Tim Gowers's excellent blog posting focused the long-standing discontent of the research community with Elsevier and nucleated a boycott which may prove to be a historic moment in scholarly publishing. I urge others to join the thousands of researchers who have signed on at the website Tyler Neylon created at thecostofknowledge.com. Recent articles in publications like Forbes and The Economist indicate that Elsevier and the rest of the business community are taking note.

The arguments Gowers laid out focus on Elsevier's high prices, their bundling arrangements and subscription agreements, and their support for new laws that seem aimed at increasing publishers' profits at the expense of wide dissemination of scholarly research. These are all very convincing reasons for the boycott. The fundamental mechanism of capitalism, that prices are contained by consumers' choices not to pay high prices for what is available elsewhere more cheaply, is augmented in this case by the fact that the consumers—researchers and their institutions—also freely donate the most crucial ingredients in Elsevier's products, that is, papers, referees, and editors.

However, there is another reason for researchers to disassociate from Elsevier, which I find even more compelling: their many lapses in ethical and quality publishing practices. Here are some examples:

The Elsevier journal Chaos, Solitons and Fractals published more than 300 papers by the journal's Editor-In-Chief (58 in a single year). That these papers were not subject to peer review was later confirmed by the EIC's declaration that "senior people are above this childish, vain practice of peer review." Although the copious self-publication had begun nearly 20 years earlier, the EIC's retirement from the journal occurred only in 2009.

Elsevier journals have repeatedly published plagiarized work and duplicate publications. A search turns up over thirty papers in Elsevier mathematics journals published in the last decade which have had to be formally retracted, mostly for these reasons. On more than one occasion, the same paper has been published in two different volumes of the same Elsevier journal (presumably by accident).

Elsevier math journals have published a number of papers that make me doubt that they were subject to any peer-review whatever. An egregious example is the 2-page paper "A computer application in mathematics" in Computers and Mathematics with Applications, vol. 59 (2010) pp. 296-297, which purports to prove the parallel postulate (!) with no formulas, no references to other published works except for two papers by the same authors, and, as best as I can tell, no meaningful content whatever. This paper remains unretracted and available for sale on the Elsevier web site.

On several occasions, entire editorial boards have collectively resigned from Elsevier, usually citing discontent with their pricing. In 1999, the entire 50 person editorial board of the Journal of Logic Programming resigned after 16 months of unsuccessful negotiations with Elsevier over pricing, and created a new journal with a different publisher. In 2003, the entire board of Journal of Algorithms did likewise. A well known-case is that of the journal Topology, whose distinguished editorial board resigned en masse in 2006, again to found a different, less expensive, journal. Elsevier has continued to publish Topology since then, even though the web page for the journal gives no indication of its editorial board or even that the journal has an editorial board!

From 2000 to 2005 Elsevier published six phony biomedical journals, with titles such as the Australasian Journal of Cardiology, in return for an undisclosed sum from a large pharmaceutical company. The journals' contents were provided by the pharmaceutical company and published without further review, mostly reporting data favorable to their products. In 2009, after the practice came to light in a law suit against the pharmaceutical company and was reported in the press, Elsevier admitted that they had "published a series of sponsored article compilation publications, on behalf of pharmaceutical clients, that were made to look like journals and lacked the proper disclosure" and expressed regret.

In 1998, the Elsevier journal Lancet published one of the most significant examples of fraudulent scientific research in recent times, in which evidence was fabricated to link autism to measles, mumps, and rubella vaccine, thereby setting off a health scare that led to deaths and severe injuries and which continues to this day. Brian Deer, the award-winning investigative reporter whose articles in the Times of London and the British Medical Journal (BMJ) exposed the fraud and led to the inquiry by the UK General Medical Council which struck Wakefield from the UK medical register, also published a detailed and damning article in BMJ about the response of the Lancet editorial staff when confronted with the evidence of fraud. Deer describes how the Lancet responded to his carefully documented accusations, not with a formal investigation, but rather with a "5000 word avalanche of denials" and "a scramble to discredit my claims". Lancet's retraction of the paper did not come until 2010, 12 years after the original publication and for reasons that a British Medical Journal editorial describes as "far narrower misconduct than is now apparent."

Of course, Elsevier produces many journals. Without a doubt they publish good articles, as well as bad, and include excellent scientists in their editorial boards, as well as others. But the number and the nature of the incidents like those listed, cause me to doubt their commitment to and/or ability to achieve the quality and ethical standards that I believe crucial. There are many other publishers, including especially scholarly society publishers and university presses, but also some commercial publishers, that have earned my confidence and respect.  I will happily dedicate my efforts of authorship and editorial work to them and not to Elsevier.

Doug Arnold

The electronic era of publishing presents an opportunity to move far beyond merely adapting the current journal model.  In my September, 2011 column in the Notices of the American Mathematical Society, I noted that, of the four original benefits offered by traditional journals — enhancement, dissemination,archiving, and validation — in the electronic era only the last remains: validation of the correctness, originality, and status of papers, thereby validation of its authors' status for hiring, promotion, grant funding, salary, etc.  However, not all fields remain in thrall to the journal system, and status in areas such as computer vision, cryptography, and computer algebra relies nowadays primarily on prestigious conference proceedings.

In the column, I propose, as a possible alternative to journal validation, establishing a ``Michelin Guide'' to mathematical works on the internet.  Thus, one's vita might contain 5 one-star, 3 two-star, and one very rare three-star ``Mathelin'' rated paper, as evaluated by the Guide's referees.  Moreover, unlike journals, Mathelin ratings can well be time-varying.  One can even envision a variety of general purpose Guides, with competing ratings, as well as specialized Guides that convey status within a particular field.  What is less clear is how such a system could be practically instituted, although it is worth noting that a large amount of money is now being fed to commercial publishers from libraries and institutions worldwide, and even diverting even a small fraction would easily support the running and compensated refereeing of such a Guide.

Related proposals include Stefan Müller's that internet-based journals be overlays of the arXiv, where posted preprints can be submitted and refereed.  Once accepted, no additional publication would be required, but the paper would then carry the imprimatur of the overlay journal and be listed therein.   Timothy Gowers, in an earlier (12.12.2011) comment on this blog, points to his proposal to implement a system "something like a cross between the arXiv, a social networking site, Amazon book reviews, and Mathoverflow", detailing its advantages over traditional validation based on unpublished, anonymous referee reports.  As I state in my column: "The time is ripe for a radical rethinking of the traditional academic model for scholarly communication within mathematics.  ...  If we are not properly engaged, the future will be decided for us and, almost certainly, will not be to our liking."

Peter J. Olver

Following the suggestions from Richard Taylor and from Teresa Krick, we would like to gather your statements pertaining to pricing for mathematical journals.

Quoting Richard, "Maybe instead the IMU could consider what if anything it can do about the unreasonable prices of some mathematics journals that are taking badly needed money from our universities? And, Teresa Krick wonders "how to fight against the outrageous rates publishing houses are charging the scientific communities, for a job the scientific community does for them from the beginning to the end."

This topic has been discussed at great length in librarian circles, and also written about in professional math periodicals. Purchasing journals is nowadays done mostly at the national license level: bigger deals, larger aggregations of content. Are mathematicians having any say on prioritization, or usage of funds in this way? Where will this all lead as time goes on? Please comment on this issue. ?

Mathematical journals are an important and precious resource for our community. They embody, to some extent, our joint "intellectual property". They provide a way for mathematical researchers to distribute the results of their work. The peer reviewing process, when it works well, gives some guarantee to readers that the published papers are worthwhile and (most likely) correct.