Concepts

Stich:

There is a venerable tradition ... which assumes that the concept or mental structure underlying the use of most predicates is actually a mentally represented definition - a set of necessary and sufficient conditions. In deciding whether or not a term applies to a given case, this "classical view" maintains, we are either consciously or (more typically) unconsciously determining whether the case at hand satisfies the conditions of the definition. If it turns out that there just are no definitions for most terms, then obviously the classical account of the structure and use of concepts will have to go. In recent years, there has been a growing realization that the classical account of concepts is in deep trouble, and a number of interesting alternatives have been proposed. Perhaps the best known of these are the prototype and exemplar accounts of concepts developed by Eleanor Rosch and her associates. On the prototype theory, concepts are weighted lists of features that are characteristic of the most typical members of the category that the concept picks out. The list will generally include lots of features that are not necessary for category membership. On the exemplar story, concepts are, in effect, detailed mental descriptions of particular members of the category. Thus, for example, the concept underlying your use of the word 'dog' might include detailed descriptions of Lassie and Rin Tin Tin. In determining whether to categorize something as a dog, this theory maintains, you assess the similarity between the target and the various exemplars stored in semantic memory ... Fodor has proposed a very different alternative to the classical account of concepts. On his view, the concepts that underlie most of our one-word predicates have no structure at all - or at least none that is relevant to the semantic properties of the concept. Of course, if this is right it is very hard to see how these concepts might be learned. And that's just fine with Fodor, since he thinks they are all innate (see Fodor 1981a). There are [in fact] lots of interesting theories about concepts on the market which are compatible with (and which might well explain) the finding that most of our concepts appear to have no intuitively acceptable definitions.

Stephen P. Stich, Deconstructing the Mind, New York, 1996, OUP, 173.

END