLIOGRAPHY ...... 246
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF ILLUSTRATIONS
1. The System of Human Intellectual Organization. . . . 88 2. The Formal Structure of Language ...... 128 3. The Relation of the Semantic and Phonological Components...... 135
111
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PART I
THE DEVELOPMENT OF A THEORY OF COGNITIVE STRUCTURE
AS A TRADITIONAL CONCERN WITHIN PHILOSOPHY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I
ARISTOTLE
The attempt to examine whether there is an underlying
basic structure to human thought and to determine on that basis
how information about the world is received, organized and
processed has traditionally been an area of significant
philosophical inquiry. In working toward an adequate account
of what knowledge is, and the extent to which we are capable
of obtaining it, philosophers have of necessity sought to
explore the means by which we obtain knowledge and determine
what serves as the criterion of valid and invalid knowledge.
They have also tried to explain how we originate complex ideas,
pursue intricately reasoned thoughts and define the limits
beyond which we can no longer say that we are able to know.
In philosophy, properly such questions fall within the
province of epistemology, for they are concerned with how we
obtain knowledge and organize our ideas. They have at the same
time significant implications for the study of metaphysics,
since in examining the kinds of knowledge which we are capable
of obtaining and what form knowledge takes, we become involved
in questions of what exists and of how what we call reality is
organized or structured. The adequacy of answers to these
crucial metaphysical questions is directly a function of the
answers we have given to the prior epistemological questions. 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Any attempt to analyze the structure of thought through
language also finds itself firmly grounded within this
tradition of philosophical inquiry. Whether language was
regarded as providing a picture of the world or as mirroring
thoughts with the mind, its study has been viewed as one means
by which we can uncover underlying assumptions and progress
toward a better solution to the philosophical perplexities
we are considering.
In order to establish the precedent for this study as
a legitimate concern within philosophy and to pinpoint its
precise place within the range of philosophical inquiry, it
is necessary to first consider some early attempts which
examine how the mind organizes its concepts concerning reality.
Keeping in mind that while the goals of earlier philosophers
were in many respects quite different one from another as
well as from those held today, the approach which they took
and the insights gained as a result of their explorations
have much in common with current work being done in the
cognitive sciences -- in psycholinguistics, cognitive
psychology and aritificial intelligence and in recent
philosophical research in language and mind. It will be my
purpose in Part One of this study to examine the theories
presented in the writings of two philosophers, Aristotle and
Kant, viewing them both as attempting to analyze the structure
of reality from within the context of our understanding of it.
I"Jhereas Aristotle was concerned with setting forth a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4
comprehensive account of the actual structure of reality
which he believed was mirrored in and could be determined by
an analysis of language, Kant's emphasis was on describing
the structure of our thought about a world which we can never
encounter face to face. The goals in each case were not the
same and at many points the assumptions made were clearly
contradictory, but the results of each enterprise were in many
respects remarkably similar. Wh.at does emerge in each of their
writings is an attempt to delineate the categories by whicn we
structure our thoughts about the world.
P.F. Strawson, in drawing a distinction between what
he saw as his task as a descriptive metaphysician and that
of earlier revisionary approaches, considered Aristotle and
Kant, like himself, to be engaged in the attempt "to lay bare
the most general features of our conceptual structure,"^ and
by thus doing, provide insight into the actual structure of
our thought about the world. Irrespective of whether or not
Strawson's interpretation is accurate in every respect, the
significant point for us here is that both Aristotle and Kant
can be viewed as engaged in a common exploration of the very
foundations of our conceptual structure.
It is in virtue of this common link that Aristotle's
and Kant's writings are relevant to this study. The contrasts
we see between them, however, will also prove instructive to
Ip.F. Strawson, Individuals: An Essay in Descriptive Metaphysics ( fhw York: Doubleday and Co., 1959), p. xiii.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. our purposes here. Aristotle uses a focus upon language as
one vehicle by which he arrives at a determination of his
logical categories. He makes the assumption that the
distinctions we make in language reflect distinctions in the
physical world. This assumption that language reflects in an
important respect our understandings of the physical world is
maintained in the more recent research we will consider.
Despite the legitimate claim that a distinction can be drawn
between what exists in the physical world and what can be
said to exist in our understanding of the physical world,
the grounding of the analysis in language unites Aritotle's
and more recent accounts.
Kant is much more concerned with the structure of our
ideas about the world than with the ontological structure of
reality. Reality for Kant is in itself unknowable. He wishes
to give an account of the ways in which our minds filter and
select the information which serves as the basis of our
knowledge. What he provides is an analysis of the operations
of our minds in receiving, organizing and processing ideas.
Kant's orientation, then, is clearly within the same tradition
as recent studies including those of Chomsky, Katz and Fodor
in philosophical linguistics, and Piaget and Bruner in
psychology.
It was not solely a consequence of intellectual
curiosity, nor of the pressing desire to explore and clarify
for its own sake the perplexing question of what is actually
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. implied when one says of something that it exists, that
Aristotle wrote the Metaphysics. Rather Aristotle was
carrying one step further the attempt dominant in Greek
philosophy since its earliest inception -- in the theories
of the early cosmologists, of Thales, Démocrates and of Plato
-- to find some solid basis on which to ground the inquiry
for truth.
Fitting within the schema of Aristotle's general
philosophical position, metaphysics is one of the three
theoretical sciences, all of which deal with some aspect of
being. Physics deals with matter in the form of separate,
independently existing entities which are subject to change.
These can never be fully known since they are continually
undergoing the process of transformation. Mathematics deals
with being as number abstracted from matter, but still
dependent on it. Metaphysics on the other hand, deals with
that which is unchanging, with universal substance and the
causes and principles associated with it. The subject
matter of metaphysics is being qua being, the study of being
not in terms of its characteristics as a physical object, or
a numerical entity, but in its most general sense in virtue
of its own nature as being itself. Through knowledge of the
general principles of being Aristotle believes we are able to
establish a firm basis on which to ground the inquiry for
truth.
Metaphysics, Book I opens with a short, but significant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. r\ affirmation, that, "All men by nature desire to know,"^
Aristotle's task in the Metaphysics will be to seek after
wisdom, to ascertain the most basic principles and causes
which underlie and determine the structure of that which we
call reality. This task Aristotle undertakes, and knowledge
of it he considers to comprise the wisdom of the philosopher.
It is important to note at the outset that Aristotle
makes a serious assumption in the approach which he chooses.
For Aristotle what we call reality and reality itself are
equivalent. What we see is what exists in the physical world.
Consequently, the presupposition made throughout his writings
is that the human mind is able to obtain genuine knowledge of
the physical world and this knowledge is gained through our
senses. All that we experience with our senses, the sounds
which echo through our ears, the sights which pass before our
eyes, all these are experiences of actual events in the real
world. Thus the conclusions which he draws will be based on
the belief that we can make legitimate knowledge claims about
the objective world.
Sight is for Aristotle the primary sense. As he states
in the opening passages of Book I, it is sight which "makes 3 us known and brings to light many differences between things."
Sight serves as the primary source of our knowledge and
^Aristotle, ''Metaphysics, in The Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 1941), p. 6Ô9 ^Ibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Aristotle appears to accept it's testimony as the determining
criterion upon which we base judgements concerning the nature
of reality and the manner in which it is organized. This
assumption, too, plays a decisive role in how Aristotle will
undertake hi-v exploration of reality and will have a strong
influence on how that exploration is to be resolved. He has
commenced his inquiry with this adherence to an empiricist
frame of reference. Unlike Plato who distrusted the senses
and found them to be the source of misinformation and
deception, Aristotle placed great confidence in their
reliability. Consequently how the world is presented through
the senses will be the major factor in determining the form
which his ontology will assume.
What might we expect from such an approach? One very
strong possibility which immediately comes to mind is that
from such a commitment to knowledge gained through the senses
physical objects -- the very things we encounter upon immediate
apprehension of the physical world -- might play a dominant
role in the treatment which follows. And this possibility
is indeed born out. At the very beginning of his treatise,
Aristotle has made a commitment, and that commitment results
in the elevation of objects to the most prominent position in
the metaphysical analysis which follows. We are a part of an
object world and it is the sense of sight which "brings to
light" the many differences between the entities which make
up that world.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Yet, if we consider the question more closely, for a
moment, we see that this approach is not the only possible
choice available. We might call attention to the contingency
of Aristotle's conclusions by stepping back temporarily and
asking how our perceptions and corresponding ontological
assignments might differ if we were, for instance, sightless
beings? It is possible to conceive of the world as a network
of odors, as a world structured in terms of similarities of
scents? Would we in this case categorize all similarities
and dissimilarities on the basis of whether they smelled
alike? And further, would an analysis of our conceptual
schema reveal that it was structured in terms of categories
of scent rather than classes of things? These questions are
interesting to consider. When sight is assumed to be the
primary source of our understanding of reality it is not
surprising that similarities in visual attributes become
basic categories by which the sum total is structured. Yet
when we shift our orientation we can conceive of other
possible forms of analysis which result in very different
pictures of what we term reality.
Having established this empirical bias at the outset
with its attendant presupposition Aristotle moves on to
intimate what we will later attempt to defend. In proposing
a definition of knowledge he begins to build a bridge between
his empiricism and a belief in the reality of universels.
For Aristotle true knowledge is not of things, although for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10
him things are the primary ontological entities. We cannot
say in any real sense that upon perceiving an event we have
gained knowledge. Since experience only gives us partial
insights of things within a limited temporal perspective, it
is inconceivable that any real understanding can be gained
from the immediate, intuitive experiencing of reality. Thus
we cannot like the Eastern mystic, accept things as they are
presented us in raw experience, immersing ourselves in the
beauty of that experience and emerging with deepened under
standing. Rather understanding comes after this, when we
perform the act of analysis, of critical restructuring in
terms of the more general features of reality. It is. in the
subsequent act of categorizing, of drawing general conclusions,
of seeking underlying principles that knowledge consists.
Only persons, among all living animals have the ability to
possess such knowledge and wisdom to any significant degree
since only persons, as rational animals, do more than merely
respond passively to the physical world.
Experience, then, is the experience of objects and it
is here that knowledge begins. However, the human intellect
is able to move beyond our limited experience to what
Aristotle refers to as the development of knowledge as an
"art," understanding in a fuller, broader sense. That
knowledge is the knowledge of universels, of first causes
and principles. It is based in experience, but moves beyond
it to wisdom. Once wise, we are capable of learning and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11
understanding even the most difficult of subject matter and
we are able to pass this on to others. The knowledge which
we pass on, being the knowledge of first principles, is in
Aristotle's word the most exact and most knowable of all. It
is to obtain this knowledge of the first principles and
causes, to understand the why behind events that occur in
the physical world which is the goal of Aristotle's metaphysical
enterprise, and it is this which Aristotle examines for us in
the Metaphysics. Let us look then at Aristotle's treatment
of being qua being as a delineation of the categories into
which all those things to which we attach the word, existence,
are divided.
As has been stated, the subject matter of metaphysics
is being as such. In the opening pages of the Metaphysics
Aristotle sets himself to the task of determining what is
entailed in a study of being. A direction is indicated in
Aristotle's distinction between metaphysics and the other
sciences, both theoretical and practical. Whereas the
practical sciences cut off a part of being and study that
aspect alone, as a geologist might study only the geological
features of a continent, ignoring its other aspects, only
metaphysics treats being as a whole, in its universal aspect.
This it does by determining what are the most basic principles
which being in all its various senses shares. The proper task
of metaphysics then is to study the most basic principles and
ultimate causes of that to which we apply the term, real.
This will be his method of approach.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 ?
The second step in accomplishing this task is to
determine the nature of being itself. There is, Aristotle
insists, more than one sense in which things are said to be.
In our language we encounter many uses of the term, all having
various meanings. But despite this the meaning for us is not
ambiguous. We understand what it means to say that something
exists. Within the range of meaning there is a common sense
in which anything "is." Being in its primary sense refers to
substance. All that exists in some sense derives its being
from the concept of substance.
Having established already that the primary task of
the metaphysician is to determine the first principles
underlying being, Aristotle proceeds.
But since we are seeking the first principles and highest causes, clearly there must be some thing to which these belong in virtue of its own nature.4
One might think that Aristotle, being concerned as he
was with the principles of proper argumentation, might be
conscious of a weakness at this exceedingly crucial stage
in his reasoning. The question we are concerned with is what
is being or what does it mean to say that something exists.
Aristotle treats these as the same question. They may not be.
Having established that the task of metaphysics is to seek the
first principles of being as such, here he appears to be
making the undemonstrated assumption that there must be some
4 Aristotle, Metaphysics, p. 733.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13
thing to which these principles belong. The assertion that
there are principles was posited on the existence of being as
a central concern of metaphysics. But then once the existence
of these principles is established, Aristotle uses this as the
basis on which to ground his rationale for seeking that thing
to which the term, being, can be applied. At the least the
approach seems somewhat circular. TÆiy must we assume that
there is something called being which can be identified or
defined. Could we not as well be seeking the principles
underlying existence? If we grant that there are basic
underlying principles which can be applied to all aspects of
reality solely in virtue of their existence there is no
logically binding reason why we must conclude that those
principles must in some manner be linked to something called
being which is located in time and space. There are other
equally viable conclusions one might draw. For instance,
existence might be a state in which everything which we
consider real shares? Could it not even be the case that
nothing in the sense of things, or objects exists at all or
that they exist only in aderivative sense as they did for
Plato? Why does Aristotle choose the tact he does? It might
be convincingly argued that it is in part from an enchantment
with the overriding power of our visual experience that
Aristotle seems led into this object centered approach in
dealing with the question of what exists.
This tendency on Aristotle's part is further strengthened
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14
and given support by the medium in which he works. He viewed
language as a musician views his music, as expressive of a
deeper, fuller reality. He saw the distinctions made in
language as indicative of distinctions existent in the physical
world. If objects are given primacy in the language, then
they are also treated as the primary ontological category.
One might at this point respond that it is actually
the reverse that is the case. It is because objects hold a
primary ontological position that they function as they do
within the language. However true this may or may not be,
it remains the case that in looking at the full range of
Aristotle's categories one can see that the criterion used
to draw distinctions one from another was a linguistic one.
Renford Bambrough in his commentary on the writings
of Aristotle sees Aristotle's metaphysical doctrine as
essentially an account of the world as we know it, in contrast
to the physical world. It is not surprising that Aristotle
should consider the world as we know it to be synonymous with
the world itself. Bambrough proceeds.
He set himself to reveal and describe the systematic relations that can variously be understood as relations between terms in a language, between the concepts expressed by the terms, or between the things or phenomena described in the l a n g u a g e5 .
^Renford Bambrough, The Philosophy of Aristotle (New York : Mentor Books, 1963), p. 36.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15
Such a concern grounded in and using the subject-predicate
relations of language as a starting point for an analysis of
reality might legitimately be subject to criticism. As
Bambrough continues,
He /AristotleT" has been accused with some plausibility of reading into the world itself a structure that is suggested by the structure of his language, but that need not be supposed to be the actual metaphysical structure of the world.6
Aristotle, Bambrough points out, did not see himself
as engaged in a description of language, but rather of reality.
If Aristotle were to enter into the modern dispute to which I am now referring /that over whether a distinction is logicaT or ontological, linguistic or metaphysicalT there is little doubt that he would describe his metaphysical work as ontological rather than as linguistic or l o g i c a l .7
Aristotle's concern is with how things are, and this he
believes to be revealed in our linguistic descriptions of
reality. Bambrough defends Aristotle's failure to adequately
draw this distinction. He finds it a difficult matter to
know and to explain at what points and in what respects our
language does and does not represent the world that it is
used to describe. The fact that language has the structure
it has, claims Bambrough, is not from mere accident, but
because it allows speakers to deal effectively with the world.
There seems to be a fit between language and the world which
Gibid., p. 3
7fbid., p. 36.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16
makes it more than coincidence that language takes the form
that it does. Implied in Bambrough's defense of Aristotle's
approach is the belief that there is some validity to
proceeding with an analysis of language as a means of learning
about the world. For our purposes the fact that Aristotle
does not appear to be concerned with distinguishing between
these two levels allows us to look at his analysis as an
examination of the network of our conceptual view of being,
as well as of the structure of reality itself.
Granting that Aristotle is looking at language as
mirroring our thoughts and then assuming that we can
legitimately infer from these to facts about the real world,
what do we learn from the study of language which will satisfy
our primary concern with delineating the categories into which
reality as we know it is divided? The grammatical structure
of Greek as well as of English is organized around the subject- g predicate paradigm. They both employ the subject as the
dominant term in the sentence, while the other terms realize
their function in reference to the subject, either locating it
in time and space, describing it, qualifying it, referring to
it in terms of the actions which it is performing, or
designating the interrelations between two or more subject
terms. Since on this view words are representing concepts in
our minds and these in turn correspond with the entities which
^Bambrough, p. 34.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17
make up the physical world, underlying the purely grammatical
distinctions are logical distinctions, reflecting our under
standing of reality. The picture which Aristotle draws us of
reality is in most essential respects a picture of the logical
relations which hold between the concepts in our language.
It is formulated around this paradigm in which the subject
is the dominant term and the other terms derive meaning from
their relationship to it. In the physical world substance is
the primary ontological category and all other categories
derive their existence through their interaction with it.
The concept of being, though grounded in substance,
is applicable to all the categories. All that exists either
is a substance or is dependent on the existence of substances
and cannot exist independent of them. Taking reality as
comprising all that exists and inclusive of all that we
theoretically can ever know, we can divide it into ten distinct
categories, each in terms of its function in reference to the
concept of substance. Either something is said to "be"
because it is a substance, because it is a quantity defining
substance, because it is a quality of substance, because it
is position substance can assume, a place in which substance
can be located, or a time in which substance exists, because
it is a condition of substance, because it is an activity or
absence of activity of substance (passivity), or because it
designates a relation between substances. Aristotle alludes
to these senses of being in the Metaphysics. However a more
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18
detailed treatment is provided in his logical treatise, the
Organon.
It is difficult to understand precisely what Aristotle
accomplishes in his delineation of the categories. By his own
testimony the categories comprise all that can possibly exist.
This in itself is a surprisingly strong assertion. Given the
complexity of the physical world, is it possible to subsume
it completely within ten categories? And if it is, on what
basis should we accept these as being the categories into
which it is divided. We must question how adequately
Aristotle's analysis accounts for reality as we know it.
Leaving unsatisfied the question we raised as to whether
substance should be given that position of dominance which it
has, Aristotle proceeds to examine more closely this first
category, laying out for us specifically what it entails.
"Substance is thought to belong most obviously to bodies."^
Book VII presents a list of the types of entities which this
might include. Among them are animals and plants and their
parts, fire, water and earth, and the physical universe and
its parts. Thus a person would be a substance. That person's
hand would fall into the same category as well assumably as
would his finger, his fingernail, and so forth. Everything
which qualifies as a part of that person would qualify as
equally a substance as would the whole. Aristotle has begun
his inquiry with the assumption that what we understand as
^Metaphysics, p. 784.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19
substance in some essential sense is equivalent to physical
objects and the matter of which they are composed. However,
he goes on to open up the possibility that there might be
other things which are substance besides the objects of the
senses. Traditionally the term has been applied in four
different ways, to essence, to universels, to the genus and
to the substratum. The fourth possibility is essentially
equivalent to what Aristotle has alluded to above, if we
view matter as an underlying substratum out of which all
physical objects are formed. If substratum were the equivalent
to substance then only physical objects would fall into this
most basic category and indeed this comes close to Aristotle's
position. It characterizes substance in its most primary
sense. However it is impossible to equate matter alone with
substance. Where substance by definition is capable of
existing as an independent entity, matter is not capable of
independent existence. Matter can not exist independent of
its form. It has only the potential of existence and this
potential is actualized when it becomes an object. Conse
quently he is not willing to grant that this is the only
sense in which the term applies. He goes on to examine the
other three possibilities.
What does it mean to propose that substance is essence?
By essence is meant that in virtue of which a thing is what it
is, its "propter se." The term can be applies only when there
is a definition of something primary and where there is no
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 0
predication of one element by another. The crunch comes,
however, when we try to determine what is primary. Using
Aristotle's example there is no essence to a white horse.
That would be a complex concept. However, there is an essence
to horse as well as to white viewed as the white of this
particular white surface. Thus essence appears to apply more
to the general level of universels. Horses have an essence
by virtue of which all animals who look alike bear the same
name. This holds true as well for white. Essence is that
which makes a thing what it is, its formal cause. When we
seek the first causes of being we find we are seeking the
general cause which determines its form.
lÆiat does it mean to view substance as a universal?
Aristotle points out that it is inconceivable that any
universal term could be the name of a substance. The
substance of a thing is that which is pecular to it and not
belonging to anything else. Universels are shared by classes
of objects. It would not be possible for something which
exists in more than one object to at the same time be that
by virtue of which a thing is what it is. In this sense a
universal could not be a substance. At the same time
universels are predicable of a subject. Substances on the
other hand perform the subject function in a sentence.
Thirdly, if universels were substances they would exist prior
to everything else which they qualify. Finally, universels
exist in something. If universels were substances existing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21
in something, another substance, we would have two substances
occupying the same place. But as substances no universal
could exist in another substance. Something cannot be both
a universal and a substance.
What does it mean to equate substance with genera? In
regard to the first principles Aristotle finds it difficult
to determine if it is the genera that should be considered as
the ground of being or rather the primary parts of a thing.
He seems torn between viewing matter as substance or as the
form which that matter takes, the genus. In some respects
genera cannot be regarded as the underlying substance. It
cannot exist independently. Yet substances do exist
independently. On the other hand, if we know a thing by its
definition and that is made up of genera, we can see that
those genera do serve as the principles of definable things.
Any of these traditional analyses of the concept of
substance remains for Aristotle insufficient for adequately
defining the term. Aristotle poses the problem thus; If we
do not suppose substance to be separate in the way individual
things are separate, we destroy substance in the sense that
we understand it.
However, if we do suppose them to be independent and
separable, how do we conceive their elements and principles?^^
Aristotle concludes that whereas the individual thing is truly
substance in the primary sense, universels are substances in
10Metaphysics, p. 891.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22
a way which will satisfy all the logical requirements which
appear to be required Aristotle brings in three concepts, the
terms of which are reminiscent of Plato. These are matter,
form and privation. On the basis of these principles which
interact to determine the nature of substance he feels an
adequate account can be given. This account is framed in
terms of potentiality and actuality. Rather than adopting
one of the traditional approaches viewing substance as essence,
universal, genera or substratum, through looking at reality
in terms of a process of coming to be, of potential reaching
actuality, Aristotle is able to combine the best elements of
each account.
All objects are a unity of matter, the material cause
of their existence, and form, and formal cause of their
existence. Matter, a potentiality, can never exist in and
of itself. In this condition it is an undifferentiated mass
having no form. To attain actuality it must be combined with
form thence becoming instantiated as an object in the physical
world. Neither can form exist without matter. It is through
matter that it is able to realize its potential in the physical
world. The form of a tree for example is that which causes
the tree to have the form it does. That combined with matter
becomes actualized in a material tree. The human being too
is a unity of matter and form. The form in this instance
Aristotle calls the soul. Although this may not seem to
parallel the description of form and matter in the case of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23
tree, if we see soul, the rational capacity, as did Aristotle
as the essence of a person, we can see that it is what
determines the form that the matter takes. The sc“:l; then,
is the essence of what it is to be a person.
Although matter and form united in the objects of the
physical world are substances in the primary sense there is a
secondary sense in which Aristotln holds that the genera are
substances as well. Does he take this tack to satisfy the
precedent set in earlier accounts which gave to the classes
into which reality is divided a status more than just a
collection of individuals? In looking at the paradigm case
of persons as rational animals, persons as independently
existing entities in the physical world are primary substances
In Aristotle's analysis the genus of rational animal is
substance in a secondary sense and as such it in a sense takes
on the status of an entity on the conceptual level. It is
this unity characteristic of entities which seems to legislate
that it be treated as a substance rather than as something
which is predicated of a substance.
Somewhat surprising, however, is Aristotle’s assertion
that secondary substances share a higher reality then
particulars. This may reflect in part the bias of Aristotle's
concerns toward resolving the question of how we can obtain
valid knowledge. Although secondary substances are dependent
for their own existence on primary substance, they hold the
dominant position from an epistemological point of view.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24
lÆiereas things in the physical world are forever changing,
being generated and being destroyed, and thus can never be
fully known, universels do not change. Although we know
individual objects in the sense that we encounter then, it
is of universels that we have real understanding. Wisdom
is gained through the study of the universels snd this
always presents itself in terms of a potentiality. Secondary
substances viewed as the causes which determine the form of
matter of any particular object also have primacy from an
ontological point of view since they exist prior to that
object in every case.
Has Aristotle been successful in developing a
conception of substance which is both adequate and inclusive?
In order to evaluate whether the ten categories can account
for the broad expanse of reality we must examine cases which
have been problematical with any ontological account,
universals, mathematical entities, objects which are coming
to be, no longer exist or have only existed in our imagination.
How well has Aristotle dealt with the pioblem of
universals? Aristotle's initial task was a difficult one.
To avoid the scepticism of early Greek philosophy it was
necessary to provide an adequate account of the possibility
of obtaining knowledge which has the force of certainty, but
to accomplish this from within an empiricist framework. In
doing this Aristotle gave a place to universals as the ground
or our knowledge of the physical world. However, he -jas also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25
alert to avoid the tact taken by Plato, of giving universals
a status similar to things and counting them as independent,
imperishable entities. On such an account as that the universe
would become proliferated with entities. On one level we would
have things, then a second level of existence peopled by these
"things in themselves" corresponding one for one to the first
level. There would be no logical reason why, Aristotle points
out, we need stop here. It would be just as consistent to
posit the existence of a third level of entities corresponding
with the second and so on ad infinitum.
In considering alternative approaches Aristotle
maintains that it would only be right to give universals this
separate existence if indeed they were substances. However,
this could not be the case. Universals exist in individual
objects. And it is not possible that one substance can exist
in another since they are independently existing entities.
"Clearly," Aristotle concludes, "no universal term is the name
of a substance and no substance is composed of substances.
The account which Aristotle has given of reality is (as
contrasted with that of the idealist) a realistic one. It
corresponds with the universe as we know it through experience.
In this sense it is more satisfactory than the account given
by Plato. However, has Aristotle really fully avoided the
problems in the Platonic view? He still maintains the status
of universals, in a secondary sense at least, as substances.
^^Metaphysics, p. 810.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26
The orientation is still a "thing" orientation. Looking at
the problem in a more contemporary perspective, are there ways
to avoid a substance interpretation of universals? Could
universals have been dealt with for example from the point of
view of the process involved in processing information? Then,
rather than having a somewhat ambiguous nature, as substances
yet not fully as substances, they could theoretically be
handled from the point of view of how we classify information.
On this view what causes two or more objects to have the same
name would not be some common content which they shared --
the substance view -- but what we deem to be an important
similarity between them which causes us to group them
together.
A secondary area of extreme complexity involves the
ontological status of mathmatical entities. Do such
mathematical concepts as the number two share an independent
existence? Are the objects of mathematics substances? Just
as treating universals as substances would result in the
proliferation of an infinite series of entities, which are
not objects of our immediate experience but which hold a
special status as derivative from the objects of our immediate
experience, so too Aristotle feels the treatment of the objects
of mathematics as entities within our ontological schema
would share a similar result. He sees his own approach to
the defining of substance in terms of a unity of matter and
form as providing the possibility of an alternative to this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27
knotty ontological problem as well.
Aristotle chooses to examine the logically possible
modes of explanation which might account for the existence of
objects of mathematics. If objects of mathematics exist, it
will be either in the sensible objects to which they refer
(e.g. in the case of two books, the two would exist somehow
in the books as whiteness in a sense exists in the color of
their pages) or as separate from sensible objects. It is
impossible for mathematical objects viewed as substances to
exist in sensible things since, as has already been shown, one
substance cannot exist in another. They also could not exist
as independent entities which are separate from sensible
objects since, according to Aristotle's reasoning two solids
cannot exist in the same place at the same time (e.g. we
would have the two books and the two existing at the same
place although the two would be separate and independent of
the books).
Aristotle continues, if mathematical entities do not
exist either in or separate from the sensible objects to which
they refer there remain only two possible alternatives.
Either they do not exist at all or they exist in some special
way. Aristotle chooses the second course of explanation.
If we were to view objects of mathematics as separate
from the sensible things to which they refer the consequence
would result in many confusions. Wherein would be the unity
which holds together the number and that to which it is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28
attached? If the mathematical entities are prior to the
sensible object how could we imagine them existing apart?
Having shovm to his satisfaction that the objects of
mathematics are not substances in a higher degree than bodies
and not prior in being, nor existing somewhere apart from
bodies, and having shown that they could not alternatively
exist in sensibles either, he concludes that they must exist
in some special sense and thus only with qualification. When
we state propositions and make demonstrations about sensible
magnitudes these are made not about the object of reference
as sensible, but as possessed of certain definite mathematical
qualities. Objects of mathematics exist in a special sense.
They in fact are the sensibles, but as possessed of mathematical
qualities they temporarily can be regarded as mathematical
objects. Hence, we are justified in saying in this limited
sense that mathematical entities exist. Aristotle feels that
by treating mathematical entities as sensible objects viewed
in a special light he is able to avoid the problems earlier
accounts were unable to resolve.
The third problematic category is made up of entities
which do not exist, or no longer exist. If we can speak
meaningfully about such entities it has been suggested that
they must in some sense exist. (See Quine, From a Logical
Point of View for a discussion of this issue). Yet, how do
we account for their ontological status? If, as Aristotle
states, "non-being . . . non-being"!^ is a proper
IZMetaphysics, p. 732
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o9
application of the term, 'is,' what account can we give of
the use of existential terms which will include reference to
non-entities?
To satisfy this difficult case we must return to
Aristotle's discussion of the meaning of 'being' in Book IV
of the Metaphysics. He states.
There are many senses in which a thing may be said to 'be,' but all that 'is' is related to one central point, one definite kind of thing, and is not said to 'be' by a mere ambiguity . . . . So, too, there are many senses in which a thing is said to be, but all refer to one starting-point; some things are said to be because they are substances, others because they are affections of substances, others because they are a process towards substance, or destructions or privations or qualities of substance, or productive or generative of sub stances, or of things which are relative to substance, or negations of one of these things or of substance itself. It is for this reason that we say even of non-being that it is non-being.
Thus, there is for Aristotle a sense in which we can
say that non-entities exist. Where most of the things we talk
about could be said to exist in a positive sense as does the
continent of North America, according to Aristotle something
which exists in our imagination, could be said to exist in a
negative sense, as a privation of substance.
In the preceding pages we have examined Aristotle's
account of the ontological structure of reality. Proceeding
from this understanding of the role of metaphysics as the study
of being qua being, we have examined the nature of being and
^^Ibid., p. 732,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30
then of substance -- the core concept from which the meaning
of 'being' in all senses is derived. Finally we have seen how
Aristotle interprets those categories of existence which are
not substances but derive their meaning from the primary
category of substance. As we have pointed out, Aristotle's
account is ontological. However, because Aristotle bases his
ontological distinctions on distinctions made in language his
description of the ten categories of existence can be
considered as an inventory of our conceptual categories as
well.
Two major criticisms can legitimately be leveled at
Aristotle's overall approach. The question as to whether or
not one can draw any conclusions about reality through an
analysis of language was raised at the onset on this chapter.
Bambrough has defended such an approach noting that it is no
coincidence that language is an effective tool in describing
the world. Language captures the distinctions which exist in
the world and functions adequately in communicating information
about the world from one person to another. If there were not
a close parallel between language and reality this would not
be the case. P.F. Strawson also defends such an approach from
another perspective. There are basic categories, he says,
which are the "commonplaces of the least refined thinking;
and are yet the indispensable core of the conceptual equipment
of the most sophisticated human beings.It is with these.
^■^Strawson, Individuals: An Essay in Descriptive Metaphysics, p. xiv.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31
their interconnections and the structure that they form, that
the metaphysician should be primarily concerned.
The second cirticism is aimed at the feasibility of
the task which Aristotle has undertaken. In their essay on
Aristotle, Anscombe and Geach raise the question of whether
it is at all possible to compile a complete list of the
categories into which reality divides.The complexities
of this task, as they point out, are complicated by the
problem of determining which predications are simple enough
to be a category. It is their opinion that Aristotle has not
provided us with conclusive evidence that reality is as he
has described it. Where the approach presents interesting
possibilities, it cannot be taken as an accurate rendering of
the facts. "On the other hand," they point out, "the idea of
a category-difference, which is suggested by the contrast
between predications in one or the other categories is
certainly a useful one. Where it does not tell us how the
physical world might best be analyzed, it might be the case
that a similar approach will prove exceedingly useful in
broadening our understanding of the nature of our conceptual
structures. This, however, will remain to be seen in what
follows.
^^See G.E.M. Anscornbe and P.T. Geach, Three Philosophers (Ithaca, N.Y.: Cornell University Press, 1963).
l^Ibid., p. 15.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32
In Chapter II we will compare Aristotle's approach
with that of Immanual Kant. We will then go on to develop
a theory of cognitive structure basing our analysis on
conclusions of research in linguistics and cognitive
psychology.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I"
KANT
It is possible to imagine kinds of world very- different from the world as we know it. It is possible to describe types of experience very different from the experience we actually have. But not any purported and grammatically permissible description of a possible kind of experience would be a truly intelligible description. There are limits to what we can conceive of, or make intelligible to ourselves as a possible general structure of experience.^
Thus P.R. Strawson sets forth on an exploration of
the intricate passageways of Immanuel Kant's Critique of
Pure Reason. It will be our purpose in the pages which
follow to look at the epistemological theory presented in
the Critique within the context of the most generally accepted interpretations which attempt to reconcile the exceedingly complex and often contradictory passages inter lacing the work. Our concern will be to disentangle from what is not always fully explicit a true -- or at least coherent -- rendering of Kant's own mature thinking relevant to the questions of concern in this study. Thus we may
reveal more clearly the topographical features of Kant's cognitive geography and the grounds which it establishes for what will count as such a "truly intelligible descrip
tion" of the general structure of our experience and
-p.p. Strawson, The Bounds of Sense (London: Methuen and Co., Ltd., 1966), p. 15 33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34
the world as we know it.
Following the lines of interpretation of Norman Kemp
Smith's 1923 commentary our attention will be directed primarily
to Book I of the Critique of Pure Reason, "The Transcendental
Doctrine of Elements," Part 2, Division I, "'The Transcendental
Analytic," where Kant is concerned with his development of the
categories of the understanding. The passages which will be
cited comprise what appears to be the third of four stages in
the development of Kant's own theory. The first two stages
have been termed pre-critical. In the first, Kant's attention
is directed primarily to an elucidation of the concept of
transcendental object. At this point in the evolution of his
theory the categories have not yet been developed and
integrated. In Stage Two the categories are introduced, but
as yet Kant has not included his theory of synthesis through
the productive imagination which will play a primary role in
the mature theory. The position which is enunciated in Stage
Three comprises the core of Kant's Critical Philosophy. Stage
Four, which appears in the first edition but is dropped in the
second, revolves around the three-fold transcendental synthesis.
This concept proved so out of harmony with earlier sections 2 that Kant apparently decided it could not be retained.
Like Aristotle, Immanuel Kant in the Critique of Pure
Reason represents one of the heights to which the human mind
^Norman Kemp Smith, A Commentary to Kant's Critique of Pure Reason (New York: Humanities Press, 1923), p. 224.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35
has strained in its attempt to utilize its full capacity in
the contemplation of deeply complex philosophical issues.
As Aristotle was fueled by the inadequacies of earlier philo
sophical treatments to fully account for the manner in which
we compile our wealth of knowledge about the world and our
place as knowing subjects within it, so too Kant was balanced
on a point of seemingly irresolvable tension between the
polarities of eighteenth century rationalism and empiricism.
He served as the fulcrum which drew these forces into balance
and provided a possibility for harmonious interplay utilizing
the energies bound within each.
The Critique of Pure Reason is an investigation into
the powers of reason viewed as a pure faculty of the human
mind. Drawing aside for the sake of scrutiny the incidental
conditions of temporal experience, Kant was attempting to
uncover and to examine in a systematic manner what could be
accepted as those pure conditions within the human subject
for knowledge. Where in formal terms the Critique was
specifically addressing questions of the manner in which a
priori knowledge can be obtained and the limits beyond which
such principles cannot be held to apply, for our purposes what
we are most concerned with examining are Kant's attempts to
analyze the structure of what he terms pure reason. By drawing
aside the particular temporal conditions of human phenomeno
logical circumstance what he wishes to accomplish is that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35
O "laying bare" of the most general features of our conceptual
structure.
Where traditionally it had been assumed that knowledge
conformed to the object and thereby our knowledge claims gained
validity as reflections of a real world, Kant introduced what
he, at least, considered, a revolution in philosophical
thought. He turned this epistemological assumption on its
head and tried to show that on the contrary it is the object
which must conform to our knowledge of it. In describing
his revolution in metaphysical orientation he stated in his
Preface to the second edition.
We should then be proceeding precisely on the lines of Copernicus' primary hypothesis. Failing of satisfactory progress in explaining the move ments of the heavenly bodies on the supposition that they all revolved round the spectator, he tried whether he might not have better success if he made the spectator revolve and the stars to remain at rest. A similar experiment can be tried in metaphysics, as reg^ards the intuition of objects. If intuition /sense perception/ must conform to the constitution of objects, I do not see how we could know anything of the latter a priori; but if the object (as object of the senses) must conform to the constitution of our faculty of intuition, I have no difficulty in conceiving such a possibility.^
Thus Kant leads us forward on the exploration of the
realm of pure reason by first setting out what are determined,
a priori to be the necessary preconditions to sense experience.
3p.F. Strawson, Individuals: An Essay in Descriptive Metaphysics, p. xiii. ^Immanuel Kant, Critique of Pure Reason, tr. Norman Kemp Smith (New York: St. Martins Press, 1965), p. 22.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37
Drawing together the threads of his theory we must begin with
the view Kant held of reason. Reason appears to be a core
concept representing the functioning of the human mind on all
levels. Subsumed under it are sensibility, or intuition, in
which the mind is acted upon by external things, and under
standing which is the capacity of thinking either about the
objects of sensibility or of ideas which are produced in the
understanding. A third faculty of reason which has also been
suggested as a subcategory of Reason^ is reason in the sense
of the capacity for uniting the manifold of sensations through
reference to an unconditioned principle. This function,
however, might better be seen as one task performed by the
understanding.
If we are to have knowledge of objects both sensibility
and the understanding must interact together. There can be no
knowledge without first establishing a foundation from which
that knowledge is derived. This foundation provides the
matter or content of our knowledge. So, too, there can be
no knowledge without understanding. We can have no awareness
of objects until they become subsumed within the categories
of our understanding. This provides the framework or form
within which the content is organized.
The mind is the faculty in which the activities of
sensibility and understanding are integrated. What is derived
from sense experience is a bare representation of an object.
^Frederick Copieston, A History of Philosophy, Vol. 6, Part 2 (New York: Doubleday, 1962), p. 102.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33
Eut the object of which we become aware, while empirically
real, is real only within the context of our experience of it.
This understanding of what we mean when we label something
"real" must be clearly differentiated from any reality which
exists independent of our experience. When Kant says that an
object is real, it is real in this sense -- that it is a real
object of experience. For the empiricist, empirically real
has a very different and more objective determination.
Every object of experience is such a harmony of matter
and form. The matter is that which conforms to sensation and
is derived from experience. The form consists of the neces
sary conditions of all experience. It is that which in the
act of sensation structures the sensory manifold while en
abling the manifold of sensation to be arranged within a
fixed set of relational elements. These relations which the
manifold obtains comprise the a priori element of our
experience. They find expression through our experience, but
are not perceived by the senses.
Kant derives the concept of a priori from out of the
rationalist tradition. However, his use of it appears to func
tion on at least two levels. On the first level he speaks, as
indicated above, of the a priori elements of experience. These
elements are a priori because, not derived from sensation, they
find their source outside experience. Hence, in this first
sense a priori refers to the source of such concepts. Kant,
however, is suggesting a second use of the term with he
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39
addresses the nature of a priori knowledge as being
distinguished by universality and necessity. The use relates
to a distinct feature of a priori knowledge rather than to its
source. We have certain ideas which are known to be true
necessarily and universally. Since we cannot derive necessity
or universality from experience these ideas must have some
other source. That source is not derived from the matter of
experience, but from its form. They are prior to experience
or a priori.
It is the second meaning of the term which functions
in the more substantive sense for what is significant about
a priori ideas is that they are necessary and hold true
universally. However, it is the first use of the term which
is primary, for an idea is a priori if its source is prior
to experience. A priori ideas derived from sources other
than experience are, then, "marked" by universality and
necessity.
All experience comes to us structured in terms of the
forms of the intuition and the categories of the understanding.
It is these sets of preconditioned forms and categories which
make up the a priori element of experience. As such they are
prior to experience and hold true universally and necessarily
of all experience whether simple or complex.
Since we are concerned here with the relational
structure of ideas, it is the categories of the understanding
which will prove more relevant to our task. However, before we go on to look more closely at these a priori characteristics
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40
of knowledge as they serve to structure how we think, let us
first contrast them with the forms of the intuition as they
serve to structure the form of our sensibility in terms or
space and time.
Both the forms of intuition and the categories of the
understanding have this aspect in common, that they are "laid
upon" the matter of experience. However, where the forms of
intuition are necessary features of sensory experience and
determine the form of our perceptions, the categories of the
understanding are the a priori features of the understanding
and determine the logical structure of our conceptions. They,
in a sense, provide us with a catalogue of the logical types
into which we could sort and categorize our ideas. If in
fact Kant was able to provide such a list he would have
provided the key to our major concern here, to have laid out
a schematic blueprint of the content of our cognitive
capabilities.
Looking to the forms of intuition we see that all
experience is structured in terms of space and time. These
interact with the manifold of sensation to produce the content
of our experience. Rather than being relational features of
the content of the manifold, they serve as preconditions of
experience itself. The first, space, is the form of all
appearances of the external senses. When we perceive objects
we always necessarily perceive them in terms of location in
space. Time, the second, is the form of the internal sense.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41
It is the a priori condition of all experience, both external
and internal. Since all experience is filtered through our
own self-conscious awareness of it, this form provides a
temporal framework where one sensation always stands in
temporal succession in reference to those proceeding and
following it. Where we can imagine empty time, we cannot
imagine any perception not in time.
It is Kant's conclusion that the a priori nature of
space and time is an unavoidable consequence of the nature of
experience. Since space and time are preconditions for
sensibility they cannot at the same time be derived from it.
They must exist a priori. Space and time are empirically
real since they are necessary for empirical reality to exist
at all. However, they are also transcendentally ideal, for
viewed in respect to things considered "in themselves" they
are not real, but ideal.
Kant's conception of sensibility is basically a passive
account which, contrary to possible expectation, does not
approach sensibility as that which gives us knowledge (or
what we believe to be knowledge) of objects, but rather as
that brought about by the object. The object then serves the
active role in the course of experience despite the fact that
Kant has engendered it with an exceedingly tenuous ontological
status. One might have expected to find in a theory such as
that presented in the Critique the view that sensibility
caused the object or apparent object to exist. But Kant has not chosen this route.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42
Having now established these invariant features of
sensory experience, Kant concludes the "Transcendental Aesthetid'
and moves on to what has been considered by Smith to be the
major portion of the Critique^ the "Transcendental Analytid'
this comprises the first division of Part 2, the"Transcendental'
Logic. The Analytic is concerned specifically with the
development of the concept of pure understanding (as contrasted
with sensibility) and with the derivation and enunciation of
the categories of the understanding as the mode through which
the understanding correlates ideas.
In viewing Aristotle's metaphysical program we saw that
the conclusions to which he was drawn might be considered a
direct consequence of the view which he held of the nature of
logical statements of fact. Logic is the science of the laws
of the understanding. Dominated by a view of prediction based
on a subject-predicate paradigm,Aristotle was led to a
substance based mode of categorization in which the content
of our conceptual matrix could be analyzed down into categories
deriving from the major category of substance. In a real
sense such a view of predication, that which characterizes
traditional logic, circumscribed the lengths to which Aristotle
was able to move in giving an adequate account of our cognitive
structure.
Kant too was working within the context of the general
logic of categorical propositions. However, as a part of the
^Smith, A Commentary of Kant's Critique of Pure Reason, p. 191.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4:
revolution of transcendental metaphysics Kant introduced what
he felt would serve as an addendum to traditional logic. This
he termed (not surprisingly) transcendental logic. General
logic has as its subject matter the absolutely necessary
(a priori) laws of thought in abstraction from all content.
In its pure form it takes no account of the empirical,
psychological conditions under which the understanding
functions. Rather it is formal, discursive and purely analytic.
Transcendental logic on the other hand abstracts only from the
empirical content, thus restricting itself to the a priori.
Where the content of general logic is comprised entirely of
truths whose a priori features are also analytic and thus carry
us no further than the content included within the concepts,
the propositions of transcendental logic are synthetic.
Through them we gain a more adequate understanding of empirical
reality. Where the central problem of general logic is, "How
does the understanding gain knowledge of itself?" transcendental
logic is concerned with the understanding in its interplay with
sensibility. The question asked is, "How can the understanding
possess pure a priori knowledge of objects?" It is the answer
to this question which occupies the 'Transcendental Analytic."
The final section, the’’Transcendental Dialectic'applies these
conclusions in determining the proper limits to which a priori
knowledge must be applied.
Thus Kant saw general and transcendental logic as two
cooperative and mutually supporting ventures. Norman Kemp
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44
Smith finds Kant's derivation of a new form of logic to be
of exceeding philosophical interest. Rather than standing
beside traditional logic, Smith views it as an important
anticipation of developments which have been made in
nineteenth and twentieth century logical theory. Looking at
the conclusions of Kant's dialectic, we could conclude that
what he is essentially doing is responding to the inadequacies
in the traditional treatment and in the categories, providing
a much broader spectrum of logical elements. As Smith noted.
Modern logic, as developed by Lotze, Sigwart, Bradley and Bosanquet, is, in large part, the recasting of general logic in terms of the results reached by Kant's transcendental inquiries.1
If this is the case, then transcendental logic is not sup-
Q plementary to general logic, but is "its tacit recantation."
Such a conclusion, if valid is relevant to our interests
here, for what we then have in Kant's theory is a more
adequate extension of Aristotle's categories of conceptual
thought. That innovations in modern logic support ideas similar
to those Kant proposed would lend strength to the value of the
categorizations Kant proposes and would lead us closer to an
adequate rendering of the organization and logical structure
of our ideas. The success of this enterprise will have to
await evaluation in what follows. However, if it does have
value, this stems from the logical rather than psychological
^Smith, A Commentary, p. 181.
^Ibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45
nature of the conclusions proposed. In addressing the
question as to whether his transcendental logic is based on
psychological principles Kant concluded.
But in Logic the question is not of contingent, but of necessary laws; not how we do think, but how we ought to think.9
It has an obligatoriness not possible in a psychological
account because it deals with the a priori.
Looking back for a moment at our progress to this point,
knowledge as defined in the Critique is derived from the
interplay of sensibility and the understanding. We receive
the material content of that knowledge from sensibility and
this is structured in terms of the a priori forms of space
and time. A formal structure is then further impressed upon
this by the understanding. The forms which this impression
takes are determined by the a priori categories of the under
standing. The remainder of the Analytic is concerned with
filling out this picture. It attempts to establish the
objective validity of the categories in general and of each
separate category through the use of the principle of synthetic
a priori judgements. This proof of the objective validity of
the categories must be determined a priori in order that it
hold true of all knowledge in all circumstances. In doing
this Kant must also delineate for us the synthetic a priori
judgements which arise when we apply the pure categories to
sensible objects. These judgements are the principles of the
^Immanuel Kant, Logic: Einleitung, i, quoted in Norman Kemp Smith, A Commentary on Kant s Crigique of Pure Reason, 2nd. e d . (New York: Humanities Press, 1923), pT 170.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46
pure understanding. Finally, he must determine what conditions
must be the case in order that the categories can be so applied
These conditions are the universal, necessary characteristics
which must belong to all objects known by means of our senses
and are termed the transcendental schemata. We have within
the "Transcendental Analytic," then, an interplay of at least
three levels, the categories, their justification through the
analysis of judgement and the schematic conditions for
application of these categories to the content of sensibility
which is structured in terms of space and time and consequently
to the objects of experience.
It is in the activity of the understanding in judgement
that Kant feels he can determine the key to a systematic and
complete account of the categories of the understanding. By
analyzing the forms of judgement derived from formal logic
Kant establishes a somewhat artificial and rigid set of
categories which he "boldly asserts"'^ exhausts all the
logical possibilities by which the mind can unite and
interrelate ideas. As Smith concludes.
Formal logic, Kant would seem to hold can supply a criterion for the classification of the ultimate forms of judgement just because its task is relatively simple, and is independent of all epistemological views as to the nature, scope, and conditions of the thought process.
In tying his guarantee of the completeness of the
^^Smith, A Commentary, p. 181.
^^Ibid., p. 185.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47
categories so closely to the logical forms of judgement Kant
places himself in a vulnerable position. "If not all his
synthetic a priori judgements are absolute, if some are but
relative to changing subject matter, then Kant's claim to
have provided a complete system of these judgements can not
be justified.
Judgements are of three types, analytic a priori,
synthetic a posteriori, and synthetic a priori. Where
synthetic a posteriori judgements are so closely aligned in
contents to the material of perception that they hardly
deserve the designation of judgement, synthetic a priori
judgements are the basis on which the Categories of the
Understanding are derived. These objective but empirically
based judgements unify perceptions by means of explicitly
used perceptual categories which form into a synthesis with
the appropriate category of the understanding. Such synthetic
judgements are practical in the sense that they concern
experience. However, they are derived not from experience,
but from the possibility of objective experience. VJhen the
empirical content of the judgement is abstracted what remains
is the pure category itself. Thus for each of the ways in
which an objective empirical judgement confers its objectivity
onto the corresponding perceptual judgement there will be
corresponding an elementary a priori category.
Korner, Kant (Baltimore: Penguin Books, 1955), p . 26.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48
An enlightening analysis of this stage in the argument
of the Critical Philosophy is given by Norman Kemp Smith.
The connection of the categories with the act of judging is
for Smith "pregnant with many of the most valuable results of 13 the Critical teaching." Where the parallels between the
table of categories and the table of judgements provides
fruitful possibilities for the development of an adequate
classification, it at the same time supplies conclusive
evidence that the judgements were rearranged in a highly
artificial fashion to yield a more or less predetermined
list of categories.
Having established this basis for development of the
categories Kant goes on to derive them. If we take any
objective empirical judgement such as (to use an example from
the Critique) "This stone is heavy." and from that judgement
abstract the subjective judgement, "This stone seems heavy to
me." what remains is the skeleton or a priori form. In this
case the category concept which is implied is that of a
substance with accidental properties or the categories of
inherence and subsistance. The categories which Kant
delineates are organized under four headings as follows:
1 Q Smith, A Commentary, p. 191.
^^Ibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49
Of Quality
Unity
Plurality
Totality
II III
Of Quantity Of Relation
Reality Of inherence and
Negation subsistence
Limitation Of causality and
dependence
Of reciprocity bet.
agent and patient
IV
Of Modality
Possibility - Impossibility
Existence - Nonexistence
Necessity - Contingency
The adequacy of Kant's categorical account is a serious
problem to Kant's analysis. If we are to accept Strawson's
evaluation of this stage in Kant's argument we would be led
to throw out this entire section of the Critique as artificial,
misconceived and outdated. However, if we grant for the
moment the categories as Kant details them, a second problem
of interpretation arises. How are we to view the categories?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50
A t many points in his analysis Kant uses the terms concept and
category interchangeably. Such mixing of the two terms, Smith
finds exceedingly misleading. Where categories are relational
functions which provide a unity through which we can interpret
the contents of a judgement, concepts denote a common quality
found within members of a class of similar entities. They are
the content. Again Kant's preoccupation with the subject-
predicate form of logic has apparently affected his terminology.
Smith sees this as further evidence that Kant was unaware of
the revolutionary consequences of his viewpoint.
Even in the very act of insisting upon the relational character of the categories, he /_Kant^7" still continues to speak of the concept as if it must necessarily conform to a generic type.^S
Thus in sifting through the four stages of development
of the Critical Philosophy all of which are intermingled in
the pages of the Critique, we want to maintain the relational
nature of the categories which, in the spirit of the
transcendental logic, abstracts out from any particular
experience to determine the structure of the faculty of pure
reason itself. The categories as the organizing principles
of the understanding are devoid of specific content. Their
. . . connection supplies a rule by which we are enabled to assign its proper place to each pure concept of the understanding and by which we can determine in an a priori manner their systematic completeness.T6
T^Smith, A Commentary, p. 181.
IGjbid., p. 175.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51
If the transcendental logic deals with categories and those
categories are relational then what we find is an interesting
supplement to the logic of categorical propositions and to
the attempt to schematize our conceptual structure.
The function which the categories play, therefore is
one of unity, but it is misleading to interpret this unity
in terms of the coming together of a class of similar entities
unified under one concept. Each type of analytic judgement
involves the coming together under one specific functional
relation of conceptual factors which become unified with the
other elements in the judgement. The category itself does
not provide the content. Rather every pure category contains
a principle of synthesis. This principle of synthesis is
derived from the form of the judgement.
The term applied to this unifying function is synthesis.
Through synthesis the understanding generates a manifold of
complex ideas out of the elements of sensory experience. It
then organizes and interprets the contents of that manifold
through the categories of the pure understanding. These
categories serve as the form which is synthesized with the
matter provided by the senses. The essential premise of the
Analytic, according to Strawson, is the necessary unity of our
consciousness. It sets a minimum standard of what is to ac
count for experience and that standard is that the contents of
experience should yield a picture of an objective world which
is unified by its existence within a framework of physical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52
space and time and demonstrates regularity in its operations.
Without this we would have no experience at all. How Kant can
go about proving more than just the possibility of such a
premise is exposed, Strawson feels, to the objection that we
have no empirical knowledge of its truth. If there is unity
Strawson feels it could be a part of experience as produced
by our faculties out of impressions of sense rather than
being a feature of the pure understanding. The unity found
in our experience, however, might also be interpreted as a
function of our own self consciousness. This relationship
between the subject and the manifold of experience is pure
apperception, the highest and purest form of unity. The very
fact that "I think" accompanies all our representations,
itself imposes a unity onto the content of our experience.
Just as the world always viewed through rose colored
glasses would take on a thoroughly unique and colorful aspect,
so too the categories can be viewed as "epistemological
sunglasses" through which we view the world. In removing
them we would see things in an unfamiliar light, but with them
in place the world takes on its familiar forms and we confront
experience in terms of categories of cause and effect, quality,
quantity and relations, and so forth. Where we feel these
elements are part of the objects of our experience, in reality
they are a part of our own cognitive make up. This is not to
say they are any less interesting. In fact it is this which
makes Kant's approach so revolutionary and worthy of attention
for the insights it might inspire. Its significance lies in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53
providing us with a completely unique possibility for
interpreting the function of the understanding. Its greatest
weakness, Strawson suggests, lies in Kant's assumption that
a listing of all the logical forms of judgement was a possible
task. This assumption revealed his dependence on traditional
logic. The possibility of completion of such a task is in
Strawson's view highly doubtful.The second assumption which
Kant makes derives from the first. It is that no new a priori
concepts could be formed. Korner points to Whitehead's four
dimensional concept as providing a clear counter instance to
this claim.
However, Korner sees value in Kant's categorical
interpretation which is over and above the adequacy of the
categories as a whole. Even if we were to grant that Kant's
list of categories were incomplete, the significance of Kant's
approach is to have established that we do apply categories
in objective judgement.
A final issue in regard to the categories which must
be addressed is whether the justification which was provided
by Kant is adequate. Korner claims that what Kant has done
is to concoct a circular defense of the categories of the
Understanding. He justified the use of the categories by
demonstrating that the unity of pure appreception, the
applicability of the categories and the possibility of
^^Strawson, The Bounds of Sense, p. 21.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54
objective experience mutually imply one another. Thus we
are left with what may be a justification of the categories
based on an implication which is not synthetic a priori but
analytic. The difficulty of this conclusion is that by Kant's
own principles analytic propositions only elucidate the
meanings of their terms and go no further. If this is the
case, Korner points out, the transcendental deduction may be
a mere tautology.
Within the intricate organizational structure of the
Critique the categories are just one element which makes up
the pure understanding. As abstract relational forms they
are empty of content, but for each category there is derived
an a priori principle whose statement involves the application
of these concepts. Following the deduction of the categories
Kant moves on to delineate the Analytic of Principles in which
he gives a series of demonstrations of synthetic a priori
principles that are directly derivative from the categories.
These principles state the conditions of possible objective
experience of objects. The table of principles is comprised
of rules for the objective use of the table of categories.
Corresponding with the category of quantity are the axioms of
intuition. The general principle is "All perceptions are
extensive magnitudes." As a consequence of the axioms we are
able to apply mathematics to experience. Corresponding to
the category of quality are the anticipations of empirical
perception which hold that in all appearances the real, which is an object of sensation, has intensive magnitude or degrees
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55
of quality. The category of relations has corresponding with
it the analogies of experience in which experience is possible
only through representation of the necessary connection of
perceptions one with another. The analogies include the
relations of permanence of substance and inherence, of
causality and of reciprocity. Corresponding with the category
of modality are the postulates of empirical thought,
possibility, actuality and necessity.
At this point in Kant's argument we have seen the
development of the forms of intuition as the limits to which
sense experience must conform and of the categories which
serve to restrict the manner in which the content of the
understanding can be organized. We have seen how Kant
applies the categories to experience through the synthetic
a priori principles which define the characteristics required
of objects if they are to be objects of experience. Now Kant
begins to erect a third layer of complexity into his already
somewhat unwieldy structure.
The Schematism approaches the same subjects from the
point of view of the imagination and the characteristics
required of it. How Kant conceived the function of the
imagination is uncertain. In most passages which comprise
Stage three it is treated as an auxiliary function of the
understanding. In later portions it is considered to be a
distinct faculty. In order for experience to occur the
imagination must combine the manifold in certain preestablished
ways. These ways must conform to a principle of synthesis
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56
derived from one of the pure categories. The function of the
imagination is thus central to the reasoning process. It
produces the schemata.
All temporal objects in order to be the content of
perception must demonstrate that they have characteristic
ways of combination which fall under the pure categories.
The transcendental schemata are defined as the universal
characteristics which must belong to all objects as objects
in time. Thus they refer to objects as such rather than to
each individual object. The schemata of the categories
determine the specific conditions under which the category
is applicable to any manifold which has synthetic unity.
Only one thing is common to everything -- being in time.
Thus the schemata of a category also determines the temporal
conditions under which it is applicable to objects of
experience in general. In performing this function the
schemata can be viewed in one sense as the rules or procedures
for the production of images which within temporal existence
delimit and allow for the application of a category. They
link the categories to perception. Without the schemata the
concepts would be mere logical shells with no reference.
Before we can apply the concept of a dog to a particular dog
we must be able to reproduce a schematic representation in our
minds. At the same time without the schemata the perceptual
object would not be synthesized within the transcendental
imagination.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57
The pure category of the understanding when applied
and restricted to its corresponding schema becomes the
schematized category. For example corresponding to the pure
category of ground and consequent we have the schematic
category of cause and effect. Correspondingly the trans
cendental schema relevant here is necessary succession. The
pure category has no reference to time and space. The
schematized category is the concept of the synthesis of a
concept in time, and is the product of the synthetic activity
of the imagination.
By all indications it is probable that Kant's primary
concern is with the schematized categories rather than the pure.
It is Smith’s claim that when Kant speaks of the categories he
usually has in mind the schematic. Thus the question has been
raised as to why Kant introduced the concept of schemata at
all to his system if there is so little difference between it
and the level of category. Smith hypothesizes that Kant felt
the necessity of postulating a third level to bring together
the pure concepts and the senuous intuitions which he wished
to subsume under them. He claims there is no need for a move
such as this since the introduction of this third level has
only resulted in theoretical clutter and confusion. In fact
the results from a philosophical perspective are unfortunate
for it reinforces a significant misconception of Kant's theory.
At the same time Kant is led to present what ought, consistent
with the core, to be viewed as a "creative synthesis, whereby
contents are apprehended in terms of functional relations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58
/^rather than as^T" subsumption of particulars under universels 18 that are homogeneous with them . . . This is ultimately
what Kant means by the schematism of the pure forms of
understanding. Smith suggests that if Kant had used ’category'
to stand for pure forms and 'schemata' to signify concrete
counterparts with the table of categories distinguished from
the table of logical forms and entitled the table of schemata,
much ambiguity and confusion would have been avoided.
How well has Kant carried out the task which he set for
himself in the Introduction to the second edition of the
Critique^^ to establish a new method of thought which will
completely revolutionize the field of metaphysics? How
successful has he been in his exploration of the scope and
limits of human reason and in his portrayal of che general
structure of experience and the world as we know it? Norman
Kemp Smith concluded that there is a significant shift in the
argument of the introduction from asking how synthetic a
priori judgements can be possible, a question which is primary
to any investigation which grounds its conclusions on their
existence, to asking what are the principles that qualify
as synthetic a priori. If this is the case, Kant has surely
made too quickly a leap into the content of his transcendental
metaphysics. It is a much easier task to generate statements
which fit the description Kant gives of synthetic a priori
than it is to adequately defend the legitimacy of the concept.
ISsmith, A Commentary, p. 340, ^°Kant, p. 23.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59
If there are judgements which are both synthetic and a priori
then how can their necessity and universality be supported
beyond the question of doubt? Smith feels Kant is not fully 20 aware that he has made this shift. Where a shift in emphasis
itself raises questions of the tightness of Kant's argument,
the more serious problem arises in not having settled the
question of the possibility of synthetic a priori prior to
the enumeration of them in the principles of pure reason.
P.P. Strawson sees the interpretation which dominates
the Smith commentary and other traditional accounts to be
less controversal but more limited in scope than his own. He
feels the themes which Kant enumerated in the Critique reject
ing both speculative metaphysics and the empiricist dilemma
presented by Hume, while significant contributions to philo
sophical dialogue, are interwoven by much more questionable
doctrines closely related to the concerns of this paper. Kant
saw himself as involved in an exploration of the general
structure of experience. This he conceived of as an inves
tigation into the organization and workings of the cognitive
capacities of the human mind. However, in conceiving
this task Kant sought the a priori necessities implied
in our conception of experience in the nature of our
cognitive faculties. Thus Kant approaches what is a
philosophical query in a psychological idiom. An empirical
approach cannot satisfy questions which demand the force of
20Smith, A Commentary, p. 49.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60
universality and logical necessity. What we have is "an essay 21 in the imaginary subject of transcendental psychology."
Strawson is also sceptical of the legitimacy of Kant's
categories as a definitive inventory of the possible logical
operations of our cognitive capacities.
It requires only moderate acquaintance with formal logic to be both critical of the list of forms which is to be the basis of Kant's derivation in the metaphysical deduction and sceptical of the conception of the derivation itself.22
Kant with surprising enthusiasm and lack of critical perspec
tive adapted the forms and classifications of traditional
logic and used this as the source of his "boundless faith" in
the questionable structural framework which he imposed on his
material. Strawson concluded rather strongly.
The artificial and elaborate symmetry of this imposed structure has a character which, if anything in philosophy deserves the title of baroque, deserves that title.23
Korner criticizes the limitations imposed by tradition
al logic on Kant's classificatory schema from another perspec
tive. In listing the three types of judgements, analytic a
priori, synthetic a posteriori and synthetic a priori Kant
gives adequate interpretation on].y to judgements of the
subject-predicate form neglecting other possibilities.
^^Strawson, The Bounds of Sense, pp. 31-49.
22%bid., p . 31.
23 Ibid., p . 44.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61
This strange omission on his part might be explained by his acceptance of almost the whole of traditional logic, which concerned itself mainly with subject-predicate propositions.24
Here what Korner is raising is the question of the legitimacy
of the analytic-synthetic distinction. The belief in its
validity is a consequence of a paradigm view of propositions
as forms of logical predication of attributes in subjects.
Where the attribute is inherent in the subject, we have an
analytic proposition; where it is not, synthetic. Korner
feels this problem can be overcome by not restricting the
distinction between synthetic and analytic judgements to
sentences of the subject-predicate form and by allowing a
judgement to qualify as analytic if and only if a denial
would be a contradiction in terms.
These are serious difficulties in the Critique and may
obscure what there is of substance in Kant's analysis.
Strawson feels this is the case. He concludes,
Kant's major positive achievement in metaphysics was to be sought in his attempt to articulate the general structure of any conception of experience which we could make truly intelligible to ourselves.25
If we are able to untangle these more fruitful contributions
from the stilted and baroque architecture of the categories
and the psychological idioms in which they are defended, there
is much of value here.
2^Komer, Kant, p. 23. 25strawson, The Bounds of Sense, p. 49.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 2
How might Kant's theory be reconstructed to avoid some
of these difficulties? In analyzing this question Paton
outlined what he believes can be retained in a coherent and
accurate rendering of Kant's position?^ Kant wished to prove
that the form of our experience of the world was determined
not by the objects which are part of that experience but by
the nature of our cognitive capacities themselves. Kant's
argument rests on two foundations. The first is on his
treatment of the forms of judgement and the second is on his
conception of the transcendental synthesis of space and time.
The forms of judgement derived as they are form the schema of
traditional logic are not fully adequate in presenting a
comprehensive treatment of all possible logical forms. Their
derivation is part of the formalism which tends to limit the
usefulness and adequacy of Kant's approach. In the schemata
Paton feels Kant breaks away from this formalism and begins
to make really significant progress in his analysis of
judgement. He derives the categorical characteristics of
objects from the fact that all objects must exist within the
framework of space and time "and in this there is surely a 27 considerable measure of truth." While the structure of
Kant's categories which were derived from formal logic could
not withstand the test of time, "The connection of the
2^Herbert James Paton, Kant's Metaphysic of Experience (New York: Macmillan, 1936), pf 767
27 Ibid., p. 76.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63
categories with the synthesis of imagination and the form of
time is the most important and the least artificial, part of
the Critical Philosophy."^®
We can still retain the central assertion which Kant
is making that judgement is a necessary precondition to
experience of objects. Judgement presupposes a unity of
apperception, a self-conscious awareness of unity within our
stream of consciousness. It demands a corresponding unity in
the empirical manifold. That unity is imposed by the
transcendental synthesis of the imagination within the bounds
of space and time. The synthesis combines the manifold of
perception according to the necessities imposed by the fact
that the manifold is the object of thought. This approach
still implies the necessity of certain categorical
characteristics which would characterize judgements, but
these would rest on the unity of time rather than on the forms
of judgement. It thus still allows us to approach Kant's
principles in hopes of discovering a proof of the necessity
of certain categorical characteristics and serves to enrich
the concept of the categories.
It is this interpretation which appears to be most
fruitful for our interests here. If the Critique is viewed
as an examination of the necessities impressed upon and
formative of our experience of the physical world then it
serves as a possible prototype to more recent inquiries which
2®Ibid., p. 76.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 4
are concerned with similar questions but are approaching it
from within a psychologically more sophisticated context.
The significance of the view interpreted here, if we attempt
to characterize its crucial element, is to treat thinking in
an active rather than a passive mode. This view of thinking
as the connecting and relating of ideas according to certain
content-devoid logical categories is one which should prove
fruitful to our task here. In perception we ascribe unity to
an unorganized manifold of sensations. This unity is quite
different from other characteristics which are perceived in
sensation. It is produced by the understanding through the
synthesis of the manifold. Kant's point is "to analyze the
structure of a connected manifold by distinguishing its
characteristics and exhibiting their logical relations."29
The synthetic unity which obtains entails a thinking, perceiv
ing subject. Kant uses the transcendental unity of appercep
tion, the unity which coheres between the manifold and the
thinking subject, to justify his application of the categories
as necessary conditions of objective experience, since without
it no knowledge would be thinkable. Thus just as knowledge
of an object presupposes and requires both space and time,
the forms of intuition, it presupposes and requires, on the
level of the understanding, the unity of pure apperception.
29 Korner, p. 60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65
Is, then, Kant justified in proposing this
transcendental metaphysics as a viable philosophical
alternative to transcendent or speculative metaphysics?
In Kant's words.
The Transcendental Analytic leads to this important conclusion, that the most the understanding can achieve a priori is to anticipate the form of a possible experience in general. And since that which is not appearance cannot be an object of experience, the understanding can never transcend those limits of sensibility within which alone objects can be given to us. Its principles are merely rules for the exposition of appearances ; and the proud name of an Ontology that presumtuously claims to supply, in systematic doctrinal form, synthetic a priori knowledge of things in general . . . must, therefore, give place to the modest title of a mere Analytic of pure understanding.
30Kant, Critique, p. 303.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PART II
ATTEMPTS WITHIN LINGUISTIC THEORY TO DEVELOP
A THEORY OF COGNITIVE STRUCTURE
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER III
THE SYNTACTICALLY BASED THEORY OF NOAM CHOMSKY
For those whose minds are most comfortable when every
object, person and fact within their purview has been neatly
tucked away into its proper corner, great geniuses like
Galileo, Beethoven or Einstein are an unsettling phenomena.
They defy the attempt to be labeled and will not be placed
upon a proper shelf there to sit motionless for history to
comtemplate. Their activities cannot be restricted within the
narrow confines of one precise discipline, but instead their
insights and interests extend to and draw together from many
sources. They will not accept those arbitrary boundaries
inscribed by less creative minds, the clerks and draftsmen
of the academic profession, but rather lead us on to discover
new and fertile interconnections between areas of knowledge.
In such striving, as these intellectual progress is made and
new inroads carved into the rocky and previously insurmountable
precipices lying at the outer limits of our present under
standing .
It is in this light that we might look upon the work of
Noam Chomsky. Writing from the perspective of one trained in
technical linguistics, Chomsky grounds his theoretical treatment
of language within the framework of traditional philosophical
inquiry thus drawing together the concerns of both philosophy
and linguistics. In so doing he exposes new, more fertile 67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68
paths which might be pursued in answering those questions -o'f
vital philosophical interest which have already been posed
for us in the \fritings of Aristotle and Kant. Part I of this
inquiry examined how these two philosophers, Aristotle and
Kant, attempted to satisfy epistemological questions regarding
the source, nature and scope of knowledge. In Chapters 3. and
4 of Part II we will examine what can be viewed as an exten
sion of this inquiry, a continuing of the attempt to
adequately define the origin, nature and limits of human
knowledge approaching these questions as in a sense both
Aristotle and Kant did, through an examination of the logical
structure of language.
Underlying both the empirical research and theoretical
exploration of Chomskian linguistics is the conviction that
the division of philosophy, linguistics and psychology into
separate, autonomous disciplines has been thoroughly arbitrary.
Chomsky argues that the barriers which have until recently
been rigidly drawn between these three areas of study are at
last breaking down. Linked by a common concern for the
classical problems of language and mind each discipline can
support and contribute to the progress of the other, and
through such mutual cooperation more adequate answers can be
found to the significant questions which scholars in each
field are seeking to satisfy.
Thus in his introduction to Language and Mind (1972)
Chomsky sets forth as his task "to show how the rather
technical study of language structure can contribute to an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69
understanding of human intelligence."^ Through the study of
language we can begin to be able to delineate the properties
of mind that underlie the exercise of human mental capacities.
Clearly it is Chomsky's belief that human language provides
us with the key to an understanding of human cognitive
processes. Such a presupposition rests on the assumption
that language plays an essential role in thinking as well as
in human interaction. One would then expect, as does Chomsky,
that human language, by directly reflecting the characteristics
of human intellectual capacities is in a sense a "mirror of
the mind." Once we can determine the ways in which language
is structured and its elements interrelated, we can begin to
describe the structure of knowledge and to formulate some
plausible hypotheses about the intrinsic human capacities
that make this possible.
And so, throughout his writings Chomsky uses the
insights of language in seeking a fuller understanding of this
subject of concern to both philosophers and psychologists.
However, in attempting to progress toward a settlement of
traditionally thorny philosophical issues from a linguistic
perspective Chomsky first wishes to draw a clear distinction
between his own view of language and the use he makes of it,
and another very dominant tendency in philosophy which is also
linguistically based, that of ordinary language philosophy.
^Noam Chomsky, Language and Mind, 2nd ed. (New York: Harcourt, Brace, Jovanovich, Inc., 1972), p. viii.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70
This tendency attempts to solve philosophical problems through
a clarification of the meanings of the terms involved. Where
peripheral issues regarding confusions in surface grammar are
concerned, Chomsky finds this approach useful. However,
where really significant philosophical problems are concerned,
Chomsky is clearly skeptical of the contribution such studies
of language can make. As he concludes,
I think a case can be made that certain well- founded conclusions about the nature of language do bear on traditional philosophical questions, 2 but in ways rather different from those mentioned.
Rather than viewing the sole and proper role of philosophy as
"a battle against the bewitchment of our intelligence by means 3 of language" and seeing traditional philosophical questions
as pseudoproblems, the mad ravings of philosophers who "...
like savages, primitive people . . . hear the expressions of
civilized man, put a false interpretation on them and then 4 draw the queerest conclusions from it . . . ." Chomsky is
interested in language as it is isomorphically related to
thought. Through developing an understanding of the structure
of language he believes we can begin to gain an understanding
of the structure of thought and the significant philosophical
issues associated with it.
2 Chomsky, Language and Mind, p. 167. ^Ludwig Wittgenstein, Philosophical Investigations, translated by G.E.M. Anscombe 2nd ed. (New York: Macmillan Co., 1958), p. 109.
^Ibid., p. 194.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71
In establishing a philosophical base for his conclusions
Chomsky is also concerned to show that traditional rationalist
and empiricist approaches to problems in the philosophy of
mind have been unsatisfactory in adequately portraying and
accounting for the nature of our cognitive abilities. By
restricting themselves to that which is observable in the
narrow sense of environment and overt behavioral factors,
empiricists have severely limited the range and significance
of their conclusions. Perceiving of the mind in terms of a
passive receptable of knowledge which receives impressions
through the senses, such accounts as that of Locke give
inadequate attention to the contribution the mind makes to
the learning process. More recent empirical approaches which
attempt to account for mental phenomena in terms of disposi
tions are equally inadequate. If we wish to develop a theory
which can give adequate explanations for linguistic facts
we cannot restrict our attention to "vague talk about
'habits' and 'skills' and 'dispositions to respond.'"^
Chomsky finds such an empiricist orientation not only
dominant in current work in philosophy, but characterizing
much current research in linguistics and psychology as well.
In linguistics, empiricism is associated with descriptive
linguistics and the dominent tendency to view the linguist as
being concerned solely with describing linguistic behaviors.
^Chomsky, Language and MTind, p. 36.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72
Seeing the environment as the input into a language acquisition
device, the linguistic facts which make up the output on this
model are interpreted as responses to prior stimuli from the
environment.
Chomsky’s attack on the inadequacy of empiricist
approaches is most energetic in his treatment of behaviorist
theories in psychology. Insisting on what Chomsky considers
"certain arbitrary methodological restrictions that make it
virtually impossible for scientific knowledge of a nontrivial
character to be attained"^ behaviorists have achieved a
transition from what they consider speculation to science.
The scientific character of their findings in Chomsky's
view is only gained through the restriction of their subject
matter and a concentration on periperal issues^ which have
little significant bearing on the understanding of human
behavior. "A study of human behavior that is not based on
at least a tentative formulation of relevant systems of
knowledge and belief," Chomsky concludes, "is destined to g triviality and irrelevance." By shifting their emphasis
toward the evidence itself and neglecting the deeper underlying
principles and abstract mental structures that might be
illuminated by the evidence, the behaviorist can only address
questions of performance. The very important task of
accounting for competence remains untouched. Chomsky draws
^Chomsky, Language and Mind, p. ix.
^Ibid. Sibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73
the characterization of a behaviorist as a scientist whose
only concern is with meter readings rather than with the
phenomena which lie behind these readings.
Chomsky's most notable attack on behaviorism took the 9 form of a review of B.F. Skinner's Verbal Behavior in 1959
Chomsky views Skinner's book, which was published in 1957, as
the first large-scale effort to incorporate the major aspects
of linguistics behavior within a behaviorist framework
. . . In it Skinner was attempting to give a fully
adequate account of verbal behavior in functional terms be
lieving that such behavior would be completely and accurately
described. This task would involve the identification of
those variables which controlled verbal behavior and the
specification of how they interacted to determine a specific
response. Controlling variables were conceived in terms of
notions of stimulus, reinforcement and deprivation.
Chomsky identifies the point of contention between
himself and Skinner to be grounded in their differences
concerning the particular character and complexity of the
specific contribution of the organism to learning and its
performance, and in the kinds of observation and research
each views as necessary to arrive at a precise specification
^Noam Chomsky, "Review of B.F. Skinner's Verbal Behavior, Language (1959): 35.
l^chomsky. Language and Mind, p. 26.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74
of it. It is Skinner's position that external factors
consisting of stimulation and the history of reinforcement
of the subject are most important in the specification of
verbal behavior, whereas the contribution of the speaker is
not what is the focus of scientific understanding. Thus in
order to predict verbal behavior it is only necessary to
specifiy the relevant external factors. Chomsky finds such
an evaluation "surprising."
One would naturally expect that prediction of the behavior of a complex organism (or machine) would require, in addition to information about external stimulation, knowledge of the internal structure of the organism .... These characteristics of the organism are in general a complicated product of inborn structure, the genetically determined course of maturation, and past experience.
It is the notion of the scientific character of
behaviorist research that most inflames Chomsky in his attack
on Skinner. Chomsky sees the behaviorist caught in an
uncompromisable position. If they accept a broad definition
of stimulus as any physical event which impinges on the
organism and of response as any part of the organism's
behavior, then it becomes impossible to establish any clearly
lawful connection between the stimulus - response in specific
circumstances. If on the other hand narrow definitions are
followed, little of what we consider behavior will satisfy
the requirements. Skinner's attempt to avoid these problems
^^Chomsky, Language and Mind, p. 27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75
is equally unsatisfactory. Rather than choosing a narrow set
of definitions for human behaviors he uses strictly run
experimental results on a non-human level as evidence for the
scientific character of his own interpretation of human
behavior. On this basis he makes "analogic guesses" to extend
the technical laboratory results which did deal with narrowly
defined behaviors, using terms in his descriptions which
become "mere hononyms with at most a vague similarity of 12 meaning " when applied to human action. In this way he
creates the illusion of having presented a scientific theory
which nonetheless remains broad in scope. If we are to
approach Skinner's book giving literal reading to the
definitions of technical terms we find, says Chomsky, that
it covers almost no aspect of verbal behavior. If we view
the use of technical terms as metaphorical we see that his
theory is neither scientific, clear nor careful. Chomsky
concludes that the elimination of the independent contribution
of the speaker and learner . . . can be achieved only at the
cost of eliminating all significance from the descriptive ,,13 system.
If the behaviorist is misguided in his appraisal of the
proper direction of psychological research, how then ought one
to approach the study of human behavior? Chomsky sees the
task of the human psychologist to be the development of a
^^Chomsky, Language and Mind, p. 30.
13lbid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76
theory of mind. Psychology at its deepest level should be
concerned with the human ability to construct cognitive
structures.It should study in detail the actual character
of stimulation and organism-environmental interaction which
sets the cognitive mechanism into action. Hopefully, Chomsky
believes, this will lead to a theory made up of a succession
of maturational stages. It would be the task of the learning
theorist on this interpretation to account for how the mind
fills in the detail within a structure which is innate.
In speaking of innate cognitive structures Chomsky
comes much closer to a rationalist position and, as he grants,
his interpretation of psychology is at many points highly
rationalistic. Quoting from an article which appeared in
1968 Chomsky reinforces his view that "contemporary research
supports a theory of psychological a priori principles that
bears a striking resemblance to the classical doctrine of
innate ideas.
However, it is Chomsky's belief that traditional
rationalism too falls short of providing an adequate account
of the operation of the human mind. Choosing as a paradigm
the rationalist theories of Rene Descartes, Chomsky sees the
general weakness of such approaches to lie in their failure
to appreciate both the degree of abstractness of those
^^Chomsky, Language and Mind, p. 54.
^^Noam Chomsky, Reflections on Language, quoted from "Recent Contributions to the Theory of Innate Ideas," 1968 (New York: Pantheon Books, 1975), p. 218.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77
structures that are "present to the mind," and the length
and complexity of mental operations. Such operations relate
mental structures which have semantic content to the physical
utterance which is produced.
Thus while vitally interested in the possibility of
developing a linguistic theory along Cartesian lines, Chomsky
begins by laying out those elements within Cartesian rationism
which on his view must be revised. Most notably, what Chomsky
seems compelled to reject is the basic dualism of the Cartesian
approach.
Chomsky sees Descartes' adoption of a dualistic position
to have grown out of his attempt to account for the nature of
persons as thinking beings. In considering the theory of
corporeal body, even when sharpened, clarified and extended
to its limits, he found it "still incapable of accounting for
facts that are obvious to introspection and that are also
confirmed by our observation of the action of other humans.
As a consequence Descartes concluded that the study of mind
presents us with the problem not merely of degree of complexity,
but of quality as well. An adequate theory for Descartes could
not be framed without evoking an entirely new second substance
whose essence is thought.
We, however, are not constrained as was Descartes to
postulate the existence of a second substance, differing in
^^Chomsky, Language and Mind, p. 6 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78
quality from inert matter. "It is an interesting question,"
Chomsky muses, to consider "whether the functions and evolution
of human mentality can be accommodated within the framework
of physical explanation, as presently conceived, and whether
there are new principles, now unknown, that must be involved
perhaps principles that emerge only at higher levels of
organization than can now be submitted to physical
investigation.
Thus Chomsky, while in important respects closely
alligned with the rationalist tradition, departs somewhat in
this view that the innate or a priori system which determines
the structure of both language and thought is biologically
determined and can be accounted for in physical terms without
the necessity for evoking a category of substance which is
itself non-physical. How our minds are "wired" becomes for
Chomsky a question of compelling interest. And so where in
Aristotle the main determinant of the structure of our
thoughts was a world itself structured in a definable manner,
for Chomsky the burden is shifted to the structure of our
minds. What we can know and the forms which that knowledge
takes are determined by how we actually receive and process
information, by the modes of conception inherent within the
human mind. Chomsky concludes that, "We interpret experience 18 as we do because of our special mental design."
^^Chomsky, Language and Mind, p. 98.
^^Chomsky, Reflections on Language, pp. 7-8.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79
Within the context of this very dynamic interpretation
of the role of the mind in organizing and creating thoughts
let us now move on to examine in detail the possible contri
butions of Chomskian linguistics to our inquiry into the
logical structure of thought, by examining Chomsky's theory
within its historical context in the development of linguistic
science. In Reflections on Language (1975) Chomsky analyses
the process which is involved in providing an explanation of
language acquisition and examines the theoretical choices
available to those involved in studying language and
constructing a theory of learning. In developing a theory of
language it is necessary to take into account two sets of
structures which are shared by all persons, (1 ) a system of
beliefs and expectations regarding the behavior of objects
which surround the person in the environment. These make up
the content of common sense, and (2 ) a system of language,
a grammar. On the basis of these two postulated structures
and our assumptions regarding information processing we can
go on to attempt to account for what people actually do and
how they think.
If we were constructing a theory of language what
conclusions might we draw? We might on the one hand determine
that each person possesses a set of innate concepts which
determine the character of the language they possess. We
could then develop complex assumptions as to the interaction
of maturation and experience. Adopting a theory of develop
mental stages which determine the quality as well as complexity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80
of verbal and cognitive ability. On the other hand we might
suppose the mind is a tabula rasa. We would then try to
discover what property of mind enables a child endowed with
language learning ability to acquire the grammar of the
language.
If we were to proceed as an empiricist supposing the
mind to be a blank tablet which can only record and retain
impressions and construct associations among them, then we
would attempt to design procedures of association and habit
formation and through induction would conceive of the grammar
as output. If on the other hand we took the rationalist
position we would postulate and attempt to specify a
schematism which is innate, later to become refined and
articulated by experience.
Two traditions have dominated the technical study of
language since its inception and continue in tandum today.
The first is the position which characterizes modem struc
tural linguistics. An outgrowth of concepts which emerged in
the nineteenth century, it views language as a system of phono
logical units that undergo systematic modification. The
work of one of the early structural linguists, Ferdinand
de Saussure, was formulated on the principle that once an
analysis of sentences has been completed in terms of
segmentation and classification, the structure of language
has been completely revealed. On this view language is
composed of units of sounds, words and phrases. Sentences
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81
conceived as units of meaning play no part in the analysis.
Because of its ability to provide a compatible format
for the study of comparative grammars modern structural
linguistics has dominated the field during the twentieth
century. Its major achievement according to Chomsky has been
to provide a factual and methodological basis which now makes
it possible to return to more substantive issues and the study
of language as a formal system. Rather than rejecting the
methods and accomplishments of this descriptive approach,
Chomsky wishes to make use of the methodological advances it
has made.
The moral is not to abandon useful tools ; rather, it is, first, that one should maintain enough perspective to be able to detect the arrival of that inevitable day when the research that can be conducted with these tools is no longer important.^
I-Jhat we have gained through structural linguistics is an
enormously broadened base of linguistic data which possesses
a considerably improved degree of reliability. Viewing
language as a network of structural interrelations which are
capable of being studied abstractly with a high standard of
precision, its failure lies in the restriction of its
analysis to the surface structure of language. It is Chomsky's
suggestion that this tradition should be coupled with a
recognition of its essential limitations.
19 Chomsky, Language and Mind, p. 19.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82
Whitehead once described the mentality of modern science as having been forged through the union of passionate interest in the detailed facts with equal devotion to abstract generalization. It is roughly accurate to describe modern linguistics as passionately interested in detailed fact, and philosophical grammar /_a position toward which Chomsky looks with more favor/ as equally devoted to abstract generalization. It seems to me that the time has arrived to unite these two major currents and to develop a synthesis that will draw from their respective achievements.20
The conception of linguistic theory which Chomsky
identifies as philosophical grammar developed in self-conscious
opposition to descriptive linguistics. It can be traced as
early as 1660 to what is known as the Port-Royal Grammar.
The significance of Port-Royal lies in its recognition of the
importance of the phrase as a grammatical unit which,
corresponding with a complex idea, carries meaning. Sentences
were divided into phrases or units of meaning rather than into
phonological units. On this interpretation surface structure
corresponds only to sound. Analogous to surface production
is a corresponding "mental analysis" into what Chomsky calls
deep structure and which relates directly to meaning rather
than to sound.
The approach initiated in Port Royal was continued in
the work of Wilhelm von Humboldt. Humboldt held that
underlying any human language there is a universal system
expressing mental attributes which are unique to human thought.
o n Chomsky, Language and Mind, pp. 22-3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83
Grammar is conceived of as a finite system of rules which
generates an infinite number of deep as well as surface
structures. Language is not really learned but "develops
from within in an essentially predetermined way, when 21 appropriate environmental conditions exist."
It is the concept of generative grammar centering in
the conception of "the form of language," which characterizes
Humboldt's theory. One doesn't teach a language, but rather
we "provide the thread along which it will develop of its own 22 accord" by processes like maturation more than learning.
Humboldt presents an essentially rationalist interpretation
of language acquisition.
The form of language of which Humboldt speaks can be
understood as an internal model of language which encompasses
its grammatical form and system of sound. The speaker of
language developes an internal representation of this form.
Generative grammar is the attempt to sketch the form
of a language in a concrete way. An important distinction
between this view and that of de Saussure is that on this
view a word stands for a concept rather than a thing. This
contrasts with de Saussure's identification of a word with its
object of reference. Understanding for Humboldt, then, would
not be equivalent to choosing an appropriate conception from
a store of concepts, but rather a verbal sign activates within
21 Chomsky, Language and Mind, p. 76. Z^Ibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84
the listener a link with a concept causing a corresponding
but not identical concept to emerge. "When a key of the
mental instrument is touched in this way, the whole system
will resonate, and the emerging concept will stand in harmony
with all that surrounds it to the most remote regions of its
domain."^3 A network of concepts is activated and the
placement of a concept within this system determines the
meaning the expression will have for the hearer. Our concepts,
then, function together as an organic whole. Language reflects
this conceptual unity. Its instrumental use as a
communication device is derivative and secondary.
While maintaining the comprehensive scope which
characterizes Humboldt's approach and the conviction that the
really meaningful and interesting discoveries regarding
language lie beyond the external surface phenomena, Chomsky
at the same time recognizes the necessity of preserving
the high degree of clarity and precision of methodology which
modern linguistics has achieved. And so he moves on to define
within the context of modern technical linguistics a modified
rationalist theory of both language and thought. Through the
techniques of structuralist linguistics and phrase structure
grammar Chomsky sets out to analyze the syntactic structure
of language. But in seeking those abstract principles which
govern its structure and use by biological necessity and are
9 9 ^^Noam Chomsky, Current Issues in Linguistic Theory (London: Mouton and Co., 1964), p. 21.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85
derived from the mental characteristics of the species,
Chomsky in his most recent writings makes it clear that his
concern in examining natural languages is the insights they
give into the specific characteristics of human intelligence
in general. As he states in his introduction to Reflections
on Language (1975), he wishes to investigate those areas of
a theory of language which have more general intellectual
interest, establishing as his final goal to construct a theory
of human nature based on his view of the appropriate
explanatory framework within which a theory of language can
be built. In so doing, he wishes to "isolate and study the
system of linguistic competence that underlies behavior,
moving beyond the narrow confines of taxonomic approaches.
Given the resources available to us at present Chomsky
concludes that the most hopeful approach would be to describe
the phenomena of language and of mental activity with strict
technical accuracy developing a theoretical apparatus to as
far as possible account for these phenomena and reveal the
principles of their organization and functioning.
It is with the language faculty itself that the study
of language should be concerned and this study should
concentrate on finding answers to three areas still open to
question. First, it should determine what kinds of cognitive
structures exist. These are rich and varied, but are found
with considerable uniformity among individuals of every
^^Chomsky, Language and Mind, p. 4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86
culture. Secondly, it should account for the acquisition of
these structures. What is their source? How do they develop?
Finally, the study of language should examine both the initial
and final states in the development of cognitive structures.
The issues involved here are problems for the student of
language and as such they are matters open to empirical
investigation. They are not mysteries whose only answers
are the product of hypothetical speculation. Thus Chomsky
feels their investigation should be conducted in the same
manner as the study of the physical body, through observation,
experimentation and testing. In this way we can eventually
determine very precisely the exact nature of the cognitive
faculties including that which determines language. We
must hope that in the future it will become possible to
relate these postulated mental structures and processes to
physiological mechanisms and to place them within a 25 physiologically grounded causal framework.
Let us, then, follow Chomsky as he carries out this
task, evaluating whether his description of language and mental
activity do, indeed, satisfy the requirements of scientific
accuracy while providing a theoretical apparatus which will
account for these two phenomena and reveal the principles of
their operation taking as a model of Chomsky's most recent
position the views presented in Reflections on Language (1975).
25 Chomsky, Language and Mind, p. 14.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87
The elements of Chomsky's theory as drawn in
Reflections on Language are represented in Figure I, page 88.
For the purposes of this study we will treat the account out
lined here as a framework within which the earlier elements
of Chomsky's theory from their introduction in Syntactic
Structures (1957) to the present theory explicated in 1975,
can be drawn together, analyzed and brought within some
consistent and unified whole.
Chomsky presents the language faculty as one of a
number of mental faculties which together comprise the system
of human intellectual organization. The language faculty is
charged with the task of generating sentences which have
formal as well as semantic properties. His choice of a
faculty model for describing the organization of the human
cognitive abilities seems in itself curious for the concept of
faculties in psychology has had a very definite and not uncon-
troversal history, associated as it is with the attempt during
the late ninteenth and early twentieth century to try to
identify human emotional dispositions and cognitive functions
such as memory, imagination and understanding with specific
locations in the brain. Such approaches were considered highly
speculative and have not been able to withstand the scrutiny of
empirical investigation. As a consequence the term, itself,
is considered highly circumspent and would assumably be
avoided by anyone attempting to delineate a technically
sound model of human cognitive organization. Since Chomsky
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Common Language Sense Memory imagination Faculty Belief UG
Determl
Autonomous System of Formal Grammar
Generates Associates with
Abstract Further Logical Linguistic Principles Forms Structures of Grammar
Generates
Surface Structure of State 1 to ---- ^ SS Language Initial State Steady State
Fig. 1. The System, of Human Intellectual Organization
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89
is surely well aware of this controversy we must assume either
that he would support a somehow modified faculty view which
he does not develop here, or that his use of the term is
purely casual. Unfortunately he does not go on to define in
any detail what might be included within the category of
"other mental faculties,” so his choice between these
alternatives is difficult to ascertain. Since, however, the
language faculty is discussed, let us look at this aspect of
the theory as it is integrated within the overall model,
attempting to clarify the nature of the faculties,
specifically the language faculty, and how they interact to
gether to accomplish the task of verbal production.
As one of a number of mental faculties, we understand
the task of the language faculty to be concerned specifically
with language use. The fact that there is a specific "language
faculty" implies that it is responsible for this ability of
verbal communication. Conceivably a species of animal could
exist which possessed a number of other mental faculties, say
imagination, understanding and memory, but was not endowed with
the human ability to communite. It is even possible that such
a being could be equipped with some abilities which we our
selves do not possess. However, in looking at what is implied
in the ability to use language, what is it that is central to
the possession of a "language faculty?" In one sense what
might essentially characterize such an ability would be the
use of sound for communication. If this were the case then
it would seem that a description of this faculty would be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 90
concerned to show how a propensity to communicate is
converted into something which exists on an audible level.
Such a description would be close to a physiological model
of speech production. It would include our own language use,
but might include other possibly non-logical or non-symbolic
verbal abilities as well. It seems clear that what we
are concerned with is not a description of a physiological
process and what is essential to our attempt to describe
the language faculty is not its verbal character. What is
significant about and central to the function of the
language faculty is the ability to develop a logically
structured, rule-governed symbolic system which represents
the products of thought, memory and imagination through
clearly defined generative lines which are rooted in abstract
linguistic structures. In our case this logical representa
tion of thought takes the final form of a symbolic notation
which can be either written or verbal. That it takes this
final form is not a function of the nature of the language
faculty for all that is provided by this faculty working alone
is ". . . a n abstract framework, an idealization that does
not suffice to determine a /^surface/ grammar." Once this
abstract linguistic framework exists the final form which it
takes is determined by other characteristics of the species
(e.g. vocal apparatus). It seems not inconsistent with such
a model that a being possessing the language faculty could
communicate in some entirely different form than verbal.
26chomsky, Reflections on Language, p. 41.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91
How we conceive the mode of communication is an open question.
We might portray this verbal ability as one of the faculties
as well, or we might attempt to define it in terms of a con
tent of our memory which we apply to the abstract linguistic
stem. Whether or not this would agree with Chomsky's view,
it does seem clear that the essential nature of the language
faculty is that of providing an abstract schematic represent
ation of thought, a core from which language as we know it,
as a verbal or written form of communication, can be derived.
If this is the function of the language faculty, to
create a grammar which through the process of transformations
will generate the language we speak and understand, how then
would we want to characterize it? The key to the nature of
the language faculty appears to be the concept of innateness.
If we were to ask what specifically is the nature of the
language faculty as conceived in Chomsky's theory, what we
would be asking is what are the innate factors which determine
the structure which language takes. Chomsky seems to be
pointing to those inborn characteristics of the human species
which determine the structure of human thought and language.
It is this use of the term, innate, which has received
a great deal of attention by critics of Chomsky's theory.
What Chomsky is referring to when he uses the term, thus,
should be made clear. The concept of innate ideas within
philosophical literature has been used to refer to ideas which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92
exist in the mind at birth, prior to existence. They are a
priori. Often the concept of a priori is equated with that
of analytic, a term ordinarily predicated of statements
rather than the knowledge which they may convey. For a
rationalist such as Descartes the ideas we possess within our
minds prior to experience are considered to be of unquestion
able certainty and serve as the legitimate ground upon which
knowledge can be based. Since this certainty is not something
we gain through experience of a fact, it must derive from
some other source. The only other type of certainty is that
which results from the fact that the definition of a concept
is contained within the concept itself; it is analytically
true. Thus the terms a priori and analytic have been
collapsed together. IVhat is a priori or known prior to
experience is considered true analytically. What has never
been established as regards innate ideas is, granted they do
exist, that they are necessarily true at all. All that can
be established is that if they are true, they are true analy
tically. We might conceivably have ideas which are innate,
i.e. inborn and prior to our experience, but false. There
is no reason to conclude prior to their being examined that
they are true, analytically or otherwise.
Chomsky, however, makes it clear that by innate he is
not referring to this sense of the term. There is no built in
epistemological or logical guarantee of truth implied in that
which he wishes to so designate. Chomsky's concern is not at
all to determine, as were the rationalists, what qualified as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93
cetain knowledge. Rather by innate Chomsky is referring to
"certain abstract and in part universal principles governing
human mental faculties /which/ must be postulated to explain ? R the phenomena in question . . . -- in this case language.
These organizing principles provide a highly restrictive
schema to which any human language must conform, as well as
specific conditions determining how the grammar of any such 29 language can be used." Thus they determine the basic
nature of any language and the conditions for its use.
What is the nature of these innate principles? They
are a set of preconditions which determine the form of all
language, and as such incorporate those conditions which must
be met by such grammars, a skeletal substructure of rules to
which those grammars must conform and principles that determine
their interpretation. According to Chomsky they comprise an
"intricate system of rules that involve mental operations of 30 a very abstract nature . . . ."
The innate structures to which Chomsky refers
throughout his writings seem to be conceived as basically
physiological. In Language and Mind he refers to the future
task of relating postulated mental structures to physiological
mechanisms and interpreting what are considered mental
O -J Justin Leiber, Noam Chomsky: A Philosophical Overview Boston: Twayne Publishers, 1975), p p. 170-1. ? s Chomsky, Language and Mind, p. 48.
29%bid., p. 63. 30lbid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94
functions in terms of physiological causes. In Reflections
on Language the universality of those abstract principles
that govern the structure and use of language are seen as a
consequence of biological necessity. According to Chomsky
they derive from the mental characteristics of the species
which here are spoken of as somehow basically physiological
in nature. I'Jhat Chomsky has done, if we accept the
interpretation in Reflections is to equate mentalistic
concepts with physiological phenomena. Thus in attempting to
explain the function of innate structures he says, "We
interpret experience as we do because of our special mental
design.We can rephrase this to read we interpret
experience as we do as a consequence of the physiological
make-up of our brain. The language faculty on this
interpretation is that aspect of brain physiology which
controls language use. As physiological it is necessarily
common to the species as a whole, thus placing restrictions
which are innate, but physiological in origin and delineate
quite narrowly a certain class of grammars which will function
in interpersonal communication. Chomsky feels the question
he has raised is not, "Does learning presuppose an innate
structure?" -- of course it does for we are subject to the
conditions of our biological makeup in any case. The
important question to consider is what these innate structures
are and how do they function.
^^Chomsky, Reflections on Language, pp. 7-8.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95
Conceiving of the language faculty as at core
physiological, let us attempt to answer these questions and
examine the other elements involved in language production.
Looking again to Chomsky's model we see that universal grammar
is characterized as a component of the language faculty. If
the language faculty is equivalent to the physiological
mechanisms of the brain responsible for the production of
language then what is termed universal grammar must be a part
of this network of mechanisms wired into the brain as a
physiologically grounded set of conditions which will determine
the nature of language.
It is on the basis of research into the area of
language acquisition that Chomsky feels justified in positing
the existence of a universal grammar. Human language is in
Chomsky's view a unique phenomena without any true analogue
in the animal kingdom. It is characterized by an infinitely
large range of signals which express indefinitely many new
thoughts, intentions and feelings. Where animal systems
allow communication of an infinite range of signals in a
manner similar to the human gestural system, human communica
tion is discrete. It involves an "entirely different principle o 2 of organization." When a person acquires language what
they internalize is a system of rules that relate sound and
meaning.
32 Chomsky, Language and Mind, p. 70.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96
Recent behaviorist theories suggest that knowledge of
grammatical systems arises entirely through a child's
exposure to adult speech. Despite the variations in language
experience correct patterns are reinforced and language
handling ability is gradually developed. It is Chomsky's
conclusion that any attempt to account for language
acquisition solely in terms of stimulus and reinforcement is
inadequate.
The heart of Chomsky's argument is that the syntactical core of any language is so complicated and so specific in its form so unlike other kinds of knowledge, that no child could learn it unless he already had the form of the grammar programmed into his brain, unless, that is, he had 'perfect knowledge of a universal grammar
In experience a child encounters only a finite variety of
sentences, but somehow develops the ability to produce and
understand an infinite variety. Such an ability cannot be
accounted for within a behaviorist framework which is
concerned only with the input and output of the system. What
is crucial to an explanation of the learning process is what
occurs between the points of input and output. Chomsky wishes
to focus on the specification of the learner in the ability
of language acquisition.
This activity in which a subject develops the use of
language could be described schematically in terms of a
3 3 Gilbert Harman ed.. On Noam Chomsky: Critical Essays Garden City, N.Y.: Anchor Press/Doubleday, 1974), p. 23.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97
language acquisition devise. Utilizing primary linguistic
data consisting of the sentences, sentence fragments and
non-sentences encountered in adult speech as input into the
system, the language acquisition device searches through
possible hypotheses of what a sentence means or of how a
particular thought might be communicated and selects an
appropriate grammar on the basis of its compatibility with
the data. At this stage in the process the information
contained within the LAD is not restricted to the input,
(i.e. speech samples) but rather as a consequence of the
act of processing and inductive generalization it moves
beyond the input to include within itself a highly abstract,
intricately structured, set of rules which evolve through
experience. How the acquisition model develops these systems
of grammatical rules is determined by its internal structure,
by the methods of analysis available to it and by the 34 constraints it imposes on any possible grammar. Thus a
LAD theory is a theory of linguistic universels because it
establishes sets of conditions which limit possible grammars
and possible phonetic transcriptions. What is universal
is not the content or meaning of the language but its logical
form.
34 Chomsky, Language and Mind, p. 119.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98
The fact that all normal children acquire essentially comparable grammars of great complexity with remarkable rapidity suggests that human beings are somehow specially designed to do this, with data-handling or 'hypothesis-formulating' ability of unknown character and complexity. The study of linguistic structure may ultimately lead to some significant insights into this matter.35
What Chomsky then concludes is that the concept of a
universal grammar is necessary to the explanatory accuracy
of a theory of language acquisition. This universal grammar
takes the form of a system of principles, conditions and rules
that are properties of all human languages and are invariant
among all members of the species by biological necessity.
As a consequence the linguist must attempt to derive a
theory of language which includes within it the concept of
linguistic universels. This theory must be able to on the
one hand account for the actual diversity of languages while
on the other be sufficiently rich and explicit to account for
the rapidity and uniformity of language learning and the
remarkable complexity and range of the generative grammars
that are the product of language learning.
The most challenging theoretical problem in linguistics is that of discovering the principles of universal grammar that interweave with the rules of particular grammars to provide explanations for phenomena that appear arbitrary and chaotic.35
33chomsky, "Review of B.F. Skinner's Verbal Behavior, pp. 57-8. O ^ Chomsky, Language and Mind, p. 48.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99
Moving on to the mediate level of Chomsky's model and
looking at the function of universal grammar we see that it,
as part of the language faculty in general, determines what
Chomsky designates the autonomous system of formal grammar.
This formal grammar in turn generates abstract structures which
correspond through further sets of principles of grammar with
logical forms. The surface structure of an actual language
results through the interaction of this language faculty with
other mental faculties which make up the system of human
intellectual organization.
This portion of Chomsky's model raises several puzzling
questions. In attempts to integrate this most recent
characterization with Chomsky's earlier theory how do we
distinguish the autonomic system of formal grammar from what
Chomsky terms the abstract linguistic structures? Where does
the concept of deep structure so characteristic of his early
theory fit into the more recent model? What does it mean to
say that abstract linguistic structures are associated with
logical forms? Before evaluating the usefulness of Chomsky's
model let us attempt to determine the interactions between
various levels and how they function together to provide an
in-depth characterization of human language-handling ability.
It appears that the autonomous system of formal grammar
is in some sense parallel with Humboldt's form of a language.
Thus through the interaction of the innate structures
responsible for language acquisition we are provided with an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100
abstract formal characterization of a language. It appears
this characterization can be conceived of in terms of sets
of rules which in turn form the basis for the generation of
abstract linguistic structures.
If we had to place deep structure within this model it
seems most appropriate to place it at the level of the
abstract linguistic structures for it is at the level of
deep structure that the logical relations underlying language
are expressed. This would coincide with the association of
the abstract linguistic structures with logical forms.
Through a series of transformations the grammatical elements
contained at the level of deep structure are related to
those which make up the surface structure of the language.
The grammar of a language consists of a series of
rules through which all the acceptable sentences of a
language can be produced. A grammar is adequate if it gives
a full and accurate description of the language and is cap
able of producing all the acceptable sentences of the
language. A finite state grammar attempts to account for
variety in speech in terms of sets of limiting conditions.
Every choice the speaker makes establishes limiting condi
tions for the succeeding state. Rather than portraying the
speaker as actively and creatively involved in making
linguistic choices, it represents the choices as each pre
determined by prior choices as in the following example:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101
Thus once the speaker has chosen as the subject 'a man' they
are forced to choose 'comes' rather than 'come' as the verb
of the sentence. Finite state grammar is limited as well in
that it does not provide an adequate tool for representing
all the sentences of English. Chomsky feels that the nature
of language requires a more powerful model to incorporate the
diversity of meaning within language.
The alternative to finite state grammar has generally
been considered to be phrase structure grammar in which the
structure of a sentence is analyzed in terms of tree-like
diagrams representing the logical and grammatical structure
of the sentence. Chomsky sees limitations as well as this
type of analysis. Since it reflects only the surface structure
of the language, it is not capable of adequately specifying
all the demonsions of meaning inferred by the sentence. He
concludes, ". . . any theory which, like the theory of phrase
structure grammar, assigns a single phrase marker . . . to an
utterance, is incapable of expressing deeper structural
relations and must therefore be ruled out by considerations 37 of descriptive adequacy." Chomsky believes these problems
37chomsky, Current Issues in Lingusitic Theory, p. 64.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102
might be alleviated if phrase structure grammar were used in
conjunction with sets of rules linking deep with surface
structures. These rules would make up what Chomsky terms
transformational grammar. It would consist of an unordered
set of rewriting rules whose structural description could also
be represented with a tree diagram. These rewriting rules
would apply in a prescribed sequence to generate a restricted
set of base strings. This base would be divided into a
categorical component which would define the basic grammatical
or logical relations which are found at the level of deep
structure but not always obvious on the surface rendering,
and a lexicon or vocabulary. This base in turn would generate
a sequence of base phrase markers which underlie a sentence
and are mapped into the surface structure by transformational
rules, thus automatically assigning a derived surface level
phrase marker in the process. It is the deep structure which
determines semantic interpretation. The surface structure
determines the phonetic form. Language viewed in this manner
involves a particular relation of syntax, semantics and
phonetics which can be characterized as transformational in
nature. At the level of deep structure we find the core
logical relations are established. Deep structure provides
the semantic component of grammar. Through syntactic rules
this semantic component of grammar is transformed into the
spoken (or written) form which provides the phonological 38 component.
3&Chomsky, Language and Mind, p. 125.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 103
As we can see it is at the level of deep structure that
the logical form of language can be determined. Where the
logical relations implied on the surface are often ambiguous,
by analyzing the sentence in terms of its underlying structure
it is possible to identify unequivocally the meaning inherent
in the sentence.
In his model Chomsky specifies that the abstract
linguistic structures are "associated with" logical forms
through further principles of grammar. Thus these seem to
be equivalent with what was earlier referred to as deep on structures. But what does Chomsky mean when he says these
structures are associated with logical forms? Again despite
the relevance of this association to any adequate under
standing of language, Chomsky's account does not begin to
satisfy our curiosity. In being "associated with" logical
forms it appears that Chomsky sees the base phrase markers
at this level having in most if not in all cases a corres
ponding logical form the sum total of which make up the
meaning of the sentence. If a sentence such as "The owl and
the pussy cat went to sea in a beautiful pea green boat."
for example, were analyzed to the level of base phrases, we
would have a set of core semantic elements each reflecting
a logical relation. These would include:
^^In Reflections on Language Chomsky notes that he felt a change in terminology was warranted because of the misunder standings which arose in the use of the term, 'deep structure.'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104
(1) The owl went to sea.
(2) The pussycat went to sea.
(3) The owl was in the boat.
(4) The pussycat was in the boat.
(5) The boat was pea green.
(6a) The boat was beautiful.
or possibly
(6b) The pea green was beautiful.
Whether 6a or 6b contained an accurate rendering of the
meaning of the sentence while not clear at the surface level
could be determined if we knew which was contained in the
deep structure.
The fact that Chomsky draws a parallel between the
linguistic structures which underlie language and logical
forms raises the important question of the extent to which
language can be completely analyzed in terms of logical forms,
If the parallel were one to one and a logical form could be
identified for every semantic element at this level, then
presumably it would be possible to rewrite sentences in
symbolic notation which would fully and unambiguously
represent the meaning contained in the sentence. In other
words, language would be reducable to symbolic notation.
Chomsky goes no further in pursuing this possibility.
Having looked at some of the major accomplishments
Chomsky has made in his research on linguistic and mental
structures, what are the strengths of this analysis and what
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 105
contribution does Chomsky's view of language give to our
understanding of how we organize and process ideas?
One difficulty in the past with theories which attempted
to give an account of mental processes was their lack of
confirmability. As a consequence studies in that area were
often considered of questionable validity and weak in respects
similar to Freud's theory of the subconscious. What Chomsky
presents us with, though certainly open to criticism on a
number of fronts is in Searle's terms, a revolution in
linguistic theory from at least two perspectives. First
Chomsky opens up the study of mental phenomena to empirical
research applying the methods of strict experimental procedure
to studies of the human mind. Going beyond that he moved
from the methodological principle of confining research to
observable facts to using these observable facts as clues to
laws of mental operation which are at present inaccessible to
direct laboratory observation. Chomsky feels that the worth
of his approach is that now we have "something explicit to
investigate in place of vague and near vacuous discussions of
the child's unspecified 'dispositions' which are not uncommon 40 in the literature." He points as well toward future research
which will begin to draw the connections between what we have
called cognitive structures and physiological phenomena.
^^Noam Chomsky, The Logical Structure of Linguistic Theory (New York: Plenum Press, 1975), PP- 11-12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106
Related to this strength which Chomsky's research
represents from an empirical perspective is a second major
advantage of his approach. Where in the past laboratory
studies of learning and specifically of language acquisition
could only speak in terms of the performance of a subject in
a particular learning task, Chomsky is concerned to present
an analysis of language learning in terms of linguistic
competence. By competence Chomsky is referring to the
ability of the idealized speaker/hearer to associate sounds
and meanings strictly in accordance with the rules of the
language. It can be expressed as a system of rules that
relate verbal (or written) signals to semantic interpreta
tions of these signals. Knowledge of a language on this
view would be equivalent to linguistic competence and would
involve the ability not just to produce and understand
language, but in Chomsky's terms to assign deep and surface
structures to an infinite range of sentences and to relate
these structures appropriately thereby assigning a proper
semantic interpretation at the level of deep structure and
phonetic interpretation to the surface phenomena.
The question which must be raised in this regard is
how does one study competence when all that we can observe
is the performance of the subject. It is obvious that we
must look to performance as a clue to the understanding of
competence. However no clear parallel between performance
and competence can be drawn. For instance a child may
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 107
approximate adult patterns at a low level of frequency at
first despite the fact that they possess a high degree of
competence in that specific language handling skill. But
once we have been able to observe how the subject uses
languages we should move from recording and documenting
speech to an examination of the subject's underlying
abilities to use and understand language. By abstracting
away from conditions of use and dealing on the level of formal
structures and the formal operations that relate them, we
see that it is possible to study linguistic competence in
abstraction from the problems of use.^^ This approach is
opposed to the view that all that is involved in the study of
language is the study of performance and that knowledge of a
language can be seen as a system of habits and S-R connections.
A final very significant strength of Chomsky's approach
and specifically of his model of the language faculty involves
the attempt to integrate those cognitive elements specifically
and uniquely associated with language into the overall
organization of the human intellect. In looking again to his
model of human intellectual organization what we see is not
a static portrayal of the processes involved in language
production. Chomsky's account demonstrates dimension on two
levels. Although formal, the system Chomsky presents is not
static. The structures which are generated in a continuous
^^Chomsky, Language and Mind, p. 111.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108
process of change, beginning in their initial state through
maturation or experience, moving toward equilibrium or
steady state, are thus established and subsequently revised.
In specifying that the language faculty is only capable
of providing an abstract framework within which spoken (or
written) language can fit Chomsky also points to a dynamic
interaction which is occurring continuously between the
language faculty and the other mental faculties. Language is
finally produced by this interaction between the faculties
rather than being functionally isolated within one faculty
alone. The significant consequence of this interpretation of
which Chomsky only hints in Reflections is that if no sharp
division can be drawn between the semantic properties of the
language faculty and those of common sense understanding then
the possibility remains that these might be tied together by
a network of common sense beliefs which operate as a unified
whole with the activities of other parts, reflecting and
effected by the activities of other parts. Such an interpre
tation would lend support to any attempt to examine the
structure of language as to some degree representative of the
structure of thought. That Chomsky has broken down the bar
riers between research in language and cognitive studies opens
up new and interesting directions for research. In discussing
the single most important value in Chomsky's work Harman, in
his introduction to the critical essays on Chomsky's works,
notes that there is nothing which has had a greater impact on
contemporary philosophy than Chomsky's theory of language.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109
"More precisely Chomsky has let us see that there is a single
subject of language and mind which crosses departmental
boundaries. As we learn more about language we begin to
better understand the workings of the mind. And in this
process we are engaged in an inquiry which has its roots well
grounded in philosophical enterprise.
Despite the obvious strengths of Chomsky's approach,
his work has sustained serious criticism from a number of
quarters. The first question which must be raised is whether
Chomsky is engaged in a legitimate philosophical inquiry. One
criticism that has been leveled against Chomsky's work is that
it represents an empirical rather than a philosophical approach
and does not begin to solve traditional philosophical problems
in the philosophy of mind. The question which is being
addressed here is just where the line should be drawn between
the proper sphere of lingusitics and of philosophy. Are
linguists encroaching on a territory which has already been
staked out by philosophers and should their activities be
restricted to the collection of data on language use? On the
other hand as techniques are developed for a more precise
understanding of the phenomena of language is it inevitable
that this field will eventually become the domain of
linquistics? Is the proper sphere of philosophy limited to
conceptual analysis.
^^Harman, Noam Chomsky: Critical Essays, p. vii,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110
In attempting to draw a line between linguistics and
philosophy Chomsky recognizes the empirical nature of his
research. However, to the extent that this empirical
research has a bearing on significant questions in philosophy
he stresses that the conclusions he reaches are philosophical
in nature. Chomsky does not wish to restrict philosophy to
conceptual analysis, recognizing that significant theoretical
inquiry is in essence philosophical. In discussing the
relevance of Chomsky's research to traditional problems in
philosophy Justin Leiber in his examination of the philo
sophical implications of Chomsky's research concludes:
The extention of Chomsky's linguistics into semantic theory that Katz applied to the problems of analytic philosophy were less alien to recent philosophy and better understood, though it has generally been regarded with great and I think deserved scepticism.43
A second criticism of Chomsky's approach rests with his
use of the concept of innateness. Nagel criticizes this use
as lacking the force of analyticity which was central to the
meaning of the term in traditional philosophical literature.
The importance of all this is that the innate factor, which Chomsky argues must underlie our language-learning capacity, bears no resemblance to the sort of unquestionable epistemologically unassailable foundation on which some philosophers have sought to base human knowledge, and which is generally referred to as a priori or innate knowledge.
43 Leiber, Noam Chomsky: A Philosophical Overview, p 166.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I l l
A mere innate tendency to believe certain things or perform in certain ways, no matter how universal, is not a priori k n o w l e d g e . 44
Putnam criticizes the use of the concept of innateness
from another perspective.^" If there are significant
uniformities in language they can be adequately accounted for
on a much simpler model. Putnam suggests the concept of
"general multipurpose learning strategies" as one possibility.
Following an empirical investigation of sepcific areas of
human competence, an investigator would devise hypotheses to
account for human competencies. If the learning strategies
were found to be the same in several areas then the existence
of general multipurpose learning strategies would be supported.
Putnam feels that invoking the concept of innateness only
postpones the problem of learning rather than solving it for
it really tells us very little about how language is learned.
In his review of Reflections on Language, Scinto
criticizes the manner in which Chomsky supports a belief in
innate structures. He claims that Chomsky argues from the
fact that structuralists are concerned with discovering the
grammar of a language to the need to establish a theory of
mind as a precondition to the writing of a grammar. From this
Chomsky argues that there must be a number of universal
44Ernest Nagel, "Linguistics and Epistemology," in Harman, On Noam Chomsky: Critical Essays (Garden City, N.Y.: Anchor Press/Doubleday, 1974), 45see Hilary Putnam, Mind, Language and Reality: Philosophical Papers (Cambridge: Cambridge University Press 1975.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112
features common to all adequately determinate grammars and
these must derive from innate features of a language
acquisition device.
A final and quite serious criticism of Chomsky's work
involves the question of whether Chomsky has given adequate
attention to non-linguistic factors that are involved in
language acquisition and the place of language within the
larger context of human conceptual activity. In being
concerned with examining language and accounting for its
origin and structure Chomsky does not satisfy the concerns
of those who wish an adequate explanation of the nature of
thought, of how we come to understand concepts through our
interaction within the human sphere. He does not provide an
analysis of meaning which accurately defines the extent to
which our ideas derive through the continuous growth in our
understanding of the world around us. This question is raised
by Partee in "Linguistic Metatheory.Here she states that
she finds Chomsky's autonomous syntax approach in Syntactic
Structures to be "surprising." In defending his separation
of syntax and semantics in The Logical Structure of Linguistic
Theory (1975) Chomsky states that it is not necessary to
appeal to meaning in the development of linguistic theory.
Chomsky's objection to meaning as a criterion of analysis is
^^Barbara Partee, "Linguistic Metatheory," in Harman, On Noam Chomsky: Critical Essays (Garden City, N.Y.: Anchor Press/Doubleday, 1974).
Reproduced with permission of the copyright owner. Further reproduction prohibitedpermission. without 1 1 3
based on what Chomsky considers the obscurity of semantic
notions. Because of vagueness and unclarity, meaning is an
ineffective tool for grammatical analysis. Chomsky goes even
further and is willing to assert that semantic notions are
not only unnecessary, they are "irrelevant." It is possible
to account for grammaticality solely in terms of syntax.
John Searle criticizes Chomsky's separation of syntax
and semantics as both pointless and perverse. It is Searle's
view that language should be properly viewed as an institu
tional fact governed by constituative rules. Linguistic
competence would be interpreted as an ability to use sentences
to accomplish goals or intentions. These could only be
understood within the context of interpersonal communication,
as speech acts. Such an interpretation is, according to
Searle, entirely consistent with Chomsky's view of the
autonomy of formal grammar. Why, then, does Chomsky insist
on studying syntax independent of semantics? The explanation,
according to Searle, can be found in Chomsky's view of man
as essentially a syntactical animal.
The structure of his brain determines the structure of his syntax, and for this reason the study of syntax is one of the keys, perhaps the most important key, to the study of the human m i n d . 4 7
Thus in studying language for the insights which it gives into
the human cognitive processes Chomsky restricts his interest
to syntax. The question this raises is whether language can
^^Harman, On Noam Chomsky : Critical Essays, p. 15.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 114
be understood solely in terms of syntax? Can we adequately
account for language by examining its form rather than its
use? Recent approaches while following the direction of
Chomsky’s research have disagreed with Chomsky on this point.
In stressing their connection with the overall thrust of
Chomsky's approach, generative semantics stresses that the
generative aspect of linguistic theory lies in its semantic
rather than its syntactic component. A grammar starts with
the meaning of a sentence and subsequently generates the
syntactical structures through the introduction of syntactic
and lexical rules. On this view syntax is seen as a collec
tion of rules for expressing meaning. Rather than establish
ing a boundary between syntax and semantics, the generative
semanticists are concerned with how form and function
interact together to establish meaning. According to Searle,
. . . there is no way to account for the meaning of a sentence without considering its role in communication, since the two are essentially connected.48
A second attempt to integrate semantic and syntactic
components is referred to by Partee as the Katz-Postal 49 hypothesis. Katz and Postal support the view that Chomsky's
deep structure as described in Aspects of a Theory of Syntax
does provide a sufficient basis for semantic interpretation
“^^John Searle, "Chomsky's Revolution in Linguistics, in Harman, On Noam Chomsky: Critical Essays (Garden City, N.Y.: Anchor Press/Doubleday, 1974), pf 3ÏÏ 49partee, "Linguistic Metatheory."
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 115
and that a careful syntactic analysis will lead to derivations
at the level of surface grammar in which transformations are
meaning preserving.
It appears that Chomsky in Reflections on Language
provides a more significant role for semantics than was the
case in earlier versions of his theory. Here, as we recall,
Chomsky locates the semantic component at the level of
abstract linguistic structures and considers the syntactic,
semantic and phonological components as all essential to an
understanding of language. Chomsky also implies in his
analysis that verbal (or written) communication is a function
of the interplay of several non-linguistic factors associated
with other mental faculties, the imagination, memory and the
understanding, all of which operate together with the language
faculty in the production of language.
If Chomsky were to exclude a whole range of questions
regarding meaning, performance and style from his account, its
usefulness would be significantly reduced and its practical
implications would surely be limited. Given however that
Chomsky feels he has developed a theory which can be integrated
into a general theory of learning, what impact will this
interpretation have on the development of an integrated theory
of learning? In Language and Mind (1972) Chomsky states.
Insofar as we have a tentative first approximation to a generative grammar for some language, we can for the first time formulate in a useful way the problem of the origin of k n o w l e d g e . 50
50chomsky, Language and Mind, p. 78.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 116
We are now in a position to ask "What initial structuresmust
be attributed to the mind that enable it to construct such a
grammar from the data of sense?We have seen that
universal grammar is one element which would be contained in
a theory of language acquisition. So too we see that
parallels can be drawn between linguistic structures and
specific logical forms.
In judging the adequacy of a theory of learning we
must determine if the theory is capable of accounting for the
degree of sophistication which is characteristic of the
cognitive organization of mature organisms. Chomsky has
been led to the conclusion that the cognitive organization
of the mature human is a complex, integrated system which
includes cognitive structures. Familiar learning theories
have failed to account for these cognitive structures. If
we are to develop learning theories which do adequately
describe the level of complexity of human thought, Chomsky
finds it essential to move beyond the limitations of the
S-R model. For our purposes what we would like to see
developed further in Chomsky's theory would be a specifica
tion of the exact nature of the relationship between abstract
linguistic structures and logical forms. If abstract
linguistic structures have their roots in forms which are 52 logical in character, and if, as Chomsky seems to imply
^Chomsky, Language and Mind, p. 79. 52 See Chomsky, Reflections on Language.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 117
there are links between the semantic component of linguistic
structures and those of common sense understanding, this would
have significant implications for a theory of learning.
Chapter IV will examine the progress which has been made in
this regard looking to work in the area of philosophy of
linguistics and generative semantics for answers to these
questions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PART II: SECTION B
AN ANALYSIS OF RESEARCH IN LANGUAGE AND THOUGHT
AND ITS IMPLICATIONS
FOR THE DEVELOPMENT OF AN INTEGRATED THEORY OF LEARNING
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER IV
SEMANTICALLY BASED THEORIES
Spoken words are the symbols of mental experience and written words are the symbols of spoken words. Just as all men have not the same writing, so all men have not the same speech sounds, but the mental experiences, which these directly symbolize, are the same for all, as also are those things of which our experiences are the images.^
The revolution which was touched off in linguistics
with the writings of Noam Chomsky has set off sparks in the
field of philosophy as well, for the questions which, as we
saw, plagued Aristotle in his attempt to provide an account
of how we organize and structure our knowledge about the
world, though mollified in the conclusions of Kant's Critique
of Pure Reason, were never fully satisfied. Although Kant
provided a model of the process which occurs in knowledge
acquisition, he left unanswered (or unanswerable) questions
involving the relationship between the world of ideas and the
world of facts, such questions as to what extent can we infer
from our knowledge of the world to the nature of the world
itself; to what extent are we justified in designating some
ideas true or valid and others false. We were provided with
a structural design for knowledge, but we were not able to
connect that model in any meaningful sense with the real world.
^Aristotle, The Basic Works of Aristotle, p. 40.
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 120
Kant's account suffered a second, crucial inadequacy.
While establishing a formal classificatory schema of categories
into which we separate out and characterize ideas, Kant
provides us with no adequate defense for choosing the very
categories which he does. His system, thus has no firm
empirical base upon which a defense of its structural design
can be mounted. Rather Kant, in the classicalist tradition,
returns to Aristotle and the organizational form of
prepositional logic to provide an epistemological framework.
There is a clear parallel between the Kantian categories of
the understanding and the elements of Aristotelian logic.
That this is a possible way to classify ideas does not lead
us to the unarguable conclusion that this is the way to
classify ideas. What is needed if we are to establish a
productive theory of the nature of cognitive structure is to
ground our conclusions in a meaningful empirical base. If
we want to talk about the structure of thought, we must look
to ideas and determine just what that structure is. It is
here that the recent extensive examinations of language viewed
as a reflection of thought become relevant, for they provide
a firm theoretical base from which empirically verifiable
conclusions can be draxmi. At this stage in the development
of linguistic theory we are beginning to uncover information
which is directly relevant to the solution of significant
epistemological problems. If we view written words as the
symbols of spoken words, and spoken words as the symbols of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121
mental experience, we can continue the task which Aristotle
began, to give form and substance to our knowledge of the
structure of thought.
Chomsky's work in transformational grammar and syntax
might be viewed as an initial step in this foray. It provides
us with a rationale for expanding our understanding of
grammar to encompass the logical as well as empirical (surface)
features of grammar and a technique for formalizing the
connection between surface phenomena and an underlying logical
structure. Where the currently dominant taxonomic approach
in linguistics viewed grammar as an elaborate system of
classification and segmentation of language into syntactically
based units which are assumed to contain all the relevant
phonetic and semantic information, transformational theory
is concerned to treat a theory of grammar as a theory to
explain how speakers associate acoustic signals with meanings.
In such an interrelation the syntactic component generates a
multidimensional analysis for such sentence, with a series
of levels at which any sentence can be broken down and
analyzed. At the level of surface grammar all the information
necessary for a phonological description can be identified
while a semantic description requires the derivation of deeper
level grammatical categories.
The theory-rich implications of Chomsky's work had
inevitable consequences for the philosophy of language. In
the twentieth century the concern with defining the bounds
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 122
within which language could function as the expressive medium
for philosophical thinking had generated the theories of
logical positivism and ordinary language philosophy. Reacting
against the excessive spiraling of speculative metaphysics
at the close of the nineteenth century, logical positivism
argued that the sentences of ordinary language were too
irregular and unsystematic for use in the solution of
significant philosophical questions. The attempt was made to
replace ordinary language with an artificial or logical
language in which every term was clearly referenced and in
which the relationships between terms were fully specified.
Ordinary language philosophy, on the other hand, argued for
the validity of studies of language use as the only way in
which an untangling of conceptual problems could be achieved.
For the proponent of this approach philosophical confusion
lays not in the unsystematicity of language, but in the
failure of those who study language to use it conventionally.
Thus we recall Wittgenstein's stinging attack against
philosophers and the conclusions which they draw as the work
of "savages, primitive people, who hear the expressions of
civilized men, put false interpretation on them, and then 2 draw the queerest conclusions from it."
o Ludwig Wittgenstein, Philosophical Investigations, trans. by G.E.M. Ans combe, 2nd ed. (New York; Macmillan, 1958), p. 47e.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 123
However, within some philosophical circles it was felt
that the ordinary language approach, while useful in clarifying
specific terminological confusions, did not provide an
adequate base for the overall solution of philosophical
problems. Philosophers familiar with the work being done in
linguistics and transformational grammar, have turned their
attention to the insights which have been gained about language
viewed as the mode through which thoughts are transmitted
from one person to another, see it as a base upon which a
productive theory of language as well as of mind could be
built. In the following pages we will look to the writings
of Jerrold J. Katz and Jerry A. Fodor, two major proponents
of this approach, and consider the extent to which they
provide insight into the nature of mind and of the logical
structure of thought.
Fodor and Katz reject as inadequate the approaches of
both positivism and ordinary language philosophy. Where
positivism views language as irregular and unsystematic, they
argue that studies in language aquisition demonstrate results
that can only be explained on the assumption that language
is highly systematic and regular. Consequently what is
needed is a theory which represents the full complexity of
language, rather than the inherent limitations of an artificial
language. The attempts by Carnap, Russell and Whitehead in
the early years of the twentieth century to develop a language
of logic proved incapable of expressing the complexities
inherent in natural language.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 124
However, Fodor, and Katz are equally dissatisfied with
more recent approaches.
One must agree with the positivist’s charge against the ordinary-language philosopher that any account of a natural language which fails to provide a specification of its formal structure is ipso facto unsatisfactory. For it is upon this structure that the generative principles which determine the syntactic and semantic characteristics of a natural language depend.^
The ordinary language philosopher has failed "to appreciate
the significance of the systematic character of the composi
tional features of language."^ As a consequence any hope of
penetrating beyond the surface into the underlying logical
structure of language is slight.
In discussing the philosophical import of studies in
empirical linguistics and psycholinguistics Katz sets forth
his view of the approach which philosophy of language should
take. "I shall argue that it is possible to infer the form
of the thought beneath the outward form of surface grammar
. . . ."^ In Katz' view linguistic theory incorporates
solutions to significant philosophical problems. By studying
the outward surface phenomena of language and its grammatical
structure we are led to an understanding of its inner logical
form. Thus in The Philosophy of Language (1964), he states
^Jerrold J. Katz and Jerry A. Fodor, The Structure of of Language: Readings in the Philosophy of Language (Englewood Cliffs, N.J.: Prentice-Hall, 1966)7 p. 11.
^Ibid. 5Jerrold J. Katz, The Underlying Reality of Language and its Philosophical Import (New York: Harper and Row, 1?71), p . 12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. .25
as his goal, . .to understand conceptual knowledge on the
basis of discoveries in empirical linguistics about the manner
in which such knowledge is expressed and communicated in
natural languages."^ In The Underlying Reality of Language
and its Philosophical Import (1971) he continues.
The special task of philosophy of language, which distinguishes it from other branches of philosophy, is that it seeks to shed light on the structure of conceptual knowledge on the basis of insights into the structure of the languages in which such knowledge is expressed and communicated./
A theory of language is an attempt to understand the
universal characteristics of language, to give an account of
language which holds true for all forms of human communication.
"It formulates the principles that determine the necessary
form and content of natural language ..." while giving g definition to the notion of a 'natural language.'
Abstracting from sets of empirically adequate linguistic
descriptions, it generalizes to a theory of linguistic
structure. The theoretical constructs used by the linguist
thus serve to represent these features which are found to be
invarient from language to language. Expressing the essential
features of syntax, meaning, and sound, they provide the
apparatus upon which solutions to philsophical puzzles can be
built.
^Jerrold J. Katz, The Philosophy of Language (New York Harper and Row, 1966), p. x.
^Katz, The Underlying Reality of Language, p. 183. ^Katz, The Philosophy of Language, p. x.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 126
Drawing from the data of descriptive linguistics, then,
Katz suggests that the philosopher of language will be able
to draw conclusions regarding the form and content of
conceptual knowledge from information on the form and content
of language. His approach combines the logical positivists
conception of a formalized theory of linguistic structure
with the ordinary language philosopher's demand that the
conventions of natural language are the proper sphere of
philosophy and will prove adequate for dealing with problems
of a philosophical nature. Since the basic premise of the
philosophy of language on the Fodor/Katz account, is "that
there is a strong relation between the form and content of
language and the form and content of conceptualization^
Katz sees the tasks of theoretical linguistics and that branch
of philosophy which is concerned with the logical structure
of language to be closely interrelated.
The linguist whose aim is to provide a statement of ideal linguistic form unadulterated by the influence of such extraneous factors can be compared to the logician whose aim is to provide a statement of ideal implicational form unadulterated by extraneous factors that ,„ influence the actual inferences men draw.
Russell,the early Wittgenstein and Carnap, assumed that grammar
was not sufficient to reveal the logical form of propositions.
^Katz, The Philosophy of Language, p. 4.
Jerrold J. Katz, "The Relevance of Linguistics Philosophy," Journal of Philosophy 62 (1965): 590-602.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 127
However, in Katz' view, "Suitably extended, grammar might well
reveal the facts about logical form, too."^^ In considering
the function which Chomsky's generative transformational
grammar might play in this activity, he concludes,
. . . the distinction between underlying and superficial syntactic structure is a significant step toward the philosopher's distinction between logical form and grammatical form . . . .
Again in his outline of semantic theory in 1972 Katz reiterates
this claim that the study of grammatical structure in
empirical linguistics will eventually provide a full account
of the logical form of sentences in natural language.
The idea underlying this conception is that the logical form of a sentence is identical with its meaning as determined compositionally from the senses of its lexical items and the grammatical relations between its syntactic constituents.
In this activity Katz points out the similarities between his
own approach and that of Frege. It was Frege's position that
anything which is thinkable is communicable through some
sentence of a natural language because the structure of
sentences are a mirror of the structure of thought.
How, then does the Fodor/Katz analysis move from the
syntactical approach of Chomsky to a semantically based
l^Katz, "The Relevance of Linguistics to Philosophy," p. 594.
12-Ibid.,- p . 597.
*JerroldJerrold J. KatKatz, Semantic Theory, (New York: Harper and Row, 1972), p. xxiv.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 128
theory of lingusltlc description? Katz proposes what he
conceives to be an accurate model of the formal structure
of language^^ as reflected in Figure 2.
SYNTACTIC COMPONENT
generates abstract formal objects
underlying phrase superficial (derived) markers nhrase markers
input Input
SEMANTIC PHONOLOGICAL COMPONENT COMPONENT
Outout Output
Meaning Sound
Fig. 2. The Formal Structure of Language
Beginning with Chomsky's account of syntax phrased in terms
of superficial or derived phrase markers at the level of
surface grammar and underlying phrase markers at the level
of deep structure, Fodor/Katz develop further the conception
14Katz, "The Relevance of Linguistics to Philosophy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 129
of the semantic component of language and its relationship
to the corresponding phonological component. The syntactic
component is equivalent to a system of rules which generate
a string of formatives (words) having one or more structural
descriptions (sets of labeled bracketings which define the
syntactic relationships holding between the formatives).
These bracketings are phrase markers. The rules are phrase
structure rules which define how derivations of specific
sentences are to be made from the grammatical components of
the syntactic structure. An example of a set of phrase
structure rules would be the following:
1. Sentence______NP + VP
2. NP = Philosophers
3. NP = Truth
4. VP = V h NP
5. V = Love
The corresponding labeled bracketing or phrase marker would be,
Sentence
NP VP
Philosopher + s V NP
love truth
The output of the phrase structure subpart of the syntactic
component is a finite set of phrase markers. These correspond
with what Katz identifies as underlying phrase markers. At
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 130
this point the rules of the transformational subpart of the
syntactic component come into play and on being applied to
the underlying phrase markers, derive new phrase markers.
The terminal symbols of this process which may consist of one
or a series of applications of transformational rules are the
superficial phrase markers of surface structure. On this
interpretation,
The logical form of a sentence is a set of its semantically interpreted underlying phrase markers; the grammatical form of a sentence is its superficial phrase marker with its phonetic representation.15
The syntactic together with the semantic components comprise
a theory of logical form while the syntactic together with
the phonological components comprise a theory of grammatical
form. Thus we see that a comprehensive theory of
language universels contains three subtheories. Syntactic
theory expresses the grammatical universals of language,
phonological theory states the formal substantive and
componential universals which make up the sound structure of
language, and semantic theory expresses those features which
are common to the sense of language abstracted from any
specific grammatical form.
A significant point of difference exists between the
account of Fodor/Katz and the Chomskian analysis. Fodor and
Katz see the semantic elements of language to be fully
l^Katz, "The Relevance of Linguistics to Philosophy," p . 599.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 131
autonomous and related to the syntactic elements only by
those abstract connections we have noted which tie the three
systems together.
Grammatical and semantic markers have, then, different theoretical import. Grammatical markers mark the formal differences on which the distinction between well-formed and ill- formed strings of morphemes rests, while semantic markers give each well-formed string the conceptual content that permits it to be a means of genuine verbal communication. They are concerned with different kinds of selection and they express different aspects of the structure of a language. We can justifiably regard semantic markers as theoretical constructs distinct from the markers employed in grammatical description.16
We can see that the relation between form and meaning is, then,
entirely arbitrary. The semantic component operates
exclusively on the underlying phrase markers whereas the
phonological component operates on phrase markers at the
level of surface phenomena.
Having separated out semantics as an autonomous area
of study Katz and Fodor together in "The Structure of a
Semantic Theory" and Katz in his major work. Semantic Theory,
published in 1972 move on to give substance to a theory of
semantic description. At the least a semantic theory must
explain how a speaker is able to interpret sentences from
their context. It should explain the selective effect of
setting upon how a speaker understands sentences. At most a
l^Jerrold J. Katz and Jerry A. Fodor, "The Structure of a Semantic Theory," Language 39 (1963); 210.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 132
semantic theory should be expected to serve as a theory of
interpretation, defining how a speaker processes the
information which he encounters in his environment and
integrates it into his cognitive structure.
The components of a semantic theory together provide
an explanation of how speakers are able to interpret sentences
which in most cases are entirely novel. These include a
dictionary of the language which enumerates the possible
meanings for each word of the language as well as projection
rules which define how the dictionary definitions are to be
applied. The dictionary entries have two parts consisting
of a grammatical section which identifies the syntactic
classification of each item in the lexicon and a semantic
section which represents each of its senses. The projection
rules assign semantic interpretations to the formatives of
the syntactic component in the following manner. Each lexical
item receives a meaning as specified by the dictionary entry.
The projection rules combine the meanings of the lexical items
according to the syntactic description of the sentence (i.e.
as required by its grammatical form). Thus the output of the
syntactic component acts as an input to the semantic component.
The semantic interpretation of a sentence is defined by Katz
as an understanding of the full set of statements that can be
made about the meaning of a sentence. "Given definitions for
each semantic property and relation . . . it will be possible
to enumerate the full list of semantic predictions about the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133
semantic properties and relations of S. This list is the
semantic interpretation of S.”^^ The account which Katz has
provided is of interest in that it defines meaning in terms
of properties and relations. This approach could have
significant consequences for the development of a formal logic
of natural languages and although this tact is not pursued
in any deliberate way in either Katz or Fodor it remains
a possible consequence of their interpretation.
In attempting to delineate the semantic interpretation
of any sentence, Katz/Fodor utilize the tree-diagram
characteristic of transformational grammar. A semantic
analysis of any term would include semantic markers, elements
in terms of which semantic relations between items in the
lexicon are expressed. Semantic markers, are similar to
logical categories and parallel grammatical markers (e.g.
noun, verb, adjective, etc.). The analysis would also include
distinguishers which define any specific distinguishing
characteristics of an element within the analysis. Katz/
Fodor provide as an example the semantic analysis of 'bachelor.'
The terms in ( ) are the semantic markers of the analysis
while those in j_ _ J are the distinguishers. Such a tree-
diagram provides us with a full interpretation of the semantic
character of the term being analyzed.
^^Katz, Semantic Theory, p. 47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 134
Bachelor
I noun
(human) (animal)
(male) /who has lowest (male) academic degree/”
\ /who has (young) (young) never married/
/knight serving /fur seal without under standard a mate during of another knight/ breeding season/
We will recall that on the Katz/Fodor analysis the
semantic component while differentiated from the syntactic
component was generated from the output of the syntactic
component, the underlying phrase markers (see diagram on
page 128). This approach is now identified as interpretive
semantics to distinguish it from what has been labeled
generative semantics which rejects the position that underlying
phrase markers serve as the full input to the semantic
component and that the semantic component is an interpretive
component which operates on phrase markers which have been
independently generated by the syntactic component. Rather
than seeing semantic representations as interpretations of 18 syntactic structure, George Lakoff argues the alternative
ISceorge Lakoff, "Linguistics and Natural Logic," Synthèse , 1971.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 135
position that they are independently generated. The semantic
component then provides the input to the transformational
component and the relation between the semantic and syntactic
components are reversed. These inputs into the transforma
tional component are in turn mapped onto the surface level.
Thus generative semantics relates representations of sentences
to superficial phrase markers and ultimately to phonetic
representations by transformational rules with the interven
ing level of underlying phrase markers. A model of this
conception of the relationship between the semantic and
phonological components could be drawn as follows :
SEMANTIC COMPONENT
LEXICON COMPONENT
TRANSFORMATIONAL COMPONENT
PHONOLOGICAL COMPONENT
PHONETIC REPRESENTATION OF S
_ „ . , Fig. 3. -he Relation of rhe Semantic and Phonological Components
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 136
Generative semantics stands as a possible alternative
conception to the semantic theory presented in the writings
of Katz and Fodor. However, Katz in Semantic Theory sees
this approach as leading to cases in which general metho- 19 dological constraints are violated.
We turn now to the implications of the theory of
language set forth within interpretive semantics. If the
objective reality of language is seen in terms of chunks of
physical sound then empirical investigations will be geared
toward the classification and segmentation of language into
grammatical units. This interpretation will determine the
nature of the practical applications of the theory. However,
if the objective reality of language is seen as consisting
of an internalized system of grammatical rules, then the
conception of the nature of relevant investigation as well
as the practical implications which it offers are quite
different. It is Katz' contention that the commitment of
empirical linguists to the external material side of language
had prevented their recognizing the mental reality that deep 20 structure represents. The concern within semantic theory
on the other hand, is with the discovery of hypotheses about
the "mental capacities that underlie the complex chain of
operations by which structures expressing the meaning of a 21 sentence are related to its physical exemplifications."
l^See Katz, Semantic Theory, p. 412. ZOlbid., p. 13. Z^Ibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 137
O O Fodor in The Language of Thought attempts to account for
the nature of the processes which underlie the phenomena of
language, basing his account of how precisely the mind works,
on the information which has emerged from recent empirical
studies of language and cognition. Research in progress in
linguistics and psychology provides in his view the best
answer to the epistemological questions which are of concern
in this study.
Fodor's argument has four main premises.
(1) The only plausible psychological models represent
cognitive processes as computational.
(2) These models presuppose a medium of computation,
a representational system.
(3) This system must share a number of characteristics
of a language.
(4) It is a reasonable goal for research to characterize
this representational system.
The existence of a prelinguistic representational
system is prerequisite to the existence of language. In
Fodor's words,
Either it is false that learning L is learning its truth definitions, or it is false that learning a truth definition for L involves projecting and confirming hypotheses about the truth conditions upon the predicates of L or no one learns L unless he already knows some
22Jerry A. Fodor, The Language of Thought (New York: Thomas Y. Crowell, Co., Inc., 1975),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 138
language different from L but rich enough to express the extensions of the predicates of .^3
Learning a first language is a matter of hypothesis formation
and confirmation. It involves establishing a set of truth
definitions which reflect the semantic properties of the
predicates of the language. One learns the semantic
properties of the predicates of a proposition only if one
learns a generalization regarding the extension of these
predicates. Learning a determination of the extension of a
predicate involves learning that it falls under certain rules
which govern its application. In order to accomplish this
task there must be a system capable of representing the
predicates and their extension pre-existent to the acquisi
tion of language.
Fodor's theory regarding a language of thought is not
a metaphysical hypothesis, but is constrained by empirical
considerations. Having argued that the existence of a
prelinguistic representational system is a necessary pre
requisite to language acquisition, Fodor moves on to show how
research in linguistics and psychology are pertinent to the
confirmation of these theories. Facts about natural language
give us some of the best data for inferences about the
structure of what Fodor labels the "internal code."
23 Fodor, The Language of Thought, p. 82.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139
The very rapid pace in which progress has been made in
linguistics and psycholinguistics over the past few years is
the only factor which Fodor can contribute to the relative
lack of attention which has been paid to the question of how
models of language articulate with theories of cognition. In
fact it is language which gives us our best insights into
the nature of cognitive operations.
Prior to Chomsky, classification of linguistic types
into groups which shared like causes was considered relevant
to an understanding of the semantic content of language.
This approach of taxonomic linguistics was consistent with the
Skinnerian view of language as "verbal behavior."
On the Chomsky/Katz/Fodor interpretation the ultimate
goal of a theory of language is to characterize the corres
pondence which exists between a speaker and hearer and the
computational process which brings it about. On this model
a speaker maps messages onto acoustical wave forms. In turn
the hearer maps wave forms into messages. The exact character
of the mapping depends on the conventions of the language.
Communication occurs because the speaker and hearer both know
the same conventions and how to use them. This theory of
communication finds support in a theory of generative grammar.
The mapping which occurs is indirect and occurs via a number
of intervening representations. Among these intervening
representations are several which correspond to the structural
description provided by generative grammar. Therefore,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 4 0
structural descriptions can be viewed as "psychologically
real" in that they mediate the communication process.
The point is, of course, that we know a good deal about the form of structural descriptions and the information they contain, and we know something -- though not much -- about the kinds of information processing that goes on in encoding and decoding the acoustic objects that structural descriptions apply to. This sort of information bears on the nature of messages since, whatever else messages are, they must exhibit a systematic relation to structural descriptions and that relation must be computable by such information-handling procedures as speaker/hearers have available.
In defining the inter-relationship which should be
developed between philosophy and linguistics Fodor points to
the relevance which an analysis of language in terms of levels
of representation has to a clarification of the concept of
meaning. A theory of meaning serves to pair natural language
sentences with some sort of representation of their truth
conditions. This can be seen in the distinction in philosophy
between surface form and logical form which is based on the
fact that although the surface form of language does not
provide the necessary base for the application of logical
rules some translation of it would. To represent the logical
form of a sentence is, for Fodor, to represent its truth
conditions explicitly in a way a sentence does not.
Linguistics, therefore, provides a means for characterizing
the set of representations computed in the course of encoding
and decoding. As a consequence.
24 Fodor, The Language of Thought, p. Ill
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 141
A theory of the structure of messages is thus constrained by a theory of natural languages in at least the sense that messages must provide appropriate input/outputs for these computational mechanisms.25
If facts about language and language processes constrain
theories about messages they constrain theories about formulae
in the language of thought since messages must be formulated
in the language of thought. If linguistic theory specifies
the form of the language of thought, psycholinguistic theories
can be seen as specifying the order in which representations
are computed and the nature of the information-handling
processes which affect these computations.
Psychological evidence also supports a model of thought
and in turn of language as having a number of levels of
representation. They point to a fundamental feature of higher
cognitive processes, the intelligent management of internal
representation. There are a variety of representations
which any given input might receive. The representation
which is given depends on the demands of the subject's task
and such factors as motivation, need, interest, etc. The
subject's ability to match the use of his representational
capacities to the needs of a situation is a form of intelligent
behavior. This process is computational.
^^Fodor, The Lanaguage of Thought, p. 115.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. .4 2
Rather, the point is that, when things go right, what the subject effects by the management of internal representations is a rational correspondence between his performance and his goals.26
The language of thought provides the medium for representing
the "psychologically salient aspects of the organisms
environment" and is thus essential to the decision-making 27 process.
Over the past ten years a body of psychological
research has accumulated which supports a view of mind as
computational and interprets higher cognitive behavior as
rule governed. However in general such theories of cognition
have not been entirely consistent in their discussions of the
nature of cognitive operations. Although they assume in
their general structure the existence both of an underlying
computational process and a representational system in which
this is carried out, there has been no real attempt to make
this assumption explicit. In order for an organism to
represent their behavior to themselves, to be self-consciously
aware of their own mental processes necessitates in Fodors
view a system in which this representation can occur. What
Fodor proposes is to "resurrect the traditional notion that
there is a 'language of thought' and that characterizing that 28 language is a good part of what a theory of mind needs to do."
2 6 Fodor, The Language of Thought, p. 52. 27 Ibid., p. ix.
ZGlbid., p. 80.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 143
. . . if our psychological theories commit us to a language of thought, we had better take the commitment seriously and find out what the language of thought is like.29
Fodor conceived The Language of Thought as an essay in
speculative psychology, an attempt to see how the mind works
by looking at recent empirical studies of language and
cognition. He grants that, "It may, after all, turn out
that the whole information-processing approach to psychology 30 is somehow a bad idea." But in any event, "It now seems
reasonably clear that the whole learning-theoretics approach
to explanation of behavior was a bad idea and that the theory -31 of mind that it proposed was ludicrous."
Having established the source of support for his
theory of language within linguistics and psychology, Fodor
moves on to characterize the nature of the inner code which
the mind uses in representing thought. What would such a
representational system be like? We can infer from the
function of the pre-linguistic representational system in
language acquisition to a hypothesis regarding its form.
One can learn L only if one already knows some language rich enough to express the extension of any predicate of L
One can learn what the semantic properties of a term are only if one already knows a language which contains a term having the same semantic properties.32
^^Fodor, The Language of Thought, p. 52. 30lbid., p. ix. 31lbid. 32ibid., p. 80.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 4 4
Thus the language of thought must be rich enough to represent
the predicates of any natural language.
There also appears to be reason to believe that there
is a direct correspondence between the semantic categories of
natural language and the language of inner representation.
The primative basis of the vocabulary of a language is equiva
lent to the smallest set of vocabulary items in terms of which
the entire vocabulary of that language can be defined. "It is
thus an open possibility that the vocabulary of the system
used to represent the messages conveyed by the sentences of a
natural language correspond precisely to the primative basis 3 3 of that language." At the semantic level of grammatical
representation we have a dictionary which pairs defined terms
in natural language with defining formula in the representa
tional system. Generative grammar treats definitions as
species of syntactic relations. It derives defined terms from
definite expressions by rules indistinguishable from syntactic
transformations. The constraints which would be necessary for
a determination of what would count as a valid definition would
then be inherited from the constraints on syntactic transfor
mations. If it can be demonstrated that such a parallel
exists between the surface form of language and its under
lying logical form then Fodor concludes there may be some
credence to Wittgenstein's insistence upon the significance
33podor, The Language of Thought, p. 124.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 4 5
of surface forms of natural languages. If the sentences
of natural language are complex this may show only that we
need a correspondingly complex metalanguage in which to
represent their logical form.
The upshot of these remarks is a suggestion that i regard as entirely speculative but very interesting to speculate about: viz., that the language of thought may be very like a natural language. It may be that the resources of the inner code are rather directly represented in the resources of the codes we use for communication.35
Can we then make inferences from the grammatical form
of sentences to the logical form of language? Katz grants
that this is indeed a difficult task.
What makes the problem of logical form in natural language difficult is that similarities in the overt grammatical form of sentences disguise their underlying logical differences.2 °
As Wittgenstein concludes in the Tractatus,
Language disguises thought so much so, that from the outward form of the clothing it is impossible to infer the form of the thought beneath it, because the outward form of the clothing is not designed to reveal the form of the body, but for entirely different purposes.3/
However, the application of the techniques of transformational
grammar to the semantic elements of sentence structure provide
us with a possible means of coping with these difficulties.
^^Fodor, The Language of Thought, p. 156. 35ibid. O fi Katz, Semantic Theory, p. xvi ^^Ludwig Wittgenstein, Tractatus Logico-Fhilosophicus, trans. by D.F. Pears and B.F. McGuinnes (London: Routledge & Kegan Paul, 1921), p. 63.
Reproduced with permission of the copyright owner. Further reproduction prohibitedpermission. without 146
Can we hope to establish a logic of natural language? Katz
sees this goal as conceivable at least in principle.
Logical, on this hypothesis, is an attempt to give a systematic theory of the semantic structures in natural language that determine valid inferences . . . .2°
Therefore, the doctrine this discussion implies is that what is common in all and only cases of necessary inference and what logic in any of its forms studies is the semantic structures defined in such definitions at that of 'analyticity,’ ’contradiction, ' 'entailment,' and 'metalinguistic truth,' together with the definitions expressing the rules of prepositional logic and quantifica tion, insofar as these are broad enough to represent all the logical aspects of the logical particles.39
However, logic has traditionally only been able to express
sentences of the declarative form. The nature of this
limitation can be seen in the case of interrogatives. Because
questions do not admit of truth or falsity it has traditionally
been held that they do not assert anything. In order to
provide an adequate analysis of language Katz must include a
logic of questions as part of the logic of natural language.
"Nothing," he concludes, "precludes the possibility of
questions having genuine logical properties and relations
under some non-truth-functional interpretation of deductive
connections. All that must be demonstrated is that deductive
connections hold between the entities which make up the
question. To do this Katz constructs a. non-truth-functional
O O Katz, Semantic Theory, p. 227.
39lbid., p. 226.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 147
interpretation for validity on the basis of the structure
of semantic theory. In form, "The underlying phrase marker
for an interrogative sentence is like the underlying phrase
marker for a declarative except that its first symbol is Q
and it contains one or more noun phrases . . . . The
shared prepositional structure of a declarative sentence and
its attendant interrogative can be seen in the following
diagrams.
Philosophers love truth.
S
NP VP 1 Philosopher + s V NP
love truth
Do philosophers love truth?
Q NP
Philosopher + s V NP 1 I love truth
We began this study with an examination of Aristotle's
theory of ontological categories. Aristotle conceived of
the categories as the most general classificational divisions
under which all ideas could be subsumed. They were essentially
40Katz, Semantic Theory, p. 205
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 148
incapable of receiving further analysis into more basic units
and thus held the status of natural kinds. However, as we
noted, Aristotle did not provide a rationale for how he
determined the categories and the method he undertakes to
classify entities into one or another category appears to be
in large measure a matter of intuitive judgement. On this
basis Kant among others has criticized Aristotle.
To search in our common knowledge for the concepts which do not rest upon particular experience and yet occur in all knowledge from experience, of which they as it were constitute the mere form of connection, presupposes neither greater reflection nor deeper insight than to detect in a language the rules of the actual use of words generally and thus to collect elements for a grammar.41
Kant agreed with the linguistic approach of Aristotle but saw
no reason why language should have just this constitution.
He concluded.
This rhapsody /the categories of Aristotle/ must be considered Tand commended) as a mere hint for future inquirers, not as an ideal developed according to rules and hence it has, in the present more advanced state of philosophy been rejected as quite u s e l e s s . 42
Katz concurs in this analysis,
Aristotle's system of categories does not itself, and is not embedded within a more general theory that might provide us with an explanation as to why all natural languages utilize just these categories; why, that is, the categories set forth in Aristotle's particular proposal ought to be regarded as linguistically universal in scope.
4lRant, Prolegomena, quoted in Katz, The Philosophy of Language, p. 22F1 42lbid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 149
His account provides no means for deciding empirically what are and what are not , genuine semantic categories of language.
But in contrast with Kant, Katz does not reject Aristotle's
approach as "quite useless."
However, since we are not here to criticize Aristotelian categories but to resurrect them, we shall argue that these difficulties can be removed in principle if the theory of semantic categories is incorporated into the ,, theory of language in the way to be suggested.
On the Fodor/Katz interpretation it is the task of a theory
of language to set forth the universal elements of language
in the form of a linguistic description of the interrelation
ship between syntactic, phonological and semantic components.
The problem before us is the question of how we are to
determine the semantic categories into which language can be
analyzed. If we look to the dictionary definitions of terms
we see that these are redundant. For example female assumes
human being. It is possible to reduce dictionary definitions
to make them more economical, excluding redundancies and
including a set of rules which account for regularities in
meaning. An example Katz provides for such a rule would be
the following:
V (Mg) V ... V (Mj^l7 ------
/_XHuman)v (Animal) v ... v (PlantT----^Physical Object
^^Katz, The Philosophy of Language, pp. 226-7.
44ibid., p. 22.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 0
Everything on the right side which cannot appear on the left
side is a semantic category. The other elements in the
analysis are semantic markers and are subsumed under the
semantic category.
Roughly the semantic categories of language are those concepts represented by the semantic markers belonging to the intersection of the sets of semantic categories for particular languages, as obtained in the manner just described.45
It is these which make up the universal categories of language.
Picking semantic categories for language now becomes an
empirical matter of running through the semantic markers
and choosing those which cannot be subsumed under any other
category. This can be accomplished at the level of universal
grammar as well.
On the above considerations, the theory of categories, in just the sense in which the notion of a category was employed by philosophers from Aristotle through Kant, becomes an integral part of the theory of language. The theory of language provides a definite answer to the question, left open by Aristotle . . . of how to determine in a nonarbitrary way that a proposed set of categories is both correct and exhaustive.46
Does Chomsky's work in generative transformational
grammar provide philosophy with an empirically verifiable means
for determining the basic logical categories which exist at
the level of thought as well as on the level of language?
^^Katz, The Philosophy of Language, p. 235
4Glbid., p. 237.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 1
Through an analysis of language and the universal features
which characterize it can we determine new insights regarding
the nature of our cognitive operations and the manner in
which we organize and structure ideas? It is Katz's con
clusion that these are the implications which research in
linguistics holds for a solution of these traditionally
philosophical problems.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PART III
COGNITIVE APPROACHES TO LEARNING THEORY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER V
JEROME BRUNER
The progress which has been made in the study of
language and linguistic structure over the past one hundred
years has been paralleled by an equally vital and productive
activity within the field of human action and in the study of
the psychological processes which underlie human behavior.
Philosophers were the first to raise questions regarding the
mind and were concerned with developing an understanding of
its operations. The history of work in epistemology is the
history of a concentrated attempt to uncover answers to this
question and to determine how the mind organizes our ex
perience into a coherent and unified whole. As techniques
of scientific discovery were developed, the techniques of
many philosophers, too, became more refined and scientifically
grounded. Where Rene Descartes developed a theory of mind
founded on an introspective examination of his own cognitive
experience, John Locke represents the beginnings an attempt
to provide a more clearly empiricistic approach to the study
of sensation and thought.
By 1850, what is today called psychology was part of
a rapidly expanding branch of philosophy. Cross fertilized
and stimulated by research in the physical sciences, its buds 153
Reproduced with permission of the copyright owner. Further reproduction prohibitedpermission. without 1 5 4
were beginning to swell and in the warmth of the late after
noon of the nineteenth century its branches burst forth in
blossoms developed whereby it could be subjected to quantita
tive measurement. In 1885 Ebbinghaus was also attempting to
apply measurement techniques to learning and retention, and
by 1905 Binet was engaged in the quantification of intelligence
in school children. It is these events that have determined
the shape of psychology in the present day. One hundred
years from its inception, psychology has established itself
as an experimental science whose methodology has been firmly
grounded in the principle of confirmability through
experimentation.
The dominant mood of psychological research in the
twentieth century has been that of scientific behaviorism.
Stimulus-Response (S _ R) theory has laid at the cornerstone
of most psychological experimentation and has served as a
fundamental explanatory principle of human action. Watson's
early research has served as a paradigm in its standards of
observation, testing, measurement and confirmation and
represents a rejection of introspection as a legitimate
source of knowledge about the operation of the human mind.
While not denying the reality of mind and consciousness,
Watson found no evidence that such conceptions had meaning
from a scientific perspective. If the rationale for their
existence was introspection and introspection alone, with no
attempt at controlled self observation and confirmation, the concept of "mind," as Hebb concurred, ". . . had to be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 5
which were the harbinbers of the new era in psychological
research.
Writing in his Principles of Psychology in 1890,
William James foresaw the significance of the new methods
which were being applied in psychological research. Discussing
the changes which were taking place, he noted.
But psychology is passing into a less simple phase .... The simple and open method of attack having done what it can, the method of patience, starving out, and harassing to death is tried; the Mind must submit to a regular siege, in which minute advantage gained night and day by the forces that hem her in must sum themselves up at last into her overthrow. There is little of the grand style about these new prism, pendulum and chronograph -- philosophers. They mean business, not chivalry. What generous divination, and that superiority in virtue which was thought by Cicero to give man the best insight into nature, have failed to do, their spying and scraping, their deadly tenacity and almost diabolic cunning, will doubtless someday bring about.^
By 1875 the revolution was well underway, and with the
development of quantitative experimental techniques the seeds
of experimental psychology already had been firmly sown. In
that year Weber and Fechner, for example, were attempting to
put to measurement the degree of sensitivity which could be
elicited in response to a stimulus. The significant point
of this experiment was that sensation had in the past been
treated as a mental event. But now a technique had been
^William James, Principles of Psychology, Vol. I, (New York: Dover, 1950), pp. 192-3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 6
discarded from scientific consideration until better evidence
could be found.
From the point of view of the behaviorist there was
no longer any need to allude to mental phenomena in order to
explain human action. However, this limiting of the scope
of behaviorism has had what Hebb views an unnecessarily
detrimental effect on S-R theory. By omitting thought,
intelligence, insight and expectance from psychological study
he accuses behaviorism with being seriously inadequate for
dealing with the complexity of human action. The rejection
of introspection, should not necessarily rule out the study
of mind and mental operations. Rather the value of behaviorism
lies in the clearly therapeutic effect it has had on the
psychology of human action through the laying out of strict
standards for its investigation. As Hebb points out.
Paradoxically, it was the denial of mental processes that put our knowledge of them on a firm foundation, and from this approach we have learned much more about the mind than was known when it was taken for granted more or less uncritically.4
The standards of observation and confirmability have served
to clarify the problems which had resulted from introspection
and to place the study of mind within a more productive context.
Rather than seeing the approach of experimental psychology as
^Donald Hebb, A Textbook of Psychology (Philadelphia.- W.B. Saunders Co., 1966), p. 5. 3lbid., p. 314.
4lbid., p. 6.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 7
necessitating a denial of mental processes, it established a
requirement that any discussion of mental processes had to be
couched within an acceptable scientific framework emphasizing
testability and confirm.ability. For Hebb this is accomplished
through laboratory study of neurological functions. This
perspective provides an alternative both to introspection
and the psychological speculation which it generated and to
be narrowly defined concerns of early behaviorism. Hebb finds
evidence for the conclusion that carefully controlled research
supports the contention that mind and consciousness ^ have
scientific meaning and can be investigated empirically. If
they are viewed from a physiological perspective as identi
fiable, measurable physical phenomena we no longer have to
violate standards of scientific accuracy in considering them.
Such an interpretation as this allows us to treat mind and
mental concepts as processes which occur inside the brain
and that determine complex human behaviors.
Such psychologists as Tolman^ and Lashley^ have devoted
their research to a demonstration of the inadequacies of S-R 7 8 theory. However behaviorists as C.L. Hull and K.W. Spence
5s. Tolman, Purposive Behavior in Animals and Man (New York: Irvington Press, 1967). ^K. Lashley, The Neuropsychology of Lashley: Selected Papers, ed. F.A. Beach et al (New York: McGraw Hill, i960). ^C.L. Hull, A Behavior System (Westport, Ct.: Greenwood Press, 1974). O K.W. Spence, Behavior Theory and Conditioning (Westport, Ct.: Greenwood Press, 1978).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 158
have attempted to develop more adequately the explanatory
powers of S-R theory.
It could be argued that behavior1st theory does not
deny such unobservable processes occur and that in fact S-R
theory is grounded in a physiological conception. On such an
interpretation neural processes would be traced from their
inception (within for example a sense organ) along neural
pathways to the muscle where the overt response takes place.
This possibility for a physiological explanation of what occurs
between the stimulus and the response has often been forgotten
by behaviorists themselves. What Hebb, as a neurophysiologist
suggested was one alternative to a rigidly structured
behaviorist account of human behavior. Such an approach
would allow for discussion and investigation of thought and
cognitive operations whereas early varieties of behaviorism
did not, and at the same time it would account for the
complexity of cognitive operations which the mind is capable
of performing. In placing emphasis on this aspect of human
behavior, studies in brain physiology mesh with S-R theory
rather than opposing it. They provide one possibility for
developing a comprehensive and satisfactory account of human
action within the context of scientific behaviorism. Such a
concern with the study of the cognitive processes involved
in human thought and action is referred to in general terms
as cognitivism and can be seen as a major alternative to the
approach of strict behaviorism. Where the emphasis of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 159
behaviorist is on the overt behavior of organisms, the
cognitivist examines how the mind organizes and structures
ideas. The emphasis is placed upon what goes on between the
point of input into the system, and that at which a response
is made. There is concern with explaining the process of
mental operations, as well as their form of logical structure.
Cognitivism's concern in the development of a theory
of learning has been toward providing an explanation of how
the learner assimilates information and inteprets it into
a unity which we call knowledge. Until very recently all
scientifically respectable theories of learning depended on
the concepts of stimulus, response, reinforcement and
association for their explanations and insisted that within
this framework every form of learning could be accounted for.
The cognitivist claims that such a model is only adequate for
explaining a very limited range of learning situations. Where
learning involves simple association the model of stimulus
and reinforcement is adequate. But what the cognitivist finds
of greatest theoretical interest is what is referred to as
"meaningful learning." Meaningful learning occurs in a
situation where the learner is involved in incorporating new
knowledge within the structure of the knowledge which already
exists. The new information is related in a "non-arbitrary,
substantive fashion" to what the learner already knows.^ It
^David P. Ausubel, Educational Psychology: A Cognitive View (New York: Holt, Rinehart, Winston, 1968;, p. 24.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 6 0
can be representational, the learning of labels or names;
conceptual, the learning of single words which stand for
concepts; or propositional, the learning of the meaning of
ideas. By incorporating potentially meaningful verbal
materials within a cognitive structure, a new, differentiated
cognitive content is generated. The new knowledge is
assimilated and stored in linked fashion with the old.
One of the earliest of the cognitive schools was
Gestalt psychology. In its emphasis on the importance of
viewing experience as a unified whole. Gestalt psychology is
representative of the cognitivist attempt to examine not just
overt behaviors, but to develop a scientifically productive
theory of the nature of cognitive processing. The theories
developed by Max Wertheimer, founder of the Gestalt school,
viewed the human mind as actively engaged in the process of
construction. From the raw materials of experience one
constantly absorbs, integrates and constructs a body of
knowledge which is uniquely one's own. Concurrent with new
experiences this knowledge is continuously growing and
becoming more complex. What develops is, in a sense, a
schematic diagram or. cognitive map around which experience
is organized. This diagram represents a very distinctive
patterning of logical interrelationships which are established
between the elements of experience and which differs from
person to person. lÆien new facts are encountered they are
not just added to the existing body of knowledge. Rather
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 161
they are "tied on" to some related piece of knowledge and
stored in an appropriate location within the blueprint of
the mind.
Because this operation i^ not merely one of "dumping"
new knowledge in with the old, retrieval of information is
greatly facilitated. Material which is to be stored in
memory is properly coded and stored. Because this theoretical
filing cabinet is logically structured, it would allow for
easy recall. The task of recall or memory operates something
like a data search in which the desired material has first
been located within the system is then easily retrieved. As
new material is added on to old, the older memory traces are
replaced by new and the structure of the cognitive content
is thereby altered.
A motivating interest of the cognitive theories is to
develop an understanding of concept learning. This would
include not only learning of perceptual concepts like red or
square, which have been the primary subject of laboratory
observation, but would include more importantly the learning
of conceptual or abstract concepts which have less identifiable
criteria of application. The results of studies of concept
learning are relevant both to how we perceive the world and
how we think about (conceive of) the world. Where the
principles of conditioning and simple associative learning
explain how people identify concepts by association, they do
not explain the more interesting and complex forms of concept
formation. "Therefore, contemporary theorists of concept
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 162
learning have abandoned simple learning theory as the
fundamental basis for concept learning.Rather they have
turned their attention to an analysis of the logical structure
of concepts and of concept-learning tasks.
Crucial to this analysis is a clear account of the
idea of 'concept' itself. According to Hulse, a concept is
a set of features or. identifying characteristics which are
abstracted from any specific instance, are interdependent on
one another and are linked together by a rule. The rules
define the concept by specifying the relation between the
defining characteristics of the concept and this relation
depends in turn on the semantic structure of language. In
learning a concept, then, one learns not only a set of
characteristics which by convention are associated with the
concept, but how these are logically interrelated with one
ano ther.
Concept attainment is viewed within the cognitive
tradition as essentially a process of determining those
criteria which are relevant for placing something within a
specific category. The features or attributes which serve
as the defining characteristics of the concept are the general
descriptive properties characteristic of the thing being named.
They include public criteria (e.g. that being a female is an
attribute or defining characteristic of all mothers) but may
^Ogteward H. Hulse; James Deese; and Howard Egeth, The Psychology of Learning (New York; McGraw Hill, 1975), p. 259.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 163
also include personal criteria as well (e.g. that mothers are
patient/mothers are impatient). Given the great diversity of
experience concept formation seems essential to give order to
the world as we perceive it. Rather than responding in terms
of uniqueness we organize in terms of class membership, we
engage in the process of categorizing or rendering discrimin-
ately different things equivalent on the basis of some
feature(s) which they share. By organizing in terms of
categories we are able to group together mentally, objects
with differences on the basis of what they have in common.
In doing so our ability to deal with diversity is facilitated,
while at the same time we are enabled to move beyond thought
of particulars to thinking on the level of abstract concepts.
Where there is a network of logical interrelationships
which hold between the various attributes contained within
a concept, there is also a logical interrelationship which
becomes established between the concepts which we learn. This
the cognitivist refers to as the cognitive structure. The
form which each individual's cognitive structure assumes is
a function of what Hulse describes as certain substantive,
organizing properties of concepts and propositions, the
cognitive structure variables. These anchoring ideas function
in the transfer of learning both in lateral transfer, the
Bruce Joyce and Marsha Weil, "Concept Attainment; A Model Developed from a Study of Thinking," Models of Teaching (Englewood Cliffs, N.J.; Prentice Hall, IncTl 1972).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 164
use of what has been learned in a more elementary set of
subordinate concepts serving as a prerequisite to the
understanding of more complex higher order concepts.
Closely related to the cognitivist's study of concept
learning is an interest in the process whereby we develop
the ability to use and understand language. In the past
decade the field of cognitive studies has been focused on
examining the structure of language and the process of language
learning, seeing this as a clue to how we organize and
structure our conceptions about the world. In a study
completed in 1970, McNeill concluded that none of the basic
principles of elementary S-R learning theory were applicable 12 to the learning of language. In the same vein Fodor, Bever
and Barrett found evidence in support of the contention that
the abstract notions of subject and predicate, noun and verb,
which are necessary to the ability to use language are not
responses in the S-R sense and cannot be fitted into a 13 conditioning model.
Where animals may be capable of elementary modes of
communication involving labeling and simple association,
human language on this view provides us with the capacity for
constructing an infinite number of propositions from a finite
number of elements. In labeling one is able to generate simple
12d . McNeill, "The Development of Language," in P.H. Mussen, ed. Carmichael's Manual of Child Psychology, 3rd ed.. Vol. 1 (New York: John Wiley, 1970). 13Jerry Fodor et al. The Psychology of Language (New York: McGraw Hill, 1974).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 6 5
although meaningful phrases representing a set of facts about
the world. In contrast, however, human language enables us to
deal in terms of logical relations, implications and judgements
which reflect logical interrelations in the world. Such
communication requires an ability to deal with abstract
concepts and to utilize an abstract system involving rules or
principles rather than one made up of signs or tokens which
represent facts in one to one correspondence. The fundamental
characteristic of human language is not that it is composed
of speech sounds that represent elements within the world, but
that it is a system of communication which expresses logical
interrelations and is regulated by rules. In learning
language a child learns not only a vocabulary, but a set of
grammatical principles or rules which govern how this
vocabulary is combined to form the complexity of thoughts
which we call language.
The rules which the child utilizes are the grammar of
the language. How they are learned is a topic which the
cognitive psychologist is concerned to investigate. Slobin
in 1968 reported that his studies showed this ability is
not a consequence of imitation.The research of Piaget
supports the view that a child's conceptual language abilities
emerge in an invariant sequence regardless of such factors as
Slobin, "Imitation and Grammatical Development In Children," in Endler, Bochter and Osser, Contemporary Issues in Developmental Psychology (New York; Holt, Rinehart, Winston, 1968).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 6 6
social class and nationality, although environment does
significantly determine how these abilities develop once they
reveal themselves. Learning language is essentially a
cognitive learning process. In what Ausubel calls State I
we perceive new information through our senses and, in the
case of language determine its meaning. Then, in State II,
we relate new meanings to the relevant and related existing
structure of ideas which is uniquely our own, reconciling
this new material with our existing knowledge.What we
learn is influenced not only by the cognitive structure
variables, the characteristics of previously learned
materials, but by such personal factors as developmental
readiness, intellectual ability, motivation, attitude and
personality and by such situational factors as the arrange
ment of materials, extent of practice and, in a classroom,
the class climate and characteristics of the teacher.
Having now established the general outline of cogni
tivism and the contribution this strain of psychological
research has made to our understanding of the nature of
human cognitive operations, we will look in detail at the
work of Jerome S. Bruner as representative of a cognitively
based theory of learning and then to Jean Piaget for his
more detailed treatment of how the child constructs a logical
model of reality.
^^Ausubel, Educational Psychology: A Cognitive View, p. 24.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 167
Bruner and Piaget share the view that in cognitive
development and specifically in learning what occurs is a
substantive change in the manner in which our thoughts are
organized. This change is qualitative. It involves the
restructuring of ideas in new patterns involving unique sets
of logical relationships. Such a process stands in contrast
to quantitative changes in which new materials are passively
absorbed and linked onto old in sequential fashion. The
concern of Bruner and Piaget is with the role which the mind
(or what we refer to as the mind) plays in the development of
thought. The question which they wish to raise is what is
going on in learning -- how do we go about constructing our
understanding of the world. Part III will examine the
contributions which the writings of Bruner and Piaget have
made on an understanding of this process.
Bruner and Piaget are interesting also for their
differences. Where Bruner's concerns are psychological in
nature, Piaget uses the data he has collected from extensive
observations of children in various stages of intellectual
development to speak to more broadly philosophical as well
as psychological questions. It is his belief that hard core
empirically based research is the only meaningful approach
to the solution of the philosophical problems of knowledge
and mind. Thus we will violate chronology beginning first
with Bruner then moving on to Piaget's research into the
formation of logical structures for the insights which it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 168
might provide to the questions which are of interest to us
here. We will then move on to examine the implications of
these approaches to our overall goal, the development of an
integrated theory of cognitive structure.
The approach of Jerome Bruner and others within the
cognitivist tradition to the analysis of human behavior
stands in stark contrast with that of other movements in
psychology over the past hundred years. Where behaviorism
had emphasised the primacy of the environment and its action
on the individual and had seen learning as a passive activity
controlled by laws of stimulus and response, Bruner struck a
new note turning to the subject as the locus of interest and
investigation. As Jeremy Anglin in his introduction to
Beyond the Information Given points out.
In contrast Bruner's view of man as an information processor, thinker and creator emphasizes both the rationality and the dignity of which human beings are capable.16
Bruner's theory is based on the view that persons are
creative, thinking beings dignified by their ability to use
reason in the solution of complex problems. In this he
echos the views held two thousand years ago by Aristotle
who viewed person as a rational animal distinguished by the
ability to think, plan for the future and to philosophize.
^^Jerome Bruner, Beyond the Information Given, edited by Jeremy M. Anglin (New York: W.W. Norton, 1973) , p . xxiii.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 169
Bruner labeled his approach instrumental conceptualism.
It was his aim to build a working theory of cognitive growth
in which knowledge could be seen as developing through a
continuous recursive process of hypothesis, data check and
confirmation.17 At the basis of this theory is the idea of an
internal model or system of representation. From birth
throughout life a person is engaged in the construction of
models of reality. These models are a function of both the
genetic makeup of the individual and the environment. In the
initial stage incoming sense data is related to this internally
stored model of reality. An inferential leap is made from the
newly acquired information to a hypothesis. A confirmation
check then occurs in which the hypothesis is tested against
further sense data and finally the cycle is completed with
the hypothesis being altered or retained. The sense data
functions as the clues or cues which the subject utilizes in
the evaluation process, the hypothesis as a rule or theory to
be confirmed or denied and the internal model as a generic
coding system which establishes the standards upon which an
evaluation is made. A consequence of the recursive character
of Bruner's model is that it ". . . can only partially and 18 intermittently be tested against input." It is, therefore,
always tentative and changing, always undergoing a process
of revision.
17Jerome Bruner, Studies in Cognitive Growth (New York: John Wiley and Sons, IncTj 1966) , p . 3l9. ISibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 7 0
The process of developing a cognitive structure or
model of reality is the process of learning to organize diverse
information within a formal schemata. It is referred to as
coding.
For anybody to understand action whether he be a child or adult, requires the ability to categorieze a flow of events in a complex, possibly natural way.19
Experience presents us with a wide range of elements which
must be so categorized. At the same time the human mind is
capable of drawing fine discriminations of similarities and
differences within the range of elements. To cope with this
we engage in the process of categorizing, we group things
together on the basis of their features. The categories
which we perceive to exist then become reflected in our
language. Language is made up not only of concrete referents,
but of terms which reflect abstract concepts, relations,
feelings, evaluations as well.
A coding system is ". . . a set of contingently related,
nonspecific categories; it is the person's manner of grouping
and relating information about this world, and is constantly 90 subject to change and reorganization." As a hypothetical
construct inferred from the nature of antecedent and
consequent events, the coding system, those schematic, formal
^^Jerome Bruner, Human Growth and Development (Oxford: Clarendon Press, 1978),p. 67.
^^Bruner, Beyond the Information Given, p. 222.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 7 1
features of our model of reality, are what allows us to make
use of the information given and to go beyond it. How we do
this, how we process information and the features of our
coding system are affected by such factors as learning set,
what defining attributes we are primed to look for; by our
need state or level of motivation; or our degree of mastery,
whether we have already developed an understanding of lower
level regulations; and by the diversity of our training.
Studies by Beach and Jaynes how that early and diverse train
ing of lower organisms seems to be one of the conditions for 21 producing intelligent behavior in the more mature organism.
This blueprint on which reality is constructed which
is continuously being edited and revised rests, ". . . o n
what might be called an axiomatic base -- our ideas of cause
and effect, of the continuity of space and time, of invariances 9 9 in experiences, and so on.' It is these conceptions which
together form the skeletal foundation around which we organize
our knowledge of reality and draw an interesting similarity
between Bruner's theory and that of Immanuel Kant. But where
Kant's forms of the intuition and categories of the under
standing defined how the human mind by natural necessity
organized experience, it is not clear whether Bruner's
axiomatic base is also innate. In Studies in Cognitive Growth
he raised this question.
^^Bruner, Beyond the Information Given, p. 233. ^^Bruner, Studies in Cognitive Growth, p. 319.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 172
It seems not unlikely (no stronger phrase than that is justified) that some of this axiomatic structure informing our models of reality is already given in the innate nature of our three techniques for 'modeling' reality . . . .23
However, within Bruner's theory it would be just as possible
to present the base as acquired gradually through our
interactions with the physical world. Such a position would
correspond more closely with that of Piaget who sees the 9 A development of the concepts of cause and effect and of the
continuity of space^^ and time^^ to be a consequence of our
manipulation of objects during the sensory-motor period of
development.
Bruner fits his developmental theory within this
recursive mould. How the mind represents the world and
organizes its model of reality can be seen to operate on
three distinct levels depending on the developmental level
of the subject. At the earliest stages of development the
world is known primarily within a context of the habitual
actions which we develop to cope with it. This mode of
viewing the world Bruner terms enactive representation.
Representation begins as a trace of a prior response. Where
2 3 Bruner, Studies in Cognitive Growth, p. 319.
^^Jean Piaget, The Child's Conception of Physical Causality (New York: Harcourt Brace, 1930),
Jean Piaget, The Child's Conception of Space (London: Routledge and Kegan Paul, 1956),
26jean Piaget, The Child's Conception of Time (New York: Basic Books, 1969),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 173
at the earliest stages all action is basically a matter of
immediate response to an initiating stimulus with no linking
between sets of stimulus-response situations, this most primi
tive representation forms a trace which guides a new response
and makes anticipation possible. It begins a linking or
anticipation which forms a patterning of expectations of
what will occur as a consequence of any previously occurring
event and in this sense it provides one form of representa
tion of reality. The significance of the elements which make
up the trace is determined by their relation to the larger
context of actions and the purposes which we have in perform
ing them. Tied as it is the sphere of action, how we
conceive the world in enactive representation is a function
primarily of the nature of our neuromuscular system.
Bruner referred to the mediate stage of modeling
reality as ikonic representation. At this stage the child will
represent the world to himself through an image or spatial
schema which is relatively independent of action. However,
this image is determined primarily by surface cues which tend
to control attention. There has been no generalizing beyond
the surface features of objects, nor movement into deeper
structures like heirarchy and relations. Thus the features
of visual perception, that it is organized around a minimal
number of cues, that it is egocentric, concrete, non-transfer-
able, closely related to action and subject to affect, tend
to determine the nature of the child's imagery and are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 7 4
exaggerated in this form of representation.
The final stage of representation is referred to as
symbolic. Symbolic representation stems from a form of
primitive and innate symbolic activity that through
acculturation gradually becomes specialized. In this stage
action (enactive representation) and image (ikonic representa
tion) are translated into language. The learning of the use
of language for representation is a gradual process of
cumulative development of ability which is determined not
only by our cultural environment, but by our native endowment 27 for mastering particular symbolic systems.
We have, then, in Bruner's theory a developmentally
based explanation of knowledge attainment which is formulated
around the hypothesis that the knowing subject actively
constructs models of reality, subjects them to the testing
of experience and is continuously revising them in accord
with the evidence which experience provides. These models
appear to be grounded in a set of organizing principles which
may be innate, but which might also develop gradually through
interaction with the environment. Having presented this as
the framework within which knowledge evolves, let us now look
more closely at what occurs in more specific instances of
cognitive activity, in perception, conception and language
acquisition.
Jerry A. Fodor and Jerrold J. Katz, "The Structure of a Semantic Theory of Language," Language 39 (1963): 170-210
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 175
A significant feature of Bruner's approach is his
emphasis on the continuity between perceptual and conceptual
activity. The recursive process which we go through in
acquiring knowledge operates on all levels, in the perception
of an event, the attainment of a concept and in the solution
of a problem. Bruner viewed perception as ". . . a n
essentially inferential process in which an individual con
structs his perceptual world on the basis of the information
provided by his senses.In this sense his treatment
diverges from traditional approaches and has been referred to
as the "New Look in Perception. Perception is seen not as
an isolated, independent system, but as integrated with other
psychological factors such as experience, motivation, person
ality and social environment. Bruner characterizes this
approach as representing a movement of confluence in psycholo
gical theory. Growing out of an interest in blending percep
tual theory with what has been learned about the function of
need, interest, past experience and organization of perceptual
field, Bruner was of the opinion that whereas needs, interests
and past experience don't influence perception directly, they
create structures or rules of operation that mediate in a 30 subtle, indirect fashion to regulate cognitive activity.
In perception the perceiver is actively engaged in
28Bruner, Beyond the Information Given, p. xiv.
29lbid. 30lbid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 176
selecting information, forming hypotheses and in a sense
creating reality. It would not be misleading to claim that
on such a view as that presented by Bruner, objects as such
are constructed by the observer. What we perceive is a
consequence of the cognitive structure variables which make
up our representational model of reality. A child's prefer
ence for certain featrues or experiences results in a biased
exposure to them. A correlation is built up among these
features in the form of cell assemblies. This correlation
is a consequence of the correlation of the input. However,
it is reinforced when these features are correlated in
reaching and manipulative experience and it inevitably
determines the manner in which the child organizes and
segments the world.
The "New Look" stresses that perception is both
veridical and categorical. The veridical nature of percep
tion relates to its representational function. What is per
ceived is somehow a model which represents the physical world.
Its validity is determined by the extent to which the abstract
model appears to adequately portray reality. Perception is
an integrated system in which experience in one sensory area
serves to support the evidence derived from another and thus
there is a confirmability from one level of sensation, for
example the kinesthetic, to another, possibly the visual.
What looks like an apple also proves to feel and smell like
one. Thus the meaning of a thing is made up of the placement
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 7 7
of that object within a network of hypothetical inference
concerning its other observable properties. The form which
this network assumes is a consequence of the second feature
of perception, that it is categorical.
The rather bold assumption that we shall make at the outset is that all perceptual experience is necessarily the, end product of a categorization p r o c e s s . 5 1
In perception as in set theory we are involved in the place
ment of an element in the universe within a subset of that
universe. Motion, causation, intention, identity, true
equivalence and space are for Bruner categories that have
primitive counterparts in the neonate and function to
determine our perception of the world.
Studies in neurophysiology have served to support this
belief in the existence of categories of thought which are
innate. Hebb, like Kant, argued for the existence of certain
primitive, innate unities or identities within perception
which are referred to in Hebb as the mechanism of perceptual
readiness.32 However, Hebb conceives these innate features
of perception as neurophysiological processes, such as that of
grouping and integration in which established neural associa
tions serve to facilitate perception of events experienced
together in the past through the establishment of expectancies
31 Bruner, Beyond the Information Given, p. 8 .
32gee Donald Hebb, The Organization of Behavior (New York: John Wiley, 1949).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 178
described as learned frequency integrators which may be viewed
from a neuro-anatomical perspective as synaptic knobs. Also
falling within this category of innate perceptual mechanisms
would be the tendency toward access ordering in which the
, degree of accessibility of coding categories to
stimulus inputs is related to the regulation of the number
of preactivated cell assemblies that are operative at time
of input."33
Perceptual learning, then, consists ". . . i n the
learning of appropriate modes of coding an environment in
terms of its objective character, connectiveness, or
redundancy, and then in allocating stimulus inputs to
appropriate categorial coding systems.Representation
depends on the creation of a system of categories of relation
ships that fit the world in which the person lives. How we
process our perceptions of the world is a function of both
personal and social factors. Personality processes are
"indispensable intervening variables" for perceptual theory.
A theory of perception should account for the differences in
perceiving which characterize different personality
constellations. We develop highly generalized expectancies
to respond to events in the environment. These expectancies
or hypotheses are dependent on our model of reality or what
33Bruner, Beyond the Information Given, pp. 22-3,
3^Ibid., p. 1 2 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 179
might also be referred to as an established habit family
heirarchy. Social or environmental factors, too, have a
significant effect on the process of perception. It is
Bruner's contention that there are important institutional
pressures which develop within technological societies and
which lead to a demand for confirmation between the three
modes of knowing, for a correspondence between what we do
and what we say.
The hypothesis I would like to set forth is that there is a greater push toward hierarchical connections in technical cultures than in those that are less technical.
It is not that one sees "better" or represents what one has learned in habit patterns 'better,' or even talks or thinks in language 'better.' Rather what seems to be the case is that there is an insistance on mapping each of these systems into another, with a resulting increase of the translatability between each of t h e m . 55
In his theory of concept attainment as in that of
perception, Bruner is reacting against the techniques of o r behaviorism and also of psychoanalysis. He feels that
neither of these approaches can adequately account for the
exquisitly creative forms of problem solving which we see
demonstrated in human intellectual achievement. In contrast,
Bruner stresses that all thinking processes are essentially
rational. As in perception the knowledge gained in concept
formation is generative. Conceptual development is an active
O C •^•^Bruner, Studies in Cognitive Growth, pp. 324-5. 3 6 Bruner, Beyond the Information Given, p. 122.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 180
process of data collection, hypothesis formation and
confirmation.
After working on concept attainment -- the strategies by which people discover equivalence in the things around them -- I was enormously impressed at the logic like or 'rational' quality of adult human conceptualizing. While the conceptualizing efficiency of our subjects was not notably high -- they wasted information in most unpuritan ways -- nonetheless, they seemed to go about the task of searching for information in a manner that reflected recognition of complex environmental regularities, of their own limited capacities for processing information, of the risks involved in making certain kinds of guesses and choices of course. One could discern systematic strategies in behavior that had the quality and creases of well practiced rule-governed r o u t i n e s . 37
How we study concept attainment has been made
problematic by the inaccessibility of the subject to
controlled observation and testing.
It is perhaps because of the inaccessibility of reportable experience that psychologists have produced such a relatively sparce yield of knowledge when they have sought to investigate concept attainment and the thought processes by techniques of phenomenological analysis.38
It is exceedingly difficult for one concerned with more than
behavioral criteria to state just what it is that leads a
person to claim that a concept has been successfully mastered.
One possibility, suggested by Bruner is to determine if the
attributes that are criterial for the subject in the
37Jerome Bruner, Toward a Theory of Instruction Cambridge: Belknap Press! 1967), p! 27 38 Bruner, Beyond the Information Given, p. 132.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 181
categorizing judgement are also the attributes that define
the concept. If this can be ascertained then a better
determination of the extent of concept attainment might
result.
The process of concept attainment involves a series
of complex decisions. These necessitate a determination of
what precisely is the nature of the task being undertaken,
of which attributes and of how many should be attended to,
and of how to change or alter any voided hypotheses.
For any given concept-attainment task, for example, there is an ideal strategy that can be constructed having the property that by following it one can attain a concept wûth a minmimum number of encounters. . . .39
But also involved is the question of the cognitive strain
involved. Where there are ideal strategies for a minimum
number of encounters, there are also ideal strategies for a
minimum of strain. We learn to compromise our strategies
and to alter them in accordance with the nature of the concept
sought. Bruner finds evidence to support the conclusion that
western society seems to prefer conjunctive classifications,
classifications in which there is a joint presence of several
attributes (e.g. long haired cats). This, he conjectures, may
stem from the tradition of Aristotelian logic. He raises
the question, "Does the difficulty of dealing with disjunctive,
relational, and probablistic concepts reflect the difficulty
^^Bruner, Beyond the Information Given, p. 136.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 182
of such concepts or does the difficulty perhaps reflect
certain cultural biases in problem solving.
Creativity is an important feature of concept attain
ment. It grows out of a combinational activity in which ideas
are placed in a new perspective and in turn serve as the basis
for the generation of new insights. This capacity for
creativity is the subject of Bruner's attention in Beyond the
Information Given.
Perception and conception are the first two levels of
cognitive activity. The third level involves the development
of the ability to use language. Having argued for the position
that in perception we are actively engaged in building a
model of reality and that in conception we construct a net
work of concepts about the world, Bruner also fits language
acquisition and symbolic representation within his recursive
mold. In the acquisition of language we see the development
of a symbolic system of representation which is related
isomorphically to a reality comprised of objects and ideas.
The objects of perception and of thought are translated into
the domain of language and consequently become subjected to
manipulation and higher level abstraction. In language we
encode experience. What we learn is not only the rules of
grammar, but how to organize our thoughts categorically and
heirarehically. The syntactic system of language makes
possible the use of words as category place names and
^^Bruner, Beyond the Information Given, p. 138.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 183
facilitates the process of abstraction. In this sense
language is necessary to thought. However, Bruner rejects
the notion suggested by Vygotsky that language is an in
separable part of cur apparatus of thinking and prerequisite
to it.^^ Where there is an interaction between thought and
language it is reciprocally supportive. While "... language
provides a kind of temptation to form concepts of objects and
events that have a structure comparable to those contained in
words . . . " language at the same time presupposes certain
underlying cognitive processes required for its use. Experi
ence and mental operations must be present before language
can develop. Then, once language is applied we can in turn
use it as an instrument to scale to higher more general
levels of thought.
All languages have a base grammar which is character
ized by subject-predicate relations (x is a function of y) ,
verb-object relations (cause and effect), by modification and
by rules for transformation. In language learning,
. . .the child is learning to use his words and their semantic markers for the picturable or ikonic aspects of his world, but the words themselves are immeshed in a highly abstract and heirarchical system of categories used formally to signal causation, predication and modification through sentences that the child can 'rewrite' according to transformational rules that he early m a s t e r s . ^3
^^Bruner, Studies in Cognitive Growth, p. 39.
^2 Ibid. 4^Ibid. , p . 45 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 184
This process reveals itself in language before it appears in
thought. Given cultural support, by the age of five to seven
these fundamental rules of categorization, hierarchy and func
tion begin to be applied to the child’s conception of the
world as well as to the elements which make up the language.
It is the rule-boundedness of language which assures
that the child will be able to produce an endless number of
syntactically legitimate utterances. Chomsky's LAD (language
acquisition device) in Bruner's view, ". . . helped produce
a new way of seeing what is involved in acquiring language
. . . . It portrayed language acquisition as a discovery
procedure whereby the rules of acceptable grammar were
discovered through practice. Chomsky presents this procedure
as based in the language learner's innate grasp of the
universels of language.However, Bruner concludes that
Chomsky's theory contains within itself its own downfall.
Chomsky conceived of this innate ability to grasp the
universels of language as independent of any knowledge of the
non-linguistic world. He isolated language learning from
thought and then was able to support a theory of innateness
which was necessitated by the independent character of
language. In this, Bruner claimed, Chomsky was clearly wrong.
The child's mastery of concepts about the world is deeply
44 Bruner, Human Growth and Development, p. 53. ^^Noam Chomsky, Reflections on Language (New York: Pantheon Books, 1975).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 185
dependent upon a prior mastery of concepts about the world
to which language will refer.A theory of tne innateness
of the patterns of language was also rejected by Bruner.
We move, perceive and think in a fashion that depends on techniques (ways of acting or utilizing implements of a technological society) rather than on wired-in arrangements in our nervous system.47
There is no doubt "... that a child is equipped with some
means for generating hypotheses about language that could not
simply be the result of learning by association and reinforce
ment . . . There is indeed something pre-programmed about our
language-acquiring capacity."^8 But it is not innate. It
takes the form of a readiness to grasp the rules for forming
sentences and may be based in patterns of pre-linguistic
learning as well as in part determined by neurophysiological
factors.
Having laid out the psychological principles involved
in perception, conception and language Bruner moves on to
determine their practical implications for a theory of
instruction. When we learn, what we are essentially doing is
developing a network of representations which corresponds to
the world of our experience. This has important implications
for education. The adequacy of any set of instructional
techniques can be measured by the extent to which they permit
^^Bruner, Human Growth and Development, p. 64,
^^Bruner, Studies in Cognitive Growth, p. 56.
^^Bruner, Human Growth and Development, p. 64.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 186
the maximum reconstruction of reality through abstract
cognitive structures and transferability of learning. The
heart of the educational process "... consists of providing
aids and dialogues for translating experience into more
powerful systems of notation and ordering.Since meaning
ful learning is an active process in which the individual
selects information, forms hypotheses and alters them in the
face of evidence, Bruner argues for the incorporation of
"Discovery Learning" into the educational curriculum.
Instruction should take place in a hypothetical mode in which
the activity of the learner is encouraged. The student
should be involved in a continual process of organizing
evidence, and participating in active dialogue with the
instructor and with other learners. It is crucial that
solutions to problems are thoroughly worked through rather
than being provided in their final form.
In the hypothetical mode, the teacher and the student are in a more co-operative position with respect to what in linguistics would be called speaker's d e c i s i o n s . 50
The child is encouraged to participate in the speaker's
decisions and, as a consequence, the child is in a vastly
improved position to be able to generate new, insightful
ideas. In active learning a person relates incoming informa
tion to a previously acquired psychological frame of reference
^^Bruner, Toward a Theory of Instruction, p. 21.
SOsruner, Beyond the Information Given, p. 403.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 187
which involves a cognitive structure, a generic coding system
and an internal model of representation. This model evolves
qualitatively with growth. Thus the instructor should tailor
the information presented to the developmental level of the
learner in order chat it can be best assimilated and used.
Our aim as teachers is to give our student as firm a grasp of a subject as we can, and to make him as autonomous and self-propelled a thinker as we can -- one who will go along on his own after formal schooling has e n d e d . 5 1
Bruner's proposals for improving educational metho
dology culminate in what he referred to as the "Spiral
Curriculum." Courses of study should emphasize a continuity
and development which is continuous throughout, from the
earliest stages of learning to the more advanced stages of
intellectual development. A child should be introduced at an
early age to the ideas and styles that in later life
characterize an educated person and emphasis should be placed
on the development of a pattern of critical thinking. It is
Bruner's belief that "... any subject can be taught
effectively in some intellectually honest form to any child 52 at any stage of development."
Does Bruner present a viable theory of human learning?
Do his proposals for maximizing learning provide us with a
workable framework within which we can develop a theory of
instruction which takes into account the conclusions of
3^Bruner, Beyond the Information Given, p. 403 32ibid., p. 398.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 188
studies in cognitive development? The work of Piaget in his
studies of human growth and development may enable us to
determine how we wish to respond to these questions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER VI
PIAGET
The difficulty of delineating precisely the boundaries
that lie between philosophy, linguistics and psychology, and
looking at each discipline individually for the insights it
contributes towards a common goal, as a task becomes even
more complex in the work of Piaget. Aristotle as a scientist
chose a philosophical mode in which to answer questions which
derived from a scientific curiosity. In contrast we might
portray Piaget as a philosopher, the intellectual heir of
Aristotle and Kant who, writing in the twentieth century,
has chosen science and the methods of experimental research
as the proper mode for answering questions of an epistemolog-
ical nature.
In his introduction to Insights and Illusions of
Philosophy, Jean Piaget describes for the reader how the book
came about and what his purposes were in its writing.
It does not claim to be more than the testimony of a man who has been tempted by speculation and who almost devoted his life to it, but who has understood its dangers, its illusions, and its many errors and wishes to communicate his experiences and justify his painfully acquired convictions.1
Ijean Piaget, Insights and Illusions of Philosophy, trans. by Wolfe Mays (New York ; World Publishing Co., 1971), p . xvii.
189
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 190
Although Piaget has been known throughout the world
as a developmental psychologist and educational theorist, it
was in philosophy that he received his initial exposure to
questions regarding the mind and how it constructs a model
of reality. Raised within the intellectual environment of
continental philosophy, it is not difficult to understand
how the currents that were stimulating thought in the
universities of Europe during the period between the two
world wars, the exuberance of activity generated initially
by the writings of Immanuel Kant, by Hegel and subsequently
by Husserl, would not only attract him by the heights to
which they dared aspire in their search for understanding,
but finally would cause him to reject conclusions grounded
on what for the careful inquirer were seriously weak
foundations.
In an important sense the work of Kant has served as
a point of synthesis in the history of philosophical thought
of the divergent strains of rationalism and empiricism. Where
on the one hand rationalism was unable to push beyond the
sterile certainty of a priori truths to achieve a meaningful
grasp of reality, empiricism was never able to guarantee that
its own knowledge claims were more than contingently true.
Through the recombination of ideas in Kant's Critique of Pure
Reason we are provided with some possibility for achieving
both certainty and truth and are thus brought to the close
of an era of continuous andseemingly irresolvable debate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 191
But philosophical inquiry did not lie to rest with
Kant. As certain issues were resolved, new ones arose and
these were the seeds of what we call continental philosophy,
notorious for its speculation and, as Piaget points out,
honeycombed by intellectual pitfalls and illusions.
Despite his deep involvement in the study of philosophy,
Piaget’s attitude toward it was, it appears, not a friendly
one. The strength of the great philosophies of the past was
that they had grown out of the heat of scientific discovery.
Piaget sees recent philosophy as totally lacking the scientific
foundation required for a legitimate search for truth. In
our quest to understand how philosophy in union with
linguistics and psychology plays a role in the development
of an understanding of the human mind and its operations, we
should look to Piaget and his evaluation of philosophy in
the twentieth century for the understanding it gives of
whether there is a proper role for philosophers in this
inquiry.
Recognizing the shared interests which philosophers
and psychologists alike have in understanding the mind and
its functions, Piaget believes it is only through psychology
and its empirical approach that this understanding can be
satisfied, Piaget's criticism of philosophy derives from
what he considers a refusal to take into account the empirical
evidence of scientific research. Where early philosophers
were moved by scientific curiosity and developed their
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 192
philosophical theories as a result of their own empirical
investigations of the universe, philosophers today in Piaget's
estimate fail to ground their philosophical positions on an
empirical base. Aristotle's philosophical ideas grew out of
his own biological interests. 7Jhen he was involved in
discovering fact he was a scientist, but when he began to
theorize he became engaged in a philosophical enterprise.
For Piaget scientific investigation is a necessary precursor
to any meaningful philosophical thought.
. . . it seems undeniable that the most important systems in the history of philosophy . . . have all arisen from a reflection on the scientific discoveries of their authors themselves or on a scientific revolution occuring in a period in which they lived or immediately preceding it.2
Piaget's criticism of recent philosophy is that it has turned O away from the empirical world.'' It has concentrated its
attention on introspective description and in doing so it no
longer provides us with a viable means for answering signifi
cant questions of an epistemological nature.
Piaget's criticism of philosophy can be questioned on
two counts, one of relevance and one of substance. Let us
first consider the relevance of his attack to philosophy
as a whole.
2Piaget, Insights and Illusions, p. 46.
3lbid., pp. 209-11.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 193
Ivher he rejects philosophy, Piaget rejects all
philosophy rather than any particular approach or any
specific field of inquiry. When we read the criticisms
which he sets before us, it becomes apparent, that Piaget's
view of philosophy is seriously limited. In point of fact,
his references are drawn exclusively from within the
continental tradition, primarily from existentialism and
phenomenology.^ Although he clearly attacks philosophy as
a whole as not capable of providing insights in cognitive
studies, Piaget is, it appears, reacting against the
introspective tendencies of recent work in phenomenology.
His criticisms do not address work which is being done other
than on the continent.
Piaget may be justified in criticising the orientation
of phenomenological research for its neglect of objective
scientific data. Phenomenology is inherently a subjectivist
approach and phenomenolegists do use introspection as a
means of getting at what they consider to be valuable insights
into "reality." But at the same time we might defend
phenomenology noting that it does not claim to be drawing
scientific conclusions. Piaget has, however, made a serious
oversight in his evaluation by not considering the work
which has been done over the past fifty years within British
and American philosophical circles. Throughout the history
^The name which he specifically refers to as representative of what he wishes to criticize is Sartre.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 194
of modern philosophy it is possible to identify very clear
and significant methodological differences between
philosophers in the continental (Descartes, Spinoza,
Leibniz) and what might be called the Anglo-Saxon (Berkeley,
Locke and Hume) traditions. These differences have carried
on to the present day.
In order to offset somewhat the very telling limitation
of Piaget's critique Wolfe Mays in the introduction to his
translation of Insights and Illusions of Philosophy addresses
the question of the applicability of Piaget's comments to
recent Anglo-Saxon philosophy. Mays argues that there is
indeed a similarity between phenomenology and the work in
conceptual analysis which has dominated Anglo-Saxon
philosophy. Both traditions in his view are united in the
conviction that empirical questions are "irrelevant" to
philosophical inquiry.^ In the case of analytic philosophy
the primary concern in philosophical discussion is with the
question of validity rather than with productive inquiry.
The analyst is interested in examining the logic of the
language rather than the particular casual and genetic factors
which are so important to any understanding of language use
and because of this their study remains sterile and
unproductive. Reiterating Piaget's position. May claims
that one must first collect the data and then develop theory.
Spiaget, Insights and Illusions of Philosophy, p. ix.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 195
Only the scientist has the ability to reach meaningful
theoretical conclusions.
In his criticism of philosophy Mays sets up a mis
leading dicotomy between the scientific approach which is
directed toward uncovering facts about language use through
observation and that of philosophy. He paints the philosopher
as thoroughly removed from and completely disinterested in
the world of facts. Spinning off on a slim thread of
speculation recent philosophical theory has little to tie it
to the real concerns of the study of language and little to
offer to a better understanding of its use. In May's
estimate Piaget would be equally as critical of recent work
in conceptual analysis as he is of phenomenology and would
be unsymphthetic to any defense of the needfulness of such
an approach.
The philosopher, in contrast, sees philosophy and
psychology as serving distinct but equally important roles.
As Mays grants.
The philosopher's usual counter to this is to say that he is concerned with the logic of learning, thought, etc., and not its psychology.
As a result it is assumed that philosophers can talk about the intended meanings, habits, capacities, skills, etc., of human beings without the need to elucidate such questions reference to psychological r e s e a r c h . 6
^Piaget, Insights and Illusions, pp. ix-x.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 196
However, May is still unwilling to concede a place for such
an approach and criticizes work in language analysis as both
"highly impressionistic" and not subject to experimental
control or verification. "Such an approach, he notes, "may
have had some credence when introspection was the only method
used for describing and analyzing psychological phenomena.
But he does not feel it adds to or strengthens the conclusions
of psychological research.
In reading May's criticisms of language analysis and
his attempt to extend them to a broader range of philosophy
we are still left to wonder whether they are properly directed,
It appears at some points that the work he is addressing is
that which has been done for example by Ryle and Wittgenstein
and is referred to as ordinary language philosophy.
It is true that in recent years some philosophers have appealed to behavioral skills rather than introspection in discussing psychological data.8
However, at other times he seems to be referring to the early
work of the logical positives rather than the more recent
ordinary language approach.
Piaget's experiments on concept formation have been criticised by some linguistic philosophers on the grounds that they do not show that there is anything wrong with the child's logic . . . .9
7piaget, Insights and Illusions of Philosophy, pp, ix-x, Sibid., p. X
9Ibid.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 197
Where the logical positivist in a concern for structuring
language might direct such a criticism at Piaget, ordinary
language philosophy itself would defend the logical structure
of any language including the child's as valuable in and of
itself.
Let us now look at the substance of Piaget's rejection
of philosophy. Where we would want to agree with Piaget in
his criticism of ungrounded philosophical speculation and
with the conclusion that there is an essential unity which
must be continuously maintained between scientific inquiry
and philosophical theory, Piaget's purpose in Insights and
Illusions of Philosophy is not primarily to emphasize this
point. Rather Piaget wishes to make the stronger point that
the philosopher is no longer able to serve a useful role in
furthering our understanding of mind and in providing insight
into epistemological questions. Philosophy is inherently
biased and cannot satisfy the standards of objectivity
necessary to the pursual of truth. Epistemology, once the
concern of philosophers, today would properly be seen as a
branch of psychology. Where philosophy pursues wisdom, the
rational synthesis of beliefs, psychology provides knowledge.
And where there are many wisdoms, there is one truth. The
illusion of philosophy is that it leads us to believe that
it is capable of providing that truth.
Piaget's initial claim is that those philosophies of
the past which provided significant insights into epistemolo
gical questions resulted "from a reflection on the scientific
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 198
discoveries of their authors themselves” (as in the case
of Aristotle) or grew out of a scientifically charged
environment (Hume and possibly Kant). The criticism of recent
philosophy is that it has not grounded its theory in an
empirical base. However in granting this point we are not
forced to conclude that philosophers can no longer provide
meaningful contributions to our understanding. In bis
evaluation of the usefulness of philosophy Piaget fails to
recognize the unique contribution which a philosophical
perspective provides to the understanding of epistemological
questions. Where philosophers do not engage in empirical
investigation, they are concerned with the implications of
empirical investigation. Does this mean that philosophy
does not "consider" the facts or rather that it is not its
job to research the facts? If we were to view philosophy
as science in its embryonic stage then we would see the role
of both science and philosophy as identical, the difference
lying in the sophistication of their method. On this
interpretation it is true that the philosopher's failure to
engage in empirical study would give us good reason to
question the value of the results. But the philosopher does
not view philosophy as science in its embryonic stage. The
philosopher should not be using less sophisticated scientific
techniques to obtain what ought to be considered scientific
conclusions. Because philosophy does not concern itself with
researching the facts does not mean that the philosopher is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 199
not concerned with the facts. Good philosophy is very
concerned with grounding itself in a firm factual base, for
it is the facts which form the foundation of philosophical
investigation. And so, the philosopher begins with the facts,
moves on to investigate their logical interrelations and
then is able to look to the implications of these facts for
our understanding of reality. Working with the biproducts
of empirical research, the philosopher begins by questioning
their logical consistency, whether the criteria used in
collecting evidence are consistent, and whether the conclusions
which the scientist reaches are justified by the evidence
available. Having now established a firm empirical as well
as logical base, the philosopher then moves on to draw
philosophical conclusions. The philosopher deals with the
material of scientific research examining it from the point
of view of its logical structure and validity for the
insights it provides. This work must be well grounded in
an empirical base.
It appears to be clearly the case that Piaget's
criticisms of philosophy cannot be justified as a criticism
of philosophy as a whole. They fail in two important respects.
On the one hand, they direct their attention to only a limited
segment of philosophy and that segment is not directly
relevant to our concerns here. Secondly, they base their
criticism on a faulty notion of philosophy as having the same
goals and purposes as psychology, rather than recognizing that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200
the job of the psychologist, to examine the particular casual
and genetic factors in language acquisition and of the
philosopher, to pursue the implications of these empirical
studies for our understanding of language and of mind both
have their proper place. Like Aristotle, Piaget is both a
scientist and a philosopher. When he is engaged in the
empirical study of language use he wears the cloak of a
scientist but considers the logical structure of language
and to hypothesize as to what this reveals about how we
structure ideal his work must be viewed from a philosophical
perspective. Let us then look at the philosophical conclusion
which can be drawn from Piaget's empirical research.
Within the context of history and of the ebbing and
flow of philosophical ideas Piaget places himself within the
philosophical tradition of Immanuel Kant. In Insights and
Illusions of Philosophy he states, "One can feel very close
to the spirit of Kantianism (and I believe I am close to it
. . . .) 10 Rotman also notes this similarity.
Of the three standard responses /_to the question of how we obtain knowledge/ -- empiricism emphasising (op cit) the prime importance of experience, rationalism that of pre-existing reason, and a priorism the effect of the mind's construction -- Piaget is closest, implicitly to a priorism and most antagonistic, quite explicitly, to empiricism 11
10 Piaget, Insights and Illusions of Philosophy, p. 57.
H Brian Rotman, Jean Piaget: Psychologist of the Real (Ithaca, N.Y.: Cornell University Press, 1977), p . 24.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 0 1
While such philosophers as John Locke had portrayed the human
mind as the passive recipient of perceptions, it is Kant's
constructivist theory and its dynamic view of mind which
Piaget finds most compatible with his own philosophical
position. The extensive research he has conducted over
extended periods with children at all stages of development
he believes confirms the claim of Kant that the mind is
continuously and actively constructing our perceptions of
reality. The study of this intricate process Piaget christens
genetic epistemology.
Indicative of his training in both philosophy and
psychology, Piaget believes a "deep continuity" and "fertile
interaction" should exist between the work of the develop
mental psychologist whose interest is in the strategies of
the learner and that of the epistemologist who wishes to
provide a structural analysis of the fundamental categories
of thought. Epistemology should not be considered a purely
philosophical enterprise. He views it as an interdisciplinary
study which raises questions of fact (the sphere of science)
and of validity (the concern of the philosopher). The
conclusions of research in developmental psychology serve
as the clues from which we can begin to answer the questions
of the epistemologist. They are the beginnings of genetic
epistemology, a new science which asks traditionally
philosophical question from a psychological perspective. Its
aim, in Piaget's view is " . . .to take psychology seriously
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 202
and to furnish verifications to any question which each
epistemology necessarily raises . . . The questions
which the philosopher of knowledge has asked throughout the
history of philosophy -- whether knowledge is innate, what
are the sources of knowledge, what are its limits -- are all
questions of fact and can be verified through careful
experimentation.
This, then, becomes Piaget's task, to take the questions
which epistemology has raised and to look to the process of
how knowledge is actually gained, thereby determining its
source, whether it is in some sense innate, what are its
contours and limit and how it reaches its final form within
the cognitive structure of the child. Piaget is thoroughly
of the opinion that through empirical research it is possible
to determine the cognitive geography of the human mind, and
to answer those questions regarding its source and form
which philosophy has addressed but which it now have not been
adequately described. Those issues which have been debated
by philosophers without resolve, whether there are a priori
forms, Plato's Doctrine of Reminiscence, the power and limits
of reason, are no longer the subject of speculation but are
at last accessible to conciliation. This is the goal Piaget
sets for genetic epistemology.
Jean Piaget, Psychology and Epistemology (New York: Orion Press, 1971), p. 7.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 203
In accomplishing his task the major question which the
genetic epistemologist is concerned with examining is by what
means does the human mind go from a state of less sufficient
to a state of more complete knowledge. The dominant tendency
which Piaget sees as characterizing work currently being done
in learning theory is to account for progressive adaptation
of behavior including learning by showing how one particular
action becomes modified and extended to a sequence of actions.
Examples of this approach would include Pavlov's dynamic
stereotypes and Hull's habit family heirarchies. Primary
emphasis has been placed on the environment and how it
determines actions and reactions. In Piaget's view none of
these theories has been conspicuously successful in dealing
with the more difficult and interesting problems which involve
conceptual learning. Piaget shifts his attention from the
effect of the environment to the determining influence which
the mind and mental operations have on a subject's behavior.
It is these mental operations on his view which are the one
most important factor in directing human action.
Piaget's model of human knowledge is essentially dynamic,
In order to have meaningful learning it is necessary not
merely to receive information. Information must be
assimilated, processed and stored. In contrast with a passive
model in which knowledge is viewed as a copy of reality which
we obtain in its final form through a sensory mode, Piaget
presents us with an active model of learning. Knowing is a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 204
process in which we assimilate, through a series of trans
formational stages, facts about reality into a complex
structure of ideas which has its origin in the earliest
stages of human development. As a child's actions become
diversified this structure begins to correspond more and more
adequately with what we might wish to call reality and an
isomorphism develops between the two. When we accept a fact
as true we accept it as sharing this isomorphic relationship
and we integrate it within the structure of other ideas which
we also have accepted as true. In this continuous process
our knowledge becomes progressively more and more adequate
as we build a fuller and more complete picturing of the world.
All species are born with the tendency toward adaptation
of their environment. This adaptation is accomplished through
the interaction of two complementary processes, assimilation
and accommodation. Piaget views human cognition as one
specific form of biological adaptation in which the features
of the external reality are assimilated into the learner's
psychological structure. The structure is then modified to
accommodate the new information. A balance or equilibrium
is constantly being maintained in this continuous, recursive
process and new structures are constantly being created.
Those structures are maintained which are effective in
dealing with reality and those which do not "fit" are
rejected. As Flavell explains.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 205
What you know already will greatly shape and constrain what environmental information you can detect and process, just as what you can detect and process will provide essential gist for the actuation of present knowledge and the generator of new k n o w l e d g e . ^3
On Piaget's model learning becomes an active process in which
the mind builds knowledge structures by taking external data,
then interpreting, transforming and organizing it into an
organic whole.
One of the virtues of the information-processing conception from the present standpoint is precisely that it does explicitly represent human beings as complex cognitive systems, that is, as extremely rich and elaborate structures composed of interdependent cognitive p r o c e s s e s . 14
Knowing, then, is the gradual adaptation of thoughts to
reality, and epistemology, the theory of knowledge, is the
theory of how this process takes place. It is a theory not
in the sense of an abstract speculative supposition which
proposes one more possible interpretation of how we might
view knowledge. Rather it is a theory in the scientific
sense of a hypothesis whose validity is capable of confirma
tion through empirical verification.
Piaget's studies have revealed there is a close
relationship between logical thinking and pre-logical and
sub-logical operations and actions of the child. It has been
demonstrated that classificatory criteria are tracable back
I 1 John H. Flavell, Cognitive Development (Englewood, Cliffs, N.J.: Prentice-Hall, 1977), p. 81 l^ibid., p. 15.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 206
to a primative origin in action. At the earliest stages the
child's classificatory categories are limited to functional
aggregates in which temporal and spatial continguity are
primary. Little by little the child discovers how to classify
in terms of one criterion and constructs hierarchical systems
of inclusion. Children initially classify objects by inten
sion. Intentional classification is based on a recognition
of groupings or classes of objects tied together by common
features and is directly tracable to sensori-motor assimila
tions, experience gained through the manipulation of objects.
Classification by extension begins later with the aid of
precise symbolication as language gradually allows the subject
to go beyond immediate circumstance. At this point a more
sophisticated level of classificatory skill is set in motion.
Classificatory concepts ". . . form a part of the semantic
content of language and lend themselves to mental manipulation
through the medium of language.At the highest level a
coordination of extension and intension is developed. This
stage, the stage of formal operations, is the stage at which
abstract thinking and the ability to manipulate abstract
concepts takes place.
Before moving on to consider further Piaget's analysis
of the development of cognitive abilities let us look more
closely at exactly how Piaget defines these criteria by
15Jean Piaget and Barbel Inhelder, The Growth of Logic in the Child, trans. by E.A. Lunzer and D. Papert (London: Routledge and Kegan Paul, 1964), p. 283.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 207
which we classify concepts. We have noted that, following
Kant, Piaget adopts a constructivist position as regards our
knowledge of reality. However, where Kant used reason to
develop this theory and to determine the categories by which
we structure our perceptions of reality, it is important to
note that Piaget attempts to develop a similar theory basing
his analysis of the logical structure of thought rather on
controlled observations of the activities of test subjects at
various stages of cognitive development.
In his constructivism Piaget wishes to distinguish
himself from the positions of both Aristotle and Chomsky.
On one hand he wishes to reject empiricism and the "myth of
the sensorial origin of knowledge" which began with Aristotle.
Rather than viewing human knowledge as something which is
passively assimilated in its final form through the senses,
and developing a theory (such as Aristotle's) in which the
categories through which we organize our perceptions of the
world exist in reality independently of our perceptions of
them and are themselves characteristics of reality, Piaget
sees these organizing principles to have evolved as a
consequence or our physical interactions with the objects in
our immediate experience. We develop the concept of cause
from repeatedly experiencing one event preceeding another; a
child derives its understanding of space and of extension in
space from the manipulation of objects encountered in day to
day activities. In this sense it would not be overstating
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 208
to portray childhood as essentially a period of very
seriously pursued activity in which the child is engaged in a
methodical examination of objects of the immediate environment
attempting to determine by trial and error what are essentially
the basic laws of science pertaining to physical objects.
Piaget also wishes to distinguish himself from
rationalism and the position set forth by Noam Chomsky.
Rationalism is based on the assumption that there is a
preestablished harmony between the truths of innate reason
which lie at the core of our conceptual structure and the
necessities of the world. The rationalist believes that
somehow our mind is capable of providing us with valid
knowledge of reality. However, rationalism fails to explain
why this should be the case. Chomsky extends this rationalist
hypothesis to language. He argues that all human language is
organized around a core of innate principles which determine
its structure and logical form. However, in Piaget's view
Chomsky, like the rationalist has failed to explain where
this innate schema originates and why we should believe it
has any efficacy at all.
And even with the more modern, less easily refuted, versions of the rationalist assumption, such as the linguist Noam Chomsky's assertion that all human languages spring from the same innate kernel, he /Piaget/ insists there still remains the question of beginnings, the fundamental problem of the nature and origin of the fixed innate scheme.
lôRotman, Jean Piaget: Psychologist of the Real, p. 27.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 209
Where Piaget does agree that there are organizing principles,
the cognitive structure variables, which specify the form both
our perceptions and of language he feels that the hypothesis
that they are innate is both "unnecessary" and wrong.
I agree that the structures that are available to a child at the age of fourteen or sixteen months are the intellectual basis upon which language can develop, but I deny that these structures are innate.‘17
It can, Piaget is convinced, be proved in clear evidential
terms that such a preexistent and determinate set of principles
does not form the foundation for our knowledge about the
world. Rather the structures which underlie our knowledge
of the world are in a continuous process of growth and
development which begins at conception and continues on
throughout life. The form which they assume is determinate
only in the sense that it is a function of our experiences
and interactions in the world. That we always experience
continuity of objects in time is the reason why the idea of
permanence of substance is an essential part of our perception
of reality. In the same sense because we experience the
constant conjunction of one event preceeding a second, we
develop the conception of cause and effect. In taking this
position, then, Piaget explicitly rejects the view that there
is a core of innate organizing principles which predetermine
how we organize the structure of our ideas, and in doing so
l7Jean Piaget, Genetic Epistemology, translated by Eleanor Duckworth (New York: Columbia University Press, 1970), p. 47.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 210
he draws a clear distinction between his own theory and that
of Chomsky. As we have seen, genetic epistemology is the
study of how these structures develop.
If there are no innate organizing principles which form
the fertile soil of thought and if sensation as well is not
adequate to explain the source from whence our ideas spring
forth, how then does knowledge occur. The sensorial originof
knowledge is for Piaget at least incomplete and at best
false. Where sensation is active in the elementary stages
of knowledge formation, knowledge is never the result of a
mere impression made by objects on the sensorial being.
To be more exact, we would have to speak of the perceptive and not of the sensorial origin of scientific knowledge, since perception is not composed of sensations, but is an immediate composition of them.IS
In fact it only makes sense to talk about simple sensations
from a physiological point of view. Where there are
immediate perceptions these are received as a totality in
which the individual sensations are merely structured
elements which can be analyzed out rather than structuring
themselves. As Piaget explains, "I immediately see the
house as gestalt and then analyze it in detail. At the
earliest stages we might be able to argue that in this limited
sense knowledge does have its base in sensation. However.
knowledge in any significant sense has a more auspicious
ISpiaget, Psychology and Epistemology, p. 66.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 1 1
source. It stems not ". . . from perception alone but from
the entire action, of which perception merely constitutes
the function of signalization.Intelligence is an active
process.
Only by adding something to perception do we discover the characteristic of an object. And what we add is precisely nothing but a group of logico-mathematical limits which alone make perceptive reading possible.20
We integrate our perceptions, which we receive through the
active manipulation of our environment, within a structure
of logico-mathematical principles (categories) which serve to
organize and delimit what we experience (in the Kantian sense)
while at the same time being determined by that experience.
Consequently, by the middle of the first year of life, the
child has developed a sensori-motor intelligence which
demonstrates a logical system of organization. This logic
is a logic of action, a logic of practical concepts which
they derive from their action on objects. Also developing
is a logical conception of order in which cause - effect
relationships can be identified. This will serve as the root
for the development of an understanding of conservation and
reversibility. But although there is this logical base, it
exists at the level not of thought but of action. As these
actions become internalized the child develops the ability to
19piaget, Psychology and Epistemology, p. 66.
20Ibid., p. 72.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 1 2
carry them out in thought as well as in action. The action
then becomes reversible in the sense that it can be taken
in either direction. It is seen to be invarient in that it
always involves the principle of conversation of some element
through transformation. It is interrelational in that every
operation is related within the context of the total structure.
When the child has successfully moved from action to the
internalization of action it has progressed to the concrete
operational stage of cognitive development.
The final stage of cognitive development is that of
formal thought. Where in the sensori-motor stage the child
was in active involvement with objects, in the concrete
operational stage there had been the movement away from the
sphere of action to that of thought of objects. In the formal
operational stage, however, the subject is no longer restricted
to thought of objects. Thought becomes formal and the
subject is able to move out of the realm of reality into
that of possibility. Thinking at this stage takes the form
of hypothetical-deductive constructions. It is hypothetical
because it is no longer restricted to perceived realities.
It is deductive in that it is able to pull together a set of
assumptions and draw the necessary consequences independently
of whether the fact actually exists.
As a child enters adolescence it becomes increasingly
more able to deal with a much wider range of operational
possibilities than the restrictive confines of simple
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 213
groupings of classes and relations. It develops a new set of
operational structures based on the principles of propositional
logic and moving beyond propositional logic to include such
highly complex operational schemata as combinatorial opera
tions, double systems of reference, mechanical equilibrium,
multiplicative probabilities and correlation?^ The
essential characteristic of propositional logic is not that
it is a verbal logic. It is that it is a logic of all
possible combinations whether actual or theoretical. Where
in simple abstraction from objects there is a transposition
from the level of action to that of concrete thought about
objects, here we find the beginnings of the development of
the mental process of reflection which finds its roots in
action, but which moves beyond action to the thought of
action and on to reflection on possible action. At this
stage the cognitive structure has developed into a complex
of ideas which can be seen as comprising a model of the
world from the point of view of the subject.
Throughout this discussion of Piaget's view a dominant
concept has been that of structure. While Chomsky argued
that at the core of all language there lies a common skeletal
framework upon which and around which each individual language
itself is structured, Piaget sees thought too as organized
within the context of a structural whole. As we learn we
2lFor an elaboration of these possibilities, see Jean Piaget and Barbel Inhelder, The Growth of Logical Thinking From Childhood to Adolescence. Translated by Anne Parsons and Stanley Milgram (New York: Basic Books, 1958).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 214
integrate our n e w ly gained knowledge into a pre-existing
structure of ideas which finds its genesis in the earliest
actions of the child and through progressive experience
grows, becomes altered and revised. But from an ontological
perspective what is a structure and do structures exist?
The reader may ask here whether 'structures' have real, objective existence or are only tools used by us to analyze reality. This problem is only a special case of a more general question: do relations have objective, independent existence. One answer will be that it is nearly impossible to understand and justify the validity of our knowledge without presupposing the existence of relations. But this answer implies that the word existence has to be taken to have a multiplicity of meanings .22
Piaget has proved us with a curious answer. Knowledge
presupposes the existence of relations. In the same sense
Piaget claims that knowledge also presupposes the existence
of structures. Where it is possible to consider individually
existing ideas in isolation, knowledge itself presupposes a
unity in which these ideas fall into relation with one
another within a unified whole. Thus if we ask what in
actuality a structure is, it is first a totality comprising
the sum total of our conceptions about the world. And as a
totality it is more than just the conjunction of one idea
with another. It is an organic unity which becomes more
than the sum of its parts. And as a consequence of this.
22Piaget, Genetic Epistemology, p. 23.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 215
knowledge is only made possible through the network of
interrelationships of these parts one with another.
A second feature of structure is that they are
transformational. They are tied together by laws which
govern how their parts interrelate with one another. "If
the character of structured wholes depends on their laws of
composition, these laws must of their very nature be
structuring .' 23 These transformational laws define the
logical framework which the structure of ideas assumes.
Although these transformational laws too are in a sense a
part of the structure, Piaget stresses the importance of
making a clear distinction between what we consider to be
the elements which make up the structure and the transforma
tional laws which apply to them. However, it is not the case
that there is the structure which is continuously changing
and a set of laws which are immutable. This would lead us
to a position similar to Chomsky's in which the laws of syntax
are in some sense wired into our cognitive structure, and
back to innatism.
Noam Chomsky is a case in point: for him generative grammars appear to demand innate syntactic laws, as if stability could not be explained in terms of equilibrium mechanisms
23Jean Piaget, Structuralism, translated by Chaninah Nachler (New York: Harper Torchbooks, 1970), p. 10.
24 Ibid., p . 12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 216
Finally structures are always self-regulating. The
". . . transformations inherent in a structure never lead
beyond the system but always engender elements that belong
to it and preserve its laws."25 Through the laws of trans
formation the structural unity of the whole is maintained.
This self-regulation proceeds by the application of what are
ideally perfectly explicit rules that define the structure.
At the basis of Piaget's conception of structure is
the idea of logical form. Structures are, we can see,
essentially logical networks in which groupings of elements
are tied together by the logical relationships which they
hold to one another. The transformational laws which define
these relationships are by nature logical and the system
itself is characterized by closure. In the theories of both
Aristotle and Kant a parallel could be drawn between the
structure of thought and the laws of logic. Aristotle
believed that his categories reflected natural kinds and
that reality could be broken down and fitted into these
classifications. Our thoughts were merely a reflection of
the world and its logical structure. In his interest in the
logical structure of thought Piaget, however, does not see
logic as a mode of organization forced on us by the world of
experience. Rather it is a mode of organization we ourselves
construct, ". . . by coordinating our own actions and
25piaget, Structuralism, p. 14.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 217
abstracting the relations between t h e m . "26 Piaget's aim is
to isolate and describe these mental structures on which
reasoning is based, to examine the logical structure of
thought through a practice he refers to as logistics.
It is unclear from Piaget's discussion of logistics
just how it might be distinguished from traditional logical
analysis and why he chose to initiate this term to refer to
what he conceived of as an analysis of the logical form of
thought. We might define logic as the study of the
principles of valid reasoning which is based on a systematic
examination of the structure of propositions. The logician
abstracts from the content of propositions and sets forth
for analysis their logical form. Logistics, as defined by
Piaget appears to hold a similar goal. The one possible
difference which is evident is that it is directed more
generally at an alaysis of thought rather than of propositions
alone. The task of logistics is to analyze the logical
structure of thought.
Piaget supports the view that psychological research
on the development of intellectual operations has revealed
certain structures which are at the root of natural logic.
If we abstract from the content of thought we will find that
there is a logical framework of relations which ites one idea
to another. This framework can he translated for analysis
^^Piaget and Inhelder, The Growth of Logic in the Child, X X .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 218
into symbolic form, and in doing so the operations of the
mind can be expressed in the form of abstractions (categories,
relations or propositions) which it manipulates in a purely
deductive m a n n e r . 2^ By symbolizing mental operations in
logical notation we are better able ". . . to detach them
from their immediate context and to combine them more
rigorously.
What is the status of these logical forms which
logistics is attempting to analyze? Piaget proposes three
possibilities. We could conceive of them as universels
existing in themselves, as mere syntactic features represent
ing only a tautological relationship of thought to reality,
or we could conceive of them as expressing in symbolic form
the very operations of collective and individual thought.
The first two possibilities involve us in both scientific and
philosophical problems. It is quite clear which alternative
Piaget prefers.
If we do not wish to subordinate logistics to the unverifiable hypothesis of eternal ideas, nor leave the 'language' that it forms hanging in a void without relation to living beings capable of using it, we can only conceive of this discipline as itself also concerned with the operation of thought.29
Logistics, then, provides us with a method for analyzing
the actual structure of mental operations in a symbolic
9 7 Piaget, Psychology and Epistemology, p. 113
28 Ibid. 29 Ibid., p. 112.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 219
notation borrowed from the logical analysis of propositions.
Piaget has moved from the psychological task of discovering
through observation and experimentation the process by which
knowledge is attained to the conclusion that the manner in
which we structure thought is a function of our interactions
on the level of experience. Having established to his
satisfaction the existence of a cognitive structure around
which we pattern our ideas, he begins the philosophical task
of analyzing that structure from a logical perspective.
In the theories of both Bruner and Piaget we find
psychological support for the conclusion that learning is an
active process in which the subject is engaged in a creative
process governed by the rules similar to those of prepositional
logic. Piaget has found interesting evidence to support
the conclusion that these rules are circumscribed in
accordance with the nature of our interactions with the
environment. In this sense his theory contrasts in significant
respect to that of both Kant and Chomsky. In other respects,
however, what Piaget and Bruner represent is the theoretical
support necessary to Chomsky's linguistic theory.
We are now at a point of evaluating whether this
theoretical support is sufficient to sustain a cognitively
based theory of instruction. Piaget has offered evidence to
support the contention that an understanding of cognitive
process is essential before we can begin to adequately account
for andmore importantly maximize learnings.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 220
For our purposes, if it is the case that thought is
characterized by an identifiable logical form which is
capable of formulation and analysis, it may indeed be
feasible to develop a theory of categories of thought on the
model of Immanuel Kant, basing this on an analysis of the
logical structure of language. Such an analysis should have
productive possibilities for instructional technology.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PART IV
CONCLUSION
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER VIT
THE IMPLICATIONS OF RESEARCH IN LANGUAGE AND MIND FOR THE DEVELOPMENT OF AN INTEGRATED THEORY OF COGNITIVE STRUCTURE
Thought is surrounded by a halo. -- Its essence, logic presents order, in fact the a priori order of the world: that is, the order of possibilities . . . . It must rather be of the purest crystal. But this crystal does not appear as an abstraction; but as something concrete, indeed, as the most , concrete, as it were the hardest thing there is.
The task which we have set before us, to examine
thought and to determine the extent to which it is possible
to provide insights into its form and structure through an
analysis of the form and structure of language, is now draw
ing to is conclusion. In the Metaphysics Aristotle attempted
to provide a means to empirically determine through language
what he considered to be the ultimate ontological categories
into which reality was divided. His assumption was that the
distinctions which exist in language exist in reality as well.
Consequently, rather than fitting his theory to that which
it was attempting to explain, he forced reality into the
mould set forth by language. His account provided an
ontology of reality. The extent to which it was unsatis
factory was the extent to which it assumed that language
necessarily reflects the reality which it represents.
Kant's categories of the understanding lead us much closer to an adequate rendering of the manner in which our thought about the world is organized and structured. In the
^Wittgenstein, Philosophical Investigative, p. 44e. 222
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 223
Critique of Pure Reason Kant analyzed thought, rather than
attempting to characterize reality in itself as something
which could be described independently of how it is known.
He accounted for how we perceive the world in terms of
categories of thought. These categories made up the frame
work within which we actively construct our understanding of
reality. The result was an ontology of thought, of the
basic categories into which we separate the data of
experience. The extent to which it was unsatisfactory was
the extent to which the analysis was grounded on an a priori
rather than on a firm empirical base. Where Kant may have
provided a transcendental argument deriving from what he
considered the necessary conditions which determine experience,
we still have no clear evidence to support the validity of
the categories he has laid before us.
Our aim here has been to suggest a manner in which
thought could be analyzed which would provide, within an
empirically verifiable framework, an understanding of how we
organize perceptions and structure our conception of reality.
It has been our position that an examination of the logical
structure of language provides a possible means whereby we
can determine the structure of thought and defend a theory
of categories on the model of Aristotle and Kant, but one
which is both empirically grounded and is properly conceived
as an analysis of reality as it is known to us rather than
reality in itself. Through an examination of research in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 224
linguistics and in cognitive psychology the outlines of a
philosophical theory of knowledge which establishes categories
of thought can be drawn and answers to philosophically
significant questions regarding the logical structure which
characterizes thought can be sought.
Conclusions of research in the nature of language and
linguistic behavior serve as the theoretical base upon which
an examination of the logical structure of thought through
language can be defended. Having established the precedent
for our task in the work of Aristotle and Kant we moved on
toward the completion of this task in the light of recent
studies in transformational grammar and linguistic structure,
looking first to Noam Chomsky and the techniques which he
developed for sentence analysis. We then explored the
philosophical implications of generative transformational
grammar in the writings of Jerrold Fodor and Jerry Katz.
Chomsky's contribution to the science of linguistics
provided a new method for the analysis of sentence structure.
Through application of the techniques of generative trans
formational grammar he moved beyond an analysis of the surface
structure of language to an underlying deep structural level.
It is on this level that the logical interrelationships
between the elements of the sentence are revealed. Applying
these methods it would be possible to break down language as
it appears in word or in written form into its logical
elements and to analyze how they are organized in relationship
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 225
to one another. The usefulness of this technique lies not
in the analysis of the logical structure of language per se,
but in the understanding which it provides into how the mind
collects data, organizes it into a form which can be retrieved
with relative ease, and of the network of interrelationships
which hold between the concepts which are represented in
language.
Chomsky conceived of language as an organic whole made
up of sounds, words and sentences which serve as a means
through which thoughts are communicated. The elements from
which language is composed represent a structural unity which
can only be understood from the perspective of their
interrelationships. The form which this network assumes is
a function of how the mind processes information and
consequently is determined by the physiological make-up of
the brain. Generative transformational grammar, in analyzing
the form of language at various levels from deep to surface
structure, provides us with an empirical way of accessing
the logical structure of thought as well.
Working within the framework of generative trans
formational grammar, Fodor and Katz attempted to derive
answers to traditionally philosophical questions regarding
the logical structure of ideas. They accept the conclusion
that the structure of language reveals the structure of
thought. Utilizing Chomsky's distinction between underlying
and superficial structure as a means whereby a distinction
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 226
can be drawn between logical and grammatical form, Fodor and
Katz look to language for the insights it provides regarding
the nature of thought. On this interpretation learning is
viewed as a computational process. In this process the mind
constructs models which represent reality. These models are
utilized in thought in making inferences, problem solving,
and in the accomplishment of physical tasks.
In order to account for thought in terms of a
computational process, Fodor argues for the existence of a
language of thought, a representational system which corres
ponds structurally with natural language but is prerequisite
to the learning of language. By applying these techniques
of generative transformational grammar it is possible to
determine the underlying deep structure of any sentence. In
this way we are able to gain insight into the "language"
of thought.
Deep structure corresponds with what Fodor identifies
as the logical form of language. The logical form, or what
is more generally identified as the meaning of a sentence,
is made up of the senses of its lexical items in conjunction
with the grammatic relations between its syntactic
constituents. By organizing the lexical items according to
their meanings and determining the smallest set of vocabulary
items with which we would still be able to represent all the
meanings of our language, it becomes possible to determine the
most basic semantic categories of language. In turn, if we
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 227
accept Fodor's contention that there is a direct correspondence
between the semantic categories of natural language and those
of the language of thought, we are then provided with a
means for determining how we organize the structure ideas
based on the concept of a limited set of basic semantic
categories which characterize language as well as thought.
That all persons share the same store of organizing categories
is tracable, Katz feels, to the existence of an innate
linguistic core or language acquisition device.
If we assume that the categories of language are, like other substantive universels in the theory of language, parts of the language acquisition device, then the reason why each language has just these categories as the highest concept types in its semantic structure is that linguistic experience is organized in the form of rules on the basis of these categories. This is an appropriately Kantian answer to Kant's question.2
In "Deep Structure as Logical Form" Gilbert Harman
concludes that Chomsky's theory of generative transformational
grammar provides a significant contribution to our under
standing of logical structure of thought. It is Harman's
intention to develop a theory of logical form based on the
propositions of ordinary language and to utilize this in his
description of thought.
O Katz, Philosophy of Language, p. 279.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 228
Bound up within Hannan's theory of logical form are
suppositions regarding the relationship of language to
thought. If we entertain a proposition, we at the same time
entertain a thought. This thought is a mental instance or
token of the proposition in which ideas or concepts combine
together to form the substance of that proposition. How
these combinations occur is prescribed according to a finite
number of primative abilities whose function it is to lay
out the structural framework of thought.
To be adequate a theory of logical form must assign a
logical form to every interpretation of every sentence in the
language. The rules which associate the logical form with
the surface structure of the language are the transformational
rules identified in Chomsky's generative transformational
grammar.
A theory of logical form must also state rules for
logical implication. All valid implication must be rule
governed. For example, sentence A implies a sentence B if
and only if no uniform interpretation of the non-logical
elements contained within the sentence can make A true while
at the same time making B false (without changing the intended
interpretation of the logical elements which bind the 3 sentences together.) These logical implications are then
relative to the assignment of logical forms to sentences of
the language. A theory of logical form serves to specify
3Cilbert Harman, "Logical Form," Foundations of Language (September 1972): 39.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 229
these logical implications by first specifying the distinction
between the logical and non-logical elements of the sentence.
In stating the rules of logical implication the theory should
account for all the implications of the language.
In analyzing the logical form of language and conse
quently of thought, Harman rejects an analysis in terms of
the traditional schema of formal logic. Where in quantifica-
tional theory the logical form of a sentence is revealed
through its translation under universal and existential
quantifiers, an analysis of language which penetrates to
the level of deep structure reveals that the quantifications
found in natureal language are much more varied and their
meanings much more subtle than traditional logical analyses
would be capable of demonstrating. As a consequence Harman
analyzes logical form within the context of a theory of
reasoning. Let us look at how this might be done.
In Thought Harman proceeds to develop a theory of
thought centered around the concept of reasoning as a logical
process of inference according to rules which define the
structure of language as well as of thought. Within this
interpretation a person is conceived of in functional terms
as a "system of states and processes that possess representa
tional characteristics by virtue of their role in the
functional system."^ These mental states and processes are
^Gilbert Harman, Thought (Princeton; Princeton University Press, 1973), p. vii.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 230
defined in terms of their role within that system. The theory
which is proposed, then, serves as a psychological model of
the workings of the mind, similar to a program which defines
the behavioral repertoire of an automation. Such a program,
which includes reactions to input, how internal states in
combination with input yield new internal states, and how
these internal states can lead to various outputs, interprets
instances of mental states and processes as the equivalent
of physical states and processes functionally defined in
terms of their role within the program.
For each person there is a unique patterning of states
and processes which is related in specified ways to themselves,
to their perceptions, deprivations and actions. Our conception
of ourselves in the world can be viewed most accurately in
terms of a plotting on a map rather than as a participation
in a scenario. We locate ourselves within the framework of
our experiences in time and space much as we might locate a
geometric point on the globle. In this sense our view of
reality has structural rather than just linear dimension.
It can be conceived of as a dynamic, logically structured
conceptual model of reality as we know it, of which we
ourselves are a part. The model is characterized by a unity
which ties the parts together within the whole.
In conceiving of the world, then, we develop unified
models of reality which are continuously being restructured
as new experiences are undertaken and new information is
integrated with old. Each person is involved in a process of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 231
maintaining the coherence of their world view by modifying
the way in which it represents the world.^
In attempting to define functionally the logical form
of thought through an examination of reasoning Harman
distinguishes between theoretical and practical reasoning.
Functionally defined, theoretical reasoning is the attempt
to improve one’s overall view of the world by increasing its
explanatory coherence, the result being an increase in
knowledge. In order to provide an adequate account of this
process a correlation or mapping (F) can be drawn between
the mental or neurophysiological process x and those abstract
structures of inference which can be represented as F(x).
To give such an account is to say which of infinitely many
possible mappings from processes to abstract inferences can
serve as reasoning instantiators. Practical reasoning
functions within the same framework as theoretical reasoning.
Where theoretical reasoning is concerned with what to believe,
practical reasoning is concerned with what action to take.
Since our intentions are in part determined by our beliefs,
theoretical reasoning also serves to influence outcomes within
the practical sphere.
The representational character or meaning of mental
states is determined on Harman's account by the function
In reasoning our representations are primarily linguistic. However, as in perception, they could be non- lingusitic as well.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 232
which they perform, and as such the language of thought can
be characterized in terms of its relation to spoken language.
Mental states are referred to by combining the name of the
type of state (e.g. belief, hope, fear) with a 'that'
sentence describing the state. The result would be the
description of a state such as, "the belief that snow is
white." Logical relations between beliefs hold by virtue
of the structure which the mental states have. These
structures represent the way it can be constructed out of
its constituitive elements, names, predicates, connectives,
variables and quantifiers. Mental states, then have
representational characteristics in much the same way that
sentences do. We can talk about x, as well as fear x, hope
for X, or believe x. On this interpretation mental states
can be seen more clearly to approximate a "language" of
thought.
The relation between a language of thought and spoken
(or written) language can also be phrased in terms of the
meaning of the sentence in the language of thought. Only
sentences in the outer language have meanings in the
ordinary sense. However what they mean is dependent in part
on the role of corresponding sentences in the language of
thought. Since the state expressed is an instance of a sentence of the inner language of thought, the meaning of an outer sentence is at least partly a matter of the role in thought of the inner sentence it expresses.&
^Ka^man, "Logical Form," p. 60.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 233
Belief states represent particular elements in the real world.
They result from the perception of that element in experience
and generally lead to behavior which involves it. The
character of a mental representation is determined through
the interaction of the elements of experience both directly
and indirectly and by the interaction of these groups of
elements with other elements in experience.
Karman's interpretation then, is based on the conten
tion that an integral relation exists between thought and
language. Where thought is expressed in language, language
also moulds thought. And in attempting to gain an under
standing of the structure of thought it becomes apparent
that it can be best understood in terms of a logical analysis
of the structure of sentences. Thoughts are essentially
sentences analyzed into their truth-conditional structures.
The representational character of a sentence depends on its
syntactic elements and how these are joined together. This
construction is not just a matter of the order in which the
words appear, but more fully determined by the logical order
of their elements. Truth conditions provide a clue to this
construction.^ The truth or falsity of any sentence of the
natural language can only be determined within the context of
a particular situation. In this way Harman associates his
theory of logical form with a theory of truth. The theory
of truth shows how the truth conditions of any sentence depend
7a 11 A s are B is true if things designated by A are among things designated by B.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 234
on the structure of that sentence. It states for each element
of the structure what its contribution is. The result is an
infinite set of statements of the form "x is true if and only
if s where x is replaced by the name of a sentence and s is
replaced by the sentence itself. If adequate, this theory
will provide us with a grammar of the language in terms of a
set of rules for constructing sentences out of their parts. A
theory of logical form assigns logical form to sentences and
states the rules of implication. Within this theory we
should be able to show that all instances of rules are logical
implications and that the rules account for all instances.
What is represented in language and in thought, then is
determined by its role within the conceptual scheme and the
role within this conceptual scheme is related in a significant
way to the logical implications which it implies. Thus Harman
concludes, any adequate theory of representational character
must make use of a theory of logical form as derived on the
basis of Chomsky's theory of generative transformational
grammar.
From a theoretical point of view where has Chomsky's
theory carried us through the solution of our initial problem,
to specify an empirically based theory of the logical structure
of thought which would define the categories of thought in a
more acceptable framework than that provided by either
Aristotle or Kant? Applying the methods introduced in
generative transformational grammar to our problem we find
a technique which enables us to move beyond the surface
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 235
structure of sentences to a level which reveals the logical
structure underlying language. This logical structure defines
language in terms of a series of logical elements which can
be analyzed in terms of their interrelationships with one
another. On a hypothetical level it would be possible to
design a matrix which displays the form and order of these
logical interrelationships.
What would this matrix resemble? It could be likened
to a logical diagramming of those "facts" expressed by
language. The sum total of these facts would provide a
picturing of that which language was attempting to express,
a model of reality as we express it through our words.
The reality which we express through language, however,
is not necessarily reality as it exists. But it is neces
sarily reality as we perceive it to exist. Language, then,
and the way in which it represents the world can serve as the
basis on which we are able to draw conclusions regarding the
logical structure of thought. If language expresses the
logical relationships which we conceive to exist in the
world of our inner and outer experience, it also reflects
the structure of our thought about that world. Having
penetrated to the level of the logical form of language we
have penetrated as well to the level of the logical form of
thought. What is revealed is a matrix which can be defined
in terms of a structural model of our experience of reality
which is multidimensional in character, a logical picturing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 236
of reality as we know it.
In discussing the relationship of language and reality
in the Tractatus, Wittgenstein draws a similar picture. A
proposition for Wittgenstein is a tableau vivant, a living
O picture of reality. It has meaning only in so far as it
pictures a state of affairs. The elements of the sentence
or proposition correspond to the elements of that state of
affairs.
However a pictorial view of the relation between
language and reality can be a misleading one. It engenders
the possibility of a serious misunderstanding of the essential
nature of the relationship. For when we think in terms of
a picture our immediate interpretation is that the pictorial
form is spatial. Such would be a superficial and inaccurate
interpretation of the view Wittgenstein presents. Only some
pictures represent a state of affairs by duplicating them
spatially. But this is in no way primarily the case. Such
a view of language would be to portray it as a series of
names which stand in one to one correspondence with each
element in the state of affairs which it is attempting to
describe.^ The more interesting point which Wittgenstein
^Wittgenstein, Tractatus, 4.021, p. 21. ^Although Wittgenstein in the Tractatus does assert that a proposition is a concatenation of names it seems even here in stressing the logical nature of the relationship he is introducing a complexity and flexibility into the concep tion of how propositions mirror reality. He is able to avoid the problems of such thinkers as Meinong who were forced to posit the existence of entities such as relations, classes and qualities to maintain a reference for their verbal counter parts .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 237
makes in the Tractatus is that the propositions of language
are logical pictures of reality.They are models of
reality. The concept of a model provides us with a means
whereby Wittgenstein is able to illustrate the unique nature
of the relationship between language and reality for a model
is not an exact copy of that which it models. It does not
necessarily correspond part to part with that which it
represents as a spatial picture would. Rather in its
structure it reveals the essential nature of that for which
it stands. The relations between the elements in the
picture present a model of the relations between the
elements of reality.That logical form is mirrored in the
proposition. By joining propositions together the logical
scaffolding of the world as we know it is set forth.
A second contribution which Chomsky's theory makes
to our understanding of mental processes is in his
characterization of mental operations as computational. In
knowing a language we are in a specific mental state. This
mental state consists of being in possession of a specified
mental structure consisting of "a system of rules and 12 principles that generate and relate mental representations.”
These rules define the computational aspects of language and
lOwittgenstein, Tractatus, 2.12, p. 8. l^Ibid., 4:121, p. 26 12 Noam Chomsky, Rules and Representations (New York: Columbia University Press, 1980), pt 48.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 238
determine how its syntactic, phonological and semantic
patterns are formed. They also determine the organization
of the conceptual elements in language, the system of object
reference and thematic relations which hold for example
between agents and goals, or agents and objects. Where the
first set of rules are specific to language, the second may
in fact be part of the common sense understanding and
represent a link between language and elements of the more
general conceptual structure as Chomsky had suggested in
Reflections on Language. As a consequence of this inter
pretation Chomsky in his most recent work refers to the mind
as an integrated system of rules and principles. Quoting
from Eric Lenneberg he states.
The rules that underlie syntax (which are the same for understanding and speaking) are of a very specific kind, and unless man or mechanical devices do their processing of incoming sentences in accordance with these rules, the logical, formal analysis of the input will be deficient, resulting in incorrect or random responses. When we say rules must have been built into the grammatical analyzer, we impute the existence of an apparatus with specific structural properties or, in other words, a specific internal organization.13
A third contribution which Chomsky makes to our under
standing of mind and mental phenomena is in this characteriza
tion of mental structures as physiological mechanisms.
13 Noam Chomsky, Rules and Representations in Lenneberg, Biological Foundations of Language (New York: John Wiley, T9FTTTWT-39T:r ^^Ibid., p . 5.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 239
When we refer to the mind or to mental representations we are
speaking at the level of abstract characterizations of
properties of certain physical mechanisms. It is a
consequence of these shared physical mechanisms that persons
are able to communicate on the level of language.
I see no reasonable alternative to the position that grammars are internally represented in the mind, and that the basic reason why knowledge of language comes to be shared in a suitably idealized population . . . is that its members share a rich initial state, hence develop similar steady states of k n o w l e d g e . 13
We have then from Chomsky an interpretation of language
and consequently of thought as a logical picturing of reality
which can be described by determining the logical patterns
which exist at the level of deep structure. In this manner
we can draw an ontology of thought, a description of the
basic categories into which thought divides and of the logical
structuring of these categories. We have seen mental
operations to be computational in nature. How we represent
the world is determined in accordance with established rules
which are physiologically based and are shared with all
members of the human species. Finally, we have described the
mind itself as an abstract characterization of physiological
mechanisms which occur within the brain. These mechanisms
determine the manner in which the mind conceives of reality
and how it organizes its thought.
^^Chomsky, Rules and Representations, p. 87
Reproduced with permission of the copyright owner. Further reproduction prohibitedpermission. without 2 4 0
The conclusions reached through the study of language
and language use have been supported as well in studies of
human cognitive behavior. In order to defend our approach
we sought to establish a theoretical base in the conclusions
of research into the nature of human cognitive processes as
well. If we are to develop a productive theory of how the
mind comes to "know," how it organizes and structures
reality and the nature of the structure which thoughts
assume, we must determine how one learns and to what extent
learning involves a logical structuring of ideas.
The studies of Jerome Bruner and Jean Piaget support
what might be viewed as an information-processing interpre
tation of thought. They view the mind as an active, and
efficient machine which is fed data, manipulates that data,
and then modifies it according to its own program. What
the mind is engaged in is the construction of a model of
reality.
In the construction of models of reality how our
ideas are organized is a function of what we experience and
of a continuous interplay between experience and thought.
According to Piaget's findings, children at the earliest
stages of development organize their experience on the
basis of similarities and differences which they perceive
in their interaction with objects in the environment. As
language develops the child becomes able to move beyond
immediate circumstance to consider action in thought. As
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 241
thinking reaches more sophisticated levels the subject
begins to deal in abstract concepts no longer associated
with action or objects. The result of this activity is
the gradual establishment of categories of thought on the
model of Kant's categories of the understanding. However,
these categories derive their form as a consequence of the
limitations imposed at the sphere of action. As such they
are essentially different from Kant's categories which are
necessary features of our understanding. The formal schemata
which develops provides an organic unity within which know
ledge is structured. As new ideas are acquired, material is
encoded and stored. The structures which evolve, then, can
be conceived of as logical networks of ideas, tied together
by principles which define the logical relationships of
their constituents.
Having established through observation of experimental
subjects within a laboratory setting an information-processing
interpretation of thought, and evidence for the conclusion
that learning can be described in terms of the development
of structures which are organized according to "logical"
rules determined by action, the research of Bruner and Piaget
suggests firm empirical support for the substance of the
theories set forth by Chomsky, Fodor and Katz. Where these
accounts of learning and of language use are descriptive,
they provide us with the foundation we have sought for a
prescriptive theory of instruction.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 242
In the past several years as educational practice has
undergone the scrutiny of relevance, an ongoing discussion
has been evolving within educational circles both at the
precollege level and within undergraduate liberal arts
faculties as well, as to whether any significant benefits
can be derived from an exposure to philosophical thinking.
Professional philosophers who recognize the need for such
an exposure at some point within the educational training
of the student have argued their position is a desirable
goal, it may still be asked, is it an economically expedient
one? Are there tangeable benefits which it might provide?
What has been lacking within philosophical circles
is a strong theory-based rationale on which to build an
argument for an exposure to philosophical thinking within
the overall curriculum. Research into the nature of cognitive
processing provides just such a rationale and one which could
be used to reinforce the position presented by the profes
sional philosopher for the relevance of the study of
philosophy and philosophic method.
Conclusions of research into learning by Piaget and
Bruner and into language acquisition and use by Chomsky,
Fodor and Katz point to the necessity that educational
theorists integrate within their explanatory base an inter
pretation of thought as a computational process involving the
making of logical inferences in accordance with rules. In
thinking the mind is engaged in an ongoing process of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 243
hypothesis formation and revision in order that coherence
be maintained among the elements within the conceptual
structure. The consequence of this process is a continuous
building and restructuring of what might be described as
schematic representations or models of our experience of
reality. New experiences and new information are the
initiative for this restructuring. As knowledge is acquired
and assimilated within the knowledge structure this model
undergoes a continuous process of refinement and change.
With each transformation the model attains greater explana
tory coherence and power.
As such, thought and the language through which it
is expressed, display a distinct logical form. However this
form is not derived from a logic of prepositional functions
which has proved inadequate in accounting for the com
plexities expressed in ordinary language. Rather it is
determined by the process of thought itself as centered
around the building of an explanatory model of experience
and our place within it. If we are to analyze what we might
call the logical form of language we must look then to the
nature of explanation and to how explanatory coherence is
maintained. We can interpret thought and, in its active
form, reasoning, as a process of logical inference in
accordance with rules. These rules define the logic of
thought and of language.
Educational institutions have been notorious for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 244
their compartmentalization of knowledge often to the
exclusion of more integrated studies. They have frequently
been critized for a neglect of what most would agree should
be an overriding educational goal, the development of
autonomous, critical, yet creative persons. It could be
argued that this neglect is a consequence of a lack of
attention to the more generalized functions toward which
learning can be applied.
If in learning the individual is involved in a
continuous struggle to define reality and to assimilate the
knowledge which has been gained through experience then
attention must be given, at appropriate points throughout
the educational process to how this task can best be
accomplished. What exposure to philosophy in this sense
contributes to this overall goal is to provide a cross
fertilization of the various interest areas in the curriculum.
In an important sense philosophy serves an integrative
function for the more specialized disciplines by facilitating
the synthesis of higher order generalizations. It contributes
as well by bringing under examination the function of
reasoning itself and the elements that are necessary to an
effective use of reason for the accomplishment of goals in
other areas of study.
Among the practical conclusions which are indicated
by research into how we acquire, organize and structure our
thoughts is the attention which it calls to the need for a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 245
conscious examination of the process of reasoning. As we
come to understand our thought as embodying the recursive
component of successive hypothesis generation and confirma
tion, we are then more capable of synchronizing its elements.
Overall, then, a critical understanding of the general
principles involved in reasoning, to which philosophy has
and can offer a major contribution, constitutes an essential
element of what is considered a sound educational base.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SELECTED BIBLIOGRAPHY
Allan, Donald James. The Philosophy of Aristotle. 2nd ed.
London: Oxford University Press, 1970.
Anscombe, G.E.M. and Geach, P.T. Three Philosophers.
Ithaca, New York: Cornell University Press, 1963.
Aristotle. The Basic works of Aristotle. Translated by
Richard McKeon. New York: Random House, 1941.
Ausubel, David P. Educational Psychology: A Cognitive View.
New York: Holt, Rinehart, Winston, 1968.
Bambrough, Renford, ed. The Philosophy of Aristotle. New
York: Mentor Books, 1963.
Bruner, Jerome. Beyond the Information Given. Edited by
Jeremy M. Anglin. New York: W.W. Norton, 1973.
______. Human Growth and Development. Oxford: Clarendon
Press, 1978.
______. On Knowing. Cambridge: Belknap Press, 1963.
______. Studies in Cognitive Growth. New York: Wiley
and Sons, 1966.
. Toward a Theory of Instruction. Cambridge :
Belknap Press, 1967.
Chomsky, Noam. Cartesian Linguistics. New York: Harper
and Row, 1966.
246
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 247
Chomsky: Selected Readings. Edited by J.P. Allen
and Paul Van Buren. London: Oxford University Press,
1971.
Current Issues in Linguistic Theory. London:
Mouton, 1964.
. "Explanatory Models in Linguistics." in Ernest
Nagel, Patrick Suppes and Alfred Tarski. Logic,
Methodology and Philosophy of Science. Stanford:
Stanford University Press, 1962.
. Language and Mind. 2nd ed. New York: Harcourt,
Brace, Jovanovich, 1972.
. The Logical Structure of Linguistic Theory. New
York: Plenum Press, 1975.
. Reflections on Language. New York: Pantheon
Books, 19/5.
"Review of B.F. Skinner's Verbal Behavior."
Language 35 (1959): 1.
. Rules and Representations. New York: Columbia
University Press, 1980.
Copieston, Frederick. A History of Philosophy. New York:
Doubleday, 1952.
Davidson, Donald, ed. Semantics for Natural Languages.
Dordrecht - Holland: D. Reidel, 1972.
Flavell, John H. Cognitive Development. Englewood Cliffs,
N.J.: Prentice-Hall, 1977.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 248
Fodor, Jerry. The Language of Thought. New York: Thomas Y.
Crowell, 1975.
Harman, Gilbert. "Deep Structure as Logical Form." Synthèse
21 (1970): 275-297.
______. "Logical Form." Foundations of Language 9
(September 1972): 38-65.
______. "Practical Reasoning." Review of Metaphysics.
29 (1976): 431-463.
. Thought. Princeton: Princeton University Press,
1973.
. ed. On Noam Chomsky: Critical Essays. Garden City,
New York: Anchor/Doubleday, 1964.
Hebb, Donald. The Organization of Behavior. New York:
Wiley and Sons, 1949.
______. A Textbook of Psychology. Philadelphia: W.B.
Saunders, 1966.
Hulse, Steward H.; Deese, James ; Egeth, Howard. The
Psychology of Learning. New York: McGraw Hill, 1975.
Hull. C.L. A Behavior System. Westport, Ct.: Greenwood
Press, 1974.
Inhelder, Barbel. "New Currents in Genetic Epistemology."
in Bruner and Barton, ed. Human Growth and Development
Oxford: Clarendon Press, 1978.
James, William. Principles of Psychology. New York: Dover,
1950.
Kant, Immanuel. The Critique of Pure Reason. Translated by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 249
Norman Kemp Smith. New York: St. Martins Press, 1965.
Katz, Jerrold. The Philosophy of Language. New York:
Karp er and Row, 1966.
______. "The Relevance of Linguistics to Philosophy."
Journal of Philosophy 62 (1965): 590-602.
______. Semantic Theory. New York: Harper and Row, 1972,
______. The Underlying Reality of Language and Its
Philosophical Import. New York: Harper and Row, 1971
and Fodor, Jerry. "The Structure of a Semantic
Theory." Language 39 (1963): 170-210.
. and Fodor, Jerry. The Structure of Language:
Readings in the Philosophy of Language. Englewood
Cliffs, N.J.: Prentice-Hall, 1972.
. and Postal, Paul. An Integrated Theory of
Linguistic Description. Research Monograph No. 26.
Cambridge: M.I.T. Press, 1964.
Kerferd, G.B. "Aristotle." Encyclopedia of Philosophy.
Vol. I. New York: Macmillan, 1967.
Korner, S. Kant. Baltimore, Md.: Penguin Books, 1955.
Lashley, Karl. The Neuropsychology of Lashley: Selected
Papers. Edited by F.A. Beach et al. New York;
McGraw Hill, 1960.
Larkin, Meriam Therese. Language in the Philosophy of
Aristotle. The Hague: Moulton, 1971.
Leiber, Justin. Noam Chomsky: A Philosophical Overview,
Boston: Twayne, 1975.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 250
Nagel, Ernest. "Linguistics and Epistemology." In On Noam
Chomsky: Critical Essays. Edited by Gilbert Harman.
Garden City, New York: Anchor/Doubleday, 1964.
Eaton, Herbert J. Kant's Metaphysic of Experience. New York;
Macmillan, 1936.
Partee, Barbara. "Linguistic Metatheory." In On Noam
Chomsky: Critical Essays. Edited by Gilbert Harman.
Garden City, New York: Anchor/Doubleday, 1964.
Piaget, Jean. The Child's Conception of Physical Causality.
Translated by Marjorie Gabain. New York: Harcourt
Brace, 1930.
______. The Child's Conception of Space. Translated by
F.J. Langdon & J.L. Lunzer. London: Routledge and
Kegan Paul, 1956.
______. The Child's Conception of Time. Translated by
A.J. Pomerans. New York: Basic Books, 1969.
. Genetic Epistemology. Translated by Eleanor
Duckworth. New York: Columbia University Press, 1970.
. Insights and Illusions of Philosophy. Translated
by Wolfe Mays. New York: World Publishing Co., 1971.
. Psychology and Epistemology. Translated by
Arnold Rosin. New York: Orion Press, 1971.
. Structuralism. Translated by Chaninah Maschler.
New York: Harper and Row, 1970.
and Inhelder, Barbel. The Growth of Logical
Thinking from Childhood to Adolescence. Translated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 5 1
by Anns Parsons and Stanley Milgram. New York: Basic
Books, 1958.
and Inhelder, Barbel. The Growth of Logic in the
Child. Translated by E.A. Lunzer and D. Papert.
London: Routledge and Kegan Paul, 1964.
Rotman, Brian. Jean Piaget: Psychologist of the Real."
Ithaca, New York: Cornell University Press, 1977.
Scinto, Leonard. "Review of Noam Chomsky's Reflections on
Language." in On Noam Chomsky: Critical Essays. Edited
by Gilbert Harman. Garden City, New York: Anchor/
Doubleday, 1964.
Searle, John. "Chomsky's Revolution in Linguistics." in
On Noam Chomsky: Critical Essays. Edited by Gilbert
Harman. Garden City, New York: Anchor/Doubleday, 1964.
Slobin, D.I. "Imitation and Grammatical Development in
Children." in Contemporary Issues in Developmental
Psychology. Edited by Endler, et al. New York: Holt,
Rinehart, Winston, 1968.
Smith, Norman Kemp. A Commentary of Kant's Critique of Pure
Reason. 2nd ed. New York: Humanities Press, 1923.
Society for Research in Child Development. Cognitive
Development in Children. Chicago: University of
Chicago Press, 1970.
Spence, K.W. Behavior Theory and Conditioning. Westport,
Ct.: Greenwood Press, 1978.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 252
Stigen, Anfinn. The Structure of Aristotle's Thought.
New York: Humanities Press, 1966.
Strawson, P.P. The Bounds of Sense. London: Methuen, 1966.
______. Individuals: An Essay in Descriptive Metaphysics.
New York: Doubleday, 1959.
Tolman, Stephen. Purposive Behavior in Animals and Man.
New York: Irvington Press, 1967.
Vendler, Zeno. "Reply to J. Katz 'The Relevance of Linguistics
to Philosophy.'" Journal of Philosophy. 62 (1965):
602-605.
Walsh, W.H. "Kant." in Encyclopedia of Philosophy. Vol. 4.
New York: Macmillan, 1967.
Wittgenstein, Ludwig. Philosophical Investigations.
Translated by G.E.M. Anscombe, 2nd ed. New York:
Macmillan, 1958.
______. Tractatus Logici-Philosophicus. Translated by
D.F. Pears and B.F. Me Guinness. London: Routledge
and Kegan Paul, 1921.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.