(I borrow my title from the literary criticism of James McAuley, part of whose work is featured here.)
Nietzsche held that our "beliefs" in God, Being, causality, and much else, were functions of grammar. He seemed to consider this a bad thing. It took Wittgenstein to invert this move and show us how to read the recourse to de facto assumptions, i.e., grammar, in a less pejorative way.
E.g.: the experience of feeling moral compulsion or recoil simply means the experience of a reality; not a phantom, not a preference, not a sublimation, but a simple fact. I don't pretend that the status of this fact is easy to explicate -- it is hard, like the status of Hamlet's psychology, or whether there is a series of twenty-five 0's in the decimal expression of pi, or the nature of qualia -- but when someone says "it's wrong to steal," this does not mean "I don't like it when people steal (from me)" or "In my culture, we consider stealing something not done." The grammar of "It is wrong to ---" is different from that of these other claims. This grammar is not in itself sufficient to establish the case for moral realism but it is enough to establish that naive attempts to dissolve such realism into something else are, well, naive.
Similarly, I am not a metaphysical Monist (despite great temptation in that direction), because I do not see how monism can account for the meaning of the experience of encounter between I and Thou. When I look into my friend's or my enemy's eyes or at the height of a mountain or a sum on a piece of paper, I encounter something, someone, not-me. This experience has a "grammar" to it that is simply not accounted for by any Thou-art-That-ness. Again, this is not in itself sufficient to refute monism, but all I am trying to establish here is that the experience of encounter is not corrigible in the same way that Lewis Carroll's "Mad Gardener's Song" illustrates:
He thought he saw an elephantHowever implausible such a realization might be, this is not the same as thinking one meets another person and having this meeting dissolve into being (with) oneself (however expanded a sense of oneself this entails). In the former case, one meeting has been replaced by another; in the latter, meeting itself has vanished. The point here is that "meeting" has a grammar, and Monism claims that this grammar is (ultimately) meaningless.
that practiced on a fife.
He looked again and found it was
a letter from his wife
One could multiply examples: The Anselmian, a.k.a. "ontological," argument for God's existence, also appeals to grounds one might construe as grammatical. The reason why people have such difficulty accepting quantum mechanics is that it contradicts certain grammatical conditions about what we mean when we use certain terms -- the ordinary grammar of conditions tends to imply a kind of "object permanence" which apparently fails at the sub-atomic level, so that it becomes more and more difficult to meaningfully sustain talk about what an electron, say, is doing when one isn't looking. Even the appeal of poetry often turns upon the revealing of a grammatically permissible but surprising use of language:
A stranger always hasOr:
his homeland in his arms (Nellie Sachs, "Someone comes")
is an ever-hostile tree
across the border (Bei Dao, "Midnight Singer")
Bark be my limbs my hair be leafThe acknowledged prince of such effects is of course Celan:
Bride be my bow my lyre my quiver (Susan Howe, "Pythagorean Silence")
As one speaks to stone, likeIt is not sufficiently recognized that Meillassoux's initial argument against correlationism is such a grammatical argument. This will probably raise some eyebrows. But what the argument amounts to is this: the correlationist cannot understand science in science's terms, but must always add a "codicil," implicitly or explicitly: "The formation of the solar system occurred x billion years ago, for us." This "for us" completely re-frames the scientific assertion in a way that makes it mean something else, something so different as to be almost the opposite, Meillassoux insists. In the grammar of science, an assertion means that what it asserts happens, as it were, on its own. To append the correlationist codicil is to invert the meaning of whatever scientific assertion it is attached to.
from the chasm, from
a home become a
sister to me... (Paul Celan, "Radix, Matrix")
I am still thinking over what this grammatical re-framing of Meillassoux might mean for his overall argument. As many have observed, the argument about "ancestral" assertions is not the strongest part of After Finitude (the essential portion of the book is chapter 3). But the prominence of the role of grammar, even for a thinker who is widely taken to have hurled the gauntlet down before the Linguistic Turn, ought to remind us that it is no simple matter to shake off the shadow of language, or indeed of the human practices of which it is a paradigm.
Samuel Johnson's refutation of Berkeley usually raises a condescending smile, even though few people really believe Berkely these days. But what Johnson was getting at was that there is a grammar to claims about reality -- a grammar that is often implicit, and, as Wittgenstein saw, extends to practice in general.