6

In the last decade, more and more researchers read the scientific literature on electronic copies or on printed copies, but not anymore directly from the printed journal. Nowadays, I think the most important role of journals is therefore only to offer a well-established editorial and refereeing platform.

Anyway, there is a sort of economic loophole. Let me explain. Research institutions pay both researchers and journals (to have online access and for printed copies) where researchers publish. On the other hand, journals offer a refereeing platform which is based on the unpaid work of referees, which are, again, researchers. Therefore, research institutions pay for all, and journals earn money for the research published. I could be a little naive here, but I think not far from truth.

Therefore, why are journals useful, apart from refereeing? For instance, imagine that arXiv, or another open-access archive will introduce refereeing and some sort of editorial filtering. In this scenario, will conventional journal publications still make sense?

Improve this question
4
  • 3
    This is a hypothetical question that asks for pure opinion/guesswork/discussion answers, and as such is probably going to be ruled out of scope for Stack Exchange.
    – keshlam
    Aug 13, 2015 at 23:01
  • 2
    I see your point. I made a small edit. The point is that I really ask why journals are still useful. I used this hypothetical scenario as a counterexample
    – sintetico
    Aug 13, 2015 at 23:10
  • 1
    why are journals useful? Two words: editorial rejection.
    – Cape Code
    Aug 16, 2015 at 11:22
  • Right now the main paid added value of journals is their prestige (which do influence further career possibilities); so there is a mismatch between good things for individual scientists, and for science/society.
    – Piotr Migdal
    Aug 16, 2015 at 11:23

3 Answers 3

Reset to default
3

I used to think in a way similar to the question, but I had a meaningful conversation with a prestigious scholar one time in graduate school, who pointed out that I was missing a key purpose of journals: permanent archival.

The arXiv has been around for about 25 years. We have no idea whether the papers on it will be readable in 100 years. But we know that we can read documents on paper from ancient Egypt, and we have every reason to believe that journals printed on paper and stored in archival libraries can last just as long.

The scholar I spoke with was not particularly worried about refereeing. He was worried about whether his work would still be available in 1000 years. Of course, I didn't ask him if he thought anyone would still be interested in it :)

Improve this answer
13
  • 1
    Journals are actually printed on paper? Regularly? Or just specifically so they can be archived? (in which case there would of course be no reason that an archiving entity should not do the same for everything uploaded to arXiv)
    – O. R. Mapper
    Aug 13, 2015 at 23:15
  • 4
    Moreover, if you start with the arXiv, and then have people referee selected papers, and then pay a press to print the ones that are accepted by the referees, and then have libraries archive those printed versions -- well, all you have done is to re-create the journal system we already have!
    – Oswald Veblen
    Aug 13, 2015 at 23:27
  • 2
    I don't trust the publishers to do the permanental archiving correctly. For example, if you buy Springer math ebooks from the late 90s and early 2000s you often get scans of the books, not real pdfs. Why is that? I asked one of the authors and he confirmed that Springer had actually lost or intentionally deleted the LaTeX source files that he had given them.
    – Thomas Kahle
    Aug 16, 2015 at 11:00
  • 2
    @OswaldVeblen Hmm, I rather have a general suspicion against non-electronic(!) versions of anything for permanent archival. Going forward from here, all archiving should be digital with many copies distributed all over the place. Of course the perfect sources should be archived, not the digital -> paper -> digital scan. In case the source is lost (Egyptian manuscripts) a scan is totally fine, of course.
    – Thomas Kahle
    Aug 16, 2015 at 11:21
  • 1
    Another issue is paper quality. Because of budget constraints publishers are cutting down there too. The paper used to print todays journals may be good for 100 years, but probably not much more.
    – Thomas Kahle
    Aug 16, 2015 at 11:22
3

If arXiv added peer review and editorial filtering, then it essentially would be PLOS ONE with a $1,350 discount on the publication fee.

Improve this answer
7
  • but if referees are unpaid and if there is no additional editorial filtering (the arxiv does already filter submissions, although not like, e.g., Nature...), the arxiv would essentially remain open access and with no publication fee. Did I miss something?
    – sintetico
    Aug 14, 2015 at 8:06
  • PLOS ONE doesn't pay its referees, has similar minimal editorial filtering, and is open access. The only difference is the publication fee (thus my "$1350 discount" statement).
    – jakebeal
    Aug 14, 2015 at 12:12
  • The arxiv is for physics, mathematics, and quantitative sciences. PLOS ONE is not. That said, what would prevent the arxiv to be open access with refereeing and without publication fee?
    – sintetico
    Aug 14, 2015 at 20:06
  • @sintetico I'm sorry, but you are simply incorrect about the scope of PLOS ONE: it has no restriction in scope (i.e., the topics it covers are a superset of arXiv topics). Publication fee is somewhat beside the point: arXiv isn't zero cost, it's just got a different (and currently very lightweight) business model: you're really just haggling over price there. My main point is this: if you gave arXiv peer review and editors, it would be a "conventional journal."
    – jakebeal
    Aug 14, 2015 at 20:27
  • 1
    +1. The answer is simply 'it would be a megajournal'. I suspect these fields find arxiv as a preprint repository more useful than another broad-scope megajournal...
    – Andrew is gone
    Aug 16, 2015 at 8:48
1

Like most human pursuits, research is not planned top to bottom. Things evolve in a historical context. The historical context here is that some journals have long histories and publishing in them is prestigious because important work has been published in those journals before (and consequently they can be very picky). Yes, this argument is a bit cum hoc ergo propter hoc to begin with, but still there is a whole host of consequences following from the reputation of journals. For example grants, jobs, publications in other journals, and in general evaluation of ones scientific output is measured in terms of which journals one publishes in.

As other people have noted, turning arXiv into a journal would not make journals obsolete, because that alone would not give it the filtering effect. In my field, mathematics, and maybe also other fields where arXiv usage is widespread, publication in a journal is a stamp of quality that is added to a paper. It is not publication in the sense that the paper is then public and was not before.

As I wrote in a comment, I don't trust the journals with permanental archiving. Libraries are probably better with that, but they can't afford the paper versions anymore... so there is a problem (independent of your question). Another thing is typesetting: If the authors care, then typesetting of the arXiv versions is usually much better than in the journal versions because Springer, Elsevier, etc. have all outsourced typesetting and cut down proofreading. Typically more errors and oddities are introduced than removed in the final typesetting process.

So to answer your question: If arXiv introduced refereeing, then it would turn into another online only journal with little to no impact to the remaining journal system. But, its production cost would substantially go up, potentially endangering it.

Improve this answer
7
  • Your latest argument (introducing refereeing would increase production costs) sounds strange to me. AFAIK, practically all refereeing and editorial work in respectable journals is volunteering (service) and, therefore, cannot impact production costs of a journal. In regard to formatting, with all my positive feelings to arXiv, I wouldn't say that papers there are better formatted than those, published in Springer or Elsevier journals, quite the opposite. Of course, this is more of a function of authors' attention to details & aesthetics rather then journals' formatting requirements.
    – Aleksandr Blekh
    Aug 16, 2015 at 11:41
  • @AleksandrBlekh Yes, all refereeing is done by the community, but there is still a subtantial cost for maintaining the editorial technology. Ask the people who run diamond open access journals. Usually this is subsidized in the form of Libraries providing the infrastructure. Regarding typesetting: I'm speaking mostly about my own papers. I care a lot about typesetting. Granted, an average arXiv preprint may look worse than the published version.
    – Thomas Kahle
    Aug 16, 2015 at 11:46
  • I understand about typesetting. As for the costs, you are right about editorial technology being a significant, if not the only, cost factor. However, such costs should be similar for both cases (today's arXiv vs. arXiv with editorial functionality), if arXiv would use some of the mature open source academic publishing solutions. For examples, please see my relevant answer.
    – Aleksandr Blekh
    Aug 16, 2015 at 11:58
  • I have used at least OJS, and it falls behind the commercial solutions. You need staff to run the open systems for journals publishing more than a few papers a year. Also look at the cost of arXiv now, without this whole part. It seems like it's just a website where people upload papers, but the projected cost for 2015 is slightly south of $800,000 according to confluence.cornell.edu/download/attachments/127116484/…
    – Thomas Kahle
    Aug 16, 2015 at 12:07
  • 1) Perhaps, you're right about the issues with limited functionality or not that fancy and elegant UI / UX. I definitely don't have enough experience to argue with that, though, per description, some open source software (OSS) solutions have comparable or, at least, good enough set of features. With that in mind, I'm wondering whether the differences in functionality between commercial and OSS solutions are not essential for editorial and workflow functions, but, rather, reside in the dimension of convenience functions, mostly for editorial staff.
    – Aleksandr Blekh
    Aug 16, 2015 at 12:59

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