A peer-review system for the ArXiv?

[jrlaguna]

Zz, do you even read my comments? I NEVER say that authors pay for publication. It is the scientific community, via the library subscriptions, who pay that money, according to Paul Ginsparg (I hope you know who he is). I repeat for the last time, I feel like I'm talking to a wall. The authoring cost of a paper in theoretical physics, in salaries, is around $30.000 per paper. And do you really think that editing it can add $10.000? So, if an editor makes $50.000 per year, he or she will edit 5 papers IN A YEAR? OMG!

In summary, and in very simple terms: publication COST need not be charged on authors. OK?

And about typesetting, I'm sorry to contradict you again, but the final printed form of my papers differs in less than epsilon from my LaTeX. They just changed a bit the format, and most of the work is done automatically. Have you talked to editors? I have.

About the cost of ArXiv, it was given by Paul Ginsparg himself. I guess he summed up the mantainance cost of the ArXiv, paid by many institutions (US and worldwide) and divided by the number of papers. He got $10 per paper. How is that confusing you?

 

[dcpo]

I don't understand why the idea that it is being claimed that journals charge huge fees to the authors has taken root.

Regarding a peer review process for arXiv I see a few problems. For a start arXiv arguably serves a distinct purpose from the peer-reviewed literature. This point has been made by others here and I won't dwell on it. Another potential problem is that peer-review in itself is not necessarily a strong stamp of approval. I can point to at least one peer-reviewed mathematics journal (with high impact factor) that publishes some obviously garbage papers at the whim of its editor-in-chief, who, legitimate publication record notwithstanding, appears to have some powerful crankish tendencies. In order for peer-review to be meaningful there has to be an understanding of the standards of the publication, something guided by the editors. If arXiv were to be peer-reviewed who would set the standard for review, and to whom would the reviewers be accountable? In addition to this, while I don't regard the production of printed journals, or even the collection of articles into issues (an artifact of the print system), as necessary, I would not wish to discount the role the editorial boards of journals play in the development of their fields. Rather than try remake arXiv I see more of a future in the establishment of new journal type systems, under the guidance of editorial boards and with a review process, but offering a subscription service to an archive of accepted results. Here the typesetting could be done wholly by the authors (at least in mathematics), and post acceptance production delays would be eliminated. It would probably still be necessary to charge a subscription fee, but I'd be very surprised if the operational costs for such a system were not significantly lower than those of most in the current model.

 

[jrlaguna]

@dcpo, thanks, seriously! Although I have discussed these ideas in many places, it's really getting on my nerves how these people are missing the basic point... At least you point at difficulties which prove that you understood the proposal! :)

I really regret having posted it here. I will stick to forums which are more friendly to new ideas.

 

[micromass]

I will stick to forums which are more friendly to new ideas.

You mean forums that just accept your ideas without pointing out where you went wrong?? Don't let us stop you from going there.

 

[jrlaguna]

Such ill-intentioned misunderstanding and patronizing makes me have the strong impression that this site and their "PF mentors" get funding from the publishing industry, either open or covert. Can you prove it wrong?

 

[dcpo]

You mean forums that just accept your ideas without pointing out where you went wrong?? Don't let us stop you from going there.

I don't think that's a fair assessment at all. As an impartial observer I found the criticism directed towards jrlaguna's proposal to be rather hostile and mean-spirited in its desire to interpret meaning in such a way as to maximize disagreement.

 

[micromass]

I don't think that's a fair assessment at all. As an impartial observer I found the criticism directed towards jrlaguna's proposal to be rather hostile and mean-spirited in its desire to interpret meaning in such a way as to maximize disagreement.

Well, I'm an impartial observer as well here. I did not participate in the discussion. I saw jrlaguna basically ignore the good posts of Zz, Vanadium and twofish.
He might have a point somewhere, but he should explain it better.

 

[D H]

Here the typesetting could be done wholly by the authors (at least in mathematics), and post acceptance production delays would be eliminated. It would probably still be necessary to charge a subscription fee, but I'd be very surprised if the operational costs for such a system were not significantly lower than those of most in the current model.

There are plenty of peer reviewed journal articles on the internet for which you can find both the as-published version and the pre-review arxiv version of the articles. Find some, look at both versions with an open eye. What you will see is that scientists, engineers, and mathematicians generally have rather lousy writing skills, document layout skills, and graphics skills. Fixing those problems costs money. You can also see this in a peer-reviewed conference proceedings. Typically conference proceedings bypass the technical editing process -- and you can see it. That technical editing is expensive.

 

[dcpo]

There are plenty of peer reviewed journal articles on the internet for which you can find both the as-published version and the pre-review arxiv version of the articles. Find some, look at both versions with an open eye. What you will see is that scientists, engineers, and mathematicians generally have rather lousy writing skills, document layout skills, and graphics skills. Fixing those problems costs money. You can also see this in a peer-reviewed conference proceedings. Typically conference proceedings bypass the technical editing process -- and you can see it. That technical editing is expensive.

I have no trouble reading and understanding preprint versions of papers in my field. I do not dispute that professional typesetters can do a better job, my question is whether the improvement is worth the money that must be spent on it.

 

[ZapperZ]

I have no trouble reading and understanding preprint versions of papers in my field. I do not dispute that professional typesetters can do a better job, my question is whether the improvement is worth the money that must be spent on it.

OK, before I jump all over this one, will you please clarify one thing?

Do you think that all a journal does is improve the look of a submitted manuscript? From the way you have described it here, you appear to not realize that producing a peer-reviewed journal involves a whole bunch of other things other than just some pretty typesetting!

And please note, in at least one of my earlier posts, I've already described several of the main functions of a journal editor! And someone here thinks that *I* don't read or understand his/her post? I could easily show this as one example that someone didn't read MY post!

Zz.

 

[dcpo]

OK, before I jump all over this one, will you please clarify one thing?

Do you think that all a journal does is improve the look of a submitted manuscript? From the way you have described it here, you appear to not realize that producing a peer-reviewed journal involves a whole bunch of other things other than just some pretty typesetting!

And please note, in at least one of my earlier posts, I've already described several of the main functions of a journal editor! And someone here thinks that *I* don't read or understand his/her post? I could easily show this as one example that someone didn't read MY post!

Zz.

I don't understand why you would think that I think that all journals do is typesetting. I've explicitly made a contrary statement in fact. For the sake of clarification though, I do not believe that the only role of a journal is typesetting, but I do believe that professional typesetting and the production of print editions adds significant expense to journal production, which is then passed on in the form of subscription fees. My idea is that by abandoning print editions and professional typesetting academic material could be processed, reviewed and archived in a system retaining the academic virtues of the traditional publishing model but at significantly lower cost, a saving which could be passed on to subscribing institutions (though there would likely remain a subscription fee). Obviously there are many problems inherent in changing the status quo in such a reputation based world, but that is a topic for further discussion.

 

[ZapperZ]

I don't understand why you would think that I think that all journals do is typesetting. I've explicitly made a contrary statement in fact. For the sake of clarification though, I do not believe that the only role of a journal is typesetting, but I do believe that professional typesetting and the production of print editions adds significant expense to journal production, which is then passed on in the form of subscription fees. My idea is that by abandoning print editions and professional typesetting academic material could be processed, reviewed and archived in a system retaining the academic virtues of the traditional publishing model but at significantly lower cost, a saving which could be passed on to subscribing institutions (though there would likely remain a subscription fee). Obviously there are many problems inherent in changing the status quo in such a reputation based world, but that is a topic for further discussion.

I don't think the cost of type setting is a substantial portion of a journal's expense. I also do not think that print editions is a major issue anymore, considering that many institutions have abandoned receiving print editions, with substantial savings.

In the US, a lot of expense comes in the form of human resources. The editors of these journals are themselves professionals in the these various fields. Journals such as Nature, Science, and PRL depends a lot of these editors to weed out all of the submissions BEFORE they get to the referees. In fact, I would say that less than half of the manuscript submitted to Nature and Science even make it to the referees. So the editors do not play an insignificant role, and such time and effort costs a lot of money. Add to that the rebuttal processes (which also incur expenses), etc. and one can already see why such costs escalates if you consider just the peer-reviewing process itself! I see this as THE major cost in the whole process. If you want to do it right and proper, you cannot do this in a cavalier fashion without jeopardizing quality.

And this is why I brought up the example of PRB. One needs to really look at the number of papers being published in just one month, and think about those that (i) didn't make it through the editors (ii) made it through to the referees but didn't get published in the end (but still when through the peer-reviewing process) (iii) made it through in the end. It requires a HUGE amount of effort, a lot of time, and a lot of people to handle! And this is for a journal that, to me, requires minimal amount of typesetting, as opposed to Nature and Science.

And this is before we consider that each of these journals have an editorial office, and chief editor that has a non-trivial function, an editorial board, and the fact that these editors and associate editors, if they are any good, have to attend conferences and be up-to-date on development in these fields (again, require money and time!).

Again, my point in all of this is that typesetting is such a small part of the whole cost. The refereeing process is the major cost in such publication.

Zz.

 

[dcpo]

Whether the production of a print edition is an issue is certainly not determined entirely by whether it is subscribed to, as the cost of production of such an item drives the cost of publication up even if it is not bought.

The situation you seem to be describing is alien to me. In mathematics the editors and editorial board are tenured academics and are very rarely paid for their editorial work, so whatever the cost of production it is not related to the review process in an obvious way. See here for slightly dated but relevant information.

 

[Vanadium 50]

The issue of typesetting came up because jrlaguna made the claim that because the documents are sent in RevTex, that the typesetting was already done. Anyone who has compared the redlines of their journal articles with what they sent knows that this is not true and cannot be true.

The APS makes public how much it costs them to produce journals. Last year it was $28,098,375. We can make the argument that it should cost less, but this is what it actually costs. I should point out that Ginsparg himself, in the very article that's being bandied about, says that its the cost of peer review that drives this. Yes, the reviewer doesn't get paid, but the guy who is badgering the reviewer to get the review in on time often does.

If we want the arXiv to take on this work, it will cost about this much money. That's 41x the arXiv's budget.

Let me ask my questions again:

1) Why? There are already Open Access journals.
2) The arXiv presently serves a different purpose. Why must its function change?

 

[dcpo]

"the reviewer doesn't get paid, but the guy who is badgering the reviewer to get the review in on time often does." this is not the case in maths, I can't speak for physics.

 

[Vanadium 50]

Or the person badgering the person to do the badgering.

 

[twofish-quant]

I have even discussed with editors in different journals. Somebody (I can not cite) told me there was an attempt, many years back, to start a peer-review process in the ArXiv, but it was dismissed because they thought the work was not worth, because scientists are very conservative in their professional practice, they might be scared that ArXiv peer-review might be considered second-class. I can agree with that.

I don't think that's the big reason. The big reason is that people publish to Arxiv (and SSRN) precisely because it is *NOT* peer-reviewed and you don't have to go through the expense and trouble of publishing results quickly. I'm willing to ignore a dozen papers by crackpots if it means that I get research data and ideas now instead of in six months. I know enough about the field so that I can do my own peer review. I read the paper, and I figure out what I think about it. I don't see what peer review adds to the process.

Also the communities are pretty small. There are maybe 30 people involved worldwide in my area of research, and I know all of them, so there really is no point in starting any sort of general forum for comments.

The reason that we scientists can't work ONLY with the ArXiv is that we need the "peer-review" stamp for promotion and funding. Seriously, if this reason was not there, I would never submit a paper to any journal whatsoever.

And that's the real problem. One thing that ArXiv shows is that sometimes peer-review is counterproductive. The point that people are making is that there are large costs that are *inherently* associated with the peer-review and journal publication process which means that we really need to ask if scientific journals are really the best mechanism for scientific discourse.

In astrophysics, publication is basically a rubber-stamp. The real gatekeeping comes with writing grant proposals.

The other problem is that funds are limited, so if you make it *easier* for people to do science, then you just end up with much more competition for limited funds and people will end up doing something silly to limit funding. One way of thinking forcing people to publish to dead trees is that it's a somewhat silly way of making it more difficult to do science, because society can't handle a situation in which it is easy to do science.

If the peer review system is broken, then it doesn't make sense to force something that works into that broken system.

And peer-review is done by ourselves... it's in the very name "peer", isn't it? I've peer-reviewed as many papers as I have authored. And we do it FOR FREE.

But science is fun and it's a reputation economy, so people get "stuff" for doing peer review. The type of thing that you need to pay people for is the "janitorial" stuff. People will do peer review for free, but I don't know anyone that will measure margins and do copy editing for free.

That also works with online stuff. Putting together a website is non-trivial. Under some situations I'm willing to do it for free, but past a certain point you are going to have to pay me (and pay me a lot) to write computer programs.

 

[twofish-quant]

I can point to at least one peer-reviewed mathematics journal (with high impact factor) that publishes some obviously garbage papers at the whim of its editor-in-chief, who, legitimate publication record notwithstanding, appears to have some powerful crankish tendencies.

It's even worse in economics and management. There the academics are in their own world that very few people outside take seriously any more. I don't know of any non-academic in economics and management that actually reads or cares about what's in the peer reviewed literature.

One difference is that in physics, publishing in peer-reviewed papers is the only way of building reputation, whereas in economics and management, you can get reputation in other ways (i.e. starting a business and making a ton of money), and being an academic is sometimes an anti-signal.

If arXiv were to be peer-reviewed who would set the standard for review, and to whom would the reviewers be accountable?

And what's the point in the end? I do a keyword search. If it's an interesting paper, I read it. If not, I don't. What's the point of having someone else tell me if it's a good paper or not?

If you want "prestige" then trying to graft peer review is the wrong way around. IMHO, you could do better if you have some moderately wealthy benefactor give high profile awards for the best papers published in Arxiv.

Rather than try remake arXiv I see more of a future in the establishment of new journal type systems, under the guidance of editorial boards and with a review process, but offering a subscription service to an archive of accepted results.

Or you can have public funding on the condition that the results are openly available. Someone has to pay, and ultimately the money for most of research comes from tax revenue. I think we may reach the point in which we end up with public funding for public results, which is largely the situation in astrophysics.

It would probably still be necessary to charge a subscription fee, but I'd be very surprised if the operational costs for such a system were not significantly lower than those of most in the current model.

I would be. One thing that tends to be the situation is that when you use new technology, the costs increase, and people that attempt to "save money" with new technology are usually quite disappointed. Putting together a well-run website is highly expensive. It's also important to figure out the difference between fixed costs and marginal costs. Software has very low marginal costs, but the fixed capital costs can be huge.

I don't see that the costs themselves are the big problem. The big problems are "who pays" and "who benefits"?

 

[dcpo]

@twofish-quant: I think I agree with you about most of this, though I do believe that moving away from the paper based model would result in savings. Remember that journals already maintain an e-subscription service, so new technology wouldn't need to be brought in. You're right though that costs are not the real problem. In maths at least almost everyone works for free. If you remove the need for professional typesetting then it would save a bit by removing some of the people who have to be paid, but the real problem comes with profit driven academic publishing companies. The data shows that journals put out by universities and societies tend to have much lower subscription fees than those put out by Springer, Elsevier etc. The profit turned by universities and societies is also usually reinvested in the community in some way, which is much less the case with commercial publishers.

 

[Vanadium 50]

The profit turned by universities and societies is also usually reinvested in the community in some way, which is much less the case with commercial publishers.

But that's an argument that for-profit journals shouldn't exist. That's not an argument that the arXiv needs to radically change its purpose.

Elsevier makes a 30% profit. That's certainly large, and it certainly makes me unhappy. But it also means that if we were to ban profit completely, costs would not fall by an order of magnitude. They'd fall by around 30%.

The argument that journals should not be allowed to make a profit on taxpayer-funded research is a slippery slope. Dell makes a profit by selling me a computer to write the paper. LeCroy made a profit selling the electronics to make the experiment possible, and so on.

 

[dcpo]

I'm not arguing for a change in arXiv, or that journals shouldn't be allowed to profit, only that by abandoning print journals and moving away from for-profit publishers subscription costs could be significantly reduced (at least in maths). Sometimes commercial publisher math journals do charge an order of magnitude more (per page) than journals run by societies (that still turn a profit on subscriptions). Anyway, I'm in a rush and probably won't be able to reply for a few days, but I don't think we're necessarily disagreeing about very much.

 

[mheslep]

It's even worse in economics and management. There the academics are in their own world that very few people outside take seriously any more. I don't know of any non-academic in economics and management that actually reads or cares about what's in the peer reviewed literature...

Apparently the NBER working papers are almost the exact economics equivalent to ArXiv. Apropos:

...So, the starting point for me, when thinking about how economics works as a discipline, is to realize that the traditional model of submit, get refereed, publish, and then people will read your work broke down a long time ago. In fact, it had more or less fallen apart by the early 80s.Even then, nobody at a top school learned stuff by reading the journals; it was all working papers, with the journals serving as tombstones.

 

[dcpo]

There has recently been a surge of interest in the problem of journal subscription fees in the mathematical community, with notable figures such as Terry Tao and Timothy Gowers adding their considerable weights to the discussions. Here is a recent blog post from Gowers, dealing with the subject of online review systems.

http://gowers.wordpress.com/2012/02/02/abstract-thoughts-about-online-review-systems/#more-3973

 

[twofish-quant]

I'm not arguing for a change in arXiv, or that journals shouldn't be allowed to profit, only that by abandoning print journals and moving away from for-profit publishers subscription costs could be significantly reduced (at least in maths).

I don't think you are going to vastly reduce costs. People that assume going online will reduce costs have usually been extremely disappointed. Going online usually increases costs, but you get other things.

Now you could radically change who pays the cost. For example, one thing that you could do is the "defense contractor" model in which the government pays the big publishers a big flat fee in exchange for releasing all of the copyrights. This is more or less the way that it works in astrophysics.

 

[twofish-quant]

For example, we might be on a hiring committee for a job with hundreds of applicants. If so, we need quick, and therefore crude, methods of evaluation. Journals give us one such method: we just skim through the publication list and get a sense of the quality of the journals that a candidate has managed to publish in.

If you are going to "outsource" this sort of work, one shouldn't be surprised if the people that you outsource this work to, end up demanding money to do your homework. I don't think it's possible to talk about problems in the journal system without also talking about issues of academic hiring and "publish or perish."

If you end up with a situation in which notable mathematicians say that they no longer care about what journals people publish in, and they are willing to spend the money to do independent evaluations of papers published in arxiv, then you fix the problem. If you argue that doing independent evaluations is "too expensive" then I don't see why one complains about journal fees. Right now, you are essentially paying a company to make hiring decisions for you, and it's not surprising that they will charge you through the nose to do that.

Also the way that astrophysics works is that if you want to know whether an article in a field that you don't know about is good or not, you e-mail someone that you know and they'll tell you. One problem in making the system public is that you often get much better evaluations over private e-mail or face to face.

Something that I've found interesting is that you often get much better critiques if you keep people's identities quiet since that means that you aren't defending your reputation every time you make a forum post or post something online. Something that people will tell you over private e-mail or face to face is "this paper stinks" or "this author is a crank" and that's something I don't think you can get people to say if they are being tracked.

 

[dcpo]

I don't think you are going to vastly reduce costs. People that assume going online will reduce costs have usually been extremely disappointed. Going online usually increases costs, but you get other things.

I don't doubt that this is true in general, but every serious mathematics journal that I know of already maintains an online presence, with full electronic access to their papers etc. so I'm curious what extra costs would arise from eliminating the apparatus associated with the production of the paper journals.

Now you could radically change who pays the cost. For example, one thing that you could do is the "defense contractor" model in which the government pays the big publishers a big flat fee in exchange for releasing all of the copyrights. This is more or less the way that it works in astrophysics.

It is striking how different the journal system appears to be between fields. Who pays for publication is a related but slightly tangential question. In the mathematical community the fees charged by the likes of Elsevier are being viewed by an increasing number of people as extortionate. For example see the current boycott of Elsvier by a number of mathematicians, and, as another example, the resignation of the entire editorial board of a prestigious mathematics journal in protest at Elsevier policy a few years ago. How related is the 'fairness' of the pricing structure to who pays the bill?

If you are going to "outsource" this sort of work, one shouldn't be surprised if the people that you outsource this work to, end up demanding money to do your homework. I don't think it's possible to talk about problems in the journal system without also talking about issues of academic hiring and "publish or perish."

If you end up with a situation in which notable mathematicians say that they no longer care about what journals people publish in, and they are willing to spend the money to do independent evaluations of papers published in arxiv, then you fix the problem. If you argue that doing independent evaluations is "too expensive" then I don't see why one complains about journal fees. Right now, you are essentially paying a company to make hiring decisions for you, and it's not surprising that they will charge you through the nose to do that.

Also the way that astrophysics works is that if you want to know whether an article in a field that you don't know about is good or not, you e-mail someone that you know and they'll tell you. One problem in making the system public is that you often get much better evaluations over private e-mail or face to face.

Something that I've found interesting is that you often get much better critiques if you keep people's identities quiet since that means that you aren't defending your reputation every time you make a forum post or post something online. Something that people will tell you over private e-mail or face to face is "this paper stinks" or "this author is a crank" and that's something I don't think you can get people to say if they are being tracked.

I'm not aware of any serious argument that journals do not currently provide useful services, (theoretically) anonymous review being one of them. Among the questions that are being discussed at this time on various blogs (for example Timothy Gowers' blog which I linked to earlier) are, what exactly these services are, to what extent they justify the fees that some publishers are charging, and how can alternative publication models replace or improve on them? That there is something very wrong with the current system has effectively been decided.

Even within the paper journal system, the AMS, for example, produces reputable mathematics journals, several of which are of the very highest quality, and they do so without charging anything like the same fees that Elsevier does.

 

[twofish-quant]

I don't doubt that this is true in general, but every serious mathematics journal that I know of already maintains an online presence, with full electronic access to their papers etc. so I'm curious what extra costs would arise from eliminating the apparatus associated with the production of the paper journals.

If you move from paper to 100% online then there are substantial costs. For example, if you go with a subscription pay wall, then you have to pay someone to deal with lost passwords, set up databases, monitor cheating etc. etc. Also, the costs of maintain data in "dead tree" format is low. You put dead trees into a room, keep it air conditioned with good humidity, and that's it. For electronic data, you have to pay for electricity, server upgrades, and format transfers. If 100x people want to read your dead tree book, the extra costs are low, but if 100x people start hitting your servers, they will crash.

It gets worse because the capital costs associated with dead trees are sunk costs. People have already paid for the printing infrastructure, and as long as the people are printing something, the extra cost of printing journals is nil.

Also paywalls are horrible business models electronically because of the cost of maintain the paywall, which is why in any situation where you have a alternative business model, it's used. In astrophysics, the cost of publishing is ultimately paid for by tax dollars, but it's set up so that everyone has access to the journal article at http://adswww.harvard.edu/

It is striking how different the journal system appears to be between fields.

Yes. In astrophysics the journals are not gatekeepers. The gatekeepers are the funding and grant committees. Once you have funding and have done your experiment, then people are going to be very annoyed if there is another layer of bureaucracy to get your results out. Journal publication in astrophysics tends to be "rubber stamp" in that you get results, you send them to Ap.J., and unless you are totally incompetent, they'll get published eventually. In the meantime, you've already sent out a preprint so that everyone knows about the result months before it appears in Ap.J.

It's also striking that the journals are run by the professional societies, who are non-profit. My guess is that for-profit publishers realize that they can't make money from astrophysics journals, because if journals started monopolizing publication, people would just start sending preprints to other. One corollary of this is that just because you are published doesn't mean much. Ap.J. has been known to publish nutty stuff.

One thing about arxiv.org is that you are sharing the system with people from other fields who have other workflows, and one thing that you need to be careful is not to change the system so that it messes up how things work in other fields.

How related is the 'fairness' of the pricing structure to who pays the bill?

If someone else pays, I don't care about the details. Someone's budget might be getting totally screwed over in paying for ADS or the Los Alamos preprint server, but since it's not coming from me and since I benefit from access, I don't care as long as they keep doing it. One other characteristic of astrophysics is that there are some large multi-billion projects. Since NASA spends several billions on space probes, then the cost of keeping ADS or the Los Alamos preprint servers up and running is a rounding error.

But money is still power. I suspect that most of the money that comes to fund Los Alamos comes from particle physics and astrophysics, so if you want to change things in a way that benefits mathematicians but annoys particle and astrophysicists, it's not going to happen.

I'm not aware of any serious argument that journals do not currently provide useful services, (theoretically) anonymous review being one of them.

Curiously I don't think that peer review is important in astrophysics journals. The other thing is that in practice, the reviews aren't that anonymous. If you can't figure out who is likely to review your paper, then you really shouldn't be publishing since you haven't done background research. The fact that I think that peer review at the journal level is largely irrelevant in astrophysics (and I think this isn't a wild opinion) is why I'd push back on introducing anything like it in Los Alamos.

The major benefit that I see astrophysics journals providing is long term archiving, and maybe some very light quantity control. Also, the big benefit of getting published is that your paper will forever be in ADS.

How can alternative publication models replace or improve on them? That there is something very wrong with the current system has effectively been decided.

But you have to realize that different fields have very different systems. I don't know of any astrophysicists who I respect that have serious issues with the way that things work in that field.

By contrast, the publication system in economics is widely regarded as seriously, seriously, seriously broken. It's not that the journals are expensive, but that the culture creates this sort of "tunnel vision" that has led to some really bad consequences (like the entire world financial system collapsing).

What doesn't make any sense is to me is if people in mathematics hate the for-profit journals, why not just cancel the subscriptions, and have everyone put their stuff on their websites, which is the way that it works in astronomy. Before the web started, there was a vibrant system of "paper preprints" and all the web did was to move those into the electronic world.

Even within the paper journal system, the AMS, for example, produces reputable mathematics journals, several of which are of the very highest quality, and they do so without charging anything like the same fees that Elsevier does.

So why not publish everything through AMS and walk away from Elsevier? This is something of an intentionally naive question, but it's something that just doesn't make any sense to me.

Obviously, Elsevier has found a "choke point" and like any for-profit company it's going to exploit that choke point to make a ton of money. That "choke point" doesn't seem to exist in astrophysics, and I'm trying to understand what exactly it is.

 

[dcpo]

If you move from paper to 100% online then there are substantial costs. For example, if you go with a subscription pay wall, then you have to pay someone to deal with lost passwords, set up databases, monitor cheating etc. etc. Also, the costs of maintain data in "dead tree" format is low. You put dead trees into a room, keep it air conditioned with good humidity, and that's it. For electronic data, you have to pay for electricity, server upgrades, and format transfers. If 100x people want to read your dead tree book, the extra costs are low, but if 100x people start hitting your servers, they will crash.

It gets worse because the capital costs associated with dead trees are sunk costs. People have already paid for the printing infrastructure, and as long as the people are printing something, the extra cost of printing journals is nil.

Also paywalls are horrible business models electronically because of the cost of maintain the paywall, which is why in any situation where you have a alternative business model, it's used. In astrophysics, the cost of publishing is ultimately paid for by tax dollars, but it's set up so that everyone has access to the journal article at http://adswww.harvard.edu/

The thing is that all the maths journals I'm aware of already archive their material electronically, and have their own pay wall systems for electronic access. I personally don't know anyone who does not obtain the vast majority of the articles they read electronically, though I'm sure a few exist somewhere, but even allowing for this I don't see where a huge increase in server traffic would come from should these journals cease print production. I still see print production as being a relevant source of expense, because it creates the need for professional typesetters, which I'm far from convinced do not add a not insignificant amount to the production cost of a journal.

Yes. In astrophysics the journals are not gatekeepers. The gatekeepers are the funding and grant committees. Once you have funding and have done your experiment, then people are going to be very annoyed if there is another layer of bureaucracy to get your results out. Journal publication in astrophysics tends to be "rubber stamp" in that you get results, you send them to Ap.J., and unless you are totally incompetent, they'll get published eventually. In the meantime, you've already sent out a preprint so that everyone knows about the result months before it appears in Ap.J.

It's also striking that the journals are run by the professional societies, who are non-profit. My guess is that for-profit publishers realize that they can't make money from astrophysics journals, because if journals started monopolizing publication, people would just start sending preprints to other. One corollary of this is that just because you are published doesn't mean much. Ap.J. has been known to publish nutty stuff.

Unfortunately in maths there's no real effective gatekeeper, and there seem to be a large number of people who don't let a lack of formal education or understanding of maths get in the way of writing maths papers. Also, I know of proper academic mathematicians who are rather notorious for not checking their own results. An issue in maths is that that for most people it's usually a non-trivial undertaking to read and understand a new paper, so the time lost in reading a bad or wrong paper is potentially greater. There is a separate argument to be had about the extent to which peer-review actually addresses this problem.

One thing about arxiv.org is that you are sharing the system with people from other fields who have other workflows, and one thing that you need to be careful is not to change the system so that it messes up how things work in other fields.

Of course.

Curiously I don't think that peer review is important in astrophysics journals. The other thing is that in practice, the reviews aren't that anonymous. If you can't figure out who is likely to review your paper, then you really shouldn't be publishing since you haven't done background research. The fact that I think that peer review at the journal level is largely irrelevant in astrophysics (and I think this isn't a wild opinion) is why I'd push back on introducing anything like it in Los Alamos.

The major benefit that I see astrophysics journals providing is long term archiving, and maybe some very light quantity control. Also, the big benefit of getting published is that your paper will forever be in ADS.

Again the difference between fields is striking, though the largely theoretical nature of the anonymity of the review process is similar, especially if you post a preprint on arXiv.

But you have to realize that different fields have very different systems. I don't know of any astrophysicists who I respect that have serious issues with the way that things work in that field.

By contrast, the publication system in economics is widely regarded as seriously, seriously, seriously broken. It's not that the journals are expensive, but that the culture creates this sort of "tunnel vision" that has led to some really bad consequences (like the entire world financial system collapsing).

It sounds like the system in astrophysics works well, I can't comment on the situation in economics as it's not something I'm familiar with.

What doesn't make any sense is to me is if people in mathematics hate the for-profit journals, why not just cancel the subscriptions, and have everyone put their stuff on their websites, which is the way that it works in astronomy. Before the web started, there was a vibrant system of "paper preprints" and all the web did was to move those into the electronic world.

I suppose one issue is access to archives. Maths papers remain relevant for longer than papers in most other fields, and the whole timescale for review, publication and citation is lengthened. It's entirely plausible that a working mathematician would require access to an article written by someone no longer in a position to put a copy on arXiv.

So why not publish everything through AMS and walk away from Elsevier? This is something of an intentionally naive question, but it's something that just doesn't make any sense to me.

Obviously, Elsevier has found a "choke point" and like any for-profit company it's going to exploit that choke point to make a ton of money. That "choke point" doesn't seem to exist in astrophysics, and I'm trying to understand what exactly it is.

The problem for individual mathematicians is that the current AMS journals are of a higher standard than even quite good maths papers, so if the average mathematician were to only publish when he or she had a paper of suitable standard they wouldn't publish very much. Of course there are professional society and university run journals with lower standards, but often a subfield is dominated by journals from professional publishers. I can't say anything about how this situation may have developed, but it does appear to be the case.

Here is a statement from some of the mathematicians behind the current boycotting of Elsevier. It is signed by at least two fields medalists, and the president of the IMU.

 

[Vanadium 50]

First, if there is a problem that the society journals are too "high standard", the societies can always start more journals. The APS has multiple journals that each over different needs. Nature Publishing Group has a number of journals with lower impact factors than Nature - because there is a need.

Second, I've said this before, but it doesn't seem to be sticking. If Elsevier is making a 30% profit on their journals, banning profit (perhaps by starting a new non-profit journal) won't cause prices to fall by a factor of 10, or even 2. They will fall, on average, around 30%.

 

[dcpo]

First, if there is a problem that the society journals are too "high standard", the societies can always start more journals. The APS has multiple journals that each over different needs. Nature Publishing Group has a number of journals with lower impact factors than Nature - because there is a need.

Second, I've said this before, but it doesn't seem to be sticking. If Elsevier is making a 30% profit on their journals, banning profit (perhaps by starting a new non-profit journal) won't cause prices to fall by a factor of 10, or even 2. They will fall, on average, around 30%.

The establishment of new mathematics journals is one of the many things being discussed, but is not itself a particularly simple thing to do.

I don't think your reasoning about Elsevier profits is necessarily correct, as Elsevier presumably structures its business to maximize its net profit rather than its profit margins. In any case the available data appears to show that several journals operate at some profit with prices significantly less than the 77% of Elsevier's lower bound your theoretical 'elimination of profit' calculation implies.

 

[twofish-quant]

I still see print production as being a relevant source of expense, because it creates the need for professional typesetters, which I'm far from convinced do not add a not insignificant amount to the production cost of a journal.

High end computer programmers get paid more than professional typesetters. Also for any sort of real world computer systems you'll have to hire usability experts and graphic artists, so I don't see the cost savings. The issue here is that once you've printed the journal, then you no longer have to pay the typesetter, whereas you still have to pay a rather large team of people to maintain the website indefinitely.

I just don't see the cost savings. A lot of this comes from my own experience in running a website.

Unfortunately in maths there's no real effective gatekeeper, and there seem to be a large number of people who don't let a lack of formal education or understanding of maths get in the way of writing maths papers.

There are a lot of cranks in astrophysics, but those can rather easily be ignored. One thing about astrophysics cranks is that even their output is limited. Suppose you publish one paper that claims that the big bang never happened. OK. Now what?

Also, I know of proper academic mathematicians who are rather notorious for not checking their own results. An issue in maths is that that for most people it's usually a non-trivial undertaking to read and understand a new paper, so the time lost in reading a bad or wrong paper is potentially greater. There is a separate argument to be had about the extent to which peer-review actually addresses this problem.

That does make things different from astrophysics. In astrophysics, you can usually tell within three minutes (or sometimes three seconds), if a paper is worth reading. This holds true even in fields that you aren't familiar with. Even if you aren't familiar with a foreign language, you can often tell if someone is talking gibberish.

This removes much of the need for peer review. Except in very exceptional situations, peer review doesn't add much to the process. The only time I've seen peer review be useful is if someone makes an extraordinary claim (i.e. neutrinos that seem to travel faster than light), in which case you want lots of people checking before releasing results. That way even if the claim is wrong, it isn't because of some silly error.

Again the difference between fields is striking, though the largely theoretical nature of the anonymity of the review process is similar, especially if you post a preprint on arXiv.

So if anonymity is theoretical, then what's the point of it?

I suppose one issue is access to archives. Maths papers remain relevant for longer than papers in most other fields, and the whole timescale for review, publication and citation is lengthened. It's entirely plausible that a working mathematician would require access to an article written by someone no longer in a position to put a copy on arXiv.

Astronomy papers and data also need to be relevant for a long time. People routinely look up data from the 19th century, and before (i.e. old supernova observations). As far as I'm concerned, the main purpose of astronomical journals isn't peer-review. It's archiving. And in that respect electronic media is vastly *inferior* to dead-tree media. In my department, you have paper journals that date from the 1900's and photographic plates that date from the 1940's. Those aren't going to disappear when the power goes out. What's more, you aren't having to fork over money every time you look at them.

Archival storage is going to turn into a major problem. If everything goes electronic, there's no reason to think that people in 2100 would be able to get to data from 2010 easily, and this is going to be a big problem.

The problem for individual mathematicians is that the current AMS journals are of a higher standard than even quite good maths papers, so if the average mathematician were to only publish when he or she had a paper of suitable standard they wouldn't publish very much.

So change the rules or start new journals. In astrophysics, the major journals are all "bread and butter" journals. If you discover something really weird, you'd publish in Science or Nature (although the reputation of Science has taken a hit lately).

The other thing is that astrophysics is very much data-driven rather than insight-driven. Even in theory, there are standard "bread and better results." So the bulk of the results are basically, "I looked into my telescope and this is what I saw."

Of course there are professional society and university run journals with lower standards, but often a subfield is dominated by journals from professional publishers. I can't say anything about how this situation may have developed, but it does appear to be the case.

It's important to figure out how the situation developed so that you can change it. One thing that you could ask is what would happen if people just refused to submit papers to journals and just published them on their web sites. In the case of astrophysics, things would get a lot more inefficient, since you'd have to track down papers.

 

[dcpo]

High end computer programmers get paid more than professional typesetters. Also for any sort of real world computer systems you'll have to hire usability experts and graphic artists, so I don't see the cost savings. The issue here is that once you've printed the journal, then you no longer have to pay the typesetter, whereas you still have to pay a rather large team of people to maintain the website indefinitely.

I just don't see the cost savings. A lot of this comes from my own experience in running a website.

Again, I don't doubt that any of this is true, but the cost of maintenance of the computer system is already being paid, so I can't see why there should be no saving from ceasing print production, which at this time is done alongside an electronic system.

There are a lot of cranks in astrophysics, but those can rather easily be ignored. One thing about astrophysics cranks is that even their output is limited. Suppose you publish one paper that claims that the big bang never happened. OK. Now what?



That does make things different from astrophysics. In astrophysics, you can usually tell within three minutes (or sometimes three seconds), if a paper is worth reading. This holds true even in fields that you aren't familiar with. Even if you aren't familiar with a foreign language, you can often tell if someone is talking gibberish.

This removes much of the need for peer review. Except in very exceptional situations, peer review doesn't add much to the process. The only time I've seen peer review be useful is if someone makes an extraordinary claim (i.e. neutrinos that seem to travel faster than light), in which case you want lots of people checking before releasing results. That way even if the claim is wrong, it isn't because of some silly error.

It's relatively easy to tell if someone is talking complete nonsense, but there's a lot of potential for subtle errors in maths, and reviewing papers where the author hasn't been very energetic in weeding them out is, I'm told, a pain.

So change the rules or start new journals. In astrophysics, the major journals are all "bread and butter" journals. If you discover something really weird, you'd publish in Science or Nature (although the reputation of Science has taken a hit lately).

The other thing is that astrophysics is very much data-driven rather than insight-driven. Even in theory, there are standard "bread and better results." So the bulk of the results are basically, "I looked into my telescope and this is what I saw."

It's important to figure out how the situation developed so that you can change it. One thing that you could ask is what would happen if people just refused to submit papers to journals and just published them on their web sites. In the case of astrophysics, things would get a lot more inefficient, since you'd have to track down papers.

Well, a large number of mathematicians are thinking about the situation and what to do about it. As I understand it the current boycott of Elsevier is not so much to get them to change their ways but to draw the attention of mathematicians worldwide to the problem. Any solution will presumably require some kind of general movement, and there is considerable inertia to overcome.

 

[twofish-quant]

The establishment of new mathematics journals is one of the many things being discussed, but is not itself a particularly simple thing to do.

I don't quite understand what the difficulty is. There is this weird disconnect, because sometimes it looks like that starting/running a journal is easy, and then it becomes hard.

It seems to me that all you need is a few senior mathematicians with tenure say that they are no longer going to publish in journals and that if you want to read their papers, go to their website or read their blogs. I mean, you have tenure, right? Or you can have famous mathematicians do a "Siskel and Ebert" sort of thing and have rottentomatoes for math papers.

Once it becomes "cool" not to publish, people won't. Once libraries start cancelling subscriptions and the journals aren't getting submissions then maybe the publishers will be more reasonable.

In any case the available data appears to show that several journals operate at some profit with prices significantly less than the 77% of Elsevier's lower bound your theoretical 'elimination of profit' calculation implies.

But it doesn't matter. If Elsevier's costs are 50 cents and they can charge $1 million, then they should charge $1 million. The whole point of a for-profit company, to maximize profit, and then are set up to squeeze every cent out of you. It's likely that Elsevier *could* charge a lot less, but it's there to maximize (revenue - cost).

If you think that it's a bad idea that a for-profit company does things sorts of things, then it have someone else do it.

 

[twofish-quant]

Again, I don't doubt that any of this is true, but the cost of maintenance of the computer system is already being paid, so I can't see why there should be no saving from ceasing print production, which at this time is done alongside an electronic system.

In the case of a paper publishers, they've already bought the presses and hired the people, so these are already sunk costs. Shutting down an existing business usually cost a huge amount of money.

The other thing is that if Elsevier could move to 100% electronic and fatten their profit margins, they'd do it. In any case, "reducing costs" isn't going to help you. If Elsevier reduces their costs, then it's going to go to the executives and shareholders, and you aren't going to see a penny of it. They'll still charge you the same amount of money, because they can.

Reducing costs is only going to help you if you have several competing firms that are trying to undercut each other. Personally, I'm a fan of paper, since putting things electronically gives all sorts of new ways that publishers can extract monopoly rent. Paper is surprisingly cheap, and I prefer to own books rather than to rent them.

It's relatively easy to tell if someone is talking complete nonsense, but there's a lot of potential for subtle errors in maths, and reviewing papers where the author hasn't been very energetic in weeding them out is, I'm told, a pain.

This isn't much of a problem in astrophysics. If it's a subtle and small error, then it's not likely to change the conclusion. One other thing that may make a difference is that in astrophysics, you want to see the original data or paper. You absolutely need to see what was done or what wasn't done. It occurs to me that in mathematics, once someone has come up with a proof and verified, there's no need to see the original document. (Or maybe not, you tell me). Since you want to see originals, dead trees become important.

Well, a large number of mathematicians are thinking about the situation and what to do about it. As I understand it the current boycott of Elsevier is not so much to get them to change their ways but to draw the attention of mathematicians worldwide to the problem. Any solution will presumably require some kind of general movement, and there is considerable inertia to overcome.

It looks like that Elsevier gets most of their money from the biologists so if you want a general movement, they look like people to talk to.

Also, my brief interactions with that world, is that astrophysicists could be useful because ever time I see a Elsevier flame war, you see people defending Elsevier saying that they do good stuff, and that the world would end if they didn't exist. One thing that the world of astrophysicists should is that you *can* have a world without for-profit publishers, and in fact I think that the fact that there are no for-profits makes things *better*. Something that has been interesting is that astrophysics and HEP has been at the frontiers of electronic publishing, and this is in large part because it doesn't kill anyone's business model. Los Alamos and Wikipedia got started without Congressional hearings.

The other thing is that it's possible to slay the giant. Elsevier is just another media company, and the road is littered with media companies that are in serious trouble because of inability to adapt to technology. Something that worries me a little is that the big publishers aren't quite as incompetent, and they've been suggesting a lot of things that may create new monopolies.

Finally, it's really important not to tick anyone off that you don't have to. Two of the things that have been suggested 1) peer review on arxiv.org or 2) getting rid of dead tree editions of journals are seriously going to annoy people in astrophysics, which is not a good thing if you want a united front.

 

[dcpo]

I don't quite understand what the difficulty is. There is this weird disconnect, because sometimes it looks like that starting/running a journal is easy, and then it becomes hard.

It seems to me that all you need is a few senior mathematicians with tenure say that they are no longer going to publish in journals and that if you want to read their papers, go to their website or read their blogs. I mean, you have tenure, right? Or you can have famous mathematicians do a "Siskel and Ebert" sort of thing and have rottentomatoes for math papers.

Once it becomes "cool" not to publish, people won't. Once libraries start cancelling subscriptions and the journals aren't getting submissions then maybe the publishers will be more reasonable.

At the risk of repeating myself, I believe one of the aims of the current boycott is to influence the submission habits of mathematicians away from for profit publishers. People like Tao and Gowers could probably post their research entirely in blog form, and still be tremendously influential. People like me currently need to publish in journals, and sometimes our choices are limited, because otherwise we're not going to get a job. Whatever system or systems are to replace the for profit publishers, if they are indeed to be replaced, will require some form of consensus from mathematicians around the world in order to function. Exactly what the solution will be is unclear, but movement is forming, and prominent mathematicians are taking the lead.