Culture Change in Store for SIAM Journals

April 14, 2012

From the SIAM President
Nick Trefethen

Last spring in this column, I raised the issue of the time it takes manuscripts for SIAM journals to travel from their initial submission to their final acceptance or rejection for publication. Since then, this matter has been discussed extensively by SIAM's officers, staff, Journals Committee, and editors-in-chief. I would like now to report some progress.

First, let me review the current situation. SIAM's 15 journals receive a great number of manuscripts each year, about a third of which are eventually published, usually after at least one round of revision that often strengthens the paper significantly. A tremendous amount of volunteer labor is expended in this process by referees, associate editors, and editors-in-chief. The system works. SIAM's journals have an excellent reputation and constitute a permanent record of leading research contributions. Authors know that for a paper to be published, it must contain significant results and be well written, as judged usually by two independent referees. Moreover, even the most smoothly written manuscript will be copy-edited to ensure that the final paper is of as high a standard as possible.

But it all takes so long! Authors also know that after submitting their manuscripts, they may have to wait many months before hearing anything back from SIAM. If the news is good, they will probably have to produce a revised manuscript, and after that is submitted, more months may elapse. This is not a fact of life just for SIAM journals, but for mathematics journals generally. Old-timers like me may not be so troubled by it, but for younger researchers, this can be a serious matter. Many are tempted to send their work instead to disciplinary journals that may publish in half the time.

In investigating this situation, we have looked at all kinds of data. Some SIAM journals work faster than others, for example, partly because they cater to different scholarly cultures. Medians are shorter than means. Rejected manuscripts have one set of statistics and accepted ones another. Time delays happen on desks of editors, referees, and authors, and each of these datasets has its own profile. There is plenty of complexity in this story!

But one must focus a discussion somehow, and the number we have paid particular attention to is the average time, across all SIAM journals, that it takes for a successful manuscript to go from submission to acceptance, currently 12.8 months. We have decided to try to change this. We would like to shorten the average to 89 months.

For the new policy, see below, a summary of a fuller set of procedures. Our attempt will combine (1) a tightening up of some parameters with (2) a commitment to gentleness and flexibility.

First, about those parameters. Referees will soon notice that most SIAM journals, most of the time, will ask a referee to respond within two months instead of the current three. Similarly, authors will be asked to make minor revisions within one month instead of six. (Major revisions may take longer.)

Second, the aim will always be to be flexible. If a referee or an author asks for more time, that will normally be fine. If an editor thinks a manuscript needs extra effort, no problem. And throughout the process, we will strive to be gentle in how we make requests. Many of us know how annoying it can be to receive messages from journals about "assignments" and "deadlines." SIAM doesn't want to talk like that. We will try to avoid the word "deadline," remembering that we are relying on expert labor contributed gratis by busy colleagues.

Our aim is a modest change in culture. As an example of the current culture, my co-authors and I just received a message from a SIAM journal accepting our paper subject to revisions so minor that they will take us only a couple of hours to carry out. The letter asks us to submit our revised manuscript within six months! Well, of course, one might say, if the authors are in no hurry about their paper, why should SIAM care? Yet SIAM does care, for a request like that sends a message about expectations, as do the submission and acceptance dates printed in the final published paper. We would like to change the expectations a little bit.

It will take some time for the new procedures to come into place, and some journals may choose to override the defaults. It may be quite a while before we know if the effort is successful. There will always be special cases, and some papers really do require an exceptional amount of attention. But we hope that before too long, it will be unusual to wait more than a year to get your paper accepted by a SIAM journal.

SIAM'S New Policy on Refereeing and Revision Times

SIAM's 15 journals consider more than 4000 submitted manuscripts every year, an enormous effort that relies on hundreds of volunteer editors and thousands of volunteer referees. As with other mathematics journals, this process often takes quite a while, and in recent years the average interval from submission of a SIAM manuscript to its final acceptance, averaged over all journals, has been about 12.8 months. SIAM has decided to try to improve this figure, with the aim of bringing the average down to 89 months. In particular, SIAM's journal editors will, by default, normally ask a reader to produce a referee report, if possible, within two months. This target can always be overridden for particularly lengthy or difficult manuscripts, and referees can always request extensions. In order to make the best use of volunteer referee efforts, editors also have the mandate to reject (without review) papers that in their opinion are not suitable for publication in the journal due to subject matter or clear weaknesses in scientific content or presentation style.

It is increasingly difficult to find suitable referees for all the papers submitted, and we hope that authors will also do their part to serve as referees for papers when requested by SIAM editors. We welcome comments from authors and referees on the important question of how SIAM should most effectively process its manuscripts. Comments can be addressed to

Donate · Contact Us · Site Map · Join SIAM · My Account
Facebook Twitter Youtube linkedin google+