by Dave Maier
Last month, a couple of commenters on my post on Dennett's plea for “respect for truth” asked what pragmatists I like, and for a general elaboration of my pragmatism. I had to think about the best way to respond, and this is what I have finally come up with. Sorry for the delay!
There's a famous article called something like “Thirteen Types of Pragmatism”. This is a typically pragmatist attitude: forget universal definitions, just tell me what we've got. That's what we'll do here; but even so, we'll be keeping an eye on why we want to call these things “pragmatism” at all. After all, that same attitude tells to abandon “pragmatism” if it stops being useful, and I'm happy to call myself something else if it helps.
1) Pragmatism as practice over theory
We don't need to be exhaustive here, just to link some ideas together. A good place to start in characterizing pragmatism is the ordinary not-necessarily-philosophical idea of giving priority to practice itself over any theoretical understanding of that same practice. Pragmatists of this sort say: forget the operating manual, just do what experience tells you about what works. Engineers revel in the perceived virtue of this attitude: dirt under the fingernails and all that.
2) Pragmatism as science over metaphysics
But this is stacking the deck. Naturally even theorists recognize the priority in this sense of that which is represented over the necessarily merely derivative representation of same. If the manual says the engine will explode if you use such-and-such type of fuel, but experience shows otherwise, then the manual is wrong. A related but more philosophically consequential pragmatist attitude pits the empirical world of our experience against a purported “metaphysical” world of abstractions and essences.
There's only one world, of course; the question instead concerns the best method of investigating it. Theory is okay on pragmatist grounds if it is scientifically respectable, as after all Newton's laws of motion are as theoretical as you can get, and we're not giving those up. This is better than the previous thought, but it still stacks the deck. Here too metaphysicians recognize the importance of connecting what they say about the world with what we experience. Still, metaphysics isn't science, and can't be dismissed on those grounds alone, if at all.
3) Pragmatism as anti-Cartesianism
On the other hand, “metaphysics” remains a natural term of abuse for pragmatists, as is the idea of a world beyond experience. Historically, pragmatists have attacked that “metaphysical” idea most directly in its Cartesian manifestation, for example as implicated in that version of skepticism. Cartesians are best known for mind-body substance dualism, but they don't need that particular idea to motivate their skepticism. All they need is a more general conceptual dualism of subject and object.
Now this sort of pragmatism looks like the kind I like: the kind dedicated to rooting out the many pernicious manifestations of that Cartesian idea. Unfortunately, however, in combating each of these one by one (e.g. skepticism in particular), by my lights pragmatists have often failed to stay focused on the dualism itself. I actually think this was unavoidable, and I don't want to take any credit away from our honored ancestors. I just want to distinguish this sort of pragmatist “anti-Cartesianism” from later versions (like mine, below).
So again, the problem with Cartesian skepticism is the underlying conceptual dualism; but in any case we must address it epistemologically. Most of our knowledge is defeasible: its truth is not guaranteed, by evidence or conviction or anything else. So why doesn't that mean that “it might be false” — and thus that we shouldn't believe it, but instead remain uncommitted? The characteristically pragmatist answer to this is to argue that doubt must be justified just as much as belief does. However, the Cartesian withholds belief solely on theoretical grounds, i.e. that the objectivity of the world means that any defeasible belief of ours could be false without our knowing.
The original pragmatist, C. S. Peirce, attacks this attitude as a pretence, or “paper doubt”. We may say we are in doubt, but when push comes to shove, or even before that, we find that we act just as if we knew — which for a pragmatist means that we believe after all. All we need do is recognize that our beliefs might not be fully true, but hold on to them anyway if that's what we need to do. This has been a very inflential take on the problem, even beyond pragmatism, but we can't leave it there. “Fallibilism,” as this view is called, seems to me simply to amount to skepticism in the end. How can we say we believe, if at the same time we hold back from full commitment?
4) Isaac Levi and pragmatist epistemology
Isaac Levi is a contemporary pragmatist who was a teacher of mine at Columbia (home of many famous pragmatists, including John Dewey and Sidney Morgenbesser). His views are not well known, which is a shame, as I think his take on properly pragmatist epistemology is just what we need (the pictured book rocks). Levi regards knowledge, perfectly pragmatically on my view, as “a resource in inquiry and deliberation” – something we use to decide what to believe (that is, how things are) and what to do.
A resource is a tool, and not every tool is a mirror or even a map; yet surely our resource won't be very resourceful if there's no connection to the world at all. And knowledge, as a species of belief, necessarily (if not uncontroversially) aims at truth. Here in a nutshell is the pragmatist attitude, sort of the inverse of Ronald Reagan's attitude toward arms control: verify, but trust.
Levi's main innovation in this context is a correction of Peirce's fallibilism. While fallibilists claim to believe, they allow that what they believe “might not be true” — and thus that they are in doubt after all. Levi's “infallibilism” takes belief seriously as a commitment: as far as I'm concerned, my beliefs are true with probability 1; anything less isn't belief at all.
Two comments about this before we move on. Most opposition to infallibilism, even among pragmatists, comes from the utility (naturally enough) of the idea of degrees of belief, which can be used, if quantified, in probability calculations of the Bayesian sort. Theoretically, however, this latter idea is problematic in its capitulation to skepticism (and also see below). In any case Levi deals with this in his discussion of what he calls “credal states” (which I never actually studied in any detail, so let's leave it there).
Also, infallibilists must still make sense of the corrigible nature of belief: that we (sometimes) respond to new evidence by changing our belief commitments. Levi's “corrigibilism”, however, in acknowledging this, simply registers the theoretical possibility of error, without allowing that to affect our full commitment to our current beliefs.
Okay, one more thing. I can't give this a full defense, but another important aspect of Levi's views is his formal definition of knowledge, in line with his conception of it as a resource, as merely true belief, rather than the usual definition as justified true belief (JTB). For after all, if I believe something I must have already taken it to be justified, making the latter idea an idle wheel. More precisely, Levi doesn't abandon the idea of justification, but sees it as applying not to belief itself, but instead to change in belief (and thus, in line with the original pragmatist idea, as applying just as much to giving belief up as to adopting it). Not only does this make more sense, it dissolves into nothing the famous “Gettier problem” in epistemology, which has had philosophers looking for another condition on knowledge to add to JTB rather than taking justification away.
5) Pragmatism as anti-dualism: post-Davidsonian/post-Wittgensteinian approaches
As I mentioned above, I construe the Cartesian heresy broadly as the conceptual dualism beween subject and object. We've already seen it as mind-body substance dualism, and as the epistemic gulf responsible for Cartesian skepticism, but this thing has roots leading all the way back to the Platonic Ideas (or, paradoxically, to the Parmenidean One). Here I follow Richard Rorty, who deserves a lot more credit than the abuse he got from both sides: from traditional pragmatists for changing its focus, and from traditional non-pragmatists for all the usual reasons. Even his biggest fans have to admit that Rorty made some mistakes; but he brought some crucial, even game-changing new voices into the conversation.
Reading Rorty turned me on to Donald Davidson, whom he interpreted not (simply) as the post-Quinean analytic philosopher he seems, but as a pragmatist in this extended sense. On this reading, Davidson's attack on what he called the “dualism of conceptual scheme and empirical content” is the natural next step, after Quine's naturalistic rejection of the analytic-synthetic distinction, in the continuing rollback of Cartesianism. Davidon always resisted the pragmatist label, but surely our concern is not the name, nor even Davidson's own opinion of the significance of his views, but what use we can make of his innovations for our continuing project.
What use is that? What I like here is the way Davidson's conception of inquiry (belief) as essentially connected with interpretation (meaning) shifts from seeing language as a (semantic and epistemic) intermediary (a conception which leads directly to skepticism) to seeing it as a medium of experience and thought. This “triangular” picture (speaker — interpreter — world) is the ideal stage for a resolutely anti-dualist picture.
Let's leave the details of Davidson's death blow to the Cartesian picture for another time. The other key philosopher Rorty brought into the picture was Wittgenstein (he was also keen on Heidegger, but I'm not feeling it there). Here too he got abuse from both sides: Wittgenstein acolytes not happy with Rorty's appropriation of their hero as a pragmatist (of all things), and everyone else for, again, the usual reasons. Now it's true that Wittgenstein is not a pragmatist in any usual sense; but I think he provides a few key parts to our developing anti-Cartesian conception of philosophy.
Somewhat paradoxically, we turn to Wittgenstein most effectively after we've brought Davidson into it. At this point the dualist is on the mat, but it's really easy to let him slip out of our grip. For we've already allowed that however useful our beliefs may be, as beliefs their truth can always be brought into question. So how about Davidson's philosophical theory of interpretation itself? Is that true?
Wittgenstein is usually interpreted as rejecting philosophical theory entirely (as a philosophical “quietist”). But we do better to extend our anti-dualist attitude to the next level: the dualism between philosophical theory and everyday statements of fact. Wittgenstein's early work seems to emphasize, not attack, that distinction; but even in the Tractatus he was concerned to affirm the importance of that which is left after we have restricted intelligible discourse to the latter. It's natural to read Philosophical Investigations as the appropriate correction: theoretical discourse isn't meaningless, it's just another form of language. Just because it's about another kind of language doesn't mean it's not also about the world (and can thus be considered true in a natural sense). But here too we'll have to come back to this some other time.
6) Pragmatism as pluralism: interpretivism and perspectivism
At this point (i.e. even with Wittgensteinian help) we're still grappling with the proper status of philosophical belief. It seems like it needs to be true, on pain of irrelevance, or unintelligibility as belief; yet this need seems to reinvite skeptical attack. Again, the positivist strain in early pragmatism would have us reject “theory” in favor of scientific practice, which is necessarily empirical as opposed to a priori (which, I grant, in some contexts can be just the ticket). But another way to oppose “theory” [or “metaphysics”) is to oppose metaphysical realism, which demands that using philosophical theory at all requires that it mirror reality in the disputed sense: one which means that we can't believe something in one context but not another based on our needs. (It can't be true and then not true simply because of how we feel like talking! That's idealism!)
Here again we may turn to Levi's conception of knowledge — here applied specifically to philosophical theory — as a resource or tool. We have plenty of tools in our toolbox without demanding that each of them be useful in all cases. We may appeal to Davidson at one point and Wittgenstein at another, without demanding that their views be melted down into an amalgam for universal use (as it sometimes seems that John McDowell, another philospher I admire, is trying to do). The trick, again, is to reconcile this sort of pluralism with the proper sort of respect for truth, which I discussed last month. I've also discussed this idea under the name “perspectivism” in my posts here on Nietzsche (like this one, and earlier here). When Nietzsche commends the “healthy” individual as keeping his “Pro and Con” under control, this is what I see him doing: taking control of truth, using it in service of knowledge rather than meekly attempting to mirror it as metaphysics (that is, metaphysical realism) demands.
So I hope that helps situate last month's remarks with respect to pragmatism. I'll stop here, but no doubt there will be plenty of opportunity later on, when I try to defend my quasi-pragmatist uses of Wittgenstein and Davidson, to hold my feet to the fire.