Scientific communication

« previous post | next post »

Paul Krugman, "The Facebooking of Economics", 12/17/2013:

Economics journals stopped being a way to communicate ideas at least 25 years ago, replaced by working papers; publication was more about certification for the purposes of tenure than anything else. Partly this was because of the long lags — by the time my most successful (though by no means best) academic paper was actually published, in 1991, there were around 150 derivative papers that I knew of, and the target zone literature was running into diminishing returns. Partly, also, it was because in some fields rigid ideologies blocked new ideas. Don’t take my word for it: It was Ken Rogoff, not me, who wrote about the impossibility of publishing realistic macro in the face of “new neoclassical repression.”

In most of the fields where I've worked (computational linguistics, speech technology, computer science, …) real scientific and technical communication now takes place mainly via published conference papers or arXiv preprints, usually available on authors' websites. I've joked for a couple of decades that in those disciplines, publishing in old-fashioned journals has become a sort of cultural ritual, like wearing academic regalia at graduation exercises, which has no substantive role in the actual life of the field.

This is not yet true in most areas of linguistics, because the LSA (and similar organizations) still haven't gotten on board with the decades-old practice of requiring conference presentations to be accompanied by 4-8 page papers rather than 200-word abstracts. So one-to-two-year publication delays (and some blockages of new ideas)  remain, alas, the norm.



  1. John Spevacek said,

    December 18, 2013 @ 10:24 am

    Strangely, preprints (and arXiv) have never caught on in my field – chemistry.

  2. Sally Thomason said,

    December 18, 2013 @ 10:26 am

    But it doesn't have to be that way with the LSA. When I edited the journal, eons ago (1988-1994), the average turnaround time from submission to decision was 3 months, and accepted papers were published in less than a year from date of submission if the author was brisk about making revisions. (Quick publication because with an 88-90% rejection rate I never had a big backlog in the accepted-paper pipeline.)

  3. Brett said,

    December 18, 2013 @ 12:48 pm

    It's interesting how the ubiquity of the preprint system has also affected journal publishing in areas where preprints are standard. Mostly these changes are for the better, although that's not always the case.

    I come from theoretical particle physics which is sort of the "home ground" of the arXiv. One notable thing is that, since everyone expects to be able to read papers online, the major-journals are mostly accessed electronically, rather than digitally. The Physical Review moved very early to get their entire archive online and searchable. Other fields have followed suit, but many have not taken advantage of one of the side-benefits of this system. When print publication becomes a relatively minor part of the operations (with virtually nobody except academic libraries—and sometimes not even them—bothering to get their subscriptions printed on dead trees), issue size becomes much less important. For the physics journals that I publish in, there is no such thing as a publication backlog. Each issue contains all the articles that have been accepted and proofed (the whole process typically taking a few months), however large that makes the printed form.

    The one significant downside of having all access be digital now is that if a library wants to cancel its subscription to a pricey (e.g. Elsevier) journal, in addition to not getting new issues, the library patrons typically lost online access to the archives. It's an unhappy surprise when you have to actually request a bound volume from storage to get to read the article you want.

    The fact that everyone working in the field has to be able to produce their own preprints also means that the articles reach the journals in a form that is very easy to convert into final printable copy. Everybody has to know LaTeX. This skill requirement also served as something of a barrier to prevent crackpots from getting into the arXiv system (although such passive measures eventually had to be replaced with more active means of keeping them out). In fields that require substantially more or substantially less skill in making usable page layouts, the system might not work as well.

  4. Frank said,

    December 18, 2013 @ 2:29 pm

    I remember thinking that this would be the wave of the future when the Rutgers Optimality Archive had all the important OT publications and OT looked to be the dominant paradigm in linguistics. In retrospect, it seems unfortunate that it was tied to a specific theory.

  5. Peter Erwin said,

    December 18, 2013 @ 3:01 pm

    real scientific and technical communication now takes place mainly via published conference papers or arXiv preprints, usually available on authors' websites. I've joked for a couple of decades that in those disciplines, publishing in old-fashioned journals has become a sort of cultural ritual …

    In my field (astronomy), the arXiv is very widely used, but so are the journals, which act as (if nothing else) a formal seal of approval for posting things to the arXiv; we're unlikely to take very seriously a paper that doesn't describe itself as "accepted by" or "in press at" one of the main journals. (There are debates about how acceptable it is to post something to the archive when it's still in the submission-and-refereeing stage — in which case the comment field will say "submitted to X"; some people do this, others don't.)

    (There's also an excellent archive of the print journals in the form of the NASA Astrophysics Database System, which includes scans from the major journals going back to the 19th Century; listings for current and recent papers include links to the arXiv version as well as the official publisher's version.)

    As for conference papers: more and more conference organizers in my field seem to be deciding not to bother with published conference proceedings, but instead request that the authors provide PDF or PowerPoint copies of the talks which are then posted on the conference web site (along with, in some cases, videos of the talks). There's no way to represent these in the arXiv (yet, anyway), so it's possible to imagine a future where conference proceedings become a smaller and smaller part of "the literature."

  6. Noah Motion said,

    December 18, 2013 @ 3:57 pm

    It seems to me that conferences that use abstract-only systems more readily allow for presentation of fresh, new research than do conferences that require papers. A project has to be fairly far along to get a 4-8 page paper out of it. Farther along than is needed for an abstract (at the time of submission), anyway.

    And what's the average length of a linguistics paper in, say, Language? Phonology? Journal of Phonetics? Linguistics papers tend to be long and complex. You could maybe make the case that linguistics papers tend to be too long and complex, but 4-8 pages seems too short, and it's easy to imagine a number of bad side effects if the primary vehicle for published linguistics research was this kind of conference paper.

  7. peter said,

    December 18, 2013 @ 5:45 pm

    Fields Medallist Bill Thurston had this to say on papers in geometry, his area of mathematics:

    "As long as people in the field are comfortable that the idea works, it doesn’t need to have a formal written source.

    At first I was highly suspicious of this process. I would doubt whether a certain idea was really established. But I found that I could ask people, and they could produce explanations and proofs, or else refer me to other people or to written sources that would give explanations and proofs. There were published theorems that were generally known to be false, or where the proofs were generally known to be incomplete. Mathematical knowledge and understanding were embedded in the minds and in the social fabric of the community of people thinking about a particular topic. This knowledge was supported by written documents, but the written documents were not really primary.

    I think this pattern varies quite a bit from field to field. I was interested in geometric areas of mathematics, where it is often pretty hard to have a document that reflects well the way people actually think. In more algebraic or symbolic fields, this is not necessarily so, and I have the impression that in some areas documents are much closer to carrying the life of the field."

    William F. Thurston [1994]: On proof and progress in mathematics. Bulletin of the American Mathematical Society, 30 (2): 161-177

  8. Jonathan Badger said,

    December 18, 2013 @ 6:31 pm

    Peer-reviewed conference papers and preprints haven't really caught on in (experimental) biology either. Conferences in biology are basically an excuse to network and hear what other people are working on — not actually somewhere where results are presented in a formal sense the way they are in mathematics and related fields, and while a few biologists are beginning to submit preprints to arXiv, most basically have the attitude "Why? That's as much effort as writing a journal paper with none of the benefits".

  9. CBK said,

    December 18, 2013 @ 8:16 pm

    National Bureau of Economic Research working papers are like journal articles in the sense that they are not freely available to everyone. There is a $5 charge, but paying it appears to require mailing in a check or making a phone call. NBER working papers are, however, free to journalists (see

  10. Peter Erwin said,

    December 19, 2013 @ 1:01 am

    Jonathan Badger:
    For what it's worth, conference papers in astronomy are never "peer-reviewed"; they're meant to be summary versions of the talk or poster you presented at the conference, and are typically written a few months after the conference itself. (So the conferences themselves are probably similar to what you're used to.) The most useful kinds of conference papers tend, in my opinion, to be those based on invited review talks for a particular sub-topic; papers based on "original research by me and my group" talks can be interesting, but usually lack the information needed to properly evaluate the results.

    I admit to being curious as to why people in your field think preparing preprints is "as much effort as writing a journal paper", when the traditional idea of preprints in the physics world has usually been something along the lines of "what we submitted to the journal, without the final proofreading and formatting from the journal."

  11. Bob Ladd said,

    December 19, 2013 @ 3:11 am

    Has any anthropologist or sociologist of science ever done a serious study of the cultural differences between different academic fields (the existence of which this comment thread clearly demonstrates)?

  12. Dr. Decay said,

    December 19, 2013 @ 4:40 am

    My observation of the physics community is that the vast majority submit their manuscripts in parallel to a journal and to the arXiv, sometimes before, sometimes after acceptance. Most people closely survey the arXiv to get the latest results and information. But if you look at the citations in their articles, they are overwhelmingly to peer reviewed journals. Anything that exists only on the arXiv is a little suspect. Maybe that's a good thing.
    Of course it may have also to do with citation counters like the Web of Science who tend to include only "real" journals. Since we are often judged on the basis of their citation data, we have to play the game.

  13. Jonathan Badger said,

    December 19, 2013 @ 6:21 am

    Interesting –it sounds like astronomy is a third way — As a computational biologist I'm basically familiar with the computer science world (where the conferences are where peer-reviewed papers of original results are presented and considered published) and the experimental biology world (where, much like Mark describes the mainstream linguistics world, conference papers don't really exist as such and journals are the definitive publication venue).

    In regard to biologists and arXiv, yes, it probably wouldn't be *that* much additional work objectively given that you could use the same manuscript that you submit to a journal, but the culture in biology is such that there isn't much benefit as neither tenure committees nor funding agencies would count posting to arXiv as publishing results as they haven't been peer reviewed. And if you are working in a particularly competitive subfield there is a worry that preprints could tip off competitors leading to getting scooped in getting the first journal article.

  14. Brett said,

    December 19, 2013 @ 9:17 am

    @Jonathan Badger: I don't think the "third way" description is really apt, because I don't think that what has happened in computer science is basically an unrelated phenomenon. In CS, refereed conferences have basically replaced journal publication. Whereas in math, physics, astronomy, etc., the old journals still exist, and, as Dr. Decay pointed out, preprints that are not subsequently published in journals are considered somewhat suspect. Tenure and grants are based on what actually get refereed, but what the preprints do is make the actual information freely available to anyone, as soon as the author(s) have got it in a reasonably final-looking form.

    I think the concern about being scooped is, frankly, the strangest objection to the preprint system that I hear from people in other fields. While many people aren't going to take something that is only on the arXiv that seriously, they can use the arXiv posting date to see who came up with an important result first. The arXiv date should be roughly the same as the journal submission date for establishing priority. Why would anyone take seriously a priority claim based on having the first journal article, when somebody else posted the same thing on the arXiv months before?

  15. AndrewD said,

    December 19, 2013 @ 9:40 am

    I believe Darren Naish who blogs at Tetrapod Zoology ( ) has published Palaeontology papers in/at ArXiv. One reason as I understand it, was the severe lead times in Journal publication the other was the belief in open access publishing .

  16. Alon Lischinsky said,

    December 19, 2013 @ 11:36 am

    @Bob Ladd:

    Has any anthropologist or sociologist of science ever done a serious study of the cultural differences between different academic fields?

    yes, of course. It's a huge topic, so I couldn't even begin to sketch a bibliography, but the acknowledged classic is Becher & Trowler (2001) Academic Tribes and Territories: Intellectual enquiry and the culture of disciplines, 2nd ed. Buckingham & Philadelphia, PA: OUP.

  17. Bob Ladd said,

    December 19, 2013 @ 5:18 pm

    @ Alon Lischinsky: Thank you!

  18. George Finecott said,

    February 13, 2014 @ 5:46 am

    I admit to being inquisitive with reference to why individuals in your field think planning pre – prints is "to the extent that as composing a diary paper", when the conventional thought of pre – prints in the physical science world has normally been something along the lines of "what we submitted to the diary, without the last edit and designing from the journal.

RSS feed for comments on this post