Derrida's critique of Austin isn't an unhelpful one, since it evinces just the sort of metaphysical thinking that Austin and Wittgenstein are concerned to think past. Basically, Derrida's claim in "Signature Event Context" is that quotation can't be "parasitic" on performative utterances because to engage in such an utterance requires a citation or quotation of the appropriate formula. (Like everyone else, and I do mean everyone Derrida neglects Austin's distinction between "hollow" and "void" performative utterances, but that's for another post.) So performative utterances are logically dependent on a practice of quotation.
This is wrong. It would be far better to say that the very idea of quotation arises out of performative utterances. Performative utterances -- or Wittgenstein's language games -- come first. It may be possible to formulate the rules of such games, but those formulations are descriptive, not prescriptive, attempts to formalize what we do. Performative utterances, and moves in language games more generally, are practices before they are more-or-less-successful attempts to be adequate to some set of rules governing them.
For Wittgenstein these are practices which arise out of what he calls "agreement in forms of life." The supersubtle mechanisms by which such agreement could evolve (see, for example, Robert Axelrod's Evolution of Cooperation) can't and won't have presupposed something so crude as Derridean citation. Quotation in the Paris-is-Burning Jennie Livingston sense, possibly, where quotation is something closer to biological mimesis as Roger Caillois understands it: interaction, gaming, self-exposure to the spatial world. But this has almost nothing to do with Derridean formalism.
But what we could see, and say, is that citation, verbatim quotation, the idea of the verbatim, Quinean inscription, arise from performatives. Formulae are fossilized performatives, and the idea of a formula (which is of course manifold in rituals and rites themselves dependent on prior belief in the performative power of utterance, a belief raised to a magical pitch) can give rise (see Homer and Milman Parry) to the idea of quotation itself.
So the great, Emersonian literary device of quotation is secondary to performative utterance. That's what makes it literature: the evocation of a fictive world, where the performances aren't real, and all the more haunting for that reason. Maybe I should say something about the hollow vs. the void. In Austin hollow performatives are those which are not "meant," are those which the utterer performs without any intent to back them up. Void performatives are those which have no standing, no matter how passionately they are uttered. Fiction or the literary is the region, then, of the void, not of the hollow. The poet nothing lieth because he nothing affirmeth, but instead gives us some sense of what the void is, next to which our loquacious selves are so precariously perched (to allude to Kenneth Burke).