Philosophy 159: Relevant Alternatives Theories

Relevant Alternatives Theories

Here's a schematic version of the skeptical arguments we've been looking at so far:

You can't rule out the possibility that you're a BIV, or in the Matrix, or deceived by an evil demon, or..., even though those are possibilities you recognize to be incompatible with p.
You can only know that p if you can eliminate or "rule out" all the alternatives you recognize to be incompatible with p.
So you don't know that p.

The Relevant Alternatives Theorist is someone like Austin who denies 2. He says you don't have to rule out all the alternatives to P, or even all the things you recognize to be alternatives. You just have to rule out the alternatives that are somehow "relevant."

What makes an alternative "relevant"? There are different stories about this.

One story is called Contextualism. That story goes as follows.

Contextualism

Words like "I" and "now" are clearly context-sensitive. What person or time they pick out depends on the context in which they're uttered.

Similarly, one might argue that "big" is context-sensitive. If we look at Mighty Mouse scurrying about with a bunch of other mice, we might say "He sure is big." If we look at Mighty Mouse scurrying around with a bunch of kangaroos, we'd probably say, "He sure is small." Perhaps we're saying something true in each case. In the first case, what we mean to be saying is that Mighty Mouse is big for a mouse. In the second case, we mean to be saying that Mighty Mouse is small for a land animal, or something like that. The comparison class is supplied by the speaker's intentions and the audience's expectations, and these vary from context to context.

One could also give a Contextualist account of words like "flat" and "empty." One could say that what counts as flat varies from context to context. When we want a flat surface to calibrate our machine tools, the Kansas highway doesn't count as flat. When we want a flat surface to land our damaged airplane, the Kansas highway does count as flat. What "flat" means depends on the intentions and expectations of the people engaging in the conversation.

This is different than the story we saw Stroud giving last time about words like "flat." Stroud was claiming that it's never literally true to call the Kansas highway flat, but it might sometimes be appropriate or reasonable to call the highway flat, for instance, if you're looking for a place to land your damaged airplane. The highway is close enough to being flat for your practical purposes.

The Contextualist, on the other hand, would say that in some contexts it really is literally true to call the Kansas highway flat. Just as in some contexts it really is literally true to call Mighty Mourse big. In some contexts the little irregularities in the highway surface don't count; they're just too small. In those contexts, the highway has no bumps in it that are big enough to count. So it's literally true that it's flat.

I want you to understand the difference between these two views. If you feel like, "Well how could you decide which view is right?" that's a perfectly sensible reaction to be having. It is hard to tell what kinds of evidence would show the one view to be right and the other view to be wrong. This is a topic of much controversy in contemporary philosophy of language. But we don't need to decide which of these is the right view, right now. We just need to understand the difference between the two views.

The Stroud view says it never is literally true to say that the highway is "flat." But it would be very complicated and tedious to always speak the literal truth, and still get across the things we want to get across. Suppose the copilot has control of your plane and it's about to crash and he shouts "Do you see any flat place for us to land?" and you look out the window and a bit off to the left you see an empty Kansas highway. If you just say "No," because the highway is not really literally flat, your copilot won't know to steer to the left and you'll crash and die. If you say "No, there's no place that's really flat, but there's a highway over there which is pretty clo--" Whoops. Took too long. It makes the most sense just to say "Yes, over there!" even if what you're saying is literally false. It's literally false but for practical intents and purposes it's the most reasonable thing to say. That's Stroud's view.

On the Contextualist view, on the other hand, in a context like the one I've described, it really is literally true to call the highway "flat." Just as it can sometimes really be literally true to call Mighty Mouse "big," if it's clear in the context that you mean big for a mouse. So too it can really be literally true to call the highway flat, if it's clear in the context that little bumps in the asphalt are too small to count.

So that's what Stroud and the Contextualist would say about words like "flat." They'd say similar things about words like "empty." And this is also what they say about words like "know."

A Contextualist about knowledge says that what "know" means varies from context to context, just like "big" and "flat" and "empty." The general story is: "S knows that p" means that S truly believes that p and S has evidence that's good enough to rule out all the relevant alternatives--but in different conversational settings, the Contextualist thinks, different alternatives count as relevant. When you're talking to your butcher, things like BIVs and the Matrix and perfect dreams are't relevant. So in contexts like that you can truly say that you know you have hands. When you're talking to the skeptic, on the other hand, the set of relevant alternatives is different. In a context like that, things like BIVs and evil demons may become relevant alternatives, and since you can't rule those alternatives out, in those contexts it would be false to say that you know.

Stroud on the other hand says that it's never literally true that you know you have hands. Your evidence may be good enough that you're close enough to knowing for all practical intents and purposes, it may be practically appropriate or reasonable for to say that you know, but still, it's never really true that you know. It's not true that you know because there are some alternatives, like being a BIV, which your evidence does not enable you to rule out. (In the article "Skepticism and Everyday Knowledge Attributions," Cohen calls this "the pragmatic view.")

Austin's Relevant Alternatives Theory

Stroud interprets Austin as a kind of relevant alternatives theorist. But Austin is a little bit different than the Contextualist. Like the Contextualist, Austin thinks that knowing you have hands only requires you to rule out some alternatives, not every alternative. Like the Contextualist, Austin also thinks that whether you know you have hands can vary from context to context. But the Contextualist thinks that what determines what alternatives are relevant are the intentions and expectations of the people involved in the conversation. Austin thinks that what determines what alternatives are relevant is the kind of evidence that you possess. On his view, you only have to rule out those specific alternatives you have some reason to believe do obtain. Hence, the guy in Pollock and Cruz's story, who has evidence that he's a BIV--he has to rule out the possibility that he's a BIV. We on the other hand don't have any evidence that we are BIVs, so we don't have to rule that possibility out, even if we're talking to a skeptic.

This is a respect in which Austin's view and the Contextualists' view are slightly different. But for now, we don't have to worry about the differences very much. We'll be talking about them more later.