Furnishing the Mind

            Concepts, it is said, are the vehicles of thought—the basic building blocks of our mental lives.  But what exactly are concepts?  In Furnishing the Mind Jesse Prinz formulates an impressive new theory of the nature of mental concepts.  It is too early to tell whether his theory is thoroughly successful, but he has doubtless developed an interesting and important theory of concepts that will remain at the forefront of the concepts debate for some time.  Prinz’s sensitivity to the nuances of both philosophical argument and psychological data results in a superb example of the complexity and sophistication that can be achieved in cognitive science at its interdisciplinary best. 

            Prinz’s book is written for scholars in philosophy and the cognitive sciences.  Although the first three chapters that survey the contemporary debate about concepts are very readable and do not presuppose any prior acquaintance with the debate, the uninitiated would find the remaining chapters rather difficult to follow. 

            The first three chapters of the book survey the major philosophical and psychological theories of concepts and the standard problems facing each of them.  Prinz tells us that some philosophers (e.g., Aristotle, Locke, Berkeley) thought that concepts were mental images.  Imagism provides a straightforward explanation of how concepts are acquired perceptually, but it seems that we possess many concepts (e.g., justice, truth, democracy) that have no corresponding perceptual images.  Other philosophers (e.g., Plato) have thought that concepts can be identified with sets of necessary and sufficient conditions that define the concepts in question.  Bachelor, for example, can be defined as ‘unmarried male.’  Although this view has had a long and popular career in philosophy, very few concepts can be given adequate definitions.  Moreover, people do not seem to rely upon definitions while acquiring concepts or forming categorization judgments. 

            Prinz also describes three psychological theories of concepts that have been widely discussed in recent decades: the prototype theory, the exemplar theory, and the theory theory.  According to the prototype theory, information about categories is stored as representations of the average or most typical instances of those categories.  An encountered object is judged to belong to a category if its representation is more similar to the prototype representation of that category than to any other.  Prototype theory can explain many facts about how ordinary subjects think about various categories and their instances, but it seems to have difficulty explaining how we are able to reason about atypical instances of certain categories and how people can possess the same concept if they do not always form the same prototype representation.

            According to the exemplar theory, concepts are not stored as representations of typical or average members of certain categories.  Rather, representations of particular encountered instances are stored.  Although exemplar theory can explain certain facts about categorization, it has difficulty explaining how we can possess concepts whose instances we have not experienced and how people can possess the same concept if they do not store the same exemplars.  According to the theory theory, concepts are mini-theories.  Concepts are sets of explanatory principles for making sense of different aspects of the world.  If concepts are theories, the theory theory can explain why possession of a concept always seems to be accompanied by certain kinds of theoretical knowledge.  But if theories are themselves composed of concepts, serious problems face the identification of concepts with theories. 

            Prinz aims to develop a theory that:

a) avoids the difficulties facing the dominant philosophical and psychological theories of concepts;

b) preserves the strengths of the dominant philosophical and psychological theories of concepts;

c) comports well with and explains important philosophical intuitions about the nature of concepts; and

d) comports well with and explains important psychological data concerning categorization.

In short, Prinz’s rather ambitious goal is to formulate a better theory of concepts than has ever been formulated before.

            In an age where nativism about concepts (i.e., the view that many concepts are innate) is taken as an indisputable given, Prinz defends a version of ‘concept empiricism’—the view that all concepts are perceptually derived.  According to Prinz’s theory, which he calls ‘proxytype theory,’ concepts are copies of perceptual representations that are stored in long-term memory and can be activated in working memory.  Prinz (pp. 150, 189) writes,

Tokening a proxytype is generally tantamount to entering a perceptual state of the kind one would be in if one were to experience the thing it represents....  [T]hinking is a matter of redeploying perceptual representation off-line.

Since one would enter into one sort of perceptual state when experiencing a prototypical member of a category and a different sort of perceptual state when experiencing an atypical member of a category, different kinds of proxytypes will be constructed in working memory on different occasions.  Context will determine which proxytypes get activated.  Sometimes it will be best if a proxytype represents the general tendency of a category (i.e., a prototype); at other times a unique, perhaps atypical proxytype (i.e., an exemplar) may be best. 

            Leading psychological theories of concepts focus on only one kind of mental representation that can be called up on the basis of stored information.  However, perceptually derived long-term memory networks encode a wide variety of information about perceived objects.  These networks can give rise to many different representations for a single category that can function in working memory.  By claiming that concepts should be identified with the full range of these representations, Prinz is able to incorporate the strengths of each of the leading theories of concepts into a single account of concepts.

            Prinz’s claim that all human concepts are copies (or combinations of copies) of perceptual representations may sound as if it is incompatible with any form of nativism about concepts, but he claims this is not so.  Prinz defends a weak form of antinativism but allows for the possibility that some perceptual representations may be innate.  Because of the widespread acceptance of various forms of nativism, antinativist arguments do not strike many contemporary researchers as very plausible.  As Prinz (p. 189) says, “Cognitive science was born out of arguments for nativism.”  Consequently, Prinz does not want to make his concept empiricism rely upon antinativist arguments.  Instead, he tries to motivate his position by appealing to its explanatory power. 

            A key explanatory virtue of proxytype theory is that it obviates the need to postulate separate mental representations for perception and cognition.  Prinz thinks that, “Once we have postulated a certain class of representations for a theory of perception, it is cost effective to see whether those same representations can be used in a theory of cognition” (p. 122).  This part of Prinz’s work is a development of the ideas of Lawrence Barsalou.

            Prinz’s book is sometimes challenging to read because of the amount of time he spend contrasting his own view with others and because of the amount of empirical information he includes in his exposition.  He wants to show that proxytype theory overcomes the weaknesses of other theories of concepts, that it comports well with a wide variety of psychological data, and that it can be defended against a range of standard objections to empiricist theories.  In order to accomplish all of this in a single book, Prinz does not simply spell out his own theory.  He introduces his theory one part at a time, dialectically playing each part off of other possible positions.  Prinz’s explanatory strategy requires a lot of concentration on the part of the reader, but it is well worth the effort.  I highly recommend his book to anyone engaged in research on concepts. 

© 2002 James R. Beebe

James R. Beebe teaches philosophy at Louisiana State University. 

Link: Jesse Prinz home page