About the title

About the title

I changed the title of the blog on March 20, 2013 (it used to have the title “Notes of an owl”). This was my immediate reaction to the news the T. Gowers was presenting to the public the works of P. Deligne on the occasion of the award of the Abel prize to Deligne in 2013 (by his own admission, T. Gowers is not qualified to do this).

The issue at hand is not just the lack of qualification; the real issue is that the award to P. Deligne is, unfortunately, the best compensation to the mathematical community for the 2012 award of Abel prize to Szemer├ędi. I predicted Deligne before the announcement on these grounds alone. I would prefer if the prize to P. Deligne would be awarded out of pure appreciation of his work.

I believe that mathematicians urgently need to stop the growth of Gowers's influence, and, first of all, his initiatives in mathematical publishing. I wrote extensively about the first one; now there is another: to take over the arXiv overlay electronic journals. The same arguments apply.

Now it looks like this title is very good, contrary to my initial opinion. And there is no way back.

Friday, August 17, 2012

The twist ending. 3. R. Kirby

Previous post: The twist ending. 2. A Cambridge don

R. Kirby (UC at Berkeley) is the Managing Editor of Gowers's journals. This justifies the following digression into Kirby's past achievements in scientific publishing.

In the 90ies he declared a war on the main journal in his field, namely “Topology”. Originally published by “Pergamon Press”, it was sold in early 90ies to Elsevier by late Robert Maxwell when his financial empire started to face serious problems. Of course, this wasn’t a good development, but it remained an excellent journal due, most likely, to its excellent and small editorial board. It was moderately expensive. I still fail to see any reason to single it out (as I don’t see any convincing justification for singling out the Elsevier in the recent boycott).

R. Kirby launched a new journal “Geometry&Topology” specifically intended to compete with “Topology” (and to eventually bring it down). It was published both online and in paper version. Online version was free; the paper version was very cheap initially. In contrast with “Topology”, the editorial board of “Geometry&Topology” was big and growing with time. The journal was also growing, and with the number of pages the price of the paper version was growing (the libraries were encouraged to subscribe to it; technically, for libraries the electronic access never was free). “Geometry&Topology” succeeded in diverting a lot of papers from “Topology”, and the editorial board of “Topology” was constantly pressured to attempt to lower the price (even when the individual subscription price to the paper version of “Topology&Geometry” surpassed that of “Topology”). Elsevier argued that the list price of a subscription is not relevant anymore (by the reasons I explained above using the example of “Inventiones”). The purpose of a relatively high list price, I believe, was to encourage participation in “package deals”. Eventually, the editorial board and Elsevier made a quite reasonable deal substantially lowering the price, but it was too late (and the list price already was not relevant).

On August 10, 2006 the whole editorial board of “Topology” resigned. Elsevier continued to publish the already accepted papers and managed to fill by them the 2007 volume. The subscription to 2007 volume was free for subscribers to the 2006 one. But the journal was, of course, dead.

Within a month (if I remember correctly, already in August 2006) “Geometry&Topology” closed free access to its electronic version or at least announced the imminent closing. Since then the access to the electronic version is by subscription only. Well, this is how much one can trust promises to be freely accessible in perpetuity. At the same time, “Geometry&Topology” doubled the subscription price, and invented some convoluted reason for quadrupling the subscription price for year 2007 (for any form of subscription, electronic or paper, individual or library). Being a member of our Library Committee, I attempted to understand their reasoning, but failed.

Nobody saved any money as a result of success of Kirby’s project. A slightly modified editorial board of “Topology” launched a replacement, “Journal of Topology” (apparently, Elsevier own the rights to the trademark “Topology”). “Geometry&Topology” is not a part of any package deal. I don’t know if the Oxford UP, the publisher of “Journal of Topology”, offers package deals, but our library had to subscribe to it as a standalone journal. So, the cost of subscription to specialized journals in the field of topology for our library substantially increased. If anybody was subscribing to “Geometry&Topology” or “Topology” as an individual, she or he, most likely, lost these subscriptions because of much higher prices.

During this struggle with “Topology” mathematicians gradually started to consider “Geometry&Topology” as the journal of choice for paper in topology and related fields.

So, the main result of Kirby’s ten-year effort is the fact that he now controls the main journal in his field (topology, of course). It seems that he had no chances to get into the editorial board of “Topology”. The co-author of his most famous papers, L. Siebenmann, was a member of the editorial board for decades, first as a regular member, then as a honorary one.

Next post: The twist ending. 4.


  1. Dear sowa,

    All this is very interesting. Two questions:
    In the 90s, did Kirby made an explicit promise to keep the g-t as an open access journal?
    Are there any explanations (online or elsewhere) of Kirby available about raising the price?
    In any case it would be interesting to read any statement of Kirby from the 90s regarding g-t.

  2. Dear Ayudug,

    The question is what you understand by an "explicit promise". When G&T was founded, there was, probably, even no such term as "open access", at least in mathematics (I believe that the term originates in bio-medical sciences). Of course, I may be mistaken; I never followed the births and deaths of various bureaucratic buzz-words.

    There is a technical problem. Most of Kirby's ideas (except the mathematical research papers) were posted on his website. The current version does not even remotely resemble the ones in the previous millennium. Fortunately, there is The Internet Archive and it Wayback Machine. Here is the link to the archived version of his open letter to Elsevier: Letter to Elsevier, 27 May 97. In particular, Kirby wrote:

    "We have started an electronic journal, Geometry & Topology, based at the University of Warwick (see http://www.maths.warwick.ac.uk/gt/). It is free, at least in the foreseeable future. It is off to an excellent start with a large distinguished board of editors and standards at least that of Topology. We expect to offer a paper subscription at a very low price, determined by the cost of a commercial printer using our TeX files."

    Is September of 2006 foreseeable from May 1997? I believe it is; others may argue that even tomorrow is not predictable (and I would agree). What happened between these dates? The first volume of G&T consisted of 109 pages (probably, 10 times less than "Topology"; this is easy to check). It is rather easy to offer such "journal" for free, like mathematicians offer their preprint of research papers or preliminary versions of book for free on their web pages. The 2006 volume consisted of 2158 pages, and had a lot of subscribers to the paper copy. And it had 10 years archive. This is a business on a different scale. Was this foreseeable? Well, if the new journal succeeds, it will inevitably go to this level. If fails, not.

    Also, in addition to the written documents, there was a general impression. Some things are never said directly, they are hinted at, and then the understanding is tested (like in a documentary about WWII I saw a couple of days ago).

    As of explanations for raising the price in 2006, I saw it on the paper and I can hardly expect that I will be able to find this piece of paper. I will try to find some online document, but cannot promise a success. The reason for doubling the price was obvious: running G&T started to cost more money than University of Warwick was willing to contribute. What was tricky is the quadrupling the price for 2007 (only). They wrote that in 2007 they will publish two volumes instead of one, so this doubles the (already doubled) price. But G&T is yearly journal, and a "volume" is by the definition all the papers published in a given year. And now 2007 is listed as one volume. As I said, I did not understand the reasoning, and therefore cannot reproduce it.

    1. Thank you.

      I checked the prices. Electronic only version of G&T is $385, meanwhile the Journal of Topology charges $600. So, presumably, he is really trying to keep the price as low as possible. The most interesting thing is of course your question about this "take-over of the power". Was it intentionally planned and designed or it's just happened... In fact, the filed is so huge that two high quality journals can easily coexist. I will try to find time too look at the dynamics of volume sizes of both journals.

    2. "Journal of Topology" is published by Oxford UP. Its prices are among the highest in the scientific publishing. The prices for its mathematical books are about two times higher than the prices of other publishers. Fortunately, their mathematics program is very small, but occasionally they publish a book I would like to buy...

      At the same time, Oxford UP has a very unusual "Open Access Policy" for London Mathematical Society journals ("J. of Topology" is one of them): http://www.oxfordjournals.org/our_journals/lms_open_access.html. The papers are freely accessible before publication (advanced access to fully formatted, except the page numbers, papers) and 6 months after the publication. This policy is opposite to the usual one, when papers are freely accessible after some period. Practically, this means that a topologist has no need to buy any subscription to "J. of Topology" in order to stay up to date.

      As I wrote in the post, there is no way to know the true intentions. In addition, they may change over time. In the letter to Elsevier Kirby clearly said that he expects to kill "Topology", despite some sentimental feelings toward it. If your goal is to lower prices, or just to have your own business, you do not talk about killing any existing business. You create a competitor, and the prices go down. About 2004 Elsevier offered a deal to the Editorial Board of "Topology": the price is lowered about 2 times (I don't remember the exact numbers), electronic assess is immediate, the paper version delivered once a year. Since it is impossible to check out the current issues of any journal in all libraries I have seen (except at the IHES and in the defunct USSR), the last condition is purely symbolic. But it was too late.

      I disagree that the field of topology is huge. The high days of topology are 60ies, with some spillover into 70ies and a little bit into 80ies. At that time there was only one journal in this field, "Topology", which published not much more than 400 pages per year. There is nothing like that flowering of topology nowadays. It would be much better (for topology and for the whole mathematics) to have only one specialized journal in topology, publishing only really good papers and only closely related to topology (so a paper in, say, Riemannian geometry will usually not qualify). I think that the same 400 pager per year would be enough. But G&T alone publishes much more, and the publication in G&T never was a sign of the quality of the paper comparable to the publication in “Topology” in the past, despite the claims of the editors.

      Finally, "J. of Topology" was not created as a competitor to G&T. It is a replacement of "Topology", and should be compared to it, not to G&T. $600.00 is about 2 times cheaper than the Elsevier's offer I mentioned.

    3. I did some additional search.
      The archive copy of the Subscriptions page in Volume 10, 2006. I reproduce the main part.


      Volume 9, 2005: $360 (paper).
      Volume 9 comprises 2414 pages, and will be printed and mailed by February 2006.

      Volume 10, 2006: $360 (paper).
      Volume 10 is estimated at 2400 pages. It will be printed and mailed in issues of 400-500 pages, whenever 400 to 500 pages of material are published electronically.

      An electronic subscription to G&T for 2006 is $240.

      Note that G&T is making a transition from printing and selling the previous year to the normal mode of journal publication, in which a subscription is for the coming year and the journal is printed periodically. Thus, for only this year, subscriptions for two years are being sold at the same time.

      G&T expects to keep its price below 15¢/page for the paper version, and below 10¢/page for the electronic one.

      The electronic version of G&T will be free one year after publication.

      Back issues and Starter packs

      Geometry & Topology

      Volume 1 1997 (112pp) $18 (currently temporarily out of print)
      Volume 2 1998 (335pp) $50
      Volume 3 1999 (446pp) $70
      Volume 4 2000 (579pp) $80
      Volume 5 2001 (945pp) $140
      Volume 6 2002 (990pp) $150
      Volume 7 2003 (1073pp) $160
      Volume 8 2004 (1499pp) $170
      Volume 9 2005 (2414pp) $360
      Volume 10 2006 (2400pp est) $360

      Starter Pack

      Soft bound copies of volumes 1 to 10 of GT (10,800 pp est) $1,200"

      So, I made a mistake: the price doubled from 2005 volume to 2006 volume, and even more than doubled, and not from 2006 to 2007. The trick with additional doubling is quite clear from the business point of view. It is used sometimes, but is not considered to be an honest business practice. Suppose that you are renting an apartment, and the payment for a month is due on the last day of the month. Then your landlord announces that he changes the model to another one: since September, you will have to pay the rent not on the last, but on the first day of the month. So, your payment for the August is due on August 31, and the payment for September on September 1. The landlord is pocketing forever the interests paid by his bank for September. G&T pocketed the interests for a year. And subscribers had to fit this double payment in a one year budget. I can understand a landlord, especially, at is often the case around universities, most of his renters rent only for one year and move in in September. But why they did this to fellow mathematicians?

      The promise of free availability after one year was honored only for one year. The 2007 volume is freely accessible, the 2008 is not.