Is it ok to put old papers on the arXiv?

Question

In my field (I am a condensed-matter theoretical physicist) it is such a normal thing to put papers on the arXiv, that people sometimes look at other people's arXiv list of papers instead of their publication record to get to know what they do, etc.

It just so happens that a few years ago I did a PhD with a supervisor that does not like the arXiv that much (he apparently had some conflict with some authors of a paper that was very similar to one of his papers, but one was already published while the other was just an arXiv preprint -or something like that). As a result, from my list of less than 20 publications, there are four of them from my PhD that are missing on the arXiv.

I was thinking of submitting the latest versions I have of them as arXiv preprints now, four years after their publication. For this I already have the agreement of all co-authors (including my supervisor), and all the journals involved seem to allow to put preprints of the material they publish. Do you see any problem with this? I fear that some people might take it as a way to advertise as new something that was already published a long time ago (though I would of course give the full reference to the published version). There may be other issues I am not aware of. The reason why I would like to do this is that I am applying for fellowships, positions, etc. and I am afraid some of my (best) publications may go unnoticed if somebody looks me up on the arXiv.

Thank you, Stephen

Answer 1

Yes, this is in general OK. The main limitations on posting old papers to the arXiv are (1) copyright considerations (i.e. you must either hold the copyright, or have permission from the copyright holder to post the e-print) and (2) the approval of your co-authors, or at least the corresponding author, which can put you in a hazy gray area if you can no longer find them or they explicitly deny it. You seem to be OK on both counts.

The criteria for publishability on the arXiv are lower than a journal, and there is no previous-publication requirement or disbarment. From the arXiv primer,

Material submitted to arXiv is expected to be of interest, relevance, and value to those disciplines. [...] Submissions are reviewed by expert moderators to verify that they are topical and refereeable scientific contributions that follow accepted standards of scholarly communication (as exemplified by conventional journal articles).

There is no requirement that work be unpublished; indeed, post-publication posting is an integral part of the self-archiving, green open access model in the post-print variety, often with a specific embargo dictated by the publisher.

That said, each community uses the arXiv in subtly different ways, and someone who repeatedly posts old papers may get called out on it (either positively or negatively!), so if you're nervous, find someone in your field who is very active on the arXiv and ask, or simply look for precedent - it's a bit tedious but not terribly hard. For example, arXiv:1601.06054, first published in June 2014, is currently on the cond-mat front page.

One thing you should do is make sure that your metadata is pristine and complete, and to get it right the first time. This includes the DOI and journal reference - you're not there to hide anything, you're there to make your research more accessible and discoverable, and the way you do that is by having good metadata. A comment saying that this is a late posting of an old paper would not go amiss - it will defuse the people who see the posting and think it's new and would otherwise be disappointed or annoyed.

And, of course, if you're looking to make your research discoverable online, think carefully through your online presence - arXiv, Google Scholar, ResearchGate, Academia.edu, Dissemin, Mendeley, or whatever services you choose to use - and make sure that they're all what you want them to be.

Answer 2

So long as you clearly indicate the origins of the paper, including the publication details, there should be no problem, as you're not "hiding" that the papers aren't new. Another way to help offset this impression is to make sure all the papers are posted as close to one another chronologically as possible. (That is, it's unlikely that someone would have, say, four new papers to publish on the same day.)