James Ryerson in The New York Times, 12 December 2008:
A report on DFW’s very analytic deconstruction of a paper by Richard Taylor which draws metaphysical conclusions from a simple thought experiment.
In summary,

…his own analysis of the problem “seems to warrant the following conclusion of our own: if Taylor and the fatalists want to force upon us a metaphysical conclusion, they must do metaphysics, not semantics. And this seems entirely appropriate.”

Sometime in his later college years, Wallace became troubled by a paper called “Fatalism,” first published in 1962 by a philosopher named Richard Taylor. The fatalist contends, quite radically, that human actions and decisions have no influence on the future. Your behavior today no more shapes events tomorrow than it shapes events yesterday. Instead, in a seemingly backward way, the fatalist says it is how things are in the future that uniquely constrains what happens right now. What might seem like an open possibility subject to human choice — say, whether you fire your handgun — is already either impossible or absolutely necessary. You are merely going with some cosmic flow.

Perhaps most counterintuitively, the fatalist argues that this topsy-turvy doctrine can be established by mere reflection on the simple logic of propositions about the future. If I fire my handgun, one second from now its barrel will be hot; if I do not fire, one second from now the barrel will not be hot; but the proposition one second from now the barrel will be hot is right now either true or false. If the proposition is true, then it is the case that I will fire the gun; if it’s false, then it is the case that I won’t. Either way, it’s the state of affairs in the future that dictates what I will or won’t do now.

Obviously, there is something fishy going on here. But Taylor’s highly sophisticated version of this argument makes it extremely hard to pinpoint what exactly is amiss, not least because he makes his case for this controversial doctrine using only a handful of uncontroversial assumptions about logic and language (for instance, that any statement is either true or false). What most bothered Wallace about Taylor’s paper was not the despair-inducing worldview of fatalism itself (though that was indeed worrisome); it was, as Jay Garfield recalled, that “this metaphysically troubling conclusion followed from these ordinary-seeming premises.” Taylor seemed to have scrambled the normal relations among logic, language and the physical world, detaching them from their proper spheres. There was a kind of anguish for Wallace in the prospect of a world so out of whack. “He was very level-headed in so many ways,” Willem deVries, a philosopher now at the University of New Hampshire and the principal adviser on Wallace’s thesis, told me. “He wasn’t attracted to philosophy because you could construct these weird, mind-bending arguments. He was quite wary of the mind-bending. Maybe because his own mind could bend so easily.”

But how to straighten out Taylor’s fatalism? Wallace proposed that there was a flaw in Taylor’s argument, a hidden defect. In essence, Taylor was treating two types of propositions as if they were the same, when in fact they needed to be distinguished and treated differently. Consider the sentences “It was the case that I couldn’t fire my handgun” and “It cannot be the case that I did fire my handgun.” At first they may sound similar, but Wallace argued that they involve quite different notions of impossibility. “It was the case that I couldn’t fire my handgun” refers to a past situation in which discharge is deemed impossible because (let’s say) my gun was broken. “It cannot be the case that I did fire my handgun” refers to a present situation in which discharge is deemed impossible because (let’s say) my gun is still cool to the touch. The first notion involves an earlier, physical constraint on firing (namely, the broken gun); the other involves the current absence of a necessary consequence of firing (namely, a hot barrel). An extremely sensitive observer of language, Wallace noted that there is a subtle indicator of this important distinction already at work in our language: the fine differentiation in meaning between “I couldn’t have done such and so” and “I can’t have done such and so.”

Armed with this small but powerful insight, Wallace was able to pick apart the machinery of Taylor’s argument. All the things about the “Fatalism” paper that appeared maddeningly simple started to look complex and thorny. By the time Wallace worked out all the details — the precise interactions among elements of meaning, time and possibility — it was clear that he had defused Taylor’s argument. (The formal apparatus that Wallace developed in the thesis, a so-called intensional-physical-modality system, would have been a novel contribution to the philosophical literature; deVries and Garfield each expressed to me their regret that Wallace never published the paper.) Perhaps our actions are indeed fated, Wallace acknowledged — he had nothing to say either way about the metaphysical substance of the doctrine. But if fatalism is true, he demonstrated, we are going to learn that fact only through an argument that draws on something richer and more substantive than the arid, purely logical moves Taylor made. If Taylor were to overthrow our worldview, he would have to roll up his intellectual sleeves and delve into reflection on meatier issues like cosmology or entropy or the like.

The real accomplishment of Wallace’s thesis, however, was not technical or argumentative but more like a moral victory. His demonic attention to detail in language and logic, and his seemingly limitless cognitive abilities, had set aright a world momentarily upended by a conceptual sleight of hand. “In light of what we’ve seen about the semantics of physical modality,” Wallace wrote in the closing passage, “I hold that Taylor’s semantic argument does not in fact yield his metaphysical conclusion.” He then ventured modestly that his own analysis of the problem “seems to warrant the following conclusion of our own: if Taylor and the fatalists want to force upon us a metaphysical conclusion, they must do metaphysics, not semantics. And this seems entirely appropriate.” Things, for the moment, were as they should be.