Posts from August 2010

Comment on One Libertarian Seminar Ends, Another Begins by Rad Geek

MBH:

(1)* The analytic-synthetic dichotomy implies psychologism.

Alright, but what’s the argument for that? Are (1a) and (1b) below:

(1a) If an analytic-synthetic continuum is the case, then a belief that p will not necessarily cause a belief in all that p entails. For instance, p may entail q but q may be believed because of empirical evidence that q — not automatically by the belief that p.
(1b) An analytic-synthetic continuum is a frame that disallows psychologism.

… supposed to be a lemma to establish that (1) is true? If so, then it seems that you’ve just given me the following argument:

(0) If ASC, then necessarily not P.
Therefore, (1) if not ASC, then P.

… which is simply a formal fallacy. Am I missing something? If so, what? If not, how is this any kind of argument for premise (1)?

(2) Under psychologism, if p entails q, then the belief in p will cause the belief in q.

I don’t know of any psychologicians who believe that this is true in the unqualified form you’ve just stated it. (For example, lots of people believe some stuff about probability which is sufficient to prove that switching is a winning strategy in Monty Hall scenarios. But very few of them believe that switching is a winning strategy, unless you try very hard to show them how this can be so. Psychologicians don’t expect that people who believe in a set of premises always as a result believe in everything that the premises jointly entail! In fact, they more or less always (1) qualify the claim by saying that the causal connection only holds under a set of ideal conditions for cognition (e.g., the believer has to have time to think about it, has to be paying attention, etc.) and (2) insist that the causal connection between believing that p, and believing in q, under those ideal conditions, is in any case only an empirical generalization, which may admit of exceptions, not a necessary connection. But if anything, genuine impositionists (like the impositionist reading of Kant, or like Mises) want to insist on a necessary relationship, not an empirical generalization; which is part of the reason why so many of them (Impositionist-Kant, Mises) tend to be so fiercely anti-psychologistic.

; they generally hold that this kind of epistemic closure is qualified by a set of normative cognitive conditions (the believer has to have adequate time to think, a certain degree of intelligence, etc.), and, furthermore, they hold that this causal relationship between believing that p and believing that q, under the normative is an empirical generalization which may admit of exceptions, not a necessary connection or universal law.

(3) Believing = Seeing. (You can only see what you believe is there.)

I have no idea what you mean by this. = suggests equivalence or identity; but your parenthetical gloss looks like a statement of necessary conditions. In any case, when you say “seeing,” are you actually referring to sight here? I.e., is (3) intended to say exactly that S counts as (literally) seeing O only if S believes that O exists? That’s what your later references to perception seem to suggest. But if so, then when you come to:

(4) If p entails q, then an analytic a priori judgment that p allows for the perception that q

… you’re stuck with a real hot mess. I guess this is intended to be a simple conditional within a pair of universal quantifiers over all p and all q, but in order to actually get this anywhere near being adequately formulated, you’d actually embedding a lot more in the way of conditionals and quantifiers (e.g., “an analytic a priori judgment that p allows for…” is going to mean something like, “If p entails q, then: if q is an observation report of some S seeing some O, then (it is possible for q to be perceived if: p is an analytic judgment and there is an apriori justification for p).” But in the rest of your argument, you fallaciously convert the italicitzed “if” to an “only if” (by introducing a “would otherwise be disallowed” that was nowhere in the previous premise); you also act as if you already have all the antecedents of these embedded conditionals already satisfied, and can proceed to assert conclusions that you inferred from the consequent (that a perception that q is presupposing a belief that p). But you haven’t satisfied them. You’ll need to actually give an example of (1) a p which is an analytic judgment with an apriori justification; (2) a q which is an observation report to the effect that some S sees some O; and then (3) show that that analytic truth logically entails that observation report (!). Which seems to me something you’re not likely to be able to do; that’s not the sort of thing that analytic truths normally entail. But if you don’t have at least one p and a q you can give us, then you have no basis for inferring (6); you’re just asserting something you inferred from the consequent without first having satisfied the antecedent of the conditional.

Or maybe you meant something else by “allows for the perception” than this sort of observation-report, but then it’s not clear at all how (4) is supposed to be connected with the previous premise in (3), if (3) is supposed to have something to do with the factivity of seeing. Of course, as I said, I’m really pretty baffled as to what (3) is supposed to be saying; so perhaps (3) also means something different, which is connected. But what, then?

Me:

Frege definitely endorses a version of the synthetic-analytic dichotomy. . .

MBH:

Maybe on the surface he does. I’ll take your word for it.

Well, you don’t need to take my word for it. He discusses it in Foundations of Arithmetic, among other places.

But his destruction of psychologism hammers the synthetic and the analytic together.

Maybe, but I don’t see how that pertains to the question. I think Frege also has commitments which undermine reflectionism, but the fact remains that, nevertheless, if you asked him for his views on the Third Realm, he was very ardently committed to a picture that looks an awful lot like reflectionism.