by Dave Maier
Epistemology books (in the previous century anyway) pretty much all start out the same way. Epistemology is the philosophy of knowledge, and so the first thing we have to do is to determine What Knowledge Is, before then going on to find out how best to get it, confirm that we have it, what it’s good for, and so on. By the end of the first page our author has usually decided that knowledge is a form of belief which is true and justified; the question for the rest of the chapter, or even the book, is what else we may or may not need for a belief to count as knowledge. As it turns out, there are many such “JTB+” accounts of knowledge (note: none of them work, but it’s good practice figuring out why).
But how do we know that we want a “JTB+” account to begin with? That first bit was pretty quick. (Full disclosure: my own analysis of knowledge is a “TB” account, and as you can imagine it has been rather frustrating to see one’s considered views universally dismissed as obviously false on page 1 of virtually every introductory epistemology text in the land. But I digress, as this is not my point today.) What usually happens is this: our author says something like “Let’s say Jill believes that Jack is cheating on her, but as it happens he is not. Does Jill know that Jack is cheating on her, or merely believe without knowing? Clearly, in this case we would say that she does not know. She believes she knows, but she does not. Knowledge, that is, entails truth.” And that’s that; on to the (supposed) justification condition.
However, as you may have noticed, that’s not an argument – it’s merely an expression of an intuition; and intuitions are slender reeds on which to base our philosophical edifices. Or so say a new (or perhaps not so new, by now) breed of philosopher, who hoist the banner of “experimental philosophy”. Their emblem is a burning armchair, symbolic of the movement’s rejection of the detached, unempirical intuition-mongering of last-century mainstream philosophy. What do you mean “Clearly we would say X”? How do you know? Why don’t we actually go find out what “we” would say? Let’s round up some people and ask them!
This is what experimental philosophers do. Their research is explicitly empirical, as pointedly opposed to the traditional reliance on intuition. A common response to this from mainstream philosophers, especially at first, has been to mock experimental philosophy as turning the scholarly contemplation of the eternal verities over to the untutored mob, as if a mere vote could determine philosophical truth. This is certainly unfair to at least the best experimental philosophy, but even when we grasp their methodological point, some weirdness seems to me to remain. But instructive weirdness!
Take one experimental result relevant to our example above. Philosophical intuition has it, according to our epistemologist, that knowledge entails truth. But when experimental philosophers actually asked people about this, they denied it. Presented with cases in which someone just knew that there was some milk left, but when she looked in the fridge, she found it empty, most respondents judged that such a thing is possible: she knew, even though what she knew was false.
I have no doubt that the researchers have all of their statistical and methodological ducks in a row here, as I have certainly heard people say things like this. What caught my eye instead, when I saw this, is how the researchers portrayed the results: they had determined experimentally, they said, the surprising fact that knowledge need not be true. That is, the takeaway was explicitly described as a discovery about knowledge. We had thought, based on “intuition”, that knowledge was one way, but it is in fact another way entirely. Chalk one up for experimental philosophy!
What I find striking here is this. Experimental philosophers convincingly present themselves as attempting to bring philosophy into methodological line with the empirical sciences. It’s not, if I’m reading them right, that they think philosophy should simply become an empirical discipline, but instead that it not be so cut off from the empirical world that it should take place entirely in the philosopher’s head (or, again, in the armchair).
So far, so good; but now comes the weird part. There have been many such calls over the past decades, often billed as a “return to realism”, urging philosophers to abandon their twentieth-century obsessions with the language or concepts we use to talk about the world, or airy abstractions like “meaning” or “discourse”, in favor of a return to philosophy’s traditional concern – not unlike that of science – with the world itself, i.e. the object of our talk rather than the talk itself (or the world at one remove, as reflected or represented by our talk). With its explicitly empirical orientation, experimental philosophy seems to fit right in: the result of their investigation, they claim, is not mere intuition-mongering but an empirically robust discovery about knowledge: i.e., that it can be false.
Realist philosophy universally scorns the idea that what people say can determine how things are: that’s idealism! The whole point of the realist attack on linguistic philosophy is that what people say and how things are are two different things, metaphysically speaking, and that only those with an insufficient appreciation for the objectivity of the real world would confuse them, or think that we should examine the former obsessively when it is the latter which is our real concern. Yet here in the heart of the philosophical turn to empirical, world-oriented science, we have a philosophical “discovery” based entirely on what people say about certain English sentences.
As empirical inquiry, this “philosophy experiment” bears a strong resemblance to research in lexicography, and one might expect the result to affect our dictionaries just as much as our epistemology texts. Or not: it seems that Merriam-Webster is way ahead of us, as definition 2a of “know” in the first link that pops up reads “to be aware of the truth or factuality of : be convinced or certain of” [emphasis added]. In other words, what our “philosophy experiment” told us we could have learned with a simple trip to the dictionary: that the word “know” has a use according to which what is “known” can be false.
My point here is not to condemn experimental philosophy, which is more varied than my quick sketch here might suggest, but before I draw my actual moral we should allow our traditional epistemologist to reply to the experimental “refutation” of his analysis of knowledge. Must he concede, as claimed, that knowledge can be false? It does seem that he needs to say something. However, far from conceding that his analysis of knowledge is flawed, he might instead simply make clearer what he is talking about and why. It is true that when one is using the term in the indicated sense, what is “known” can indeed be false; but even if we want to call this “knowledge,” that’s not what epistemology is about. Most of the hypothesized book, after all, concerns how best to obtain knowledge and what it’s good for; which means that the sense in which “knowledge” simply means “belief” or “conviction”, regardless of truth, is not what we’re talking about when we do epistemology. We can simply acknowledge the ambiguity and move on. Why should we care about obtaining or using the sort of “knowledge” which can be false? That’s not the kind we’re interested in. We already have more than one perfectly good word for that sort of “knowledge”, i.e. “belief” or “conviction”, and if that’s what we’re talking about we can use those words instead. Why court confusion when you can be precise instead?
It’s not like that’s the only ambiguity in the term anyway. English uses the same term for what other languages mark as two different things, i.e. wissen and kennen in German, or savoir and connaître in French (the latter word in each pair meaning “know” in the sense of “to be familiar or acquainted with”). That is, those two languages use different words to mark that distinction, while English marks it with different senses of the same word. The entry for “know” in the Oxford English Dictionary notes that the meaning of the now archaic English verb “wit”, derived from “wissen”, has been folded into the single English word “know”, derived (through the English word "ken") from “kennen”. The discussion then moves on to the consequent difficulty, given this, of assigning a primary meaning to “know”.
Anyway, just as our epistemologist can say “we’re talking here about wissen, not kennen (or “carnal knowledge”, or whatever else)” he can perfectly well say “… and not about the sort of “knowledge” which can be false, as in the sentence ‘She just knew that there was some milk left, but when she looked in the fridge, she found it empty.’” Now of course if you want to make philosophical hay from a contingently true fact about patterns of linguistic usage, by all means go right ahead – and indeed this is what some experimental philosophers do once their results are in. But when they do, they don’t seem to me to be acting much differently from certain other philosophers whom we might have thought to be their polar opposites, i.e. “ordinary language” philosophers like J. L. Austin.
Austin is sometimes portrayed as the poster child for the very excesses of “linguistic philosophy” now enduring the realist backlash I noted above. He spends pages and pages carefully examining the minutiae of the various usages of a key word rather than, opponents gripe, actually giving us formally rigorous arguments about how things are – talking about “good” rather than, you know, the good. But Austin is admirably clear about his intent. In “A Plea for Excuses” he spends some time explaining his method. Here’s a taste:
When we examine what we should say when, what words we should use in what situations, we are looking again not merely at words (or ‘meanings’, whatever they may be) but also at the realities we use the words to talk about: we are using a sharpened awareness of words to sharpen our perception of, though not as the final arbiter of, the phenomena. […] Using, then, such a method, it is plainly preferable to investigate a field where ordinary language is rich and subtle […] At the same time we should prefer a field which is not too much trodden into bogs or tracks by traditional philosophy, for in that case even ‘ordinary’ language will often have become infected with the jargon of extinct theories, and our own prejudices too, as the upholders or imbibers or theoretical views, will be too readily, and often insensibly, engaged. […] Granted that our subject [that is, excuses] is, as already claimed for it, neighbouring, analogous, or germane in some way to some notorious centre of philosophical trouble, then, with these two further requirements satisfied, we should be certain of what we are after: a good site for field work in philosophy. [Philosophical Papers, pp. 182-3; emphasis in original]
Field work! Now of course it’s true that Austin does not do “philosophical experiments” in the experimental philosophy vein, with modern statistical methods and all the trappings of empirical research. Compared to the similarity in the philosophical results, though, this difference seems minimal. After all, if you found Austin’s examples counter-intuitive, you could go do that research if you felt like it; but even if Austin’s reflections took the form of an experiment that found that people do indeed distinguish in certain cases between, e.g., doing something “by accident” and doing it “by mistake”, it seems that the rest of his article could remain exactly the same (albeit with a statistical analysis appended to the end), with exactly the same philosophical takeaway.
My point is not to say any more (or less) about experimental philosophy’s role in contemporary philosophy than that it could be very valuable if it is done properly (and in cases where the results show that it is really needed, and the consequent analysis fruitful, rather than simply in response to some other philosopher’s having relied on the dreaded “intuitions” while sitting in the dreaded armchair; for all I’ve said about the experiment about “knowledge,” it doesn’t seem that that one qualifies). Instead, it is that even in each of the two apparently opposed cases, where it seemed at first that we were looking at the one or the other of two distinct things – the empirical world, on the one hand, and the words and concepts we use to talk about it, on the other – that distinction, on which the realist backlash essentially relies, dissolves to nothing. Or not to nothing, as words and their referents of course remain different things. Rather, that distinction becomes, as Wittgenstein likes to say, “not a nothing but not a something either”. And that, just as much as one or the other, is what we need to understand.