Wednesday, October 11, 2006

A Priori Paper

I'm posting a draft of my new paper on the a priori. Comments very welcome, as always.


Joachim Horvath said...

Just a few comments on parts I and II of your paper:
- You base a priori conceptual knowledge on introspection of concepts understood as mental representations. However, for many introspection is clearly an empirical source of justification because it delivers information about highly contingent states of affairs, e.g. that I'm hungry now or that my toe hurts. What's your response to them?
- You define a priori knowledge as "knowledge which is justificatorily independent of evidence from sense experience". Two remarks here: It seems better to conceive of the a priori primarily as an issue about justification, not knowledge (e.g. to enable fallibilism and to avoid Gettier problems). And then, why not just follow Casullo's very carefully worked out account of it: A priori justification as justification by some non-experiental source?
- You list Boghossian under the rubric of non-realism about the subject matter of the apriori. Now, that seems plainly wrong to me, since the very point of his rejection of metapyhsical analyticity seems to be his realism about subject matter!
- Further, you also list BonJour as a realist proponent of the conceptual approach which also seems wrong since he spends a lot of pages on criticizing attempts to explain a priori justification in terms of analyticity (i.e. in terms of conceptual truth)!

Joachim Horvath said...

Some more comments on parts III to V of your paper:

- In part IV, you suggest the following: "I think knowledge through concept examination is possible because, due to our sensory contact with the world, our concepts come to encode trustworthy information about that world, and we can recover this information by examining those concepts."
My objection: If our concepts only encode sensory information about the world, then it seems that we can only learn something about the actual world by inspecting them. What we often aim at, however, when we're engaged in conceptual analysis is how things necessarily are, e.g. that knowledge is necessarily true belief and not just in the actual world! So, the whole modal dimension of conceptual knowledge seems to be not accounted for!

- Somewhere in IV you say: "If Gullible has no reason whatever to trust Guru, but does so anyway, then the mere fact that what Guru says is true does not mean that Gullible ends up knowing it." Yet, an externalist about justification could just claim here that if it is in fact truth-conducive to trust Guru then no further reason is necessary for Gullible to end up with justified beliefs or even with knowledge! Question: Do you tacitly assume internalism about knowledge here?

- At the end of IV you make the following rather bold claim: "On this view, to say that knowledge obtained through conceptual examination is non-empirical knowledge is rather like saying that knowledge obtained by observing that p and deducing from p that q is non-empirical knowledge because the second step doesn't involve the use of the senses." However, on any traditional account of the a priori, no conclusion from empirical premises would count as a priori justified - so this looks a bit like trying to change the subject here (in order to avoid the next problem - see below)!

- At first glance, your variety of the conceptual approach seems to be best classified together with Mill and Quine: As denying conceptual knowledge any epistemological status which is in principle different from ordinary empirical knowledge. To that charge you reply in the final section: "What is important is that it preserves the idea that the subject matters which have been called a priori do indeed have a special epistemological status: they are knowable in a special way, involving conceptual examination, which most things aren't." To this claim I have two objections:
First, this seems to be a rather weak version of "knowledge in a special way" on which even gathering evidence by examining newspapers may count as a special way of coming to know (but maybe you're willing to accept this slightly trivialising consequence)!
Second, it's not at all clear that the worldly fact that vixens are female is only or even specially knowable via examining our concepts - it seems to be equally knowable by induction, for example. And the same goes for other conceptual knowledge!
Possible Reply: But via induction we only come to know that actually, vixens are female, while via conceptual examination we learn that necessarily, vixens are female.
Rejoinder: But that sits very uneasy with a purely empirical input step with regard to the information that is encoded in our concepts. It's fairly unclear how and at which point the modal information is supposed to come in, given such an account!

By the way, despite my foregoing objections I liked your paper very much: it's right to the crucial point, very crisp and highly thought-provoking!

Carrie Jenkins said...

Thanks Joachim - lots of helpful comments - I will try and respond to some when I have time!