15

Suppose candidate A has only published alone, and in an alternate universe candidate B has an identical list of publications, most of which are with coauthors. All else being equal, would a typical hiring committee rate A higher?

I suspect the answer depends highly on the field; I'm especially interested in (pure) math.

  • 3
    Why the alternate universe? – gerrit Feb 13 '14 at 10:38
  • 4
    The answer may depend in part on who the coauthors are. In math at least, if candidate B's PhD advisor is a a coauthor on most of those publications, then candidate A would certainly look better. If candidate B's coauthors are peers, the situation looks rather different. – Mark Meckes Feb 13 '14 at 10:46
  • 2
    @gerrit: The point of this phrasing is to imagine that two candidates are exactly the same except for the soloness/jointness of their authorship. This is not possible in "just one universe"... – Pete L. Clark Feb 13 '14 at 21:46
  • 3
    Since the order of authorship is alphabetical in most math journals, sole authorship conveys more information than it does in other fields where the first author likely had the greatest intellectual contribution. Perhaps this is unrelated, but my guess is it has something to do with why lack of co-authors may matter in math. – WetlabStudent Feb 14 '14 at 0:28
  • Things clearly depend on the field. In biology, I think that single authored paper are usually only reviews and might be casually discounted; they at least may look very weird. On the other extreme, I heard that in some social sciences the only things that gives you real credit are solo papers, chapters and mostly books. Fundamental mathematic is somewhat in between, as pointed out by Pete Clark. – Benoît Kloeckner Feb 15 '14 at 10:17
16

All else being equal, solo papers can only be weighted more highly than joint papers in hiring decisions. This is probably especially true at the very top: e.g. a joint Inventiones or Annals paper makes your application look fantastic. A solo publication in either of these journals is a golden ticket for many academic jobs.

Things get much more complicated when one tries to determine how much more to weight solo papers than joint papers. To the best of my knowledge few departments or universities have hard-and-fast rules or even written guidelines about this, so much of this evaluation goes on in the minds of the individual evaluators. The truth is that in some situations jointly authored papers will count to exactly the same degree as solo authored papers, and in other situations the existence of coauthors will cause the work to be substantially discounted.

There is a dramatically increasing prevalence of joint papers in the mathematical profession. Thirty years ago they were quite rare; and they are even more common now than at the beginning of my career, which was not much more than ten years ago. There are now, for instance, certain conferences and workshops in which several people sign up in advance to work on a certain problem under the guidance of a senior mathematician. And then everyone who came to the workshop gets their name put on the paper, even if everything they did was under the guidance of someone else. This is a model much closer to that of the laboratory sciences than what used to be common in mathematics. In my opinion, it is time for the profession as a whole and various groups within the profession to put down in writing some feelings about the merits of joint papers. Of course this will be hard to do since the matter is so complicated: it matters whether your coauthors are "senior" or "junior" to you, it matters whether they have supervised you, it matters what percentage of your papers are joint and whether your papers are always joint with the same coauthors....

Sometimes I see certain publications listed on young people's CV's and think "I find it unlikely that they had a significant intellectual contribution to that work." That's a problem both ways: i.e., people may be wrongly evaluating the merits of this type of work in either direction!

6

This situation is very complicated. First, there's no hard and fast rule as different mathematicians have different opinions. Second, it really matters how individual facts (like whether one paper is coauthored or singly authored) fit into a larger picture.

People want to hire candidates who have demonstrated that they have their own research program and their own direction. Coauthored papers can hurt with this. This is especially true if a candidate has too many papers coauthored with their advisor, or almost no singly authored paper, or too many papers coauthored with a single more senior person, or all papers coauthored with the same person. (Of course, other factors like letters can counteract against this narrative.)

On the flip side, people also want candidates who are influential on their field. Coauthored papers are one way to show that other people are interested in your research program. (Of course there's other ways to demonstrate this, like giving talks at great places or letter writers say you're influential on them.)

In general my impression is that a paper with n authors counts as less than a singly authored paper but as much more than 1/n of a singly authored paper, and that in some sense the perfect situation is to both have effective collaborations and also do good work solo.

  • 1
    I agree: a joint publication with coauthors who are not obviously senior seems to count much closer to 1-epsilon than to 1/n. It is less clear that it should be counted in this way, but as you (and I) say, things quickly become complicated when one tries to work out what the rules should actually be. – Pete L. Clark Feb 14 '14 at 1:06
5

This depends not only on the field, but also on the institution doing the hiring. At some places, a joint publication and a sole-authored publication are identical in terms of "credit" toward promotion decisions. At the other extreme, other institutions consider a three-authored paper as one-third credit. There are many in between as well. Truly elite places don't care about how many papers or the authorship, but whether you've amazed the world.

Further, if the publication record is substantially more or less than expectations for the position, the authorship doesn't matter. A place that expects its hires to have a dozen papers won't be impressed by an applicant with 5, even if they're sole-authored. A place that expects one or two will be delighted with the same 5, no matter how many authors.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.