11

Do journals publish papers that cite preprints? I have an article where a good chunk of work is based on a preprint and I have cited it, accordingly. But that preprint was released on arXiv in 2021 and it hasn't been published yet. So, in that case, would it be a bad idea to send my article to a journal for submission?

P.S. There is another recent and yet to be published article on arXiv that cites the same preprint. So, possibly the work presented in that preprint is correct since there hasn't been any revisions yet.

3 Answers 3

28

In a field like math it is almost necessary to do this, both because preprints are encouraged even by major journals and the fact that the publication process is long.

Mathematicians like preprints since it gets work out early so that it can be built upon as well as vetted by the larger community. If you had to wait for official publication before citing what you have used it would slow down the advances in the field.

So, yes, by all means, cite a preprint if that is what you have used, especially if the later "official" publication has not appeared.

If the final version has appeared, however, it is preferable to cite that, since it has gone through the review process. It is, therefore, more likely to be correct in all details.

3
  • 1
    And if the final version appears after your paper has been accepted but before it goes to the publisher, update the citation. Commented Oct 9, 2023 at 23:36
  • 1
    Sometimes people even cite the preprint when there's a journal/conference version. The famous "attention is all you need" paper has a NIPS version (proceedings.neurips.cc/paper_files/paper/2017/file/…) but many papers seem use the citation for the preprint (arxiv.org/abs/1706.03762). Presumably because arxiv has an export bibtex function that doesn't mention the NIPS publication. Commented Oct 9, 2023 at 23:48
  • 1
    I've spent ages in the past tracking down the proper citation for something after reading the pre-print. Commented Oct 9, 2023 at 23:49
7

Yes, you can and should cite pre-prints. The purpose of a citation is not to make your statement more "true" or your paper more "reliable", but to make sure that you don't take credit for the work others have done. A citation can also serve as an "external appendix", e.g. interested readers can read more on this topic in (Smith and Jones 1867). In both cases there is nothing that requires that the cited work is published.

6
  • 11
    While a significant reason for citations is to give credit to other people for their work (and not claim it as your own work), the primary reason for citations is to allow readers to follow the building blocks of knowledge that were used to get to the positions presented in the work. People being able to do that is necessary for building the knowledge we have. Nearly everything is building upon existing knowledge.
    – Makyen
    Commented Oct 9, 2023 at 18:34
  • @Makyen We agree that we are all standing on the shoulders of giants, and that if we write something we should place it within the context of what has been done before. However, this is not the goal of a citation. This is what the section "previous research" is for. That text will usually contain many citations, but it is the purpose of the text to provide context, and the purpose of the citations is to make sure that it is clear who is making what statement and the possibility for the reader to read the original text (an external appendix). These are the two purposes I stated in my answer. Commented Oct 9, 2023 at 18:50
  • 1
    @MaartenBuis: I disagree about the "previous research" part. That section serves to provide context for the focus topic(s) of the paper. It does not, and cannot possibly, provide the context for all remarks and explanations given at some point in the paper, for which a citation is suitable. After all, it would make for an awfully confusing reading experience if the "previous research" section contained lots of seemingly unrelated statements on tangential topics, whose relation to the paper only becomes clear in hindsight. Commented Oct 10, 2023 at 8:31
  • If you need lots of seemingly unrelated stuff in your paper, than it is time to rethink your paper... I am of the school that an article should do one and only one thing right, and leave everything else to other articles. Commented Oct 10, 2023 at 8:45
  • @MaartenBuis: We may be following different conventions of structuring papers then. Personally, I prefer when a paper provides sources for all of its statements, and not just for the fraction that specifically deals with the primary research question of the paper. You are using a specific type of user study technique or you present your mesasurements using a particular visualization? Cite its source, where readers can find more information about it. You base your prototype on another, unrelated project? Ditto. And so on. Commented Oct 10, 2023 at 12:02
6

Yes, [some] published research papers cite preprints

Example (see Ref 16).

You must log in to answer this question.

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