When citing a preprint (which has been published on arXiv), which is the correct year to put? Options that seem reasonable include:

  1. The year the first version made it to arXiv (that's when the theorem was originally published),

  2. The year the current version appeared on arXiv (that's the closest to "official published version" of the paper),

  3. Current year (that's the best available approximation of when the paper will be published),

  4. Some way to indicate that the paper is forthcoming, such as "2017+", "forthcoming", etc.

  5. No year at all (just say it's a preprint).

Which of these, if any, are acceptable? If more than one, which is the best option?

I am especially interested in the case when the paper appeared quite a while ago, and it is by no means obvious that it will end up being published.

Some related questions:

Citing a paper with multiple versions - this concerns a paper which was already published by journal, and discusses more subtle issues such as establishing precedence. The accepted answer states:

  1. If a paper has been published, always cite the published version.

Choosing which version of an article to cite - this one concerns a paper which was published at several conferences.

When citing an arXiv paper, what year should one write? - this one is concerned with choosing a particular arXiv version to cite, but it does not address the fact that the paper will probably be eventually published by a journal. The answer accepted answer states:

I would cite the exact version you are referring to in your work.

This is not very helpful if the versions are very similar modulu irrelevant technical details, but I suppose it gives priority to (current version) over (first version).

  • The answer to this question is, in fact, the same as the last question that you cited. The goal of citation is to provide the most accurate information about the source of your information, so you should cite the version that you have used as a reference. That is likely to be the most current version, and may "upgrade" (even to a final journal version) as your own work is reviewed and revised.
    – jakebeal
    Commented Jun 8, 2017 at 23:35


Browse other questions tagged .