Friday, January 11, 2013

Fame Through Friend-Citation

Alert readers might raise the following objection to my recent post, Fame Through Self-Citation: By the 500th article of our Distinguished Philosopher's career, his reference list will include 499 self-citations. Journal editors might find that somewhat problematic.

If this concern occurred to you, you are just the sort of perceptive and prophetic scholar who is ready for my Advanced Course in Academic Fame! The advanced course requires that you have five Friends. Each Friend agrees to publish five articles per year, in venues of any quality, and each published article will self-cite five times and cite five articles of each other Friend. (In Year One, each Friend will cite the entire Friendly corpus.)

Assuming that the six Friends' publications can be treated serially, ABCDEFABCDEF..., by the end of Year One, Friend A will have 29 + 23 + 17 + 11 + 5 = 85 citations, Friend B will have 80 citations, Friend C 75, Friend D 70, and Friend E 65, and Friend F 60. Friend A will have an h-index of 5 and the others will have indices of 4. By the end of Year Six, the Friends will have 4,935 Friendly citations to distribute among themselves, or 822.5 citations each. If they aim to maximize their h-index, each can arrange to have an h-index of at least 25 by then. (That is, each can have 25 articles that are each cited at least 25 times.) This would give them an h-index exceeding that of the 20 or so characteristic of leading youngish philosophers like Jason Stanley and Keith DeRose. (See Brian Leiter's discussion here.)

By the end Year 20, the Friends will have 100 articles each and 17,535 citations to distribute among those articles. Wisely-enough distributed, this will permit them to achieve h-indices in excess of 50, higher than the indices of such leading philosophers as Ned Block , David Chalmers, and Timothy Williamson. By the end of their 50-year careers, they will have 7,422.5 Friendly citations each, permitting h-indices in the mid-80s, substantially in excess of the h-index of any living philosopher I am aware of.

But this underestimates the possibilities! These Friends will be asked to referee work for the same journals in which they are publishing. They wouldn't be so tacky as to use the sacred trust of their refereeing positions to insist that other authors cite their own work -- but of course they can recommend the important work of their Friends! If each Friend accepts one refereeing assignment per month and successfully recommends publication for half of the articles they referee, contingent upon the authors' citing one article from each of the other five Friends, that will add 180 more citations per year to the pool. Other scholars, seeing the range of authors citing each of the Friends' work, will naturally regard each Friend as a leading contributor to the field, whom they must also therefore cite, creating a snowball effect. Can h-indices of 100 be far away?

(Any resemblance of this strategy to the behavior of actual academics is purely coincidental.)

18 comments:

Anonymous said...

"Any resemblance of this strategy to the behavior of actual academics is purely coincidental." yeah, right.

Ron Mallon said...

Seems like this is a optimistic, feel-good-about-the profession post since (1) most philosophers have nothing like that level of citation and, (2) if everyone did it, no one could get famous doing it since that citation level would be ordinary.

Howard Berman said...

Yes, but what would be their Erdos number?

Eric Schwitzgebel said...

Don't get me started on the possibilities of collaboration, Howard! :-)

Unknown said...

I'm with Ron Mallon. While it's a logical possibility that self-citation and friend-citation are problems in philosophy, there's no empirical evidence that it is in fact a problem. Anne Hartzig has done some empirical work on this issue in other academic fields and has found that self-citation is insignificant for scholars who are highly cited (which is the only group of scholars where we'd care to raise the question). Also, philosophers tend to have very low citation numbers compared to other academics, and even other humanists. So philosophers tend to cite too little as it is. As for friend citations among highly cited philosophers, I think that's a real phenomenon, but not a cause of concern. The fact is that philosophy is a small profession with a lot of very specialized debates. That means there will be relatively few people driving a subfield forward. They'll repeatedly see each other at conferences, professional meetings, etc., and likely become friends, or at least acquaintances. So we'd expect to see a lot of "friend-citations" as a consequence of our discipline's social structure, but the causal arrow of the pattern is not morally concerning.

Eric Schwitzgebel said...

Clearly one can't self-cite one's way to academic fame, though I do think it introduces a bit of noise (up to about +/-4) in the h-index. It's a little hard to evaluate how problematic the Friends strategy is. Maybe it's a big problem; maybe only a little problem. I'm open to evidence. Mostly just thought it would be fun to see to what absurdities the math would lead.

clasqm said...

I know a little cabal of educationalists who have been doing something like this for decades. One of them will publish an article, then the others will start commenting on it, then the original author starts counter-commenting, then all the others get another shot ... These guys can stretch a single idea for well over 2-3 years.

Not saying they're deliberately doing it for the indexes - these are legitimate big shots in their field and they don't need it. But it does show that setting up something like this is not as outlandish as it sounds. You just need to be doing your postgrad work at the right place and the right time where you can make contact with Friends in the same sub-discipline.

But it all contradicts your post of about a week ago, Eric. Where's the fun in this? OK, there's a certain "sticking it to the Man" aspect, but can that really last for a career?

Neil said...

It strikes me as mildly offensive to say you're "open to evidence" about whether philosophers use this strategy. It is offensive to speculate about people engaging in wrongdoing when there is no evidence that they are engaged in wrongdoing, and no reason even to suspect that they are (compare: "Is Eric Schwitzgebel a shoplifter? I don't know, but I'm open to evidence either way"). There is no reason to suspect that philosophers use this strategy for the simple reason that rewards in philosophy are not distributed in a way that is sensitive to citations. To be clear, I'm not objecting to either post, which struck me as nothing more than a bit of silliness (and silliness is not here meant to be in any way pejorative, rather something to be encouraged now and then). I am objecting to the "open to evidence" claim which is smear by implicature.

Eric Schwitzgebel said...

Thanks for the comments, Michel and Neil.

@ Michel: I find quantitative explorations of the history and sociology of philosophy kind of fun, and these thoughts fit within that vein, though mostly in a light-hearted way. (I also have posts using data from jstor, the Stanford Encyclopedia, etc.) I could imagine a future in which I have among my main research strands the quantitative sociology of philosophy based on analysis of publicly available records.

@ Neil: Some of the other commenters on the post seem to think it's an issue. Nor are they alone. I know one person who sees the psychotherapy research literature in psychology as seriously distorted by something like the Friends strategy. To be clear, the worry, I think, is not so much conscious malfeasance as a foreseeable side-effect of a certain sort of problematic clubbishness. I don't think it's unreasonable to be open to evidence about how much problematic clubbishness there is of this sort in academia, playing out in something like the Friends pattern.

Neil said...

You talked specifically about philosophy. It is within philosophy that rewards are not distributed by citations. It is that fact that makes the speculation offensive. Worries about other fields, like psychotherapy, are irrelevant here, since it is within philosophy that rewards are not sensitive to citations. Other people's speculations, if they also concern philosophy, are not less offensive.

Eric Schwitzgebel said...

Neil: It's not clear to me that citation is irrelevant to reward in philosophy. I am inclined to think citation is less relevant in philosophy than in many other fields, but overall trends in academia suggest that it might be becoming more relevant. People's comments about other fields are relevant because if something like a Friends pattern manifests in other areas of academia, that legitimately raises the question of whether it might also manifest in philosophy.

Neil said...

While I think we may be heading that way, I doubt citations are currently relevant. I assess multiple grant applications every year; I have never seen citations mentioned. I was advised not to mention my own in grant applications, since people outside philosophy will not understand why the citations are so low (relative to other disciplines).

Eric Schwitzgebel said...

I agree with your experience regarding grants, Neil. Sometimes philosophers are tempted to mention citations in trying to make cases of various sorts to people outside of philosophy; but usually this is a bad idea because our citation numbers look weirdly low compared to some other fields. Partly as a result, philosophers have tended to downplay citation analysis. But I don't think that makes citation irrelevant in the field in a less formal way. If I'm reading about topic A and I keep seeing Philosopher X cited, that will tend to lead me to think of Philosopher X's work as influential, which will increase the likelihood that I will read and cite it myself. This provides a route for distortions to arise in the field via some approximation of the Friends strategy (e.g., having a kind of clubbish tendency to cite your friends more than they objectively deserve and to reciprocate when they do so for you).

In my own case, I can feel the pull of this clubbishness. When people cite me, I tend to feel more positive about them, which then might lead me to be slightly more predisposed to cite them than I would otherwise be. I'm not sure how strong this effect is in my own case. I hope it's not too strong, but it's also not the kind of thing that I think I'm in an especially good position to know.

It seems reasonable to worry that there might be some kind of problematically clubbish bias in the field, along roughly those lines -- nothing as extreme as pure Friends as described above, of course, which was intentionally absurd. The question would then be whether the problem is really negligible or whether it has a major distortive influence on the field.

Neil said...

Now you've changed the target. Your OP concerned a deliberate strategy of achieving fame through friend citation (you were explicit that the strategy involved agreement between friends, and a deliberate policy). When I said that this strategy was not likely to be employed and therefore it was mildly offensive to say you are open to evidence that people might be using it, you said that there might be factors that influence citation practices, including a certain feeling of gratitude and an implicit obligation toward reciprocity. That's something else altogether, and almost certainly occurs. But it is no defence of your original claim that philosophers might well be engaged in a deliberate gaming of the system to say that they might unconsciously be influenced by these factors. It is a little like me defending the claim that I am open to evidence that Eric Schwitzgebel might be a member of the klan by saying that I wouldn't be surprised if he exhibited a pro-white bias on an IAT.

Eric Schwitzgebel said...

Okay, Neil. I'll grant you that point. I guess I didn't think readers would think I was seriously contemplating the widespread existence of deliberately planned cabals of this sort. But now I see how what I said could be interpreted that way. I should have been more explicit about that!

To the extent there's a problematic "Friends"-type strategy, it operates by implicit mechanisms or unsystematically by means of passing thoughts. ("Hey, I see Neil cited me. I guess it's kind of embarrassing that I haven't cited him on this topic yet.")

Neil said...

And I have cited you!

Bryan said...

This and the previous post are very thought-provoking. One phenomenon that clearly is rampant, at least in my subfield of philosophy, is the "snowball effect." (There is probably already a technical term and a secondary literature on this. Feel free to remedy my ignorance.) Prof A decides (perhaps with some justification, or perhaps not) that Prof X is doing "good work" and mentions this to Prof B. Prof B tells this to Prof C. Prof D hears independently from Profs A,B, and C that Prof X is doing "good work," so she invites Prof X to give a talk, assuring her colleagues that "everyone is talking about how outstanding Prof X is." Prof D's colleagues are now predisposed to think well of Prof X's presentation, so they do. Soon, the fact that Prof X is an unparalleled genius has become a central strand of the web of belief of people in the field, and any recalcitrant evidence to the contrary is handled by adjusting other claims in the web ("he was having an off day," "you just don't recognize how brilliant he is," "obviously your objections are motivated by partisanship," etc.). ;-)

Eric Schwitzgebel said...

I do think this happens, Bryan, especially in the smaller, more close-knit subfields. One good partial check is blind review, to the extent it can be kept truly blind. But that can only go so far, since informal connections and networking are so important!