ABSTRACT
We have seen that Dummett’s criticism of Frege’s use of his sense/reference distinction was based on proving that the category most convenient for traditional representational theory – proper names – was dependent for its articulation upon an essentially anti-realist (hence anti-traditional) principle. We have also seen that Derrida’s criticism of Husserl’s phenomenology was analogous to Dummett’s criticism of Frege in this respect. The determination to avoid psychologism and formalism, the phenomenological reduction – all these are frustrated when the realist dogma, or one of its equivalents, comes into play. But at this point it is crucial to emphasise that what has been attacked by Dummett and Derrida was not Frege and Husserl’s philosophical endeavour per se. Not even the search for the a priori of language is at stake. Rather, as Derrida says, it is the fact that Husserl’s grammar (and Frege’s grammar is no different from it in this respect) ‘is not sufficient to cover the whole field of possibility for language’s a priori’ (SP: 8), since it is ‘interested in language only within the compass of rationality, determining the logos from logic’ (SP: 8, my emphasis). This last turn of phrase is revealing. Derrida, in a statement that directly associates him with Nietzsche, notes the deep-rooted tendency in Western thought to equate logic with logos – whereas the scope of logic is actually much more inclusive:
When we speak of the purely grammatical, we mean that system of rules which enables us to recognize whether or not a discourse is, properly speaking, a discourse. Speech, to be sure, must make sense; but do falsity and the absurdity of contradiction (Widersinnigkeit) necessarily make it unintelligible? Do they necessarily deprive discourse of its experienced and intelligible character,
thereby rendering it sinnlos? This grammar concerns only the logical a priori of language; it is pure logical grammar.
