LINGUIST List 3.817

Sat 24 Oct 1992

Disc: Best Theory Fallacy and Unification

Editor for this issue: <>


  1. Avery Andrews, Best Theory Fallacy

Message 1: Best Theory Fallacy

Date: Wed, 21 Oct 92 09:17:39 ESBest Theory Fallacy
From: Avery Andrews <>
Subject: Best Theory Fallacy

The fallacy is that the use of a uniform mechanism (e.g. global derivational
constraints) for a wide range of purposes does not necessarily embody
any useful claim about language. It *might*, but this depends on the
particular properties of the mechanism. What I think that unification
has going for is (a) it has reasonable mathematical properties (those of
`and' in logics where consistency is decideable, I'd suggest). (b)
it corresponds to the idea that language makes extensive use of
`long components' (a Harrisian term, I think I recollect) that get
manifested at multiple points in the grammatical structure, with the
manifestations having to give consistent information about the long

But this could all be just wrong. Maybe movement, co-superscripting , etc.,
really are so different in their empirical properties that it's pointless
to try to reduce them to a common formal basis.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue