Citational Justice

Some musings prompted by this rather negative take on citational justice literature. In fact (partly for reasons I explain below, and in line with the linked piece) I myself am pretty hostile to the whole notion of citational justice. In any case it feels so self-indulgent as to be immoral for lefty academics to be debating "citational justice" while there are still homeless people. Let's... you know, let's get a grip here? But none the less I am by profession a social epistemologist, and I do think that we learn something interesting about how knowledge is and ought be generated by thinking about what the role and consequences of citational practices are.

The linked piece is responding to an argument from Sarah Bond for more egalitarian citational practices. They take her to be saying that academics should, in order to render our community more just, make conscious efforts to redirect our citations towards people from under-represented groups -- here that's taken to mean women and racial minorities and intersections thereof especially.

The OP authors break this down into three sorts of justice claims one might be interested in. Distributive justice - citations are a good, we should aim to ensure they are shared in an equitable (here meaning: roughly equal across demographic groups) fashion. Citational fairness - citations should be awarded for the right kind of reasons, but in fact they are presently skewed towards white men for no good reason so we ought consciously correct that. And retributive justice - if someone behaves badly they ought not be rewarded with citations, the community should not give esteem to someone who violates important moral norms, and should instead give citations to more deserving underserved populations.

The authors don't agree with any of these points and so reject Bond's argument. You can read their article to see why (I will summarise one of the arguments they appeal to below, since I agree with it). Instead they envision citations playing role of acknowledging (and, they grant, rewarding) previous work, situating oneself in literature, and pointing to evidence for one's claims. Now I certainly agree that citations do all that, and the incentive effect they therefore have is important, but I would lean more heavily upon another thing than they do: distribute attention.

I take it that all else equal the probability of me reading a paper goes up if I find it cited in a work I am already reading. This because the context of citation means I have some reason to think it's relevant to something I apparently already found relevant to my interests, and while I don't think this sort of relevance is transitive I do think that appearing in sequence is some sort of guide. Also, maybe the title grips me, or at least prompts me to google the authors and see what else they have done, etc.

Crucially, all of these things still seem to me possible in a world wherein people are consciously citing people on demographic grounds. Some arguments for consciously citing people per their demographic group presuppose the social meaning of citation remaining fixed in world wherein that's a norm. I rather suspect it wouldn't, and it would change to the detriment of whoever the new norm says to cite more.

In particular (and this is noted in the linked article) I have often thought that what would happen in a world where everyone knows You Must Cite Non-(White Men) is that people would have an excuse to write off citations to non-white non-men as not really reflecting quality of argumentative support or utility for situating within a literature. It would just be ticking a box. In this case simply making non white non men get more citations would not really be giving us access to a valuable good unfairly denied - the nature and value of citations would have shifted by the implementation of this social norm, and now what we are being handed is nowhere near so valuable as what we are presently denied. So I am not entirely sold on some of the citational justice schemes that are advocated, in fact I am hostile - I think they'd hurt me! Perhaps that's selfish, but I only note that they are apparently motivated by helping people of my description so that seems more than a personal problem to have.

(You'll note this bears some relationship to common arguments against affirmative action. And I think those are right! If what is valuable about a job is the social meaning of getting to occupy the role, a change in how we decide who occupies the role changes the social meaning of that role, and you have to take that into account. But real jobs also come with, like, wages, that you can use to buy food and shelter, which may override such concerns. Academic citations lean much more heavily for their value upon their social meaning.)

However, even in the world where the kudos attached to being cited is reduced for non-whites and non-men is reduced, so long as citations still had to be relevant to what they are citing, they may still direct my attention to work that I would not otherwise have seen. Perhaps this is good on some sort of quasi-distributive justice grounds - maybe the thing we really mean to redistribute is influence or attention paid - but I am not so sure about that. But what really persuades me this is valuable is: sending people off to read stuff they would not otherwise have read (both the authors to get the citations and whoever is reading it and may have their interest sparked) is good for introducing heterogeneity into the field. Bond herself makes this argument (I think) by stressing the need to get away from citing the same old people. That is what I really think - crude or gross demographic norms about who you cite will encourage people to read beyond the canon, this will make folk spread out more among the literature, and this kinda quasi-randomisation of who knows what will ultimately serve the field well when there is a variety of evidence bases drawn from.

But I think best of all would be to have something that more directly just encourages people to distribute their attention along unusual grounds. To that end I think there are far more efficient and serviceable ways of achieving this goal than crude demographic citation norms. In fact I think that while at first it might randomise attention a bit, eventually by the well-confirmed operation of Matthew effect people would just learn to cite whatever non-white non-men were antecedently somewhat better known, and that in itself may reflect nothing good. So the crude demographic norm would soon lose whatever diversifying effect it initially had, and there's no reason to think it will direct attention to the best work elsewhere. 

If we want to change what it is people are reading and paying attention to, more direct interventions would be wise. Change what we teach, directly make the case in your own publications that particular works by non-(white men), or anyone under-appreciated really, are especially interesting and relevant to some topic of importance, build and make use of platforms that give you reading recommendations that are not tied to the antecedent fame of whoever wrote the articles. There's no clever nudge that will save us these efforts, no short cut. We must, as they say, do the work.