On the connection between a proper view on compositionality and strict finitism: Weyl’s idea that universal formulae ranging over the natural numbers are not statements, but ‘Anweisungen auf Urteile’ (as cited in Marion, Wittgenstein, Finitism, and the Foundations of Mathematics, chapter 4) is congenial. The assumption that there exists an infinite number of sentences that satisfy a pattern in quite a similar manner can be seen as the result of mistaking rules with the set of all their instances. As actual sentences of a language these simply do not (yet) exist.
Martin Stokhof from: Interpretation date: fall 2001
Davidson’s analysis in ‘A Nice Derangement of Epitaphs’ marks a goodbye to the idea of a compositional (recursive) theory of meaning. Why? To answer this question we must answer another one first: Why did we want such a theory in the first place? The answer seems to be because we wanted an a priori characterisation of semantic competence, i.e., an account which deliberately disregards factual use. For such an abstract, non-situated analysis the potential infinity of language constitutes a major problem. In other words, it is the assumption of a pure, individual-based language which creates the problem, for which compositionality provides a solution (one that is intuitive, though arguably not the only one possible). If we drop this assumption this argument for compositionality at least vanishes (there may be other ones). Of course, another issue takes its place: How are we able to create ever new passing theories? Here Wittgenstein’s rule-following considerations come to bear directly on Davidson’s approach. It seems that Davidson has managed to back himself into a corner by not dropping the individual bias: meaning tends to get locked up inside each individual speaker, and a serious threat of semantic solipsism arises.
You must be logged in to post a comment.