When publishing a paper, I normally go through the following stages (which I imagine are fairly universal):

  1. Figuring out the solution to a new problem and writing it down.

  2. Polishing up the manuscript until all of the coauthors are happy.

  3. Putting the preprint on arXiv and any number of other repositories.

  4. Submitting the paper to a journal (or several journals, if the first one rejects).

  5. Revising the paper until the referee is happy, and getting it published.

Suppose that at some point in this process I move from University A to University B. When is the exact point where I should change the affiliation listed on the paper from A to B? This type of question has come up several times for me, but luckily in all cases I was either fairly early or fairly late in the process and the answer was clear enough. Still, I feel like I should have a clear idea of when the transition happens, and, much to my embarassment, I don't.

(If it matters, the field is pure mathematics.)


3 Answers 3


Number 3 and 4 are both points on the time line (whereas the others are time periods) so I would use the current affiliation at that point in time for both of these.

If you affiliation changes during stage 1 or 2 I would usually put the final affiliation in the paper. An exception might be if the first affiliation is through a grant and the paper is essentially what the grant paid for.


Affiliation serves (at least) two purposes: to give credit to institutions that support you in the research and to enable readers to find you.

Some institutions are also possibly sensitive about affiliation and, if you leave, don't consider you affiliated with them any more.

With this in mind, I find a general, one size fits all, solution impossible.

My own preference, provided that other concerned parties don't disagree, is to list my current affiliation, but note in the paper when and if the research was carried out at another institution, even if only in part.

It is also useful, since affiliation statements usually include an email address, that you provide one that will be honored into the future. My own former employer (retired more than a decade ago) has been good enough to keep me in the email system.

I'm trained as a mathematician but have published mostly in CS where my "preference" seems to work. It might be different in a lab science (chemistry, physics, biology,...) where the support is much deeper and more explicit.


Check the journal guidelines to see whether they say anything about this. I have certainly seen journals that say affiliation should reflect where the bulk of the work (by which I assume they mean your stage 1) was done, journals that say the affiliation should be your current institution, and cases where I couldn't find any specific guidance.

Unless the journal says something to the contrary, I would update my affiliation to be current every time a new version is produced, i.e. at point 3, point 4 (every time), and at each revision during 5. (I wouldn't attempt to make changes later than this, such as at proofing, and I wouldn't make pre-emptive changes based on upcoming moves.)

Some people give a fuller explanation by listing their affiliation as something like "University B (research conducted while at University A)".

Not the answer you're looking for? Browse other questions tagged .