notes in the margin of error

I was impressed by "On Truth and Falsity in their Extramoral Sense" when I was 17. Looking at it again, I see why: that was my first encounter with the paradox of analysis--i.e. the tradeoff between informativeness and truth in analysis. (If you define A in terms of B, this will be informative, but will probably not seem to account completely for A's A-ness, to put it generically.) I first encountered it under that name in John Wisdom, I believe, much later. Not that I'm surprised to turn up more consistency in revisiting my own past.....

But I'll let the eloquent, ornery German speak for himself. Page numbers from one of those Cambridge chronologically organized collections.

"[Underlined by my young self] If [man] does not mean to content himself with truth in the shape of tautology, that is, with empty husks, he will always obtain illusions instead of truth" (90).

"Every idea originates through equating the unequal. As certainly as no one leaf is exactly similar to any other, so certain is it that the idea 'leaf' has been formed through an arbitrary omission of these individual differences, through a forgetting of the differentiating qualities...." (95)

And I can't leave out this delightedly excoriating passage:

"What therefore is truth? A mobile army of metaphors, metonymies, anthropomorphisms: in short a sum of human relations which became poetically and rhetorically intensified, metamorphosed, adorned, and after long usage seem to a nation fixed, canonic and binding; truths are illusions of which one has forgotten that they are illusions; worn-out metaphors which have become powerless to affect the senses; coins with their images effaced and now no longer of account as coins but merely as metal" (92).

The essay's rather haphazard according to my standards these days, but I dearly miss this sort of writing in philosophy. And I wish I could take my 17 year-old self and say, "Hey! That's the paradox of analysis you're interested in! Go Alta-Vista that!"

"There might be extraordinary facts, even about our everyday experience, which plain men and plain language overlook" (J.L. Austin, "The Meaning of a Word," 69).
One of the most common objections to understanding philosophical problems by comparing them with ordinary ways of speaking is that not all of our ordinary ways of speaking are correct. We may have language games that need to change, because they're based on empirical or ethical mistakes (blaming animals, for example; or talking of some people as if they are less human than others). I very much doubt any ordinary language philosopher would have claimed that how we do speak is always how we should speak; but it is not obvious why. The answer, I suspect, will also explain the way the appeal to ordinary language is supposed to work.

How do we discover that it's wrong to blame animals? It doesn't work, primarily. It causes unhappiness and fails to facilitate anything positive, other than the expression of anger. If it's a language-game (is it, even? we pick it up from other people, but it's practiced on non-speakers), it doesn't do much. It might as well be a "wheel turning idly."

The most important concept in the PI has long seemed to me to be the "very general facts of nature" that form the background to our language-games. If certain very general facts of nature were otherwise, our language games would be different too. That is, "the way we speak" does not exist in a vacuum. It depends on much else. Thus "the way we speak" cannot be the only evidence we consider. What, then, are "very general facts of nature?" Physical facts, certainly, but also the reasons for which we speak (the kinds of aims we have--grounded in physical facts).

But that is all, so to speak, from a third-personal point of view. How does it look from the inside?

The way we should speak depends on a great deal. It depends upon what we're talking about, and what the best description of the case is, given our knowledge, and purposes. The fact that we do sometimes blame animals thus does not imply that we should, given other facts to which we may appeal. It's not the appropriate response. Finding the best description of a strange case, a misunderstanding, or a clash of concepts involves finding the best, most appropriate response. To do this, we must be open to other ways of considering the case in question. For this purpose, many other language-games all have to be functioning too; we must accept sensory evidence, understand and consider competing theories, and empathize. We cannot step outside of language even in bringing this evidence to bear on what we say.

The activity of doing philosophy, in my own practice, is an attempt to find the best description of things; however, the 'things' I want described, and the descriptions I find more salient are those which are "closer to home." I'm no Quinean because I do not give primacy to the same conceptual schemes. To do so would be to abandon my grounds. Our language is crucial to the evidence we already have for things--to what we are committed to; we can't simply remove it, as Quine imagines we can, to see the world as (Humeanly) only behavior and "stimulations of the nerve endings." That may be an appealing minimalist picture, but once we've done that we're no longer talking about ourselves. That picture is wrong because that is not how we are. (Kant was right about this.) Philosophy needs to start in the middle of things, or else it will not address philosophical problems, which arise in us there; throwing out the distinction-richness of the language we ordinarily help ourselves to is a way of not listening to or answering the person asking the question. There is, then, a given; but it is not as simple as sense-data. It is the body of language we acquire and the distinctions we learn to make within, and through it.

I do hold that philosophical problems, to the extent that they are properly philosophical, arise from conflicts among concepts. Some of these conflicts are superficial, and can be undone. (And the wonderful thing about Wittgenstein's method of dissolution, at least in theory, is that it works from the inside. The way it works, in the ideal, is that you find that you are committed to these ordinary ways of speaking, and discover for yourself that the problem arises from a confusion. You discover commitments you already had--not unlike a kind of recollection.) Some of these conflicts are, I think, not at all superficial, and cannot be undone (in this category fall all philosophical problems describable as "conflicts of perspective (1st v. 3rd)). But they can be best described, and perhaps the best description will work some kind of knot-untangling magic; I doubt it, but we will be braver and less idle if we try. Not least because the countless philosophers out there not working in these frameworks keep describing things with abysmal impoverishedness.

James Conant distinguished between "Cartesian skepticism" and "Kantian skepticism" ("Varieties of Skepticism"). Each has many criteria, but roughly, the Cartesian skeptic considers what ought to be required to have knowledge and finds it lacking in even the most ordinary case, whereas the Kantian skeptic looks at, say, knowledge [though Kantian skepticism can apply to a vast range of topics], and is simply baffled as to how it is even possible. For the Cartesian "it looks as if there is something we cannot do," whereas for the Kantian it "looks as if there is nothing to do (not even dream), where we had previously thought there was something" (107/111).

I'd like to put a finer point on Kantian skepticism. I think the simplest incarnation of the Kantian skeptic (sorry, Kant) is the Tortoise, from Lewis Carroll's "What the Tortoise Said to Achilles". Presented with a simple inference, the Tortoise asks for further justification, to which Achilles can only respond by providing another inference, for which the Tortoise again demands further justification, ad infinitum. The Tortoise accepts nothing as basic. This is not unlike the difficulty of justifying an inductive inference without appealing to induction. Our Tortoise might as well be asking the Kantian question, how is modus ponens possible?



If this is the right way to think of Kantian skepticism, or Kantian questions, they start with an inability to see something as 'basic' (as 'bedrock') which must be seen as basic, or, perhaps, with a question that cannot be answered or phenomenon that can't be explained from within the framework in which it's asked. More modus ponens inferences will not enable the Tortoise to understand modus ponens inferences. What will help the Tortoise? --Some other kind of explanation. Probably exercises in aspect-perception. Phenomena as basic as modus ponens invite puzzlement; it's natural to want to know why they work. That which must be taken as basic invites the need to understand it the way we understand everything we understand through it.

Other predicaments that inspire Tortoise-like responses:

• basic actions--how do we make them happen?


• induction--how do we justify inductive inference?


• how do we know our own mental states?


• in what does following a particular rule consist?

There is more to each of these, of course, even by way of large patterns. For instance, I think it's helpful to think about both the problem of justifying induction and rule-following as drawing on an inevitable switching between foreground (a given instance or something that conforms to the inductive law or of following a rule) and background (the entirety of the inductive law, the entirety of the rule--which both cover an infinite number of cases).

I will not pretend otherwise: for me, all interesting philosophical questions are Kantian questions. I'm not sure I can defend this preference, except to say that a concern for what can be explained and how concerns inquiry in all domains--not simply philosophy; though wherever it arises, it is philosophy.