We've all been to philosophy talks where the person asking a question at the end of the talk just goes on and on about something tangential, and it is not even clear what the question really is, and by the end of it you just think that person is showboating. O.K. If you were at the Taylor Carman talk about Heidegger and ontological difference last week at the APA, that was me. I was the irritating question guy.
[O.K. I will say this. Even though you really feel like an a-hole when you realize that you are that guy asking the rambling question, I'm not really apologetic, since I think my question foundered by having falsely presupposed that my interlocutors had read Saloman Maimon (apologies to anyone there who has read Maimon!), which in some ways was a charitable assumption. And I should note that by far the biggest surprise I've had since recently reading some of this Heidegger stuff is that phenomenologists (both "analytic" and "continental" Heideggerians) are almost always more ignorant about the period from Kant to Hegel than my logic type friends are. This continually blows my mind, given the stereotypes about the virtues and vices of relative kinds of philosophers. This is probably some combination of (negatively) phenomenologists understandably being marks for Heidegger's own dreadful history of philosophy and (positively) Graham Priest having turned the logic types onto the time period, many logic types finding history of philosophy a fun area where they don't have to worry about publishing anything, and the excellent current work both in the history of logic and in logically reconstructing past arguments that gets logic types fascinated by history. I don't know. But I would like to suggest a philosophical moratorium where nobody is allowed to write anything or present a paper until they have read all of Frederick Beiser's books. Obviously this is a self serving suggestion though, because in that possible world, I'm less likely to be the irritating question guy.]
But in any case one can (and I should have) frame(d) the criticism of Carman's position without reference to Beiseriana, because then relevant aspects of Beiseriana have been recapitulated in the last century. I think that Carman made a fairly common mistake in responding to Quine's critique of the a priori as a way to defend transcendental approaches. But Quine's critique isn't the one relevant to transcendental approaches anyhow. Actually there are two: (1) the Putnam/Wilson/Stich critique of scheme/content (analogous to Maimon's main argument in his Essay on Transcendental Philosophy), and (2) the Russell/Priest argument concerning the inconsistency of philosophies of finitude (analogous to Scholze and Reinhold's arguments). In my question I tried to discuss Maimon and Scholze, but would have done much better to discuss Wilson and Priest. Here I'll just explain the Putnam/Wilson/Stich point, mostly because I'm not confident I can reconstruct the dialectical context in which Carman said that there is nothing incoherent about positing a limit without transcending the limit (which is what Priest shows Russell's paradox to decisively undermine). I'll try to tie it back to Priest at the end though, if I can.
Carman was responding to Cristina Lafont's critique of Heidegger, where Heidegger's views are shown to founder on an equivocation. On the one hand Heidegger's hermeneutics pushes him to a kind of holism, but on the other hand he requires a strong form of the a priori/a posteriori distinction inconsistent with that holism.
Since Carmen's own work reads Heidegger as engaged in a kind transcendental psychology, it is really important for him to answer this critique. He did in two ways, one which I agree with and the other which I think involved a fundamental misconstruing of the twentieth (really the eighteenth and ninteenth!) century critique of the a priori. The first way was to criticize Lafont for taking Heidegger's notion of the a priori to be too linguistic. This seemed really important to me and right as far as it goes. But the second was to argue that the majority of Heidegger's insights can be kept if we allow the distinction between a priori and a posteriori to be a vague one, with a priori claims in principle revisable, and with the difference between a priori and a posteriori to be one of degree rather than kind (Heidegger's "ontological difference" becomes one of degree then).
This is I think a very important project, and Carman is a very good philosopher, so he had lots of interesting things to say about it. I don't know if he mentioned Friedman's Dynamics of Reason, but Carman will find a good ally there.
Unfortunately though, the project does not rehabilitate transcendental philosophy in the way Carman's Heidegger would like it to, because the problems with the a priori relevant to the justifiability of transcendental methods have nothing to do with the supposed sharpness of the distinction or with the supposed unrevisability (without changing the meaning!) of a priori claims.
I think the best place to start in thinking about this is by considering Mark Wilson's classic paper "Predicate Meets Property." Wilson's thesis of the "underdetermination of meaning" doesn't just argue that there are truth value gaps in cases where the meaning of a sentence doesn't make it clear whether the sentence is true or false. Rather, he argues that paradigm instances that are later taken to verify or falsify a claim are often of this type. His hokey example is of people who see an airplane for the first time. They could say, "Lo, a metal bird" or the could say "Lo, a flying house" and the circumstances in which they see the object first would probably determine which is correct. But in both evolutions of usage a bird is not a house, in neither do the people take their meanings to have changed, and moreover the object becomes a paradigm instance of either a bird or a house. I realize this seems hokey, but in Wandering Significance Wilson gives lots of examples from the history of mathematics, classical physics, and the study of folklore that instance this very thing. The world surprises us in various ways, and it's not determinant ahead of time how usage will evolve to deal with this, even when the new usage becomes canonical and we don't say that the meaning has changed (a result of Wilson's arguments is that sameness of meaning is not transitive, which is an interesting point in its own right; to me making vastly more plausible metaphysical views where identity of objects across time is not transitive; at some point I'm going to map the two literatures onto one another and get some cool results).
Putnam and Stich have noted that in a lot of Wilson type cases, it is completely indeterminate whether ontology (in the old fashioned sense of the inventory of the universe) changes or ideology changes. In one of his papers, Putnam argues caloric actually does exist, we just call it negative valence electrons (crap, maybe it's phlogiston? I forget). His point is that we could have responded by saying this was what we were talking about all along, and we just had some false beliefs about it. And in many if not most cases there are strong pragmatic pressures to do this very thing, such as successive models of atoms. I mean we could have said that quantum physics shows that the atom doesn't exist, but instead we say that we were previously talking about atoms but had some important false beliefs about them.
But the world doesn't force us to say one thing or the other. And, as Wilson shows in his book and Stich argues in the first essay of Deconstructing the Mind, this semantic slack is really important. We have to communicate with language, which involves presupposing falsely that there isn't so much slack, but we also have to keep communicating as our theories and the world change, which requires a lot of slack. But a constitutive part of this slack is slack about whether these changes involve a change in what we take to exist or a change in our beliefs about what we previously took to exist.
And now we can see why the Putnam/Stich/Wilson point was part of the resurgence of metaphysics in analytical philosophy (Putnam I think caused massive confusion in twentieth century philosophy by erroneously attributing his own arguments to Quine; people go back and read Quine's own writings because they are canonical and then get mired in much worse arguments than the ones actually presented by Putnam; Stich and Wilson do not make the mistake of reading their own arguments backwards onto Quine, though I have actually heard others wrongly do it on their behalf sometimes). Transcendental approaches, either linguaform or phenomenological, only work by walling off the empirical from the transcendental, and giving the philosopher some special access to the transcendental (whether via analysis or phenomenology). Philosophers don't concern themselves with what exists, only with a subset of what we say about what we take to exist. But the very idea of a distinction between what exists and what we say about it is what has been called into question (Sellarsians and Davidsonians have different arguments to this conclusion). So there is simply no way to wall off metaphysics (concerning what exists) from something more pure and transcendental. Whenever we pretend to back up to just talk about preconditions for our talking about or experiencing or using things we end up transcending those limits and talking about the things themselves.
Carman's point was that within the realm of what we say we can make distinctions that are close to traditional distinctions made by philosophers concerning the a priori and a posteriori. And this is all good and true, but that wasn't the challenge to transcendental anti-metaphysicians like Carman (or Carman's Heidegger). We make all sorts of analogous distinctions to traditional versions of the a priori/a posteriori for all sorts of reasons (though it is highly contextual; talk to a lexicographer who has worked for two different dictionary companies and I guarantee that you'll never see these things the same again), and that is fine. But they won't sustain the anti-metaphysical stance implicit in Carman's ontological anti-realism/ontic realism approach to Heidegger, or any such transcendental methodology, because the phenomena we have to react to is such that current practices (linguistic and non-linguistic!) radically underdetermine whether our reactions should involve revising our view of what exists (content) or keeping the content and revising our beliefs about that content (scheme). And again, Wilson shows convincingly that classical mechanics, mathematics, and the study of folklore simply would not have been possible if our practices didn't have this radical slack. So as philosophers we just can't transcendentally wall ourselves off from content, for example, with Heidegger talking not about objects themselves but just with how we are forced to engage with them.
I should close by reiterating that I do think Carman's Heidegger's Analytic is both a great piece of Heidegger scholarship (though I'm no expert) and a really important work in the philosophy of mind in its own right (AOC here at least). I do, however, think that Phillipse is much closer to the truth when he writes that early Heidegger suffers from a basic incoherence concerning what "the problem of the meaning of being" is supposed to be, stemming from Heidegger's own confused conflation of transcendental psychology and Husserlian regional ontology. The end result is that Heidegger himself was inconsistent, and those of us who want to learn something from him have to be selective about what we take (though to be fair, Mark Okrent has a really nice story about "the turn" which comes pretty close to positing that Heidegger was aware of this inconsistency). So at the best you get really good metaphysics such as Graham Harman's or really good philosophy of mind such as Carman's, each selectively reading.
But at the end of the day, with Harman and Putnam and Stich and Wilson and Graham Priest (and against Carman and Brandom and McDowell), I think philosophers do much better to talk about the being of real things like musical pieces, numbers, atoms, economic classes, video games, paintings, etc. and not wall ourselves off by trying just to talk about ourselves. The post Kantian moral was that no such walling off is defensible, and it is depressing to that it was so badly forgotten in the "back to Kant" movement that gave birth to both Carnap (Marburg School) and Heidegger (Southwest School) and that it remains forgotten in places as diverse as Berkeley and Pittsburgh.
[I must again note that above I didn't even touch on the important anti-transcendental argument scheme that Graham Priest elucidates in Beyond the Limits of Thought, where he is able to tell the history of philosophy in terms of the inability to describe a limit without transcending that limit. Carman's talk also contradicted Priest's arguments. At one point Carman flatly stated that one can describe a limit without transcending that limit. Priest actually locates an instance of his argument against this very claim in Hegel (though I don't remember if Priest traces their genesis from the early critics of Kant through Fichte to Hegel). It is interesting that the Russell/Priest argument scheme is independent of the scheme/content considerations I presented above. Again, in the early Kant critics, Maimon first problematized scheme/content and Scholze and Reinhold first made something like the Russell/Priest argument. I only say this to note that defenders of transcendental approaches need to deal with both scheme/content considerations and the Russell/Priest argument to the conclusion that any positing of a philosophically interesting limit presupposes (in the intelligibility of being able to state the limit) the transcendence of that very limit.]
Recent Comments