9
$\begingroup$

This is a soft question aimed at establishing what people think is the professional best-practise for submitting unoriginal work on the arXiv.

There is a draft of an article [1] by Robert Szelepcsényi, at his webspace at the University of Chicago, apparently written more than a decade ago during his graduate studies. The work seems to be correct, modulo some very minor errors, and contains a result that I would like to refer to, independently of what the University of Chicago intends to do with that webspace. It seems to be available only at that single location; and there doesn't seem to be any publicly available works which duplicate his results.

Partly for the sake of ensuring its correctness, I have taken it upon myself to write notes recapitulating the results and motivations of that draft, acknowledging that it is essentially a recapitulation of [1], and which hopefully smooths over some of the roughness present in the existing draft. I can certainly put these notes online on webspace of my own, but at present the arXiv seems to be the most responsible place to submit this draft for future reference.

Question.

Assuming correct attribution, and granted that these results are not widely available otherwise, is there any particular reason that I should not submit my draft of Szelepcsényi's results on the arXiv?

[1] Logspace MOD Classes with Composite Moduli (Preliminary Version) — Postscript File

Improve this question
$\endgroup$
15
  • 6
    $\begingroup$ Seems like a perfect question for the new academia.SE; it is not at all specific to TCS. $\endgroup$
    – Raphael
    Jul 26 '12 at 22:15
  • 5
    $\begingroup$ The question doesn't seem to be particular to TCS so IMHO Academia is a more appropriate place for it. $\endgroup$
    – Kaveh
    Jul 26 '12 at 22:31
  • 3
    $\begingroup$ isnt this kind of obvious-- have you tried contacting him directly and asking for his opinion on that? $\endgroup$
    – vzn
    Jul 26 '12 at 23:48
  • 2
    $\begingroup$ @Raphael, Kaveh: it's not a bad point, though it's something where attitudes are more likely to differ from field to field. Perhaps not for those fields for which www.arXiv.org routinely takes submissions, though. $\endgroup$
    – Niel de Beaudrap
    Jul 27 '12 at 11:26
  • 1
    $\begingroup$ @vzn: Perhaps you might have more luck than I in finding a current email address (or even a curent affiliation) for him. It's hard to find any addresses which might concievably reach him, and the ones I have tried get notices of failure to deliver. $\endgroup$
    – Niel de Beaudrap
    Jul 27 '12 at 17:29
10
$\begingroup$

This should be perfectly acceptable if it is framed as an exposition article. There is a long tradition of expository articles in mathematics (one example that comes to mind is Thurston's exposition of a method developed by Conway to determine if the plane can be tiled by a given shape, see here). Just make sure that the abstract and the introduction convey that all you're doing is to explain Szelepcsényi's proof.

Improve this answer
$\endgroup$
3
  • 1
    $\begingroup$ Possibly a difference here is that there is no expository intent. $\endgroup$
    – Suresh Venkat
    Jul 27 '12 at 23:42
  • $\begingroup$ @Suresh That's true and it was bothering me when writing this. I feel that once Niel starts writing, he will inevitably come up with modifications and simplifications or just a different way to present things..which maybe goes towards an exposition. In hindsight, I agree with a comment here that incorporating the result in an extended version of the paper that uses the result, maybe as a clearly marked appendix, may be better. $\endgroup$
    – Sasho Nikolov
    Jul 27 '12 at 23:47
  • $\begingroup$ I do have an expository intent, but his article isn't in dire need of exposition; his gaps are implicitly filled by context. I do write an account of why Szelepcsényi seems to have chosen to take the approach he does: that's essential for readability, if one is going to deliberately and explicitly ape another person's paper with repeated citation. But it isn't clear to me that playing "mathematical archaeologist" was a sufficient condition to make it an acceptable practice. (Of course, I'm refactoring some of his analysis to try and make it easier to approach, so maybe that concern is moot.) $\endgroup$
    – Niel de Beaudrap
    Jul 30 '12 at 13:32
7
$\begingroup$

Depending on what you want to do with it, a possible alternative is to just fold it into the work you're doing, i.e. recapitulate the results, with clear attribution, but inside your work. The results in Szelepcsenyi's paper don't seem to take up too much space to do this with a journal paper. It'd be a bad idea with a conference paper. Of course if you're not sure about where this might be published in any reasonable time frame then a purely expository paper is a better idea.

Improve this answer
$\endgroup$
1
  • 2
    $\begingroup$ I like this idea, and let me add that even if you're submitting to a conference, you can still have an extended arxiv version of your work, that includes an exposition of Szelepcsenyi's result $\endgroup$
    – Sasho Nikolov
    Jul 27 '12 at 23:48
1
$\begingroup$

I think this is tricky territory. The author of any creative work legally owns the copyright for that work unless he/she cedes it to another party. Notwithstanding the legal issues, any researcher would naturally feel that they have ownership of their work. So, publishing it (and I believe putting it on arXiv is a method of publication) without their permission would be unseemly. Putting it on one's own web page would be less confrontational.

It is perfectly acceptable, however, to cite a draft paper that sits on somebody's web space. It does not need to be on arXiv to be cited.

Improve this answer
$\endgroup$
9
  • $\begingroup$ I'm concerned, in part, for the permanence of Szelepcsényi's draft. (The presentation could also use a little polishing, though.) I have indeed deliberately copied much of the techniques of the draft, and even the structure of his presentation, but with constant citation to hammer home that it isn't original to me, and with some simplifications both for the sake of charity and concision. Is it your opinion, then, that the fact that it contains no original content (i.e. that the "added value" of my article is essentially only in the presentation) means that it is tantamount to plagiarism? $\endgroup$
    – Niel de Beaudrap
    Jul 27 '12 at 21:34
  • $\begingroup$ @NieldeBeaudrap No, plagiarism is not the issue. Publishing somebody else's work without their consent is the issue. $\endgroup$
    – Uday Reddy
    Jul 27 '12 at 23:08
  • 3
    $\begingroup$ I am a little disturbed by the assertion that relating someone else's proof in your own words with proper attribution is some sort of copyright infringement. Certainly if I was writing a monograph, I wouldn't have to contact every paper author whose work I am describing for permission, would I? I hope citation would suffice $\endgroup$
    – Sasho Nikolov
    Jul 27 '12 at 23:52
  • 1
    $\begingroup$ @Uday Reddy: Copyright laws are quite different from country to country. Which country’s law are you talking about? $\endgroup$
    – Tsuyoshi Ito
    Jul 29 '12 at 0:23
  • 1
    $\begingroup$ whether he's writing a monograph or not, he's doing what the writer of a monograph does: narrating someone else's proof in his own words, with proper citation. i seriously doubt that that constitutes copyright infringement (but i am willing to be disproved), and i fail to see what's unprofessional about it. $\endgroup$
    – Sasho Nikolov
    Jul 29 '12 at 12:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.