Rich Lusk

“Making analyticity depend on logic raises some difficult problems. We must ask about the status of the logical laws used to prove analyticity. Are these laws analytic or synthetic? It is unlikely anyone would claim that laws of logic are derived from experience. Even die-hard empiricists like Hume have recognized the a priori character of logical laws. Besides, if these laws are derived from experience, they are contingent and dependent on fact and therefore cannot serve as a basis for making a distinction between analytic and synthetic judgements. On the other hand, if they are analytic, all we have done is explain analyticity in terms of analyticity. In this case, to say that analytic truths are those that are purely logical is merely to say that analytic truths are those that are analytic! This is hardly helpful.”
― Rich Lusk

“As Quine points out, when we are told that the judgement 'All bachelors are unmarried males,' is analytically true, we must immediately ask how we came to know this. Somone may appeal to a dictionary, claiming that 'unmarried' and 'male' make up the definition of the word 'bachelor.' However, it is simply not accurate to view a dictionary as a book of analytic truths. First of all, as Quine goes on to show, the work of the lexicographer is largely that of an empirical scientist. The lexicographer bases his definitions on synonymy observed in a natural language. Languages are learned through experience; we are not born with a built-in a priori vocabulary. Children (usually) learn words not by looking them up in a dictionary, but by listening to others communicate. In other words, the lexicographer works on a Wittgensteinian principle, generally determining the meaning of a word from its observed use in a language. But can the fallible, limited observations of dictionary writers serve as the basis of absolutely certain analytic truths? Why should we accept a dictionary as law? Lexicography is far from being a perfect science...Clearly then, definitions cannot serve as the basis for analyticity. Besides, synonymies are the result of observed patterns in a language. Because they are based on empirical factors, on any normal account of the analytic-synthetic distinction, synonyms would qualify as synthetic judgements subject to revision. Thus, trying to root analyticity in synonymy ends up making analytic truths depend on synthetic truths! The observations of a lexicographer are, of course, fallible, and therefore cannot provide the key to certainty that analytic judgements are supposed to provide. Any account of the analytic-synthetic distinction based on word meanings is bound to fail.”
― Rich Lusk