No, we can't censure people for ad hominem attacks in scientific discourse.

“Individuals engaging in ad hominem attacks in scientific discourse should be subject to censure.”

That quote is taken from an editorial about errors in science that recently appeared in the Proceedings of the National Academy of Sciences (Issues with data and analyses: Errors, underlying themes, and potential solutions. Andrew W. Brown, Kathryn A. Kaiser, and David B. Allison. PNAS March 13, 2018. 115 (11) 2563-2570; published ahead of print March 12, 2018). It followed from the authors’ suggestion that critics of a given piece of research should “Comment on studies, data, methods, and logic, not authors.” This is good advice. However, assuming the authors are using the word censure as it is defined (a formal statement reflecting severe disapproval), suggesting it as a response to ad hominem attacks in scientific discourse is quite remarkable.

Being censured would likely have serious consequnces for most academics, especially given how tricky their job market is. Of course the seriousness of any punishment is also relative to the crime. Being censured for committing massive scientific fraud that misleads the public as well as wasting their money, for example, would seem like a slap on the wrist. In this case, we are supposed to consider formally punishing people for “engaging in ad hominem attacks in scientific discourse.”

Now, I think most of us agree that there is no place for ad hominem attacks in scientific discourse (or do we?) – but do we all share a definition for what such an attack is? Given what I’ve seen among my peers, the answer is no.

For example, the editorial includes three examples of ad hominem attacks, none of which I find compelling. One is about “attacking based on perceived expertise”, and the offending statement is apparently this one: “Dr. Erickson and co-authors are not acknowledged experts and, as far as we know, have no record of clinical or research experience in this area.” A low blow? Maybe. But surely the credibility of a source of information is relevant. Does this deserve censure?

The second is about calling out conflicts of interest. Now, I can be convinced that maybe we put too much stake in such conflicts, and that science should be debated on merit (and I have a lot of time for Ken Rothman’s opinions, one of which is cited in the editorial). But am I really looking to censure someone over it if they are factually correct? I am not.

The final example (which was actually the first one noted in the editorial) is a blog post written by Jordan Anaya about the “work” of Brian Wansink (about whom more has been written than I care to summarize). In it, he strikes a tone that I have to admit is not my cup of tea. He calls Wansink the “Donald Trump of Food Research” (ouch), but then, in my opinion, goes on to actually back that analogy up. The rest of the article however (again, in my opinion) stays focused on facts, and clearly notes when what is being said is an opinion. There is nothing there that I think qualifies as an example of “going beyond commenting on the work itself to criticizing the person in extreme terms.” If a colleague or student sent me the post for an opinion, I’d strongly suggest they remove the Donald Trump bits, but that’s it. If this is the example of “extreme”, then this argument is over.

Given this subjectivity in defining an ad hominem attack, who then makes the decision? Who decides it was made in the context of scientific discourse (two professors walk into a bar…)? Will there be a committee? Who will bring forth the complaint? How do we assure there are no conflicts of interests among the adjudicators? And finally, why would anyone critique a piece of science and risk being accused of engaging in ad hominem attacks?

Beyond this issue of subjectivity, there are practical concerns. Frankly, I don’t think academics are very good at self-policing professional standards. I’m sure we have processes in place to deal with things like sexual harassment, fraud, and abject incompetence - but I don’t get the sense they work very well. I can’t see how it would work any better for this.

For these reasons, the very idea of censuring scientists for ad hominem attacks is half-baked at best.

But my problems with the PNAS article that included this suggestion go beyond this. As I mentioned, the overall editorial is about data errors and contains a fair amount of useful content on the topic. However, despite this focus, there is only one active suggestion (that I can see) about punishing people – and it’s not about the people committing the errors, but rather the people critiquing them. This gives the impression that these ad hominem attacks are highly prevalent and damaging, but as noted above, just three examples are provided.

What isn’t noted in the editorial, however, is that Prof. Allison, the corresponding author, is a friend of Brian Wansink (this was instead divulged at this talk by David Allison). So here we have a powerful academic singling out a critique of his friend that was written by someone who is at a much earlier (and thus more precarious) stage of their career, and doing it in a particularly influential outlet**. This is a problem in my opinion, and a much larger one than the so called ad hominem attacks noted above. So here I am, calling out this highly questionable choice, given that there are other critiques using questionable tone the authors could have chosen, including some written by senior people, and that aren’t about their friends.

Further, the framing of Jordan Anaya’s essay, in both the talk and editorial, is incomplete. It gives the impression that the questionable blog post was an initial salvo aimed at Wansink, when in truth, it came much later in an ongoing process that in fact started with the kind of private-channel communications and traditional forms of scientific discourse that Prof. Allison advocates for. This should be immediately apparent to anyone with an internet connection and a search engine, so the omission here is startling to me.

I pointed out some of my concerns on Twitter, but had a hard time expressing myself in that medium, hence this blog post. I would like to disclose that I have used the example of Brian Wansink in talks about research integrity, and wrote an early critique of the Wansink blog post that started things off (meaning his post, not mine). I follow, I think, all of the people involved in the initial critiques of Wansink’s work on Twitter and admire what they do, but don’t know any of them personally. I was also irked by the PNAS editorial’s use of the phrase “Trial by blog”, but that will have to be a topic for another day!

** For those that don’t know, the National Academy of Sciences (USA) is composed of senior scientists who wield considerable power in their fields and institutions. While I can’t hold individual members responsbile for the idea we should censure for so-called ad hominem attacks, its appearance in PNAS is not trivial.

Other related posts, from other people, that I think are worth reading if you made it this far:

Error Detection Week — Thursday. Some Points On Bullying, Attacks and Criticism

How Not To Be A Crank

The Inside Story Of How An Ivy League Food Scientist Turned Shoddy Data Into Viral Studies