Precis of Dividing Reality

Abstract

What I call the Similarity Principle says that a (general) word ought to denote a class of things that are more similar to each other than to other things (DR, p. 36). closely related formulation, which I'll here take to be equivalent, is that a word ought to denote a class of things having something in common with each other that they don't have in common with other things. The Similarity Principle is an example of an intuitively rational constraint on the lexicon of a language. If we imagine such strange words as gricular (green or circular) or carple (car or apple), we feel intuitively that there would be something absurd about having such words in our language, and one explanation of this-indeed the explanation that is likely first to come to a person's mind in some rudimentary form-is that these words violate the Similarity Principle. My book is a study of rational constraints on lexicons, constraints like the Similarity Principle. The central problem of the book is to articulate and (especially) try to assess such constraints. I call this the division problem, for it is a problem about how words ought to divide up reality. Other putative constraints examined in the book have to do with projectibility, causality, natural properties and things, salience, important properties, and economy. The serious difficulty of the division problem emerges when we focus on examples in which ordinary words are replaced by words like gricular and carple in such a manner as to yield strange languages that have the same descriptive content as English; that is, any sentence in English is (a priori necessarily) equivalent to a sentence in the strange language, and vice versa. These languages strike us as intuitively absurd, but this seems to mean that there are lexical constraints with the puzzling effect of disqualifying languages in which we could say everything we ordinarily say (modulo equivalence). It is difficult to defend such constraints. This characterization of my project, I have found, is sometimes a philosophical turnoff. A study of rational constraints on lexicons? What's next,