Saturday, January 25, 2020

The Cofnas Controversy

There was a bit of a storm in online philosophy over the past week. It was mostly just a mess, but has some interest as an example that raises professional ethics issues in the context of academic philosophy. Nathan Cofnas, a graduate student at Oxford, had recently published a paper on research ethics on genetic study of racial differences in the journal Philosophical Psychology; Mark Alfano, a professor at Delft specializing in moral psychology, worked up a petition and a blogpost and went to Twitter:



The hubris of this is rather remarkable. Philosophers have never gotten together and elected Alfano the Grand New Aristotle to be able to call boycotts; it's one thing to boycott a journal oneself, and make known why, and another entirely to think a general boycott is something you can call. In any case, what we have here is an attempt to escalate immediately out of the starting gate. The normal and ordinary professional course in dealing with a paper that is in some way seriously bad is to publish a responding paper that shows it. A paper that is significantly bad might conceivably lead one to petition the editors to invite a formal response in the journal itself. Because philosophers are so diverse, and use such diverse methods, retraction in philosophy is a relatively unusual thing, and the most common reason for it is discovery of plagiarism; it typically involves some sort of clear violation of professional ethics. Calling for a retraction rather than a response is already quite strong, but, of course, Alfano's call is not even as modest as that, since it is in fact a call to punish the editors for publishing it, on the grounds that it showed a failure of peer review. Public boycott is a punishment of reputation; since philosophical journals exist as a part of the reputational economy of academic philosophy, public boycotting is an attempt to black-mark both a journal and its editors, and this is explicit in Alfano's characterization -- the editors are to be black-marked for incompetence, and must prove to the satisfaction of others that they will be more competent in the future. It is a mind-bogglingly arrogant demand. Academics assess each others' incompetence all the time as a matter of private judgment; they may at times be public about their private judgment. But to demand of colleagues that they admit that they are incompetent, apologize for it, and provide you with a plan to prove that they will stop being incompetent, and to call on the rest of the philosophical community to force them into a position in which they have to do this, is something that no one with any sense would ever, ever do. Moral turpitude or violation of professional ethics might call for something like that, but even then you wouldn't escalate to that right at the beginning.

And the ground for it is absurdly weak. The claim that a paper represents a failure of peer review and that therefore the editors must apologize and retract the paper, perhaps (as the petition suggested) even resign over it, is thoroughly inadequate. Peer review is, first of all, not a system for preventing bad and inaccurate papers from being published; it's a system for raising the bar for publication in such a way as to improve the overall quality of scholarly journals. No way of refereeing submissions you could design would guarantee that all the papers published are accurate and not flawed; you certainly can't do it under current academic refereeing conditions. Thus from a single failure, one cannot in any way determine whether a journal's refereeing and acceptance practices are flawed. Even setting aside the fact that the referees and editors may have read the paper more charitably than Alfano does, and taken it to be making more modest claims than he does, editors and reviewers will make mistakes. If a very bad paper gets through, it's reasonable to suggest that the process should be reviewed, just in case; but such a review may or may not find anything. To conclude from a single case that a journal's practices are flawed is an illegitimate inference. And even if you had an excellent argument that it was flawed, you would not usually escalate immediately to a boycott; the normal and professional course would be to start by calling the attention of the editors to the problem. Perhaps you could justify, depending on the case, petitioning the editors to clarify their review process. But the petition itself merely says that the editors should at least have requested further revision; this is indeed probably the strongest that can be said on the evidence available, and it is well short of anything that would justify what was demanded by the petition. It's another example of the extraordinary arrogance in play here.

Alfano did not help matters at all by his direct behavior to Cofnas, at one point on Twitter boasting that he would destroy Cofnas's reputation, which shows, I think, that he has very nonstandard views of the professional ethics of his own profession. Graduate students are not immune to criticism, even vehement criticism, of their arguments or of their professional behavior by normal channels. In the philosophy profession, however, it's pretty widely accepted that the professoriate has an obligation to cultivate graduate students as junior colleagues, rather than bully or intimidate them, and to make an effort always to criticize in a way that allows improvement. The norms for this are not very standardized or rigid, but there's not any room for doubt that threatening to destroy their reputations and ruin their careers is a violation of this obligation. And, indeed, I know colleagues who were infuriated by Alfano's handling of the matter because it forced them, in the context of a paper with whose argument they had no sympathies, to step up and make sure that the graduate student nonetheless was not ruined by the unprofessional behavior of the professor and was given as much benefit of the doubt as could be given.

Much of the controversy, of course, and much of Alfano's self-righteousness in the pursuit of it, seems to have been due to issues of racism. As the petition says, "the paper disingenuously argues that the best explanation of differences in IQ scores between racial and ethnic groups is genetics." (Whether the paper should actually be read this strongly is part of the controversy, but the point is that the racial issue is front and center, despite the fact that the argument was rationalized in terms of professional norms of peer review.) Alfano, I think, saw himself as taking the anti-racist stance; but whether he did or not, some of his supporters very certainly did. This is in fact not true; the stance explicitly taken is that differences in IQ scores between racial and ethnic groups are due to environmental factors like lead poisoning, which is not an anti-racist position. If, for instance, "Women are overemotional because of their biology" is a sexist statement, it would not be reasonable to conclude that "Women are overemotional because of their environment" is therefore not a sexist statement; the thing that made the first statement a candidate for being a sexist statement is carried over into the second, and they are just disagreeing about whether the best sexism is biological or not. You can, of course, have alienans explanations, explanations that subvert the phenomenon. For instance, "Women are 'overemotional' because people arbitrarily redefine 'overemotional' to mean behaviors of women they don't like" gives an alienans explanation; the explanation if true would mean that women aren't actually overemotional, that their being classified as such is due to defective classification. An example of an alienans explanation in the case considered by Cofnas would be, "The best explanation of differences in IQ scores between racial and ethnic groups is that IQ testing is defective in some relevant way". But "because of environmental factors" is not an alienans explanation; if something is explained by environmental facts, it has to be really the case. Thus it seems that if "The best explanation of differences in IQ scores between racial and ethnic groups is genetics" is racist, it follows that "The best explanation of differences in IQ scores between racial and ethnic groups is environmental" is also racist, because they just disagree on how the phenomenon arises, not on the phenomenon itself.*

In any case, although it's not a situation in which anyone covered themselves with glory, the whole situation is a good example of one reason why professional norms develop, namely, to handle problem cases in ways that don't make them one-man crusades, which are often liable to bad judgment, and to prevent arbitrary escalation to extremes by individuals and small groups. It also shows that overreach and lack of professionalism can make one's actions much less effective.

----
* My own view is that the most plausible explanations in this kind of situation are alienans -- in particular, I suspect that racial and ethnic differences in IQ scores are likely an accidental artifact of which kinds of intellectual abilities we take to be easily quantifiable and how we score them relative to different broader populations, so that IQ scores are not always describing the same things in the same way; such a view would be consistent with the position that IQ does measure something that is a proxy-for-intelligence-under-very-specific-and-limited-conditions, but that no definite general conclusions can be drawn from populational differences in scores, and that especially no definite general conclusions about intelligence itself can be drawn from them, since that is talking about something well beyond those specific and limited conditions. I don't think there's much sense in asking whether a particular score on a test is due to genetics or to environment; it's due entirely to having taken the test a certain way. And I don't find plausible the view that we have already in hand any way to think coherently about these matters at the level of whole populations. But, of course, what I find plausible has very little relevance to anything.