Scepticism

3. Relevant alternatives fallibilism

As we have noted, the sceptic attempts to undermine our claims to know by calling attention to sceptical alternatives. The relevant alternatives response to this sceptical manoeuvre is to deny that these alternatives are relevant. An alternative, h, to q, is relevant just in case we need to know h is false in order to know q. So if h is not a relevant alternative, we can still know q even if we fail to know h is false. This view entails that the deductive closure principle is false.

There are two ways to argue for this view. The direct way is to cite alleged counterexamples to the deductive closure principle. Some philosophers have done this by appealing both to our intuition that we know many propositions about the external world and to our intuition that we fail to know the falsity of sceptical alternatives. So my strong intuition that I know I am looking at my computer monitor and my strong intuition that I fail to know I am not a brain-in-a-vat constitute the basis for such a counterexample.

A more indirect way to argue for this view is to construct a theory of knowledge that has as a consequence, the failure of the closure principle, as in Nozick (1981). The basic idea of these kinds of theories is that knowing requires the truth of certain subjunctive conditionals. On one (simplified) version, my knowing q requires that:

This requirement for knowledge precludes my knowing I am not a brain-in-a-vat. For I would still believe I am not a brain-in-a-vat, even if I were a brain-in-a-vat. But, this requirement allows me to know I see a computer monitor. For it seems plausible to claim that I would not believe I see the computer monitor if I were not seeing it.

A significant difficulty for the direct way of arguing for the relevant alternatives view – the appeal to counterexamples to the closure principle – is that the intuitions that support the counterexamples seem no more compelling than the intuitions in favour of the closure principle. Many think that the closure principle expresses a fundamental truth about our concept of knowledge. So much so that if a certain theory of knowledge entails the falsity of the closure principle, some philosophers are inclined to take the fact as a reductio ad absurdum of that theory.

But this presents problems for the indirect way of arguing for the relevant alternatives view: some philosophers reject theories that endorse condition (S), for the very reason that it entails the falsity of the closure principle. Moreover, there are other difficulties for theories that endorse conditions like (S). One problem for these theories is that they seem to preclude our knowing much of what we take ourselves to know inductively. Consider an example where you leave a glass containing some ice cubes outside on an extremely hot day (Vogel 1987). Several hours later, while you are still inside escaping the heat, you remember the glass you left outside. You infer that the ice must have melted by now. Here we have an ordinary case of knowledge by inductive inference. According to the theories we are now considering, my knowing that the ice cubes have melted requires the truth of this subjunctive conditional:

But (S′) looks false. It seems plausible to claim that had the ice cubes not melted, it might have been for some reason (for example, someone putting them in a styrofoam cooler) that would still leave me believing they had melted. Thus, it looks as if theories which endorse this condition are too strong. If this is correct, then the anti-sceptical results afforded by condition (S) come at the cost of scepticism about certain kinds of inductive knowledge.

We should note, however, that there is some controversy over the evaluation of subjunctive conditionals like (S′). But I think it is fair to say that standard semantics for subjunctive conditionals would render (S′) as false (see Deductive closure principle §§2–3).