Will the real "epistemic closure" please stand up?

There has been considerable discussion in the blogosphere about the "epistemic closure" of modern American conservatism. It seems to have started with Julian Sanchez, who expressed concerns about AEI's treatment of David Frum (an issue that has been disconcerting for me as well, as a think tank employee myself). Douthat, Yglesias, McArdle, Friedersdorf, and many others responded to Sanchez's thoughts. Another major engagement has involved Manzi and Levin at The National Review, over a piece that Levin wrote on climate change. Bruce Bartlett has shared his own, very well known experience in the conservative movement. Ezra Klein adds some interesting (if limited) empirical evidence (I share his criticism of the method - but as he suggests, something similar could be expanded upon). There's certainly a lot that I've missed, but each of these links will lead you to other links - so you should be able to catch up (I haven't read through it all in detail - I've been putting my free time into studying for a final this week).

Anyway, through all of these really great thoughts, I think the meaning of the initial "epistemic closure" concern has been lost. I count at least three versions of "epistemic closure" available to us - two of which are being addressed in this ongoing debate, and a third that I think is especially interesting to think about.

Version 1 (Wrong and Uninteresting): "Epistemic Closure" as shorthand for various and sundry bitching about conservatism
One of my biggest concerns, however, has been that this whole debate has degenerated into complaining about conservatism, rather than providing a clear critique of conservatism. For Bruce Bartlett, conservatism's alleged epistemic closure amounts to: "They don’t think there are any new ideas of particular interest to them. Their philosophy is fully formed. The only question is how best to implement conservative ideas in the political debate". His concern seems to be the increasing activism of conservatism, and the rigidity of their value system. Others have talked about the "conservative cocoon" phenomenon, where conservatives (allegedly) have restricted interactions with those who don't share their views (in a similar vein, a friend from college shared this this interesting research with me on the closedness of the conservative blogosphere). Douthat focuses on the point of whether there are internal debates in conservatism, a question that is especially salient in light of the Frum Affair. But none of these address the more fundamental point of epistemic closure that was originally raised.

Bartlett's concern about activism and values is beside the initial point. Marxists, for example (at least the early Marxists) were unalloyed activists with a completely closed value system, but they were very thorough in thinking through epistemological questions. As for the "cocoon", the phrase "the ivory tower" is definitive of the cocooning of a group of thinkers - and yet no one accuses the academy of having an underdeveloped approach to epistemology. Douthat's concern with internal debate also has nothing to do with epistemic closure. There is no debate to speak of that 2+2=4 or that the Earth revolves around the Sun. Does that mean it is supported by a suspect epistemology? As I say, I haven't read all of this in great detail, but a lot of the debate seems to simply be a random assortment of complaints about conservatism, with the fancy phrase "epistemic closure" thrown in as a catch-all.

Version 2 (Wrong and Interesting): "Epistemic Closure" and the way that conservatives know what they "know"
Beneath the surface of these random complaints, there is real concern with and discussion of what Julian Sanchez initially meant by "epistemic closure" - namely, how conservatives derive knowledge about the issues that they pontificate on. Sanchez argues that accuracy of the knowledge inputs for the conservative movement are increasingly unimportant. Conspiracy theories, birth certificate skepticism, and crazy Beckesque guilt-by-association chalkboard flow charts have (so the argument goes) become sufficient factual fodder on which conservatives build their arguments.

The Manzi-Levin debate on the National Review is worth looking at as an example (well, at least Manzi's initial post is an example) of a genuine discussion of epistemological questions. Climate change is one issue where I think conservatives are pretty clearly ignoring the evidence to defend what is essentially an ideological position. Other issues, such as financial regulation, health reform, and fiscal stimulus are more dubious issues. I think it's hard to make a blanket statement that conservatives "ignore the evidence" more than liberals. Often they're taking the same evidence and simply applying a different set of values to it. Regardless, it's a fruitful question to ask - and there are a lot of conservatives, particularly in Washington, that do seem to be assuming their own conclusions rather than carefully reviewing the evidence.

Version 3 (Right and Very Interesting): The "Epistemic Closure" of Julian Sanchez's undergraduate subconscious
The New York Times reports Sanchez's confession that "he probably fished “epistemic closure” out of his subconscious from an undergraduate course in philosophy". Indeed he did. Sanchez used "epistemic closure" to describe the problem of "epistemological closed-mindedness", but "epistemic closure" is also a technical term in philosophy, which I only stumbled across recently while preparing for this blog post. The short definition of "epistemic closure" is:

"The principle that, where P and Q are propositions, if we know that P, and know that P logically entails Q, we know that Q"

In other words, "closure" refers to a closed system of knowledge, not closed-mindedness. I think it's fine that Sanchez didn't use "epistemic closure" in this technical sense, but it's interesting to think about the implications of this definition. The idea of "epistemic closure" is often closely associated with skepticism, because of the very real possibility that your knowledge of your initial statement (P), somehow (implicitly) presupposes the later statement (Q). The example given in this longer treatment of epistemic closure is:

"the proposition I have a driver’s license issued by the state of North Carolina entails that North Carolina is not a mere figment of my imagination"

The skeptical position would point out that your ability to make the initial statement presupposes the truth of the second statement. It's not exactly what Sanchez had in mind (and I should point out that I think what he did have in mind, while not exactly "epistemic closure", was also very interesting), but this does relate to a problem that I'm personally concerned about: the problem of deductively extended logical systems of thought. Theoretically, deductive logic is a fantastic source of knowledge. But it relies on two fundamental pillars: the accuracy of its axioms and presuppositions, and the validity of its deductions. The skeptical objection to arguments based on "epistemic closure" are very similar to my (perhaps less formalized) reservations about deductive logic and a disproportionate emphasis on rationality and reason in general.

Logical formality can amplify the errors of bad initial assumptions, while providing the veneer of incontrovertibility. As logical systems get increasingly extended, making broader and broader claims, the initial errors (which may have been negligible) can easily multiply into substantial errors. As Nikola Tesla once said: "Today's scientists have substituted mathematics for experiments, and they wander off through equation after equation, and eventually build a structure which has no relation to reality". In economics, I think the Austrian School and Real Business Cycle School are both excellent examples of Tesla's scientists who have "wandered off", confident in their deductive systems that bear no relation to reality. At least Real Business Cycle theorists will do empirical work to check up on things. The Austrian School explicitly refuses to do this check up.

I suppose I've wandered quite far from the point of conservative "epistemic closure" over the course of this etymological exercise. Are conservatives more likely to "wander off" with deductive arguments built from fallacious assumptions? Quite possibly. I think that's an open question.