Posting to ArXiv a longer version of a submitted journal paper

Question

I have written a paper (about 14 pages long) and submitted it to a journal. The editors suggested I cut it down to a 5-page paper and resubmit. I can do that, but will end up leaving a lot of material on the cutting room floor. That material isn't strong enough to command a published article of its own, but I don't want it to be lost forever. I am considering a couple of options:

1. Upload the original 14 page version of the paper to arXiv.
2. Try to piece together the scraps that were removed from the longer piece to make a ~9 page paper and upload that to arXiv.

Which of these seems like a more sensible / standard practice?

A secondary question is whether I should do this before I submit the 5-page version of the paper (and then include in that shorter version a reference to the arXiv version), or whether it would make more sense to wait until there is a decision on the 5-page version.

Answer 1

It's not extremely common, but you can find significantly extended versions of published papers out there on arXiv, authors' websites, and the like. And at least in my filed of research it doesn't seem like that big of a problem to post an extended version after publication if there is a merit and clear reason to do so instead of cramming everything in one place or writing a weak paper out of what is left out.

For example, this paper on arXiv is an extended version of a paper published in the IEEE Transactions on Information Theory and explicitly states when it was first submitted to the journal, when a revised version was submitted, and when the extended version was uploaded to arXiv (which was after submitting the revised version to the journal but before it got accepted for publication). There is a slightly updated version of the extended one on the personal webpage of one of the authors as well.

In any case, I think it's generally a good idea to respect the editor's opinion unless you have a strong reason not to follow their suggestion. But you may also want to ask the editor if the journal has online-only storage for supplementary materials to officially attach what they asked you to cut out and/or if the journal is ok with uploading the originally submitted, longer version to arXiv. If arXiv is common in your field, I highly doubt they say no to posting the longer manuscript online; if it's not allowed to post a preprint you submit to a journal, how on earth can we use arXiv as a preprint server anyway? What would they do if you had already posted the longer manuscript on arXiv before submission to the journal, for example?

My feeling is that as long as you don't abuse the system, it's ok to post an extended, fuller version online (with a clear note that it is an extended version of a shorter paper). So, for example, if the editor is not happy about a lengthy, marginal part which you think is helpful and of interest to some but perhaps not all readers, I don't see any reason not to upload your original manuscript to arXiv. You should check the journal's policy and the norm in your field first, though.

Answer 2

I recommend the first option: piecing together scraps sounds like a fair amount of work to get something that will probably be less useful to the reader than what you currently have. If you are in a field in which arxiv publication is at all common, the practice of posting an extended version on the arxiv should be received non-negatively by everyone. (And really, unless the journal is so obnoxious so as to object categorically to arxiv publication -- in which case I hope you at least consider not doing business with them -- what could be their complaint? You brought the material to them first, and they told you to take it out.)

Answer 3

Arxiv or not, adding the material to the online supplement if possible is a good idea.

If you want the 14 page paper to be available as written, post it to the arXiv. When the short version is published, make clear that it's a longer version of the published paper.

If you want the arXiv post to act as an honest pre-print, then there's no point in waiting. However, if you only want canonized work to be publicly available indefinitely, I'd wait until after you've at least written and submitted the 5 pager. You might find that the 14 pager wasn't necessary when you're done.

I've never seen anyone mention an arXiv version in a published paper. I don't think a sensible reviewer or editor would have a problem with it, but I wouldn't be surprised if someone raised objections. Something about trying to extend the peer review approval to the unreviewed material or some such. If the paper with the same title is available on the arXiv, a lot of readers will find it anyway.