9

On Quora, I found that one of the disadvantages of uploading papers on arXiv before publishing them in journals, is that one of being "scooped". For example:

  1. Potential for Scooping: By making research freely available on Arxiv, there is a risk of being "scooped" by other researchers who may take the ideas or findings and publish them in a journal before the original authors have a chance to do so. This can lead to a loss of priority and recognition for the original work.

Being scooped (if you post papers before they are accepted for publication). If someone is sitting on similar results, they might finish writing it up quickly and submit it to a journal with a faster peer review process than the one you submitted to. Worse yet, the editor and referees might not notice that a similar paper is already on ArXiv. Thus, those people will win where it really counts (peer reviewed paper), and few people will go back in the chronology to figure out which paper was submitted and/or posted on arXiv first.

However, one of the reasons (and probably the most important one) for posting on arXiv, or other repositories, is to establish primacy. Still from the Inna Vishik comment, we can indeed read:

Establishing primacy. If you are working on something similar as other researchers, being the first to post on arXiv is an important public marker of who was first (even if the peer review process works out so that the other one gets published first).

Therefore, how is it possible that you are "scooped" (basically leading other researchers to "steal" and publish your ideas before you in peer-reviewed journals), if you have the time stamp from arXiv as a proof of primacy?

Isn't the DOI provided by arXiv enough to establish primacy?

To me, it looks like a contradiction. Or, am I misunderstanding, or mixing up things?

(A further thought: if the best solution for not being scooped is to post your manuscript only when it has been formally accepted by the peer-review journal, what is then the meaning of posting a preprint, if after a few days, the same manuscript (i.e. already after the corrections deriving from the review process) is published online by the journal?)

David Ketcheson
  • 36,304
  • 9
  • 109
  • 162
Ommo
  • 589
  • 1
  • 4
  • 13

3 Answers3

23

You're not wrong, if you upload something onto the arXiv you should establish primacy and therefore prevent someone else from scooping you. If someone does actually scoop you, you should in theory be able to force a retraction by contacting the journal's authors with the evidence for plagiarism.

How to interpret the two answers on Quora then?

  • Paul Mitchell's answer is likely AI-generated, which means you should not trust it.
  • Inna Vishik explicitly said "for the paranoid" when she listed the cons, so she's not exactly serious. If you are paranoid, then the things she mentions are possible, although you'll likely be able to publish your own peer-reviewed paper anyway. If you're very paranoid, it is further possible that someone plagiarizes you, and the journal doesn't retract it (because it closes down, because it's disreputable, because the editor-in-chief is the person plagiarizing you, etc.)
Allure
  • 127,528
  • 50
  • 325
  • 493
  • Thanks a lot @Allure :-) Your answer is reassuring :-) – Ommo Mar 21 '24 at 13:22
  • 8
    When people are nervous about getting scooped, they don't usually mean "somebody took my manuscript and submitted it verbatim elsewhere". They are more concerned that they upload their work, there are some delays (rejections etc), and in the meantime some other group sees the work, does some independent-ish work (which may well be more time-efficient than what you did because they have your arXiv paper to learn from), and then publish first. That's still not a super-realistic fear, but it's not as outlandish as this answer and the Quora guy make it sound. – xLeitix Mar 21 '24 at 14:32
  • 14
    But, independently of that, my advice regarding secrecy of results and getting scooped is always the same - don't worry about it, good groups have more than enough ideas of their own and bad groups don't suddenly produce good papers even if you give them a solid idea. – xLeitix Mar 21 '24 at 14:34
  • Thanks for your comments @xLeitix :-) – Ommo Mar 21 '24 at 19:48
  • 1
    @xLeitix some other group sees the work, does some independent-ish work (which may well be more time-efficient than what you did because they have your arXiv paper to learn from), and then publish first - would it matter? If prepared honestly, they would have to cite the original work, which still establishes primacy. – Allure Mar 22 '24 at 02:17
  • 1
    @Ommo: The situation can actually be worse than what xLeitix said. There are clever people who do scoop ideas from others, including from papers that they reject. – user21820 Mar 22 '24 at 02:51
  • 1
    @user21820 That's usually one of the arguments for using Arxiv though. If you publish to Arxiv first, it prevents a scummy reviewer from rejecting your work and then turning arround and publishing their own paper based on it. Now at bare minimum they would have to cite your Arxiv paper. – Cole Mar 22 '24 at 03:47
  • 2
    @Cole: But many do not cite the original work, nor the arXiv paper, especially if it is a rejected paper. Gate-keepers exist in many fields, even if not common. It has actually happened to me before. A common trick is to give a different solution to a similar problem, thus being able to claim that theirs is the first 'published' solution. – user21820 Mar 22 '24 at 05:16
  • Thanks for your comments both user21820 and Cole :-) Just a doubt for @user21820: With "There are clever people who do scoop ideas from others, including from papers that they reject"... Do you mean that editors and reviewers do reject on purpose to get your ideas and publish them on their own elsewhere? I mean, this would be the case once you submit a paper to a journal, it goes "under review", and either the editor or a review decide to reject your paper, since "too good", that they want to publish it (or something very similar) elsewhere? :-) – Ommo Mar 22 '24 at 08:54
  • 1
    @Ommo: That's right. There are many clever people who act in certain ways to favour their own publication chances. They may not reject a very good paper, but they will gladly find as many faults with a submission as possible (even if ridiculous) so as to justify "reject" if it benefits them to do so. – user21820 Mar 23 '24 at 16:55
8

You should completely ignore the Quora advice, most of the answers are from 7 years ago and completely out of date.

Currently, in all the fields I am aware of, submitting to Arxiv/Biorxiv is how you prevent getting scooped, not the other way around. As long as what you submit is essentially a high-quality draft of a manuscript you have established primacy.

The only way I can see this backfiring is if what you submit is so poor/error-filled that it doesn't actually support your claims (thus you establish nothing) and it prompts a competitor to submit their work to Arxiv or a journal. The main consideration of when to submit a preprint is usually a risk-reward assessment between getting scooped vs holding on to your work so you can begin following up on it before the rest of the field is made aware of it.

~7 years ago Arxiv/Biorxiv was a lot more controversial, and many people did not want to use it for fear of getting scooped since it wasn't completely accepted as legitimate. Now, however, many fields consider it completely legitimate and a solid submission carries as much weight as a publication in terms of determining primacy. Many labs now consider submitting a preprint as a normal part of the process of publishing a paper.

Cole
  • 761
  • 3
  • 7
  • 7
    “~7 years ago Arxiv/Biorxiv was a lot more controversial” Was it really? This must be a very field dependent statement. ArXiv has been the primary way of sharing research in high energy physics since the 90s. The number of fields in which it ( and its like) have become accepted has steadily grown. So it may have become less controversial in your field over the last 7 years, but that may not apply for all fields. – TimRias Mar 22 '24 at 06:49
  • Thanks a lot @Cole! (+1) :-) A very interesting (and updated) vision on this topic, thanks :-) – Ommo Mar 22 '24 at 08:55
  • 2
    -1: "establishing primacy" is not the only consideration here. As I've seen personally, it can happen that a person uploads to arxiv a very good paper, and during the time the paper is referreed at a good journal, another group generalized the results and got published. The original author gets rejected (it happens at very good journals even with very good papers) and by now more general results than theirs are published, so the paper will not get accepted in a good journal. Also, arxiv was not "controversial" 25 years ago, let alone 7. – Martin Argerami Mar 22 '24 at 11:34
1

This is field dependent, but generally the reliability of an arXiv publication is much less than that of a properly peer reviewed publication.

On the one hand, an arXiv DOI only shows that you put something on the internet. Whether it is on the same topic, let alone the same approach/methodology/results, not to say actually complete and sound, is not a given. Worse, it takes actual effort to establish all those things - effort that is already a scarce resource in proper review processes.

On the other hand, it often does not matter all that much who was first with absolute precession. If a rushed, bare-bones paper on arXiv is a few weeks or even months prior to a good, peer reviewed paper, there is a good chance people still prefer using and citing the latter. Primacy matters little for people that need to use the results.

MisterMiyagi
  • 1,602
  • 7
  • 13
  • 8
    "This is field dependent, but generally the reliability of an arXiv publication is much less than that of a properly peer reviewed publication." [Citation needed] – TimRias Mar 21 '24 at 16:42
  • 2
    Scientists seem really reluctant to risk their reputation and put their garbage out there for everyone to see for eternity and risk their reputation. (Of course, there are people that output gabage, but that is than usually also true for any of their papers that make it through the peer review roulette.) – TimRias Mar 21 '24 at 16:46
  • 6
    Even if an arXiv DOI doesn't prove much, the material you put on the arXiv proves a lot. Even if you later update your original paper, the original version and its date of submission remain available on the arXiv. More generally, even if you update the paper many times (and even if you withdraw it), all the submmitted versions and their dates remain available. So, if necessary, you could easily prove exactly what you submitted and when. – Andreas Blass Mar 21 '24 at 19:30
  • Thanks for your answer @MisterMiyagi.... However, I did not understand this part: "Whether it is on the same topic, let alone the same approach/methodology/results, not to say actually complete and sound, is not a given. Worse, it takes actual effort to establish all those things - effort that is already a scarce resource in proper review processes." :-) – Ommo Mar 21 '24 at 19:32
  • @TimRias... Do you mean that, generally, people do not post their work on repositories like arXiv, before any acknowledgement of acceptance from the peer-reviewed journal where they have submitted the manuscript? – Ommo Mar 21 '24 at 19:37
  • 1
    Thanks a lot @AndreasBlass! Also your comment is very reassuring :-) – Ommo Mar 21 '24 at 19:50
  • 5
    No, I mean people generally don’t post on the arxiv unless they are confident in the work and they are comfortable with everyone reading it. – TimRias Mar 21 '24 at 19:56
  • Thanks @TimRias :) – Ommo Mar 21 '24 at 20:05
  • 1
    @Ommo To take your DOI and prove primacy with it you require others to do at least a full review of your paper in order and compare it to the hypothetical competitor paper - i.e. establish both soundness of the paper itself and then conflict with the other. Yes, there are domains where this happens anyways (and they should disregard what I wrote) but in many others competent reviewers are a very scarce resource compared to the sheer volume of at best mediocre papers out there. One needs way more than just a DOI to get someone whose verdict actually counts to consider such cases. – MisterMiyagi Mar 22 '24 at 05:54
  • Thanks a lot for your explanation @MisterMiyagi :-) – Ommo Mar 22 '24 at 08:56