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. ?

Ivan ZezulaPost authorIf journals are paid for at national level, what about publishing at national level? Wouldn’t it be cheaper to have a national science publishing house? Maybe an economist can calculate it…

John BaezPost authorProfessorScientific journals are expensive because large companies such as Reed-Elsevier and Springer have a monopoly on them. Why are academics eager to publish in these expensive journals instead of switching to cheaper ones? Because the expensive journals are “prestigious”, and academics need that prestige for tenure and promotions. Why are these journals “prestigious”? Because academics are eager to publish in them. Mathematicians can break this vicious circle by creating boards that rate the quality of math papers on the arXiv. These ratings, if they become accepted, can serve to help evaluate mathematicians for tenure and promotion decisions. The IMU is a great position to make this happen. Please do it!

Samuel MonnierPost authorAn open review systemThere is already a database with almost all the recently written math papers: the arXiv. To get rid of journals, all that is needed is a system to assess the quality of these papers.

Now imagine a website where each paper on the arXiv would have an associated discussion board, where people could make comments, ask questions. People would also be able to write (anonymously or not) referee reports on the papers they feel they understand. To avoid spamming and incompetent reviews, one can imagine attributing to each contributor reputation points, much like what is currently done on mathoverflow.net. Similary, the message threads pertaining to each article could be rated, with the purpose of displaying the most interesting and relevant ones first.

All the technology is there and running such a website would cost a tiny fraction of the price currently paid for journals. Moreover, the discussion boards would provide an invaluable resource for the community, would make it much more easy to spot mistakes and conceptual issues and would make this knowledge instantly available to everybody. The reputation system would provide a strong incentive for the referees to write serious reports, and these would be available to anyone to consult.

Of course, the main problem would be to get a significant fraction of the community to use this system, as well as getting it to be recognized by funding agencies. Endorsement by the IMU and ICIAM could probably help a lot.

Timothy GowersPost authorI’ve recently blogged on this topic: my suggestions were similar to the ones above. However, moving to a completely different system of evaluation is not going to be easy when the current system has a considerable stability to it. In the short term, an easy measure that would destabilize it would be for the IMU to endorse a boycott of the most expensive journals. The effect of such a boycott could be that when an individual mathematician is asked to referee a paper for, say, Elsevier, he or she would be able to say no without any feeling of guilt. Since we’d all like to save ourselves a bit of work, this could be very harmful to the journals in question. If that leaves a need for journals of comparable prestige to the ones that have been boycotted, it would be an easy matter to found some. Alternatively, existing journals could simply come to be regarded as more prestigious in the new and smaller field.

It seems to me that a simple statement by the IMU that mathematicians do not have a moral obligation to provide a free service to journals that charge exorbitant prices, together with a list of such journals, would be very easy to produce, and could make a real difference. There may be legal niceties that I am not aware of, but it’s hard to see on what legal grounds the publishers could object to the withdrawal of a voluntary service.

My blog posts are here:

http://gowers.wordpress.com/2011/10/31/how-might-we-get-to-a-new-model-of-mathematical-publishing/

http://gowers.wordpress.com/2011/11/03/a-more-modest-proposal/

Matthew DawsPost authorarXiv is _not_ already usedPersonally, I think the arxiv is great– I put all my preprints on there. But it is simply not true to say that “almost all the recently written math papers” are on the on arXiv. I would think that not even half of all (reasonable quality) papers are on the arXiv. I personally wish this wasn’t so, but it is.

So there are significant barriers to just using the arXiv. It seems to me that if we cannot convince a great many of our colleagues to use the arXiv then it’s going to be even harder to get these same colleagues to abandon journals. After all, why have there been so few editorial boards willing to do as the board of Topology did, for example?

Where we publish also doesn’t just affect tenure and promotion. Here in the United Kingdom, we have the REF (http://www.hefce.ac.uk/research/ref/) which judges us all on “outputs”. While no doubt the official line is that any pre-reviewed work is fine, I can see that my head of department would be pretty unhappy if my 4 outputs were in strange new venues. Again, it’s a critical mass problem…

Peter KrautzbergerPost authorIt seems to me that the pricing issue cannot be solved as long hiring decisions are almost exclusively based on the publication track record in traditional journals (and, even worse, on the metrics of those journals).

In other words, when it comes down to it, we only value researchers by the metrics of the traditional journals they’ve published in. This gives the publishers incredible power over the development of our community and there’s no conceivable reason why for-profit publishers wouldn’t use this power to maximize their profits.

The pressure to publish is immense and getting published has turned into a game rather than an effort of communication — with many adverse effects to a functioning community.

None of the comments so far have addressed the issue that the monoculture of “publishing papers” is limiting the way a scientific community can develop. Journals used to be a necessary evil to enable a minimal degree of communication within a community that is spread around the globe. But now the community is fully connected, in real time, and there is no difficulty to stay in touch with any researcher as long as they use the internet to some degree. In turn, we can now communicate every detail of our academic work effortlessly; not just individual papers, but refereeing, student interaction, Q&A’s, video lectures, expository writing, research exchanges, live-broadcasting talks and seminars etc

I think the key problem is that we need to find ways reduce the pressure to publish the traditional way. The only way this is possible is if we find a way to publish less.

Therefore we need to spend time (and money) on finding ways to evaluate other activities of researchers.

For example, wikipedia has de facto a peer review system, especially the “Good Articles”. It’s not as easy to evaluate a wikipedia author, but an active author offers a lot of activity to evaluate them by and there is precedent for considering this a research activity.

Similarly for MathOverflow and other research-level Q&A sites: The so-called reputation points are a poor metric per se, but they nevertheless indicate an activity that is worth evaluating, i.e., a high “reputation” indicates at the very least a high activity, an amount of data that could be used to evaluate a researcher.

Fortunately, these new platforms of academic activity allow us to get started right away.

On the one hand, mathematicians can (and should) invest time in making their efforts outside of traditional journal publishing more visible, in particular on their professional homepage and their CVs (in the research section, that is!)

On the other hand, the professional societies should start taking the activity of mathematicians on the web seriously. They could support and invest in new ways of doing research online and, above all, investigate the development of standards for evaluating such activities.

===

I thank John Baez for encouraging me to make this comment.

André JoyalPost authorProfessorI would like to propose the creation a universal archive of *published* mathematical papers. With the goal of making the published papers of all times freely accessible to everyone. The papers could be transmitted to the archive by the mathematical journals. Potentially, the archive could contain and preserve the collected work of every mathematician, past or alive.

The archive could be created progressively, starting small. Of course, the for-profit journals may demand a high price for their copy rights. The true nature of their activities will be revealed. Peoples may eventually decide to stop publishing in this kind of journals.

André JoyalPost authorProfessorI would like to propose the creation a universal archive of *published* mathematical papers. With the goal of making the published papers of all times freely accessible to everyone.

The papers could be transmitted to the archive by the

mathematical journals. Potentially, the archive could contain and preserve the collected work of every mathematician, past or alive.

The archive could be created progressively, starting small.

Of course, the for-profit journals may demand a high price for their copy rights. The true nature of their activities will be revealed. Peoples may eventually decide to stop publishing in this kind of journals.

Peter McNamaraPost authorIf we can get adoption of the arxiv to 100%, then the question of where we publish is less relevant, and the only value from journal subscriptions becomes access to their archives. These are two independent problems.

For access to archives, Andre Joyal’s suggestion is worth pursuing. For submissions to arxiv, we should at least start refusing to referree (or, more importatntly, if an editor, refuse to consider submissions of) papers which are not on the arxiv.

Bruce BartlettPost authorLecturerI believe the best way to combat overpriced journals is the so-called

“Open Access First, Copyright Reform After”

movement that is championed by Steven Harnad. This is also known as the “Green Open Access” philosophy. The idea is that universities mandate staff to submit final drafts of papers on university repositories, which then show up in Google. In other words, you still submit your paper to a “normal” overpriced journal, but you deposit the final preprint version (after the referee process) onto the university repository. Harnad says that 95% of publishers already allow self-archiving. For the remaining 5%, you ensure that the university repository has a clickable “Email me an eprint copy” button. When someone clicks on this, an email automatically gets sent to the author, who then personally sends an eprint to the requester.

In this way, the dominoes of the journal publishing industry will organically begin to fall, and new systems (such as the arXiv overlay idea) will present themselves organically in time. This is precisely in line with most of the comments above, for instance that of Andre Joyal.

Steven Harnad has given many talks on this subject, and I encourage you watch his 10 minute talk on this topic at vimeo.com/32241747.

Andrew StaceyPost authorFørsteamenuensis

I must confess to getting slightly jaded by this debate. It would be nice to see some of the proposals actually tried for once, rather than just being debated. However, maybe the IMU has enough clout to actually do something so I’ll add my 2 øre.

I would say that the first thing that needs to be done is a proper evaluation of the role of journals in today’s mathematical landscape, recording both the original intention of having journals and their extended use now. Without this, it is hard to evaluate the various suggestions as to how realistic they are in solving the problem. If this has already been done, it should be publicised more!

Once that has been done, we can have a discussion as to how to meet those needs in a new system (and which of those needs are actually still needed!), or in several new systems.

However, I shall jump the gun (or shark) and presuppose that journals were originally set up for effective and timely communication between researchers, and that their use for external evaluation was tacked on afterwards.

As has been pointed out, the communicative role of journals has been technically solved by the arXiv. That it is not actually solved is no longer a technical problem, but a social one – but then any other new system would go through the same stages so this isn’t a reason to consider an alternative to the arXiv.

What isn’t addressed by the arXiv is the effective part of the communication. It takes a while to scan through all the “new” articles on the arXiv each day, and if one doesn’t happen to catch my eye then the chance of finding it again afterwards is quite slim – unless someone happens to draw it to my attention. So we need some bodies that say, “These are articles you should read.”. This is effectively John Baez’s suggestion above (I’ve made this suggestion elsewhere, as has Scott Morrison and many others). The key features of this idea that I would like to highlight are:

1. That the boards have some prestige. I would be far more interested in “The Journal of Things the AMS is Interested in” than “The Journal of Things that Andrew Stacey is interested in” or “The Journal of Things that 100 other Mathematicians are Interested in”. Not only does this give the whole thing a veneer of respectability, it avoids the “cult of the personality” and by having official boards with specific remits, it’s easy to see which boards to follow and which to ignore.

2. Selection by one board does not preclude selection by another. The remits of each board can overlap. The AMS might produce a list of the most significant works by American Mathematicians. Topology by topologists. An American Topologist might have their article in both.

The other thing that I want to say at this juncture is that I think that review sites are a non-starter. I suspect that for most papers the problem will be getting just one person to read it! Whilst one could say that maybe these shouldn’t have been published, I think that would be wrong. If it is correct mathematics, it is worth recording because we never know when it will be of use. In this age, though, not everything has to be recorded in a “traditional journal” (a certain mathematical wiki springs to mind …). These are still positive contributions to mathematics and worth taking into account when evaluating a mathematician’s total contribution.

Review sites could also drive people to publishing a particular type of paper: slightly controversial (or at least, on first sight) in a “hot topic” – just so that it gets discussed.

I think that’s all I have to say. The other suggestions here that I’ve read so far all seem good ones, but we need to actually try them out. And I agree that statements from reputable bodies such as the IMU or AMS would go a long way to helping, as would commitments to put all published works in the public domain, both past and present (I was a little put out recently at being asked to pay 30UKP for a 3-page article – and this by a publisher that I would have regarded as one of the Good Guys!).

As I say, it would be nice to see some of this in practice.

Tom LeinsterPost authorGreat as the arXiv undoubtedly is, we should require it to make some contractual commitments before we lean on it any more heavily.

When you submit a paper to the arXiv, you give it a perpetual, irrevocable licence to publish your work. But you get no commitment in return. The arXiv does not commit to keeping your work available. It does not commit to staying free. There is no guarantee that they won’t sell up to Elsevier. There is no guarantee that the next time you look at one of your abstracts on the arXiv, it won’t be surrounded by advertising.

I have no reason to think that the people who run the arXiv have anything but the highest motives. But if we’re going to make the arXiv any more central to mathematics publishing than it already is (and frankly, even if we’re not), we should ask them to match authors’ legal commitments with legal commitments of their own.

André JoyalPost authorProfThe journalists writing for the Huffington Post are working for free and they must give their author rights. Sounds familiar? The Post was sold to AOL for more than 300,000,000 dollars. The journalists working for the Post received nothing. See the blogpost of Lionel Maurel,

http://scinfolex.wordpress.com/2011/10/16/non-arianna-le-huffington-post-nest-pas-wikipedia/

Who owns our research papers?

Thierry BouchePost authorMathDocAndré Joyal’s first two comments are interesting because they call for two things that are (conceptually) available since more than a decade.

The first one is arXiv overlay journals, for which I am sceptical, the second one is the DML, of which I am an enthusiast. I will try to underline why.

ArXiv overlay journals were invented years ago based on the very obvious idea that what a journal is, from the scientist perspective, is a brand or quality stamp placed by a more or less prestigious committee on some paper. This is especially true in math where good journals are just repositories of good theorems in any area of math, a mathematician reads very seldom a journal issue cover to cover. The other aspects of journals, such as providing a smooth reader experience through common layout, copy-editing, checking that proofs are not only correct, but also readable and understandable, are vastly overlooked by the profession (and there are numerous examples where the copy-editor changed the author’s manuscript in such a way that the English is more familiar to a native reader, but says something wrong). However, this never took up, I suspect for sociological reasons, and even the Annals of math–everyone thought they could do whatever they wanted being The Annals–stepped down from arXiv overlay to a mere MSP charged access journal. I think one of the very important aspects of journals, often overlooked, is that they *regulate* the system’s entropy to a manageable one. When we are hunting for a reference for a certain result, we want that reference to be reasonably reliable, so that we can at first sight, relying on a journal reputation, assume that the found result is true enough to start elaborate over it–checking it thoroughly when we will start later on using it in a more organic way in our own research.The traditional publishing way, where a reference version of a paper is produced using many stages (refereeing, copy-editing, proof-reading) and involving a number of different people slows down publishing and insures a minimal quality check. It is not perfect and flawed papers appear in good journals, but this doesn’t seem a sufficient reason to drop the whole system, we have corrigenda for this, which is again a heavy enough process to stabilize things. And future generations will rely on this stability to build their math results. Math being math, a dynamical system like arXiv seems very appropriate as a helper for *current* research, but which version of a paper should we read, trust when journal X stamped version y and author’s version y+1 claims a stupid error was corrected while other papers whose results we use stepped from version y-1, z+12, etc.?

On the other hand, the DML (Digital Mathematics Library) is the grand idea that the mathematical heritage is a public common and should as such be faithfully archived, curated, and made accessible as widely as possible. This seems so obvious as math pervades everywhere in science and technology that it seems it should not need any selling. Making all not-so-recent (let’s say a conservative 5-10 years after publication) math papers archived and open access in an independent digital library would be almost cost neutral to publishers and provide a clear benefit to all potential users of this (carefully validated and published) knowledge. This would yield maximum visibility to all publishers achievements (and failures) and enable further research as we all know that no one can predict which math result will be useful to which scientific or technological area few decades after.

This is not so much related to math journals pricing, I confess, except that these prices raised so high that it has become profitable to close access to old articles and sell them separately, which seems quite successful so far to profit-driven publishers (some of them run by not-for-profit organizations).

In conclusion (I’m afraid, a cross-section of the three topics of this blog), there is not so much possibilities to go beyond journals while keeping the system manageable for future generations, but rankings should not prevent experimentations in this area–maybe a variation on overlay journals will finally take up? We should keep journals prices low so that mathematical knowledge be affordable to all and stay a valuable asset to mankind; and we should store this knowledge over the long term so that mankind can exploit it whenever relevant.

Besides, best wishes for a creative and affordable 2012!

Th. Bouche

André JoyalPost authorProfessorI thank Thierry Bouche for expressing his views. Let me point out the article “Cracking Open the Scientific Process” which just appeared (january 16, 2012) in the Science section of the NY Times. The author is Thomas Lin.

Moderator’s note:

http://www.nytimes.com/2012/01/17/science/open-science-challenges-journal-tradition-with-web-collaboration.html

We apologize that the comment level does not support active links at this moment.

Benoît KloecknerPost authorA French initiativeIn my lab, Institut Fourier in Grenoble, France, considering that the big deal being negociated with Springer by national partners (CNRS, INRIA, Couperin wich is a consortium representing the universities) could threaten the subscriptions of not-for-profit journals, we launched a petition to claim that we prefer not to get the access to most Springer journals rather than being forced into a big deal. You can read the precise statement here:

http://www-fourier.ujf-grenoble.fr/petitions/index.php?petition=3

(in french).

Note that the wording makes it suitable only for mathematician working in France, a choice we made because we wanted to stay focused on a concrete point (the deal under negociation).

For now, it is difficult to say what impact this initiative will have, but we can say the petition has strong support: in a little more thant two weeks, it has been signed by more than 1500 individuals (over about 3000 mathematicians in France, if I am not mistaken) and 21 laboratories (over 57 that have contract with the CNRS institute for mathematics).

It seems to me that breaking the big deals is a good way toward decreasing prices of journals; these prices are actually difficult even to determine right now, since in some bundles (e.g. Elsevier bundle) the top journals seem reasonably cheap, but are used to sell many other ones that we would never buy if not forced to. Moreover, bundling soften the threat of topology-like movements: if a journal quits a bundle, the overall price of the bundle need not decrease accordingly.

André JoyalPost authorThe time for action has arrived:

thecostofknowledge.com

André JoyalPost authorA possible initiativeA possible action of the IMU might be to invite every mathematical institute of the world to create a journal that would reflect its history and personality. For example, the Fields Institute could create a journal called *The Fields Institute Mathematical Journal* or one called *The HSM Coxeter Mathematical Journal”, since Coxeter spended most of his career in Toronto.

The editor-in-chief of the new journals may prey on the editorial board of for-profit journals to create their own.

Pierre ColmezPost authorReal prices of journalsIt seems, according to some of the posts on Gowers’s blog, that it is not so easy to determine the actual price of a mathematical journal due to the bundle policy of publishers and the secret surrounding the deals made with libraries. Maybe the IMU could try to obtain more transparency in this matter and gather all data so that we have a better view of what is truly inacceptable and decide what to do about it.