In Doing Without Concepts, Edouard Machery sets himself the ambitious task of eliminating the concept of concept from psychological research and by implication — though he does not state this directly, and his principle target is indeed the psychology of concepts — from theoretical research more generally.

The book is more than usually well structured; the argument progresses from chapter to chapter in clear and inexorable fashion; the references are generally meticulous.  (I find myself frequently annoyed with writers who do not bother with page references on anything but direct quotes.)

The argument can be summarized like this:

·         Philosophers and psychologists are talking about entirely different things when they talk about concepts, so any attempt by philosophers of concepts to respond to psychologists of concepts is a mistake.

·         For any category of x one can name, one has several concepts of x, and these concepts have little in common with one another.

·         Some of these concepts are prototypes, some are exemplars, and some are theories, and these three are entirely different things, employed in distinct cognitive processes.  Therefore any attempt by psychologists to decide between them, or indeed to come up with a unified theory of concepts, is a mistake.

How does this play out?  “When people are judging that it is simply mistaken to assert that tomatoes are vegetables, they are entertaining one of their coreferential concepts of tomato.  For these people, ‘tomato’ might express by default this concept.”  (p. 73)  Some other person might hold both “tomatoes are fruits” and “tomatoes are vegetables” without necessarily entertaining any contradictory belief — because the two propositions draw on different concepts of tomato.

Some of his statements I find mystifying.  For example:  “Once at the center of philosophy, the philosophy of concepts has now been marginalized, maybe because for a few years now, it has been stalled.” (p. 3)  My own research area is in philosophy of concepts, and I am not aware of any general sense in the literature that the field is either marginalized or stalled.  To the contrary, there is continuing excitement over the friendly competition between Jesse Prinz’s proxytypes theory and Jerry Fodor’s informational atomism.  Another popular contender is Peter Gärdenfors’ conceptual spaces theory of concepts, on which I am basing much of my own work.  A recent Journal of Consciousness Studies special issue (September-October 2007) was devoted to the interplay between concepts and consciousness.

Likewise there is his claim on the same page (unfortunately not backed by any references) that “philosophers have typically found the psychological theories of concepts to be wanting and, instead of contributing to their development, have discarded them.” Again, I would expect to be aware of this.  Instead, my experience is that most philosophers of concepts (including Prinz and Gärdenfors but, admittedly, not Fodor) richly draw upon psychological theories of concepts and adopt large portions of them for their own.

Some things I think he gets clearly right: e.g., that hybrid theories of concepts look embarrassingly ad hoc, or that much of our introspectible conceptualization resembles rule-based theorizing (of the kind so beloved of Good Old-fashioned AI, and so denigrated by its successors).

I realize that philosophers of far greater reputation and experience than I — people like Eric Margolis and Jesse Prinz — have written in glowing terms about this book, and I do not wish to detract from Machery’s accomplishments.  Still on balance I find myself disappointed, for I don’t see Machery delivering any knock-down arguments, despite the pretensions otherwise (he frequently says “I have shown” where I might have written “I have argued”).  I should say at this point that, in my own doctoral research, I am trying to put forward a unified theory of concepts, albeit from philosophy rather than psychology.

I find myself satisfied with neither his definition of psychological concepts nor that of philosophical concepts.  Psychological concepts he defines as “those bodies of knowledge that are stored in long-term memory and that are used by default in the processes underlying most, if not all, higher cognitive competences when these processes result in judgments about the referents of these concepts”. (p. 4)  Philosophical concepts he defines by their possession conditions:  “Having a concept of x is being able to have propositional attitudes about x as x.” (p. 32)

I am a philosopher, not a psychologist; but the psychological definition strikes me as too broad, the philosophical definition too narrow. Of course philosophers of concepts often do talk about concepts in terms of propositional attitudes.  But to restrict them to these terms is to deny the rich tradition, going back to Gotlob Frege and developed in diverse ways through the work of people like Gareth Evans, Christopher Peacocke, and Alva Noë, of concepts not as some kind of abstract objects but as abilities.

Of course there is a problem within philosophy, never mind between philosophy and other domains, of people talking past each other. But Machery seems too keen to emphasize the differences, too quick to assume incommensurability.

I have the same sort of concerns with Machery’s main thesis, which relies upon the differences between prototypes, exemplars and theories — especially after he acknowledges that none of them have anything like non-tendentious definitions, and goes on to say that “the core tenets of the theory paradigm are considerably vaguer than the core tenets of the prototype and exemplar paradigms despite a few groundbreaking theoretical articles….” (p. 101)

Machery defines a prototype as “a class is a body of statistical knowledge about the properties deemed to be possessed by the members of that class” (p. 83) and an exemplar as “a body of knowledge about the properties believed to be possessed by a particular member of a class.” (p. 93)  The problem is — as Machery acknowledges in a brief footnote — that “prototype” is frequently used to refer to the most typical actual or ideal member of a class (i.e., extrapolated from actual members). But an exemplar can be described appropriately as the most typical actual member. So again, Machery is emphasizing the differences and downplaying the overlap.

All of that would be fine and not necessarily tell against Machery’s point if it were clear which version of prototype, exemplar, and theory theory he is endorsing, so it could also be clear (hopefully) whether the studies he cites support those interpretations, and in particular the conclusions he wishes to draw about prototypes, exemplars, and theories being entirely different kinds of concepts. Unfortunately, he says, “I have not tried to determine which version of the prototype paradigm, which version of the exemplar paradigm, and which version of the theory paradigm is correct.  This is a project that is beyond the scope of this book and that is best left to psychologists.” (p. 196)

In the end, my feeling is that psychologists and philosophers have more in common than Machery allows — even though they often do use different language, and their focus of concern is often different.  It is true that psychologists — particularly developmental psychologists — tend to emphasize the private and personal aspects of concepts. Philosophers frequently emphasize the public and lexicalized aspects.

As for the presumed differences between concepts as prototypes, concepts as exemplars and concepts as theories, there are plenty of examples one can give of things that look entirely different depending on what perspective one is taking at the time.  Consider the way in physics that both matter and energy can be treated as wave or particle.  I believe there is a critical distinction to be made between concepts as we reflect upon them as concepts, and concepts as we possess and employ them non-reflectively — one that goes a long way toward meeting Machery’s concerns without postulating different entities, only different perspectives.  But of Machery’s central thesis, I, at least, remain a contented skeptic.

 

© 2009 Joel Parthemore

 

Joel Parthemore is a third-year DPhil student studying theories of concepts at the University of Sussex in Brighton, UK. He is a member of the Philosophy of AI and Cognitive Science research group in the Department of Informatics. In his spare time he plays with Linux computer systems. You can find him online at http://www.parthemores.com/research/.

Keywords: philosophy