grammar's autonomy and phenomenology's revelations
Let's first bring the 'autonomy of grammar' claim into clear focus. ('Grammar is not accountable to any reality. It is grammatical rules that determine meaning (constitute it) and so they themselves are not answerable to any meaning and to that extent are arbitrary.' PG 184.) Now, to take him at his word involves acknowledging that Wittgenstein is not in the business of proposing philosophical theses. The 'autonomy of grammar' claim, then, isn't well understood as asking us to accept that whilst we understand what it would mean for 'grammar' (i.e. rules of language use) to be justifiable by reference to something, to be beholden to something for its correctness, what in fact we have to grasp is that grammar is not so justifiable or beholden. What Wittgenstein is instead doing is taking a stand against a certain picture of what a grammatical truth is and pointing out the pitfalls that come along with that picture.
The picture in question would have us compare a 'grammatical' truth to an empirical truth. It makes sense to ask for the truth conditions of an empirical truth. I claim there are five snails in the garden - this is true iff the garden happens to contain five snails. It is true in virtue of something. The claim can be justified by reference to that something. But when we have a 'metaphysical' or 'necessary' or 'grammatical' truth - 2+2=4, humans are mammals, there are 4 primary colours - there is no such 'in virtue of' in play. '2+2=4' is neither true in virtue of (just picture it, pretend the following words mean something...) the 'logical superstructure of' the universe (big 'R' Realism), nor in virtue of what humans think or do (Idealism) - it is not the kind of truth which is intelligibly paired with an 'in virtue of'. A necessary truth is instead a rule or a human convention - and (in case we're now tempted by 'conventionalism') since it is a convention it also can't be true in virtue of a convention! ('One is tempted to justify rules of grammar by sentences like "But there really are four primary colors". And the saying that the rules of grammar are arbitrary is directed against the possibility of this justification, which is constructed on the model of justifying a sentence by pointing to what verifies it.' Z 331)
Now, what about the existential-phenomenological ambition of doing justice, in our philosophy, to the reality of our experience of 'the things themselves'? Simon Glendinning offers us this: 'What characterises an investigation in phenomenology is a work of convincing words which, in an age dominated by science, aims to cultivate and develop your capacity faithfully to retrieve (for) yourself (as from the inside) a radically re-vis(ion)ed understanding of yourself and your place in the world and with others.' Anthony Rudd has it that 'expressive perception is the ability to perceive patterns in such a way as to recognise, in and through the phenomena which one sees as related, what it is that they all express. ... The phenomenologist - like the ordinary language philosopher appealing to our sense of how we would use words - is not providing a chain of arguments; she is asking us to reflect on our experience and to consider "isn't this how it is?"' ... The later Heidegger, as Rudd understands and quotes him, understood our phenomenological task to be one of answering without ourselves, via a 'responsive thinking' which involves emotional attunement, to the appeal of the world's being, taking 'the step back from the thinking that merely represents - that is - explains - to the thinking that responds and recalls.'
Now, what is this 'recognising', 'responding' and 'recalling' which is not 'representing'? This offering of recognition shows itself, I think, in two ways. First in a refusal of such metaphysical thought as would assimilate the being of, say, an emotional attitude to the being of an object. (Wittgenstein, referring to King Lear, mentioned 'I will teach you differences!' as a possible motto for the Investigations.) This isn't a matter of providing a positive description of what it is to be an emotional attitude in some special philosophical vocabulary - the finding, for example of some genus and species to which it belongs. Rather it's a matter of noting differences between different concepts - here we would say this, but there we wouldn't; here such and such an inferential step would be valid but there it wouldn't be, etc. Second it's an offering of expressive voice to the phenomenon in question - allowing it, through 'poiesis', to body forth in language in its own distinctive way. This is not a matter of using language to represent an emotional attitude but rather of lending the attitude one's tongue, so to speak, elaborating and honouring its ownmost idioms.
Now, I think, we can see how Wittgenstein's conception of grammar as arbitrary and Heidegger's poietic phenomenology are complementary. Neither philosopher imagines that our concepts could be justified by reference to the facts - in fact it is in this way that the idea of arbitrariness and non-representationality can be seen to be, far from in conflict, very much of a piece. And the very idea of justification by facts here is absurd since the facts could only ever be referenced by using the concepts in question, so any putative justifications would simply be circular. The only fidelity at play is expressive fidelity; authenticity not accuracy is here the only issue. Is the expression a (not representationally, but expressively) true one? Does an expression talk over a phenomenon or instead constitute that phenomenon's voice? Does it disclose or obscure?