There’s a nice connection between Wiggins‘ paper and Davidson’s “Derangement”. Wiggins notes that Davidson, drawing on a suggestion from John Foster, convinces himself that what semanticists are trying to do is specify knowledge of a certain sort: knowledge that, given a few objective contextual parameters, would suffice for interpretation. What Davidson progressively comes to see is that no pre-existing and widely shared knowledge base will turn the trick… So he abandons the original project. Derangement, in short.
But there are two natural alternatives to the instrumentalist turn that Foster provoked. (Wiggins thinks the first of them is what Foster really had in mind all along.): 1) Describe the language as it is, rather than describing knowledge of language. 2) Describe the mental mechanisms underlying speech and comprehension as they actually function, without laying down as a condition that they will suffice for decoding literal meaning. Either way, one eschews instrumentalism and preserves the study of languages.