WMD

    Words of Mass Dissemination

WARNING: This post is really just a little philosophical essay. It is also quite inconclusive. Moreover, it would not pass muster with professional philosophers. There are, for example, no references to what other people have said about this matter. And so on. However, for all of these faults, I do in fact believe what is said here, and I think that the arguments — often merely intimated here — are not bad arguments.

Over the past few years, I have become very interested in the question

What counts as evidence in philosophy?

So suppose that I am defending some philosophical position. Let’s call that position ‘P’. (In a moment, I’ll consider a real case of a real philosopher defending a concrete philosophical position.) The question is: What counts as evidence bearing on the acceptability, reasonableness, truth, etc., of P? If you assert not-P, then what would, or should, count as evidence in favor of your position over mine?

I am purposefully being somewhat non-committal about exactly what evidence is supposed to do for us. Let E be evidence for P. Does this mean that E makes it certain that P? Does it mean that E makes P more probable than not? I am inclined to answer ‘no’ to both questions, for a variety of reasons, but this issue is difficult (and addressed at some length in the context of scientific evidence by various philosophers of science). So let’s just leave the relationship between E and P as mentioned above: E is evidence for P just in case it bears on the acceptability, reasonableness, or truth, of P. To put it in terms of ‘epistemic obligation’ (i.e., the sorts of thing one ought to do if one is reasonably going to make claims about P): E is evidence for P if E is something that should be taken into account when considering whether to assent to P. So, for example, the color of my dog’s ears is not in any way evidence for (or against) the (philosophical) proposition that ‘every event has a cause’. It makes no difference whatsoever to whether one should accept that proposition.

What counts as evidence for claims of a particular sort has a lot to do with what those claims are about. The color of my dog’s ears is not evidence relevant to the claim that all events have a cause, because, I guess, that claim has nothing whatsoever to do with the colors of dogs’ ears — it is not in any way about the colors of dogs’ ears. (’Being colored thus-and-such’ is presumably not an event, nor a cause, nor does it seem to imply much about events and causes.) Hence, I’d like to broach the question ‘what is philosophy about?’ by asking the question at the start: What counts as evidence in philosophy?

Now, one might believe that this question is too broad, that the class of all purported facts that are, or would be, evidence bearing on some philosophical question is not a class that can be delineated in any useful or illuminating way. I really do not mean to be making any commitment on that question. Indeed, I shall now simply consider an example, albeit one that I believe to be representative of a lot of what passes for ‘philosophy’.

The example is the first chapter from a classic book (The Significance of Philosophical Skepticism) by the philosopher Barry Stroud. The chapter is a defense of something like Cartesian skepticism, the view that we know nothing about the external world, because for all we know, we could be dreaming (or, in the modern context, we could be ‘brains in a vat’). Part of his defense involves defending something like the following principle:

Principle of Knowledge: In order to know that Q, you must know the falsity of everything that is incompatible with your knowing that Q.

(Stroud does not name his principle in this way — I do it here just for convenience.) For example, let Q be ‘I am now sitting in a chair’, and suppose, as Stroud does, that if you are currently dreaming that you are sitting in a chair, then you in fact do not know that you are sitting in a chair (even if, as a matter of luck, you are sitting (and sleeping!) in a chair). Then, in order to know that Q, you must also know that you are not dreaming. But, so the argument goes, you cannot know that you are not dreaming, and so you cannot know that Q.

Why should we accept the principle? Stroud makes the following case:

[The principle] cannot be avoided if it is nothing more than an instance of a general procedure we recognize and insist on in making and assessing knowledge-claims in everyday and scientific life. We have no notion of knowledge other than what is embodied in those procedures and practices. So if that requirement is a ‘fact’ of our ordinary conception of knowledge we will have to accept the conclusion that no one knows anything about the world around us.

And Stroud does think that the requirement is a fact of our ordinary conception of knowledge. But why? Here he relies on examples. I’ll mention one of them (slightly modified, but more or less the same as Stroud’s example). Imagine, that you are on a jury, and that the defendant in the case has an alibi — he was in Cleveland at the time of the crime (which occurred in London). Moreover, several witnesses corroborate the alibi. If, on the basis of this testimony, you claim to know that the defendant was in Cleveland, you will rightly be challenged about the veracity of the alibi, which, to this point, is your sole evidence. After all, perhaps these witnesses are all part a crime ring, of which the defendant is also a member. In that case, even if in fact the defendant was in Cleveland, you might not reasonably claim to know this fact, because you have not taken reasonable steps to determine the reliability of the witnesses (and if they are unreliable, then anything you claim to know in virtue of their testimony should not be counted as genuine knowledge, even if it happens, by luck, to be true). In other words, in order to claim that you know that the witness was in Cleveland, you need to satisfy the Principle of Knowledge.

OK, one may acknowledge that the jury ought to satisfy this particular application of the Principle of Knowledge, in this case. Does it follow that specifically this principle is what is embodied in our practices regarding knowledge-claims? After all, more than one principle regarding knowledge could imply that in this case the jury ought to take steps to ascertain the reliability of the witnesses. I don’t see how Stroud could ever answer this question without some really hard empirical work, and I will return to that point below. For now, let’s flag this issue as ‘the problem of empirical work’.

Stroud faces another problem: prima facie, the case that specifically Stroud’s Principle (or something like it) is in fact the one that is operative for us, here, is very weak indeed. We can agree that the jury should try to take certain steps to ascertain the trustworthiness of the witnesses’ testimony about the defendant’s alibi. But Stroud’s Principle is not just about taking certain steps. It is not just about ruling out some reasons to doubt the testimony. So let’s consider another kind of doubt one might have about the veracity of the testimony.

Imagine that it has been determined that the witnesses who testified to the alibi in fact did not know the defendant, have no interest in the outcome of this trial, no animus towards the prosecution, are not habitual liars, were able to describe the defendant accurately, and so on. Their testimony is reliable. A frustrated prosecuting attorney, upset by the situation and desperate for a conviction, spins the following tale: while these witnesses thought that they saw the defendant in Cleveland, in fact aliens abducted them yesterday, took them on board their ship, and reprogrammed their brains so that they would, today, ‘recall’ having seen the defendant in Cleveland. Is the jury obligated, in any sense, to discover whether this scenario is true? Suppose that, after taking the measures noted above to ensure the veracity of the testimony, they claim to know that the defendant was in Cleveland. Would normal people challenge them on the grounds that they had not ruled out alien abduction? Is that sort of challenge part of our “procedures and practices” regarding knowledge-claims? I suspect that it isn’t. And yet its relevance and the relevance of more radical challenges still — indeed, the challenge that we might be dreaming, or brains in a vat — do follow from Stroud’s Principle. But why then does Stroud think that the Principle, or something close to it, is operative in our “procedures and practices” regarding knowledge-claims?

Before I step back and consider my original question — remember that all of this talk about Stroud and skepticism is really just an example! — let’s take this point a step further. Stroud says that “we have no notion of knowledge other than what is embodied in those procedures and practices”. But don’t ‘we’ make, and assent to, knowledge-claims all of the time? (I object to this sort of us of the term ‘we’, which is why I’ve enclosed it in inverted commas. See below.) I claim, right now, to know that I am breathing. Would you, normally, challenge that claim? I’m confident that, in the circles that I run in, at least (which includes a lot of philosophers!), such claims are not normally challenged. But then why doesn’t Stroud conclude that skepticism must be wrong? Here is the argument for that conclusion:

Premise 1:

Whatever is a fact about our procedures and practices regarding knowledge-claims is a fact about knowledge.

Premise 2:

It is a fact about our procedures and practices regarding knowledge-claims that we sometimes make, and sometimes assent to, the claim that somebody knows something about the external world.

Conclusion:

Knowledge of the external world is possible.

These premises do not, strictly logically, entail the conclusion, but I think that we could pretty easily patch things up so that they did, and I won’t bother to try here. Note that Premise 1 is quite close to (and perhaps a consequence of) what Stroud apparently asserts. I don’t know whether Premise 2 is true, but I suspect that it is. Restricted in scope to me and the people with whom I normally associate, I am very confident that it is true. But, of course, Stroud wants to deny the Conclusion.

The problem that I have generated for Stroud comes, I believe, precisely from his endorsement of something like Premise 1, and I believe that the problem is quite general. The general form of the problem is this: for any object of philosophical investigation, O, if it is true that

Premise 1*:

we have no notion of O other than what is embodied in our procedures and practices regarding O,

then we are very likely to run into one or both of the following two problems, the first of which I called ‘the problem of empirical work’, above:

• Understanding O will involve extremely difficult empirical questions about “our procedures and practices regarding O”. (For example, what are those practices? What principles, if any, are guiding them? And so on.)
• “Our procedures and practices regarding O” entail, or suggest, or give us reasons to believe, conflicting things about O.

I believe that philosophers who adopt principles such as Premise 1* above very often run into one or the other of these problems. Very often, they appear to ignore the first problem, and simply make assumptions about how the result of such an empirical investigation would go. Often, for example, they will appeal to ‘what we say’, or even ‘what you say’ (somewhat annoyingly, as in most cases they have never even met me).

And very often, they appear to ignore the second problem as well, picking and choosing which apparent consequences of our procedures and practices to take seriously.

But wait! Why can’t philosophy criticize our procedures and practices? Why can’t it pick and choose which apparent consequences of our procedures and practices to take seriously? Indeed, isn’t it part of the point of philosophy to examine those procedures and practices, reveal inconsistencies, and resolve them precisely by picking and choosing so that the inconsistencies no longer arise?

Yes, I believe that this sort of critique is part of what philosophy ought to do. The problem here is that if our only evidence in the philosophy of O (apart, perhaps, from whatever evidence flows from purely logical considerations) is taken from “our procedures and practices regarding O”, then philosophy is, I believe, in no position whatsoever to provide this critique. (It could, perhaps, point out inconsistencies, although notice that even here the problem of empirical work will arise. But it would not be in a position to resolve those inconsistencies.)

Finally, this discussion raises the question that I used as the title of this little essay. What is philosophy about? If “our procedures and practices regarding O” count as evidence for the philosophy of O, then one of two things must be true. Either (a) philosophy is, in the end, about “our procedures and practices regarding O”, or (b) the philosophy of O is indeed about O, but at the same time, “our procedures and practices regarding O” are good indicators of the nature of O, or of the truth of claims one might make about O. In case (a), philosophy is really a branch of sociology — it is really about some of the social and linguistic practices of us human beings.

I don’t really know what to make of claim (b). It seems to me likely to be true in some cases (for some O), false in others (but I’ll grant that how things seem to me is pretty poor evidence for how they really are). Anyhow, an argument ought to be made, for any given O. And if the argument relies on some claim such as that O is itself to be defined in terms of our procedures and practices regarding O, then we are back at case (a). Moreover, it does seem that for some O, claim (b) is pretty implausible. Perhaps, in the case of knowledge, claim (b) has some plausibility about it, for knowledge is, arguably, very much a human phenomenon, or even a human construction (but perhaps not). On the other hand, Stroud’s reliance on (b) in the case of knowledge seems to have led him astray, in the way I described above. And for some other O, claim (b) seems extremely implausible to me. Consider, for example, certain parts of metaphysics, such as the philosophy of material objects (which embarks on such tasks as identifying the essential properties of material objects as such). What reason could we give for thinking that our procedures and practices regarding claims about material objects provide good evidence about the nature of material objects? Aren’t we instead (and as science seems frequently to reveal) often quite confused about the nature of material objects?

Well, I have now broached another huge topic, really a subtopic of the present one, namely, the nature of evidence in metaphysics. I should stop, though, because whenever I discuss this topic, I tend to get thinly veiled threats from some philosophers, and I have children to look after. What is philosophy about? I haven’t answered that question, but perhaps some of the considerations above are at least relevant, perhaps even important, for doing so.