After 8 years on the receiving end of academia (and a few on the providing end) I don't particularly like the scientific journal process. That's my bias. Michael Clarke wrote an excellent article on why the disruption of this process by web technology hasn't taken place yet (he argues that the web was invented to do just this). He gives 3 reasons embedded in academic culture as to why the web has only incrementally changed the world of scientific literature.

  • Validation (peer review)
  • Filtration (so much info)
  • Designation (how to get grants/tenure)

He argues that these needs are not technological and therefore not disrupted. Imagining what kind of new technology could replace the scientific journal, I immediately thought of SE.

  • Open, post-published peer review is becoming more common as data scientists blog their findings and peers comment. SE has a great set up for commenting on questions and answers as well as updating based on comments.
  • Voting in SE provides a great filter for the best stuff.
  • Reputation is becoming more and more reliable in SE and is now seen on job resumes.

I can imagine SE starting "scientific journal" sites to go hand in hand with some "Q&A" sites. Much of the functionality would be similar, but instead of questions there would be scientific articles. And instead of answers there would be reviewers' comments on the article. Some additional incremental functionality such as the ability to indicate incorporation of comments in the article, or ability to browse core articles easily, could be added as needed in the same way meta works.

What do you think?

Could/should SE host the awaited web-based replacement for the scientific publishing process?

[Update] For a bit more information on these ideas floating around see this successful journal and this idea for a statistics journal and this idea using github and this broad idea.

  • 1
    One of the things that makes SE great is wiki-style editing capability to make questions/answers shine. How would this work with published papers?
    – Geobits
    Dec 10, 2013 at 2:58
  • 3
    Yes...love this idea!! I'm personally within a few months of completing a paper and would love to be able to post it on SE. I had toyed around with the idea of trying to rework it in such a way that it would be appropriate to post it on SO, but couldn't think of a good way to make it work. Dec 10, 2013 at 3:27
  • Very good point @GenericHolidayName. This is also a function scientific journals perform, editing for clarity (especially for non native speakers). Off the top of my head, I would think a semi-wiki style with approval, but perhaps full wiki is the way to go.
    – Tom
    Dec 10, 2013 at 3:30
  • 1
    I really don't understand why you have to pay for papers?
    – Cole Tobin
    Dec 10, 2013 at 4:01
  • 6
    The technical side would arguably be the smaller portion of a project like this. Where would the community come from that powers it? What kind of partners would this need to have to get started? How would one acquire publications? How would the open peer process work, how would credentials of peers be checked? Is the engine really a suitable format for this - is peer feedback at times not incredibly long? The same goes for the publications themselves. Would their average size really fit the format of a SE question? There's a reason why content here is limited to 10,000 characters....
    – Pekka
    Dec 10, 2013 at 5:01
  • 3
    ... not trying to put down the idea, but I'm having a hard time seeing a 50-page article being discussed here in detail and the experience still being as streamlined, simple, and cool as it currently is here. I fear the more one thinks this through, the more it will turn out to be needing changes to the format that ultimately amount to a completely different product. (happy to be proven wrong, of course.)
    – Pekka
    Dec 10, 2013 at 5:08
  • 1
    Thanks for the criticism @Pëkka (no really!) Just to impart some thoughts to your questions (in order). Start with relatively new disciplines that already benefit from a reproducible research approach. No active amassing of publications, just make an attractive product as an option for researchers. Peers work same as now, credentials based on reputation. Imagine articles that fold into themselves so at first you only see an abstract and then you can expand for more detail. Perhaps feedback would be per section. Yes, this would be a different product, but perhaps not so different.
    – Tom
    Dec 10, 2013 at 6:20
  • 1
    @Tom fair enough. I guess one could even imagine linking to a PDF for the actual publication, and having only the abstract in the question. This Peers work same as now, credentials based on reputation. strikes me as super problematic though - for one, you'd have to have credentials confined to an area of expertise (right?) and looking at SO/SE rep, it is so easy to game, I'm having a hard time seeing a reputation score being taken seriously, when anyone with 50 points can freely vote... it's a big shortcoming of SE and would probably work even more badly in an academic setting. No?
    – Pekka
    Dec 10, 2013 at 6:25
  • Wouldn't PLoS have some argument to be the repository you're looking for - combining academic journals with benefits of online?
    – Joe
    Dec 10, 2013 at 6:58
  • @Pëkka good point on the reputation. This type of peer review would be different from the current pre-publication style. Instead, it would be the post-publication style similar to researcher posting findings on her blog saying "suggestions welcome in the comments". Although perhaps the article could be "frozen" after a decent amount of time. Maybe eligible reviewers wouldn't be based on reputation, but accepted points of peer feedback would generate reputation. Similar to current SO/EP the right mix of reputation points would be important.
    – Tom
    Dec 10, 2013 at 8:31
  • @Joe Referring to web-based developments including PLoS and other open access journals Clarke says "To be sure, scientific publishers have not ignored the Web. They have innovated. They have experimented. They have adapted. But it has been incremental change—not the disruptive change one would have predicted 18 years ago."
    – Tom
    Dec 10, 2013 at 9:22
  • 4
    Forget replacing scientific journals, I vote for replacing the entire national electoral process. Lets get the official Elections organizations powered by SE and vote political leaders. Seriously though, very interesting discussion so far.
    – Tom Pace
    Dec 13, 2013 at 6:41

1 Answer 1


What's wrong with arXiv? IMO it would be much better if the arXiv was improved and expanded to replace today's publishing process.

Besides, the issue is that SE as a software is ill-designed for anything other than Q&A. Also, SE is not that well known among the academia. Math Overflow is well known to mathematicians, but this isn't true for any other science SE site/discipline.

  • You may be right. arXiv still lacks some of the snazziness that SE has, but there's no reason it couldn't take a few tips from here. If it did that and expanded to all disciplines I think that would be great.
    – Tom
    Dec 19, 2013 at 2:44
  • 1
    as far as I can tell, ArXiv does not support peer review - this makes it very poor replacement for journals. As pointed in the referred article by Michael Clarke, peer review "has become a core function of today’s journal publishing system — indeed some would argue its entire raison d’etre"
    – gnat
    Jan 17, 2014 at 7:36
  • @gnat I never said it does, but I did mention "improved and expanded" for this reason. StackExchange is a pretty poor platform for this, plus it may be considered a rather frivolous place by the academia. arXiv already has a lot of support for preprints and adding a peer review system would not be as hard as adding paper support to SE. In addition, arXiv is known and liked/used by most in the academia. SE, on the other hand, not so much. Jan 17, 2014 at 8:26
  • I see, thanks. That makes sense
    – gnat
    Jan 17, 2014 at 8:46

You must log in to answer this question.

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