There’s a new movement afoot for “citation justice” a form of affirmative action in which we should cite scientists who are marginalized as a way of boosting their careers. I’m referring to citations in scientific papers, and here’s one example: “maps of chromosomes can be constructed by the pattern of recombination shown by alleles producing visible mutations (Morgan and Bridges, 1919)”.
While I still think affirmative action should be practiced in some realms, like college admissions and hiring, I don’t favor practicing it in scientific papers as a form of reparations. My philosophy (which I may not always have acted on!) is that when presenting other people’s ideas, facts, or results, you should give the most relevant citations: those that best demonstrate the phenomenon discussed. And you should be parsimonious: avoid overcitation and don’t put in too many different citations that show the same thing. In other words, I use citations based on their value to their paper—their merit, as you will.
Others feel differently, and I’m not going to argue with them except to say that if you leave out citations that are more relevant or important in favor of citations by marginalized scientists, you’re lowering the bar for citation, which could result in a poorer paper. (This of course implies that I think that science papers should function to build up the edifice of science, not effect social justice, which is better done other ways.)
However, the authors of this paper from the American Astronomical Society note that some groups are undercited:
. . . . it has also been found that when researchers cite others, they are less likely to cite women and scholars of color at rates that match their respective contributions to the field. Many reasons for these unequal citation practices have been suggested, ranging from implicit or unconscious bias to careless citation practices (such as not seeking out the original reference) to consciously choosing to exclude certain researchers and/or groups when citing others.
If it is indeed the case that women and scholars of color aren’t cited as frequently as they should be given the relevance of their work to the paper, then that should be rectified. Remember, a citation is there to document a statement or fact, not to laud somebody’s accomplishments, so what’s important here is not “respective contributions to the field” but “relevance of their work to the statement requiring documentation.” If there is under-citation in this sense, then scientists should indeed do something about it when they write papers.
But the topic of the article below is this question:
This leads to the crux of many recent discussions: is it ever acceptable to intentionally choose not to cite someone(s)?
Their answer seems to be “yes, it could be acceptable to deliberately omit a relevant citation, though there’s no cut-and-dried rule”.
Click screenshot to read:
The authors first lay out, in a good summary, why scientists use citations:
Currently, the relevant portion of the AAS Code of Ethics is found in the Publications and Authorship section of the Ethics Statement:
Proper acknowledgment of the work of others should always be given. Deliberate, wanton omission of a pertinent author or reference is unacceptable. Authors have an obligation to their colleagues and the scientific community to include a set of references that communicates the precedents, sources, and context of the reported work. Data provided by others must be cited appropriately, even if obtained from a public database.
The statement reminds us that there are several reasons why we are expected to cite others in our publications. These include citations as an acknowledgment of the contributions of others to the ideas in our work, as well as to avoid plagiarism, and we cite others to justify our methods, assumptions, and research practices. Citations are also important for maintaining the integrity of the academic record and tracing the development of ideas over time, both for the historical record as well as for a proper understanding of how a research field has evolved.
To me, this alone implies that you cite based on relevance, not as a way to effect social justice. And even if authors have done some bad things, if their research is solid and relevant to the point being made, you should cite them. Not doing so violates all the reasons given above.
But moral considerations then creep into the article of Hughes et al.:
In the case of unethical research practices, we can look to other fields outside of astronomy for some guidance. The AMA (American Medical Association) Code of Medical Ethics suggests that when researchers engage with results that were obtained in a clearly unethical way, such as Nazi experimentation on humans during WWII or the Tuskegee Syphilis Study, they should first seek to cite studies that used ethical methods and obtained the same results. If that is not possible, then the harm involved in obtaining the results should be disclosed and acknowledged, the reason for needing to cite the study justified, and the authors should pay respect to the victims of the behavior.
I’m not sure that there are any results of Nazi medical experiments that are even worth citing; I remember reading one scholar’s conclusion that these experiments were so slipshod that they never produced anything of value, even given their aims—to save German soldiers (or, in Mengele’s case, to satisfy a sadistic curiosity). And nearly everyone now knows of the Tuskegee Syphilis Study and about its unethicality. I don’t know if it generated any useful data, but to have to stop in the middle of the paper and recite a screed in honor of the victims seems to me a bit much. I’d rather just say “see X”, where “X” is a discussion of the harms produced by that study. Moral genuflection (“I will now show that I realize this work was unethical”) is somewhat demeaning in a case like the Nazis and Tuskeegee. In fact, I’m hard pressed to think of having to cite any study that requires that kind of qualification and explanation.
But the authors do find one case where citations may be properly left out without any qualification: when the scientist cited is a sexual harasser. As they say:
But the guidance becomes less clear when it comes to dealing with citations of documented sexual and serial harassers. While there have been several recent high-profile cases in astronomy, many other fields are currently struggling with this same issue. The arguments of whether we should cite these individuals boil down to two main positions:
Note that the links go to two sides of the argument, the “Yes” from my law-school colleague Brian Leiter.
This is the starting point from which the AAS Code of Ethics Committee, the AAS Publications Committee, and the Ethics Working Group are confronting the issue. There are several related questions to grapple with:
Here are the questions that, according to Hughes et al. must be answered before you can decide whether or not to cite a harasser:
- Is the research unethical, or is the person’s behavior unethical, and does it matter?
- Is sexual harassment a form of research misconduct? The American Geophysical Union says yes, and the NSF has instituted policies that require institutions to report sexual harassment findings which can lead to the revocation of grant funding. While the AAS code of ethics does not currently address this issue directly, the Astro2020 Decadal Report recommends that identity-based discrimination and harassment be recognized as causing the same level of harm to the integrity of research as is caused by research misconduct.
- How do we identify bad actors in our community? What is the threshold? By which temporal and cultural standards do we judge? Who ensures that the punishment fits the crime, and can there be a path to restoration?
- Who is harmed? What is the collateral damage? How do we limit future harm to the survivors of sexual harassment? Should we protect the junior colleagues and collaborators of bad actors from secondhand punishment, and if so, how? And when does the integrity of the scientific record take precedence?
The authors do admit that making a decision not to cite someone who’s a sexual harasser (and yes, the conclusion is that it may well be justified) is an “ethical gray area.”
But none of this stuff, to me, justifies not citing someone as a form of punishment because they engaged in documented sexual harassment.
Of course I abhor sexual harassment, and it should be dealt with promptly and properly. But why is sexual harassment the only bad act that can be punished by canceling a citation? (And yes, canceling a citation means canceling the scientific community’s knowledge of relevant science.) What about any felony: robbery, murder, or other bad acts like simple non-sexual harassment or bullying of students or colleagues? (It may be because the three authors, all women referred to as “she” or “her” on their professional webpages are more attuned to this form of bad behavior than are men.)
By all means punish those who engaged in misconduct—and apparently it doesn’t have to be “research misconduct” to make someone a “bad actor”. But remove their contributions from science? That’s a no-no to me.
I may be an outlier, but in my view there’s no good reason to not cite the scientific work of “bad actors” or harassers if the work itself is sound and relevant. Even murderers should be cited if their work is relevant. There’s no “research misconduct” worse than killing one of your students, but to me even that’s not bad enough to expunge someone’s relevant work from science.
Punishment and ostracism should be inflicted on people, not on science itself, for leaving out relevant citations because the person who did the work was bad is indeed hurting science, and scholarship in general. Being fired or punished is enough; it’s not necessary (and is indeed harmful to science) to “punish” someone further by simply refusing to cite their work. If we did that, we wouldn’t cite great literature, for many famous authors were pretty bad people, including being sexual harassers.
In the end, I agree with Brian Leiter, whose “Yes” vote for not removing citations is explained in the 2018 Chronicle of Higher Education article linked to above: “Academic ethics: should scholars avoid citing the work of awful people?” (the three people cited in his first paragraph below were accused of sexual misconduct):
Certainly, scholars should condemn Frege, Searle, Ronell, and the like. But to excise from the canon of relevant knowledge those who are appalling people is simply a further betrayal of what justifies the existence of institutions devoted to scholarship.
. . . You should not — under any circumstances — adjust your citation practices to punish scholars for bad behavior. You betray both your discipline and the justification for your academic freedom by excising from your teaching and research the work of authors who have behaved unethically. Universities would, in principle, be justified in disciplining you for scholarly malfeasance, subject to appropriate peer assessment.
Such academic misconduct is unlikely to constitute a firing offense — unlike, say, serious plagiarism or fabrication of data. But researchers or teachers who let moral indignation interfere with scholarly judgment do betray the core purposes of the university and so open themselves to professional repercussions. The foundations of academic freedom demand nothing less.
h/t: Thanks to a scientist who does astronomy for alerting me to this piece.