Sunday, November 08, 2015

Denialism

Lee McIntyre has an interesting although not, I think, entirely successful discussion of the distinction between 'skepticism' and 'denialism' at "The Stone". 'Denialism', for those who don't know, is a word that has become popular among people who like to call themselves skeptics but wish to distinguish themselves from other kinds of people that almost everyone calls skeptics. Because it has mostly arisen as a grassroots term there is, as one might expect, some difficulty in giving the distinction a consistent account: sometimes it's made in terms of how people argue, other times in terms of how they investigate, other times in terms of standards of evidence, etc.

McIntyre gives a number of characterizations of skepticism and denialism in the course of the article; it might be handy to sort them out a bit.

Skepticism

* when we believe or disbelieve in something based on high standards of evidence
* must be earned by a prudent and consistent disposition to be convinced only by evidence

Denialism

* when we are just engaging in a bit of motivated reasoning and letting our opinions take over
* when we refuse to believe something, even in the face of what most others would take to be compelling evidence
* usually have different standards of evidence for those theories that they want to believe...versus those they are opposing
* when we cynically pretend to withhold belief long past the point at which ample evidence should have convinced us that something is true
* doubting the overwhelming consensus of scientists on an empirical question, for which one has only the spottiest ideologically-motivated “evidence”

There are three interesting things about this rough sorting of the list. The first thing is that the characterization of skepticism seems to require that you cannot call yourself a skeptic unless you have good higher-order evidence that you are intellectually virtuous: your belief/disbelief is based on high standards of evidence and you are dealing with evidence in a consistently prudent way. It makes skepticism a characterization of someone's character. It's unclear what else is supposed to be added to skepticism beyond this consistent rigorous prudence in inquiry, if anything, but the characterization does require that calling oneself a skeptic requires attributing virtues to oneself.

The second interesting thing is that the characterizations of denialism are rather different. For instance, the first, third, and fourth are psychological; the second and fifth are sociological (they refer to a larger population -- "most others" in the second and "scientists" in the fifth). This is not necessarily an inconsistency, since it's clear enough that some of these are probable 'warning signs' of denialism, or specific kinds of denialism, rather than essential to denialism as such. Another difference is that the first attributes denialism to motivated reasoning -- which everyone engages in and which has to be compensated for since it can't be eliminated -- while the fourth attributes it to cynical pretense and the fifth attributes it to ideology. These are very different kinds of motivation that can be combined or separated in different ways; for instance, they do not reflect on one's character in the same way. Presumably these are just common motivations rather than integral to denialism itself, but they don't really give us much of a clue as to what causal account is going on here, beyond the fact that they are all leading us astray somehow. But it does seem clear enough that denialism, opposing skepticism as a kind of good character, has to be a kind of bad character.

The third interesting thing is that from none of these characterizations is it possible to determine whether someone is a skeptic based simply on the topic or any position they take on it. Even the second and fifth characterization depend on things that are not intrinsic to a topic -- what most others take to be compelling can change (and can even depend on who you are counting as 'most others'), and scientific consensus builds slowly and occasionally shifts. Nothing about this account prevents the skeptic from being highly dogmatic, as long as this dogmatism is prudent and based on high standards of evidence. Rather interestingly, the one time McIntyre uses the word 'doubt', he is talking about denialism, not skepticism. Nothing about this account prevents someone from sliding from skepticism to denialism and back again, although presumably the 'consistent' in the account of skepticism prevents it from happening in the course of a single argument.

Where I think it starts getting a little troubling is that a highly moralized account of the distinction, like this one, makes it problematic to use. If skeptics hold themselves to a high standard of evidence, then they can't call someone a denialist unless they have very good evidence that the other person is not reasoning virtuously, and they can't call themselves skeptics unless they have very good evidence that they are in fact reasoning virtuously, and not influenced by ideology or uncompensated-for biases. This is very much like a Kantian approach to morality: you are not allowed to act on incentive, at all, and the only way you count as virtuous is if you act for duty (here it is evidence rather than duty) alone. And it runs into all parallel issues, including the famous one noted by Kant himself: it becomes unclear how you would know you were being a skeptic rather than a denialist, because how often could a human being in most situations actually be sure, to a high standard of evidence, that he or she was driven by a disposition to be convinced only by evidence at every essential stage of inquiry?

It would be possible, of course, to bite the bullet the way Kant does -- you are unlikely ever to know for sure, and it doesn't matter whether you do -- but McIntyre began his discussion by saying that it was in fact important for us to be able to tell when we are skeptics and when we are denialists. This is precisely the one thing we don't ever learn -- McIntyre just never says how we determine that we are virtuous in this context, and while he gives some things he thinks are signs that someone is vicious, it's not very clear how general or reliable these signs are, or how we know enough, to a high enough standard of evidence, to be able to say who counts as a denialist in order to discover these signs in the first place. (Perhaps they are grounded conceptually rather than empirically? I don't know.)

None of this is necessarily fatal, of course; the distinction in the highly moralized form McIntyre presents it just never seems to be shown to be a consistent, well formed distinction that could actually be useful.