Phun With Philosophy

Patrick Belton warms my heart with another philosophy joke, this one neatly dovetailing what I've been studying in my epistemology seminar:

THE EPISTEMOLOGISTS DECIDE to field a Little League team, and three of them volunteer to serve as umpires. They go into a bar after one game, and after they've got their drinks the critical realist says "Yeah, well, I call 'em as I see 'em." The direct realist responds "Well, I call 'em as they are!" Then the Berkeleian idealist pipes up and says, "Shoot, they ain't anything till I call 'em!"

I'll leave it to readers to decide whether or not this is funny; the deep irony is that, to me at least, the Berkeleian idealist is in the best position of the three.

Why? Let's look at direct realism versus Berkeleian idealism. (Critical, or indirect realism, which attempts to straddle the two extremes, is in the unenviable position of either 1) admitting an inexplicable complexity into its epistemological system, 2) conceding space into which the Berkeleian or the Humean can drive his skeptical wedge, or 3) being pushed by skeptical worries into direct realism or some overly complicated iteration of it.)

Direct realism, a straightforward rebuttal of skepticism, argues that we directly perceive the objects of experience, that there is no mediating category of "phenomena" or "appearances" blocking such direct access to external objects, and, crucially, that mistaken, dreaming, or hallucinatory percerptions are not at all the same kind of things as valid, veridical perceptions: it's true that, were we hallucinating, or were we brains in a vat, or victims of an evil deceiver, we would not think that we were; nevertheless, the direct realist argues, when we do validly perceive the real objects of experience, we know that we perceive them.

Problems? For one thing, the direct realist and the phenomenalist-skeptic might be construed as simply talking past each other. With access to the same data, i.e., the objects of experience, the former maintains that perceptions are of really existing objects, while the latter insists that all we can or ever do perceive are appearances. Since no new evidence could ever come to light---and what datum could there be that is not an object of experience---the dispute between direct realism and skepticism might just give way to a semantic debate (which is not to say uninteresting).

I think, however, that the metaphysical/epistemological dispute is worth trying to preserve on its own terms, rather than as a reduction to philosophical semantics. If the positivists taught us anything, after all, it's that efforts to reduce philosophy to discrete semantic laws are doomed to fail.

I said at the outset that the Berkeleian is in a better position than the direct realist. Let me explain why. The direct realist has to come up with an explanation of problem cases, say, e.g., perception of light from stars so distant that they have already collapsed. What exactly, we should ask the direct realist, are we perceiving? Surely not something that does not exist? The Berkeleian, it must be conceded, faces a similar problem. How does he account for the existence of things like electrons? Does he wish to say that, since they can't be immediately perceived, they do not exist? Some Berkeleians might just do that---and argue, in effect, that electrons are nothing but convenient explanatory fictions that are useful in predicting the outcome of experiments, but are not "real," even under a Berkeleian framework. I think the more persuasive response (since I do think that electrons are real) is something like this: in order for a thing to exist it must indeed be perceived, but to be perceived is not necessarily to be immediately available to the five traditional human senses, which are contingent products of evolutionary biology, and electrons can indeed be perceived with instruments fine enough to detect them. After all, a blind man can perceive by means of other senses, and if his tactile sense is developed enough, he can use it to learn, at least topographically, what things look like. Sonar, to take another example, produces images no less valid as phenomena than the contents of a photograph.

Got off track there for a second. Whoops. Anyway, the fatal problem for the direct realist is that the best case he can make---that of Timothy Williamson against the existence of distinct internal states---dooms his project rather than propel it forward. The Williamson argument, which I will probably not do justice to, is this:

1. The appearances, or internal phenomena, are "luminous," meaning that when a perceiver is in a luminous state, he knows that he is in a lumonious state. E.g., when I feel cold, I know that I feel cold.
   
2. In order to know something, it has to be "safe." The safety requirement is difficult to characterize, except by use of examples. Here's one: If I claim, prior to any measurement that an object O weighs less than 50 kg, and it turns out that it weighs 49.9998 kg, then the proposition is not safe; it was merely accidentally correct, and I never knew that O weighed less than 50 kg. Conversely, had I claimed that O weighed less than 100 kg, my proposition would indeed be safe; my margin of error was sufficient for me to justifiably claim to have known that O weighed less than 100 kg.
3. Internal states are luminous, therefore they are known. If something is known, then it must be safe. This is the basic syllogism, which Williamson employs to set up the conditions for a reductio argument against the existence of internal states.
4. Let's run the argument: I stand outside in the middle of the night at time t1, and I feel very cold. Since feeling cold is luminous, I know that I feel cold at t1. Since I know that I feel cold at t1, the claim that I feel cold at t1 is safe.
5. Pick any interval of time. It doesn't matter how large or small. Since the claim that I felt cold at t1 was safe, it must be the case (remember, you pick the size of the interval and the margin of error) that I feel cold at time t2. By virtue of luminosity, I know that I feel cold at t2. And on and on. Until at some point in time, tx, which is, say, 12:00 pm, I definitely no longer feel cold. My knowledge that I was cold at whatever point in time preceded tx could therefore not have been safe. Hence, I didn't know I was cold. Hence, feeling cold is not luminous. Hence, there are no special, privileged internal states.

Based on this absolutely ingenious argument, the direct realist wants to put forward the notion of direct access to objects of experience. For the life of me, however, I can't understand how we can make the leap from observing a deep problem about our understanding of internal states to concluding that we have direct access to the external.

Assume, for the sake of argument, that Williamson's argument is sound as well as valid (I think it's clearly valid, and its soundness is the issue). If so, then Williamson has succeeded in pushing the skeptic further into skepticism. "Okay," the skeptic might say, "you've demonstrated that I was unjustifiably asserting too much even in my skepticism. I had merely questioned the correspondence of a conceivable external world to the phenomena of my experience. Now you've given me reason to doubt the validity even of the contents of my own internal theater. I have a final position to fall back on, which is that there are, in fact, perceptual phenomena to which I have access. That might be all I ever know. I doubt that you, the realist, would deny that I do indeed experience things. Still, though I may be forced to become more radical in my skepticism, you've certainly not given me or given yourself (if you do indeed exist) any reason to conclude that the objects of experience are real, external, and directly accessible. You've demonstrated that they are even less reliable than I, the skeptic, had originally assumed." So the skeptic might say. To borrow Schopenhauer's metaphor, Williamson has succeeded in assailing the previously impenetrable fortress of the skeptic, and indeed in tearing down its outer walls and parapets, but only by locking himself in its innermost keep.

The other possible response, which is undeniably, at least for practical purposes, more desirable than retreating into a skeptical shell that denies even the validity of subjective phenomena, is simply to reject the condition of luminosity. And that, ultimately, is what Williamson proves, and no more: luminosity is not the attribute that sets internal states apart from the external. After all, one need not have read and interpreted sophisticated philosophy in order to know that one can be in doubt even about one's own seemings. (One can draw two lines side by side, e.g., and be in doubt even about whether they seem to be parallel.) The answer, suggested by my brilliant and soon-to-be famous professor, Troy Cross, is to replace the luminosity condition with "near luminosity." When I feel cold, I know that I'm in a state that includes coldness and borderline cases of near-coldness. Now try and run Williamson's argument. It won't work. Internal states are preserved (yay!) and we've accounted for the problem entailed by luminosity.

Now, when we compare the Berkeleian to the direct realist, the relative strengths of each position are rather different. The Berkeleian maintains a one-category epistemology that leaves no brute facts behind (either God or discrete psychophysical laws coordinate the phenomena of experience), whereas the direct realist is left with both internal and external states, trying unsuccessfully to explain away the former, and having no explanation for his supposed direct access to the latter.