Discrete Analysis

You may have seen Tim Gowers’s announcement last week, but if not, here’s the news: we are launching a new arXiv overlay journal called Discrete Analysis. The editorial board consists of Tim Gowers (who will be the managing editor) and Ernie Croot, Ben Green, Gil Kalai, Nets Katz, Bryna Kra, myself, Tom Sanders, Jozsef Solymosi, Terence Tao, Julia Wolf, and Tamar Ziegler. As should be clear from this list of names, the journal will focus on additive combinatorics and related areas such as harmonic analysis, number theory, geometric measure theory, combinatorics, ergodic theory. The temporary journal website is open now, in fact we have already received the first submissions.

“ArXiv overlay” means that we will not be “publishing” papers in the traditional sense. Most of us already typeset our own papers and use the arXiv for quick, reliable, stable worldwide dissemination of our results. It is not clear that mathematical journals can improve much on that; if anything, publication in established journals is currently more likely to impede the dissemination of science through paywalls or embargos than to facilitate it. What we can provide is a refereeing and certification service where we manage the peer review and, when the outcome of the review is positive, attest through publishing the link on the journal website that the paper has been judged to be of suitable quality for publication in Discrete Analysis. Tim’s post has much more information on both the scope of the journal and the technical details of how we expect it to work. If you are finishing an article in one of the covered areas of research, I hope that you will consider Discrete Analysis as a possible publication venue. I’m proud to be on its board.

A few more inside-baseball comments under the cut.

If you have not yet been disabused of the notion that open access will be a scientist’s complete paradise, the first few comments on Tim’s post should do it. Several major funding agencies, including NSERC which funds my research, adopted open access policies in the last few years. The good part is that commercial publishers are being forced to comply. Having our papers posted on our webpages and on open repositories such as arXiv used to be a “grey zone” practice where the publishers would mostly look the other way but could clamp down on it if they wished; now it has become commonplace for publishing agreements to include explicit allowances to that effect. The sad part is that the scientists, in addition to typesetting their own papers, refereeing them, and managing the peer review as editors, all for free, are now also required to become their own IP attorneys, also for free. Evidently, publication in an arXiv overlay journal might not meet the standards for “gold” open access (look it up) if the authors choose the wrong type of CC license. An electronic journal consisting of a webpage ran by scientists and their institutional staff, and available for free to both authors and readers, might be classified as “status unclear” in compliance databases. To her credit, a representative of that particular database showed up in comments and promised to add that journal to their “list to investigate further.” Great. Let’s hope that this is sorted out before anyone has their funding suspended for noncompliance.

In a better world, a reliable, up-to-date list of open access journals would be maintained by the research office at the university where we are employed. Any questions such as those above would be resolved quickly, frictionlessly, and with no time cost to the researcher. There would be no danger that a clerical omission at an external agency would cause our funding to be suspended or revoked. Appropriate guidance on copyright would be available, and we would be able to ask for advice as needed, with the crucial provision that the research office representatives would be actively protecting our interests as researchers rather than just passing on the directives from the funding agencies. In other words, the research office would be our IP attorney, as opposed to us taking on the additional duty. This should not be too much to ask, given the current size of the administrative sectors of universities (the administration to faculty ratio at UBC is 2:1, not including clerical staff). I somehow have trouble believing that it will work that way for most of us.

I’m still all for open access. I would not be able to function the same way as a scientist without unlimited free access, from anywhere in the world, to the arXiv and to the latest research in my field. That’s why I’ve joined the Discrete Analysis board. But there should also be more pressure on our various “research support services” to actually start supporting us. We can’t just keep accepting more work, and more types of work, without it taking a toll on our research capacity.

For those of you who might remember my earlier posts about math publishing: I was skeptical of some of the proposed reforms on the grounds that they bundled open access with features I considered undesirable, such as “open peer review” aka unmoderated comment sections. That was in 2013. It’s 2015 now, and open comment sections as the default option on the internet seem to be on their way out. Many websites have closed their comment sections altogether; others are keeping them but shifting more weight to other platforms. It would be nice to think that academic comment sections should be better, but that is not always the case. Having good comments requires not only paid moderation, but also making decisions about community values that, in my experience, mathematicians are not willing to make.

I’m glad that Discrete Analysis does not have a comment section. I would not submit an article to a journal with a comment section, nor would I agree to join the editorial board of such a journal, unless that journal were to adapt moderation standards that would be acceptable to me. (Given that mathematicians seem to have unlimited tolerance for low-level sexism in comments, I do not anticipate that happening anytime soon.) That said, I’ve always been in favour of more diversity in publishing, not less. I’ll be happy to see more independent, scientist-managed journals, even if some of them do have comments. I’ll be happy to have more journals that publish negative, inconclusive, and “uninteresting” results. (That last article is long but worth the time.) I’m not making it my personal goal to put commercial journals out of business, either. I’d much rather see them adapt by changing their policies and providing additional services that might be worth paying for. For example, I’ve had to deal with papers where the proofs were essentially correct but sloppy, the constants didn’t match, or the citations pointed a bit sideways. I’ve also had to deal with papers where the English was awkward to the point of being incomprehensible. Some of these papers are valuable and worth publishing. There’s only so much that anyone can ask from an unpaid referee, but a commercial publisher could employ editors and fact-checkers who are actual trained mathematicians with doctoral degrees (and who might consider such employment as a reasonable alternative to the current academic job market).

Now that we are in the realm of daydreaming already, something else I’d love to have would be an equivalent of This. for math papers. This. is a network for exchanging reading recommendations: members can post at most one link per day and can follow other members to see their recommendations. (It is still in beta, but should be open to the public in a few weeks.) For some time now, I have not been able to keep track of all arXiv submissions that I might be interested in. The volume is overwhelming, even if I try to restrict it to research areas close to my own work. Additionally, if I neglect checking the arXiv for even just a few days while I’m travelling or on vacation, it becomes impossible to catch up. I’ve given up on trying. But if I could sign up for recommendations from my collaborators and people whose taste I trust? I certainly would. If I had a way of bookmarking papers of interest to my collaborators and researchers in my area, without mass-mailing them (too intrusive) or writing a blog post about it each time (too time-consuming)? I’d do that, too. I’m assuming that this would cost money and, as such, might not be a DIY affair for scientists. But one can dream.