Monday, January 23, 2012

What should we do?

Time for a post by tweet request.
Quite a lot on the Internets on Tim Gowers' promise not to work with Elsevier anymore. I'm not as anti-Elsevier as Gowers or many of my readers but I understand the frustrations.

It's easy to make a promise not to publish, edit or referee papers, especially when you don't need to improve your academic reputation. Still a mathematician of his magnitude really puts a spotlight on that publisher.

Because of the Elsevier stigma we've had for several years, all the theoretical CS journals of Elsevier are not nearly as strong as they have been in the past. So you don't accomplish much more just by boycotting Elsevier.

Making a difference means what you do, not what you don't do. Be sure and support journals that are worthy of support by submitting and refereeing papers and serving on editorial boards. The best attack on publishers that you don't like is to have several strong alternatives. The best way to make them strong is by having your support.

No journals is completely free of cost, they require money or time. Open access journals without page charges generally have no revenue stream and require effort to make to publish the journal. For these journals you can volunteer your time as well as submitting, refereeing and editing.

Remember it's easy to complain and say what you won't do but it is what you do do that makes the difference.

24 comments:

  1. > the theoretical CS journals of Elsevier are not nearly
    > as strong as they have been in the past

    looking at the editorial boards they don't seem weak, for example:


    Computational Geometry

    Editors-in-Chief
    J.-R. Sack
    K. Mehlhorn

    Honorary Editor
    F.P. Preparata

    Editorial Board
    P.K. Agarwal
    S. Akl
    S. Arya
    D. Avis
    H. Brönnimann
    T.M. Chan
    O. Cheong
    K. Clarkson
    R. Fleischer
    K. Fukuda
    J.E. Goodman
    M.T. Goodrich
    F. Hurtado
    N. Katoh
    R. Klein
    N. Magnenat-Thalmann
    J. Mitchell
    J. Pach
    R. Pollack
    G. Rote
    M. Sharir
    P.G. Spirakis
    S. Suri
    M. Teillaud
    T. Tokuyama
    G. Toussaint
    D. Wagner
    E. Welzl
    P. Widmayer
    J. Zaks
    M. de Berg





    Discrete Applied Mathematics

    Editor-in-Chief
    Endre Boros

    Associate Editors
    M.F. Anjos
    W. Kubiak
    V. Lozin
    D. de Werra

    Advisory Editors
    E. Balas
    T. Ibaraki
    D.J. Kleitman
    B. Korte

    Board of Editors
    N. Alon
    M. Anthony
    D. Avis
    J.-C. Bermond
    A. Brandstädt
    R.E. Burkard
    X. Chen
    V. Chvátal
    D.G. Corneil
    D. Coudert
    S. Fujishige
    F. Giannessi
    M.C. Golumbic
    R.L. Graham
    P. Hansen
    P. Hell
    A.J. Hoffman
    G. Kalai
    R.M. Karp
    R. Klasing
    S. Klavžar
    J. Lee
    L Liberti
    L. Lovász
    F. Maffioli
    S. Martello
    C. McDiarmid
    N. Megiddo
    R.H. Möhring
    G.L. Nemhauser
    H. Noltemeïer
    A. Prékopa
    R.C. Read
    F.S. Roberts
    M. Saks
    A. Salomaa
    A. Schrijver
    M. Segal
    P. Toth
    Zs. Tuza
    W. Zang

    ReplyDelete
  2. Looking at Board members is not a good criteria for how strong
    a journal is, though it helps give credibility to a NEW journal.

    The real criteria is- are there good articles published there?
    Granted this may be a hard thing to measure- one way would be
    which journals get cited the most.

    However, I agree with Lance- my intuitive feeling is that the
    Elsevier Journals are looking weaker, though I agree with Jeffe's
    inevitable comment : [citation needed]

    ReplyDelete
    Replies
    1. > The real criteria is- are there good articles published there?

      Special Issue on the 25th Annual Symposium on Computational Geometry (SoCG'09) Volume 43, Issue 8 (2010)

      Special Issue on the 24th European Workshop on Computational Geometry (EuroCG'08) Volume 43, Issue 2 (2010)

      Delete
    2. The SOCG special issues in CGTA (and in the World Scientific journal IJCGA) are long-established tradition, but like most traditions, they're maintained by inertia. A significant fraction of the SOCG community would prefer to see that tradition die; that was one of the reasons behind the recent founding of the open-access Journal of Computational Geometry.

      Delete
  3. The Theory community took two quite specific actions years ago. Under Knuth's leadership, the entire board of the Journal of Algorithms resigned, and the journal was reconstituted as the ACM Transactions on Algorithms (now published by ACM). And the FOCS/STOC special issues, which used to be published in the Journal of Computer and System Sciences (Elsevier), are now published in SIAM Journal of Computing. I think these were effective actions.

    The one bit of frustration I still have is that for someone like me, on the border between Theory and Operations Research, the most viable outlets for quick, short notes are Information Processing Letters and Operations Research Letters, both Elsevier journals. I'd love to have something equally strong that wasn't an Elsevier journal. (I don't want to write a long post about why I think simply releasing such notes into the arxiv is insufficient for me).

    ReplyDelete
    Replies
    1. Hi David
      this came up on G+ as well. I'd be interested in hearing why you thought releasing such notes on the arxiv wasn't enough ? (maybe a guest post :))

      Delete
    2. If you have any coauthors who are not already tenured professors, then going arxiv-only is not an option.

      We need journals not for their typesetting, or their binding articles into volumes, or their web hosting; but because most people working in academia still need the credentials that come from having your work accepted by a competitive venue. Virtual journals are unlikely to be competitive until they find a way to deal with the fact that they have unlimited space.

      Delete
    3. Suresh: Try convincing your dean than your arxiv uploads count as peer-reviewed publications. Also, this: http://scholarlykitchen.sspnet.org/2010/01/04/why-hasnt-scientific-publishing-been-disrupted-already/ (hat tip Moshe Vardi)

      Delete
    4. But this is a strawman argument. I'm not arguing that *all* publications should be arxiv only. I was talking about the 'short notes' that David (and AC) were talking about. if your tenure case rests on the publication of short notes only, good luck !

      Delete
    5. I recently had a similar issue as well, and found the ToC to offer a great alternative to IPL:

      http://theoryofcomputing.org/categories/note.html

      Delete
    6. Peer review helps filter, recommend, and improve manuscripts. I would value that for short notes as well as full papers.

      Delete
    7. I'd be interested in hearing why you thought releasing such notes on the arxiv wasn't enough ?

      Err...if you'd like to "count" a publication toward a grant (I'm thinking here especially of non-NSF grants), then an arxiv publication won't cut it. (And it shouldn't, since there is basically no filtering of what can get "published" there.)

      Delete
  4. In 2003/2004 the entire Journal of Algorithms editorial board resigned following Don Knuth's letter regarding Eslevier's practices. This resulted in the founding of ACM Transactions on Algorithms as a replacement for the community. Shortly after, JCSS, another Elsevier journal originally from Academic Press, lost its position as the publisher for many conference special issues, including STOC/FOCS, which took it down from its position as one of the top 3 journals in the field. (It was never a strong place for non-special-issue papers and this seems to have dropped JCSS down to 6th or 7th at best.) However, strangely enough its strong editorial board, including Don Knuth, is still in place despite the founding of ACM Transactions on Computation Theory and the open access Theory of Computing, which are two new journals covering the same range as JCSS.

    ReplyDelete
  5. Ok, yes, the cost of publishing cannot go to literally zero.

    On the other hand, no publisher was forced to support PIPA, SOPA or the Research Works Act. I think that if you support one of those, then you have chosen to *oppose* the advancement of knowledge, in favor of keeping it behind private firewalls for profit.

    We rarely have these polarizing issues that let us know where people stand. Today we do. We know the AAP is working to make the world a darker and more ignorant place with their lobbying. And yet, ACM is inexplicably still a member.

    ReplyDelete
  6. Ok, yes, the cost of publishing cannot go to literally zero.

    On the other hand, no publisher was forced to support PIPA, SOPA or the Research Works Act. I think that if you support one of those, then you have chosen to *oppose* the advancement of knowledge, in favor of keeping it behind private firewalls for profit.

    We rarely have these polarizing issues that let us know where people stand. Today we do. We know the AAP is working to make the world a darker and more ignorant place with their lobbying. And yet, ACM is inexplicably still a member.

    ReplyDelete
  7. lucky for TCS; we never cared about Journal publication. Let's boycott all journal:-)

    ReplyDelete
    Replies
    1. Conference proceedings are basically like journals, in terms of paywalls, subscriptions costs for libraries, and now, this money going in part to lobby for SOPA, etc.

      Delete
  8. I don't buy the "short notes" value of IPL at all. The quality is poor. We dropped it from our library years ago when it was $2700+ per year.

    Discrete Math is the worst Elsevier journal in CS-related fields at about $5K per year the last time I saw it a couple of years ago. By comparison, the entire ACM DL including ALL journals AND conference proceedings is $2K-$17.5K per year for the academic library price list, depending on the kind of institution. See: http://librarians.acm.org/academic/pricing. As much as we might complain about paywalls, the ACM DL seems like pretty damn good value, even when compared with the arxiv.

    ReplyDelete
    Replies
    1. I don't buy the "short notes" value of IPL at all. The quality is poor.

      This shows a complete lack of understanding of the goals of IPL. This is like saying the food in the university cafeteria is not as good as in a restaurant, which while very much true, completely misses the point.

      IPL provides a forum in which to publish "interesting results in the field of information processing in the form of short, concise papers".

      It meets this mandate very well. It has the shortest turn around time in the field and a citation index that is about half the average of the field, which is excellent, given that its mandate is not to publish long-winded breakthroughs but short and concise papers. To put it another way, its citation index per page is higher than most journals in TCS.

      Delete
    2. Dear Short and sweet, you have conceded the point by equating IPL to university cafeteria food.

      Delete
  9. You seem to believe that asserting that IPL is lower quality is an astute observation, even though I said exactly that in the first sentence of my post: "which while very much true, completely misses the point."

    The debate is not if IPL is lower quality (per paper). I'm taking issue with the first part of Paul's statement "I don't buy the value of IPL". That is where the mistake is: lower quality is not the same as "no value", indeed in some cases it is not even the same as "less value".

    For example, cafeteria food certainly has value. Moreover if you are at work on a limited budget and with limited time, it has higher value (in the sense of utility to you) than a sit down six course three star restaurant would in that situation.

    ReplyDelete
  10. Short and Sweet: Actually, when I was talking low value, I was talking per issue, per year, and per dollar, NOT per article.

    In the case of a long, possibly complex article, the value of journal refereeing is significant. That isn't the case for good quality short, easy-to-read articles of the sort that might once have appeared in IPL. E-print places like ECCC (or the arxiv) are much faster than IPL. I don't need a referee intervening for these papers. Indeed, since the mid-1990's, for computational complexity at least, ECCC has been the preferred place to publish such short notes and almost articles from IPL that are regularly cited are from the early 1990's or before.

    ReplyDelete
  11. To go back to the part of the post that I elided, I value publishing short notes in journals because I view anonymous peer-review as an essential part of the scientific process (even for short notes), and because I value the ability of journals to provide long-term archiving. I don't know if anyone will want to read what is in IPL/ORL 50 years from now, but we need to make sure that they can if they need to.

    Paul, I can agree that IPL might not be worth the price, but that's why I want an alternative not published by Elsevier. The price is the issue introduced by Elsevier; the quality I think is fine for what IPL is as a journal.

    ReplyDelete
  12. I couldn't agree more with this. For our part, here at NU's new Center for Scholarly Communication & Digital Curation, we are working on supporting new modes and models of scholarly publishing, as well as creating new and open digital platforms for publishing. We are always interested in hearing from faculty and graduate students from all disciplines (we need to know what you need), so please feel free to get in touch anytime (cscdc at northwestern dot edu), and thanks again for your post.

    ReplyDelete