Harvard Library pushes open access

I’m very glad that Harvard is doing this, as it’s one of maybe a few dozen institutions who really do have the resources to continue paying the thieves if they choose. And of course, all such institutions have enough prestige that if did so choose, they would drastically slow the process of reclaiming this market for vendors who charge reasonable rates for the value they provide (distribution/hosting/marketing), rather than charging primarily gatekeeping rents for content they have almost nothing to do with providing (the actual scholarship, peer review and editing).

Having worked most of my life in an academic library, and watched our University Librarian succumb to a journal publisher’s SLAPP suit (the university’s lawyers declined to defend him in a suit over an article he’d authored, which attacked journal publishers’ pricing policies) I’m glad to see this finally come to a head.

Machiavelli’s advice about not striking a prince without taking care to kill him seems relevant here, as does Franklin’s about hanging together lest you be hanged separately.

This is good news. Universities are an important locus of non-market values in our society but they’ve mostly been on the defensive. It’s great to see some pushing back against the commodification of scholarship.

“Universities are an important locus of non-market values in our society but they’ve mostly been on the defensive.”

This has nothing to do with market or non-market values — it has everything to do with not being a total sucker. Publishers use the carrot of reputation to get most of the work done by academics for free — and then charge the libraries they use exorbitant prices.

Substance McGravitas:

I’m skeptical that that bomb would be anything but a dud as all major research Universities have long given tacit consent to the commercial publication of their own work product by subscribing to the journals for decades, giving their content ample review by specialists and librarians, and by allowing their use in hiring, tenure and promotion decisions.

To add to GW@10, universities have long given tacit consent for those commercial publishers to determine access policies, implicitly, by determining costs. At least, in the US that is the case. Is the same true in other countries, or do prices follow the pattern of the [censored for technical reasons] industry?

Eli would not be surprised if the next move is Elsevier’s cutting Harvard off. At that point the needed rallying cry is “We are all Harvard”

Any article coming from US gov labs (even in part) cannot be copyrighted by the publisher. The copyright is owned by the gov. Not sure how that plays out in terms of being able to profit from selling the articles.

About Fucking Time!

Several universities have been pushing for this move to Open Access for years, but it doesn’t much matter if Podunk U threatens the publishers. They’ll laugh. Maybe now that the movement has the weight of someone like Harvard behind the push, we’ll get some change.

Harvard is probably nowhere near “untenable” on these subscriptions, unless the publishers are charging based on prestige and not enrollment. But many other schools below R1 in the U.S. are nearing untenable status on subscriptions. And, frankly, the current situation cannot hold. The Journal of Computer-Mediated Communication, now published by Wiley, is open-access, and has had been highly-regarded in the field for some time.

Matt, one of us is misunderstanding what is meant by “open access” here. There is no way at all that publishing will be replaced by something like the arxiv, and that’s a good thing – all the suggestions I have seen how to do this while keeping some semblance of peer review increase the already ludicrously high power a few faculty members at a few elite universities have over the job market. So I’m assuming open access means what it means in many sciences: the authors pay for being published. And that is a horrible idea.

all the suggestions I have seen how to do this while keeping some semblance of peer review increase the already ludicrously high power a few faculty members at a few elite universities have over the job market.

Ok, not getting this. (My last post got cross-posted with your last post.) Peer review, at least of the sort I’m acquainted with is already for free, as in the “peers” reviewing the papers do it for free. And the reviewers are very often not high powered faculty members (I mean I’ve peer-reviewed papers, and no one would would describe me as high-powered etc.). I imagine you have something specific in mind, but I’m not seeing the connection.

I’m assuming open access means what it means in many sciences: the authors pay for being published.

If that is what’s being proposed, then I retract my enthusiasm. But is it?

I think Christian is conflating the tangential point that Otto raises in #1 (that currently, you may bribe the mainstream journal publishers to make your paper publicly available if you can afford to do so) with the actual issue of open access. Open access journals, like the JCMC mentioned upthread, do not charge for submissions.

Ok I retract – this was my baseless assumption, possibly only based on my mistrust of anything Harvard. It’s entirely possible I misread and they meant open access including free publishing.

js. (24.), I only have seen what’s being discussed in math. Here is how I see it: if we move away completely from a publishing model that needs… well… publisher (which does not have to be Elsevier, Wiley or Springer, but still will have non-trivial overhead costs that cannot be magicked away) to something like the arxiv, then there will be no traditional editing process (editing involves a lot of administrative work by a number of people and expecting them to do this for free would be outrageous). This means, there’s nobody who can assign a suitable peer reviewer.

So we are left with two options:

1. No peer review. Everything’s posted (if from someone at a reputable scientific institution, say – the arxiv model) and the readers will figure out what’s good, what’s correct, what’s total bs. I find this problematic on its face, but one can argue that in practice I’m not waiting for new papers to be published but simply follow the preprint servers anyway – fair enough. The real problem appears when we relate all this to the hiring process, see below.

2. We replace current anonymous peer review by some kind of crowd-sourcing process, say similar to mathoverflow or dailykos. This would not be anonymous. I guarantee that as a result the opinion of Prof. X from Harvard about a paper would count much more than that of Prof. Y from UIC – even if Prof. Y is an expert on the topic. Meaning, Prof. X’s comments would be rated higher than Prof. Y’s.

Either way, currently the journals articles appear in have influence on the academic hiring process – wildly imperfect, but crucially based on the articles themselves, not merely on the opinion of elite academics about the author. Either no peer review, or the crowd sourcing process described above would, on the other hand, magnify the already too large influence on academic hiring enjoyed by the reputation of reviewers – currently in the form of letters of recommendation and graduating department. I could tell you any number of examples of candidates hired over what I (and not only me) consider better qualified ones simply due to the fact that the eventual hire graduated from elite school X and had letters from professors A,B and C at elite school X (and maybe elite school Y in the same town). I would be extremely loath to embrace any change that makes this problem worse. I don’t want to replace commodified science with feudalism.

christian_h, I think that arXiv offers the most important benefits to the public of open access without totally upsetting the publisher apple cart. It doesn’t replace the gatekeeping role (you can and many do share nonsense there), but once you see an interesting article in a peer reviewed journal you can find the almost-identical equivalent on arXiv with minimal effort. Publishers don’t go out of business, peer review doesn’t go away, authors don’t pay any more, but people who don’t have an institution or the ability to pay $30 per article still get to read what professionals are writing.

Christian: I think you’re still missing the point. Open access is, demonstrably, compatible with peer review in journals. Read Peter Suber‘s sourced and linked summary of how actually existing open access journals work.

Again, I may be missing the point but I think I don’t see that there is any. The problem (well one problem) with academic publishing is prices. These prices can be lowered three ways: more free labour; more efficient publishing processes; or lower publisher profits. Neither is in any crucial way correlated with the model of publishing (open access vs. subscription). It’s a complete distraction from the real issue. (Why would a library, on average, care if it has to pay inflated fees to Elsevier to get their institutional members the right to publish as opposed to paying inflated subscription rates?) It just so happens to be a distraction that would profit institutions like Harvard.

To be clear, J. Otto Pohl’s experience is NOT an exception. And he clearly did not count as “hardship” for Springer. So what’s the point here?

Christian: in #36, you’re still conflating buying the right from a closed-access publisher to allow the public to see your piece, as J Otto Pohl describes, with the model that’s used by open-access journals.

When implicitly ruling out the possibility that open-access could work, you’re massively overestimating both the inefficiency of, and the profits made by publishers involved in, the closed-access journal process – and hence vastly exaggerating the amount of working-for-free that is required to make up for the process. In practice, (small) grants from funding organisations tend cover the (small) editorial expenses.

One key piece of evidence you might want to consider here before assuming that open access can only profit major institutions is the actual existence of more than 7,000 open-access peer-reviewed journals (solely those listed in the Directory of Open Access Journals).

christian_h@37:
I think it depends what you mean by publishing. In Computer Science (or at least the security and cryptography subfield), nearly all important publishing is at conferences, not journals. Said conferences sometimes have paper “proceedings” (which are just printed copies of the papers) but quite often have online “proceedings”. In either case, it’s quite common for the authors to produce PDF of the papers directly, with no editorial work done other than what the authors do themselves or is done incidentally in the review process (more on this in a second). The conferences charge admission, of course, but essentially none of this money finds its way into any stage of the publication process, though perhaps a small amount goes to proceedings. Rather, it’s used to run the conference itself [hire the rooms, projectors, refreshments, etc.]

The submission and review process is run by the program chairs themselves using heavily automated tools (e.g., http://www.read.seas.harvard.edu/~kohler/hotcrp/). The review is all done on a volunteer basis, as is the program chairing. If there is an in-person program committee meeting, it’s generally held at an institution associated with one of the chairs. Otherwise, there may be a teleconference or perhaps the entire PC meeting will be held via email discussion. In any case, much of the paper discussion is done using e-mail or via the tools. Given the above, what parts of this process involve infusions of actual money:

– The servers with the conference software must be hosted, but the cost here is nominal. On the order of hundreds of dollars at most.
– A server must be found to host the accepted papers. Again, this is very inexpensive.
– If you are having a teleconference, you may need to pay for that, though things like Google Hangouts are bringing the cost down sharply. When I have chaired, I have gotten conference bridges donated.
– If there is an in-person meeting, the venue is generally donated by some institution but there may be refreshments or a PC dinner or something.

Overall, these expenses might run to a few thousand dollars for a big conference. It certainly wouldn’t be hard to float this with only minimal funding.

I’ve of course left aside the time of the reviewers themselves, but they’re not compensated, so as far as I can tell, it’s considered professional service/development, and if you can afford to do research you can certainly afford the time to do some reviewing. So, it’s less that you can’t do publishing without funding than that you can’t do research at all without some form of money.

christian_h, and others:

I wrote about this at Obsidian Wings last year, including a compare/contrast for different types of peer-reviewed publications.

Yes, in Science/Tech/Medicine, authors (and their institutions) are normally charged high fees for open-access journals. All the publications I looked at, though, had more or less sweeping policies to wave the fees for authors from certain institutions, countries, or indeed continents. Your colleagues in Latin America would undoubtably be invited to publish without a fee.

Nonetheless, I think the fees charged by PLOS and other open-access science journals are *outrageously* high and have no clear relationship to the services provided. Journals in the humanities are mysteriously able to do without such fees, so my suspicious mind says that the difference is that researchers in STM have grants which can be made to pay.

To repeat my question explicitly: let’s say the amount of revenue generated by mathematics publishing each year is X. This amount could be large, or small – doesn’t matter for the argument. Keeping this revenue constant, how does open access vs. subscription based publishing change the average amount scientific institutions have to pay for publishing? How does it not simply redistribute the costs (for example from the university library to the NSF and NIH etc.)?

If, on the other hand, you want to argue that open access publishing by itself would lead to lower total revenue and therefore costs, what’s your reasoning? Where in the process does the publishing model reduce either overhead costs or profits?

For clarity, “for every dollar spent by academic institutions on publishing” should be “for every dollar spent by academic institutions on journal subscriptions” – phrasing above is unnecessarily ambiguous.

Yes, true. So the library would profit, and the funding bodies would get hit. Since the latter have finite budgets (probably shrinking budgets in the future), this will mean less funding for actually doing science. So you have successfully redistributed costs from Harvard – endowment $30 billion, from tax-deductible giving – to tax-funded granting agencies. Most likely this will mean fewer grants – so a larger proportion of all grant money will go to the Harvards of this world since their total grants will surely not decline due to their reputation.

I can see why Harvard would support this, but why should I? I’d rather fight that 36% profit margin, and demand that my state (California) fund higher education properly.

Christian H: The problem I have with your analysis is that your premise is that managing university status is the purpose of academic publishing, and you find fault with changing publishing systems based on the impact such changes might have on hiring, tenure decisions, etc.

But the purpose of science and academia isn’t to provide jobs and status measures for scientists and academics. The purpose is to advance and spread knowledge. Open access to journals has incredible benefits to people all over the world, both inside and outside of academia.

If we aren’t smart enough to come up with some way to both spread papers freely and simultaneously make unbiased hiring decisions, that’s really not my problem, speaking as an outsider. Siloing knowledge and charging impossible prices for access, as the solution, seems like a crying shame.

I think Matt’s weaker version of the argument holds here, but Jacob’s strong version isn’t really tenable. Yes, public access to scientific journals is a major public good (this probably varies sector-by-sector; my understanding of pure maths is that a new cutting-edge paper will be comprehensible to a relatively small number of people who will hold – or are at least candidates for – maths PhDs from respected universities), but you can’t totally handwave the problem of how to generate them.

Even so, I think Christian’s concerns are somewhat overblown. Is it reasonable to assume universities will channel none of the money saved on library costs into research funding? Given that every major Anglophone university I’m aware of is seeking to cut teaching costs and spend more on increasing its research output, the converse seems more plausible.

I run and edit ( and copy edit and proof ) an open access peer reviewed journal, in the humanities. It is a part of the larger scholarly public history project I work on, and it is done in the spare moments between editing a website approaching its millionth word. It is not even close to a full time job. The Open Journal software installation is managed by a university library. If I had an active editorial board who helped me, we might have put out more than 7 issues in three years, but even that is not bad in history.
Google analytical tell me that our articles have been downloaded by thousands of readers. There is no Australian history journal published conventionally that would have close to that in subs or readers. And we have readers from all over the world, who would never be able to get a conventional journal on Australian History. Open access publishing is the way forward for academic work, and at last, academics have the means of production ready to hand if they would just stop believing that the publishing cartels offer them anything more than what they can do (and already do, often ) for themselves.
It is not hard, people. You have nothing to lose, and a world of readers and contributors to gain.

42:

I work in Africa (U. of Ghana) and I was not given the option of any waiver by Springer for Open Access publication in Human Rights Review. Maybe the sciences are different, but Springer was pretty specific that if your institution did not pay the fees that you would have to pay them yourself for Open Access regardless of location. The fee was $3,000 if I recall correctly.

Chris:

Springer put it online without any fees. But, you can only access the online version without paying if your institution subscribes to Springer. University of Ghana does not. I had to have a colleague at a University in the US who does have access download a copy for me so I could have a PDF version of it for myself before it officially comes out in print (paper).

OK… after all this discussion, what happens if we do all move to new open-access journals?

How are all the new online journals going to build their human support networks?

Many new journals are being started by individuals with access to the internet, and various off-the-shelf publishing systems, but without sufficent wider support.

New journals may be disappearing almost as fast as they are being created.

The strength of any journal lies in having a large pool of academically interconnected readers and contributors and reviewers. That way, the pre-submission processes (getting friends to read drafts, for example) can raise the standard of papers before they reach reviewers, who will respond positively to receiving good papers from the journal, and then continue to offer free reviewing.

Journals sink or swim according to whether or not they can set up positive feedback loops between the readers (who may become contributors, for example), contributors (who may become pre-submission readers and editors for colleagues), and reviewers (who may become contributors).

I have never seen any discussion of journal publishing methods that takes into account all the social interactions needed to produce a successful journal.

I am trying to address the issue at The Research Cooperative – an NPO and open access model for building support networks for any form of academic publishing.

We have over 4000 members now, but our network is still far from being large or active enough to keep up with the rapid increase in new, open-access journals.

Our system is scalable, but we do not yet have critical mass, just like many of the journals I am hoping we can support!

SCOAP3 is a beautiful model for how to support open access journals via a funding consortium, rather than publication charges (see http://scoap3.org/). These fees may be fine for the life sciences, where in many cases you really can’t do cutting-edge research without a lot of research funding, so you might as well devote a little of that funding to paying to distribute your articles. However, I’m very concerned about their consequences in other fields where this just isn’t true. Mathematicians should not have to use a publishing system designed for the needs and practices of another field, unless we decide it’s the right fit for us too. Sponsoring consortia seem to me to offer by far the best route forward. One difficulty is that they take time to set up, so we need to be working on this now rather waiting until we need them immediately, at which point it will be too late.

john b: Even so, I think Christian’s concerns are somewhat overblown.

No doubt about it. I didn’t really set out to make grand proclamations about the evils of open-access publishing – it may well be the right way to go regardless of the cost issue if it can be structured correctly – my main point was that touting open access as a solution to the combination of (a) price gouging (b) explosion in publication rates and (c) shrinking university budgets seems like a distraction. It’s not a magic wand making those go away.

Jacob: it will become your problem if you get to read even more of the same over and over again because big boring schools have taken over everything even more than is already sadly the case.

Emma: You publish 7 issues in two years. Elsevier’s Journal of Algebra, one of three rmajor journals specializing in algebra (two of which are published by Elsevier) publishes about 400-450 pages every two weeks. Would you want to be managing editor for that, for free, without any staff help paid for? Would your employer say “sure you are free to use one of the staff we pay for full time for this”?

It should be added that the Journal of Algebra – like every math journal I have dealt with from any publisher – allows authors to post (a) final versions on their personal webpages and (b) almost final versions (in this case even final ones) on the arxiv. All my articles are available for free online this way. (Otto, I would definitely post your paper on your personal webpage.) Which open up access pretty well.

@7: I’m sure there’s variation, but at least in the UNC system the policy is that faculty own their own copyrights except in very unusual circumstances (http://www.northcarolina.edu/policy/index.php?tag=500.2, see page 7-8). Works produced by non-faculty staff, on the other hand, are considered works for hire and therefore University copyright.

@Otto

Springer will usually let authors negotiate a deal whereby the author can put their paper on their own personal website. The author doesn’t pay Springer anything, and readers who download the paper from the author’s web site don’t pay Springer anything either. (Readers who want to download the paper from Springerlink, rather than the author’s website, do need to have a subscription which their institution has payed Springer for). You do need to do this negotiation before signing the copyright transfer (possibly with assistance from the journal editior) – technically speaking, you can’t just put the paper on your own website if you forgot to negotiate this versiuon of the copyright transfer.

The other two options with Springer are (b) author pays, and readers can download for free from Springerlink; (c) Reader pays for Springerlink subscription.

The first option is the better deal (while we’re waiting for a radical change to the system …); I think everyone who publishes in Springer journals ought to do it.

I share the concern that the price for option (b) seems rather high.

christian@62: for free? of course not, but for a reasonable fee for service, surely.

As a printer, I am in a position to comment on the actual costs of printing and distributing such a publication.

The marginal cost of a print edition is almost certainly less than the 315 price for non-institutional subscribers. One could produce single print subscriptions on demand for around that price and make a minimal profit. For runs in the 500-1000 range, that could be a very nice piece of business. The marginal cost of an online/email/pdf edition is obviously very low, probably less than a dollar per subscriber.

What would it cost to manage such? I could hire a very competent full time staff member for 50-60,000 a year to manage submissions and peer reviews and copy edit, maybe an additional contract copy editor for $20/hr about 1/2 time. Field expert editing checks could be provided on a contract basis for say $50-100/hr, by junior academics who would presumably be happy to get such wages. Prepress setup to create a nice finished pdf would run maybe $10 a page max including profit for me, presuming most of the math and table typesetting is already finished on the submissions.

This is not my specialty — I suppose it’s possible that I’m missing something, but it does seem to me that most of the value is coming from people who aren’t getting paid by the journal. I’m guessing that what I described above is quite a bit more than I would need.

If I were able to take over that journal, and knew it would be bought in one form or another by >500 institutions, I am quite certain that my company could produce it for a massive profit selling institutional print editions for somewhere around 20% of the Elsevier institutional price, and online access for 1/2 that. If it were bought by a lot more, the online price could go down proportionally.

This is coming from someone who owns a print shop with a completely different specialty. I don’t even make magazines or books in house. I am judging profit based on what I could cover by jobbing out large portions of the print work to small local shops I’m personally familiar with who may not be the cheapest/best options. I am 99% certain that there are companies out there who could do this for less than I could, and let me tell you I would be *delighted* to publish a journal like that for 20% of what Elsevier charges. A few contracts like that, and I would become a very rich man.

I am guessing that if this were a serious possibility and I put real effort into coming up with a tight quote based on the best suppliers I could find, or making capital investments to produce everything in house, I could get that number down into the 5% range, and still make a comfortable profit margin.

If there’s anyone reading this who wants to seriously talk about moving journals away from these rentier gatekeepers, and has a real chance of mobilizing enough academics to start serious competition for any well-known journals — I would be *very* happy to talk details.

christian_h, I have to say I am surprised and impressed by new-found voice as an articulate defender of the positive role of for-profit corporations in public life. Will this new passion translate to other questions of public policy?

No idea what you are talking about Walt. You may not be able or willing to read, but I did explicitly attack the profits of the publishing industry several times. Maybe you want to make the utterly ludicrous argument that “open access” cannot be for profit? Or the argument that Harvard must be supported by socialists because they are, after all, a “non-profit” (yeah right) institution? If so, make it. Otherwise I’d ask you to let the straw men go unmolested, they have no part in this.

Authors should not have to pay magazines.

Period.

christian, “final” and “almost final” version are definitely not “same diff”.

If libraries want to pay money to G&T t because they like them, or because they need some volumes to fill the shelves, they can do it, but to access the actual articles they don’t need to. On the other hand, here is what J Algebra says:
“Policy: Authors retain the right to use the accepted author manuscript for personal use, internal institutional use and for permitted scholarly posting provided that these are not for purposes of commercial use or systematic distribution.” and “Systematic distribution means: Subject repositories that aim to aggregate and openly distribute AAMs authored by researchers in specific subject areas.” But, again, I am sure you knew this already.

IMHO one cost being underestimated in this discussion is that of maintaining a reasonably secure web site and long-term archive. Not too hard to do for 3-5 years, but Springer has been at it for 160 years and I don’t think they are the oldest.

A big advantage of open access is that you don’t have to pay for all the “guard labor” to keep non-paying users away from the articles. You don’t need to bother with lawyers to threaten or sue copyright violators, access systems to track quotas and ensure that only licensed users are downloading content, analysis systems to look for patterns indicating illegal archive-compiling, or watermarking software to stamp each article downloaded with unique identifiers.

Elsevier claims to have roughly 7 million articles in their archives. If the average article is (say) 10 pages, call it 70 million pages of archives. By way of contrast, the HathiTrust has roughly 3.6 billion pages of archives scanned from libraries and makes the ~1 billion pages of public domain material it holds freely available to everyone with an internet connection. It has an annual operating budget of about $2 million and with this funds software development, professional system administration, content delivery to partners and the public, and redundant geographically distributed storage. It provides access to about 14 times as much archival material as Elsevier with less than 1/1000 as much revenue. The only thing they don’t have is the copy right to share the other 2.6 billion pages with the public.

christian_h, in #64 you have written the most stirring defense of Elsevier that I have ever read.

After this thread, I think that your socialism is basically a pose. There are many times where we are standing on the sidelines, and unable to have more than a tiny effect on the outcome. It’s easy to call for radical reform when there is no possibility of having any effect on the outcome, or of the reform having any negative outcome for you. But you’re apparently a tenured math professor — you’re not on the sidelines, but in the middle of the game. And suddenly you discover that radical reform might have a bad impact on you, and that conservatism is called for, and that while you deplore the costs of the status quo, you are not prepared to take any action whatsoever. If you are really serious about ending monopoly profits in publishing, you would do more than invent reasons for inaction, which is all you have done here today.

Regarding this whole argument about G&T, they definitely have a paywall keeping non-subscribers from accessing content, although they allow self-archiving by authors. Back issues become open access after five years. Another example is the American Mathematical Society, which also makes all of its content free after five years (and has a paywall before that). With a sufficiently long time window, experience shows that this is compatible with the subscription model for publishing.

Well Walt, your continuing refusal to actually read what you claim to criticise is concerning. At no point at ALL am I defending Elsevier in comment 64. That’s all in your imagination – it is clearly you who does not believe anyone but a major corporation could do the job. Because all I do in that comment is point out that a major journal will not be published on free volunteer labour alone, it will take a collective effort to actually create the necessary infrastructure.

The fact that you apparently think that “doing something” consists of changing consumer behaviour also suggests that you are a singularly bad judge of who is or isn’t serious about changing things. But hey, if you want to believe that the academic equivalent of driving a Prius (maybe with a peace sticker and a “I break for animals” on the back?) is the way to go, go ahead. Or maybe you had some other kind of action in mind? Like screwing young scientists in temporary jobs by refusing to referee their papers? How very radical of you.

Don’t trust Harvard to have everyone’s best interests at heart = defending current publication practices. Got it. Also, journals like [insert name of major publisher open access bioscience journal here] are clearly harbingers of a better future and in no way part of “current publication practices”.

Anyway, apologies I somehow managed to create a flame war by giving the impression that I am a big Elsevier/ Springer/ etc. fan – which couldn’t be further from the truth, whatever some here may be imagining.

There is probably a self-explanatory definition out there somewhere, but I confess to not knowing whether a prohibition on “commercial use” of text retrieved via a public open-access portal extends to any use not for education, charity, or curiosity. Does it extend to creating a commercial product using the ideas in the piece? To writing a book for profit?

Also FWIW, the people I’ve worked with who had the biggest obsessions around information sharing either had worked for defense contractors or spent a lot of time around people in the the open source movement. (Unless you count the people whose annoying personalities or minority status seemed to have made others abnormally unwilling to share with them.) So my experience isn’t based in issues around academic publishing per se (except to the extent the open source people seemed to share the idea, which I seem to remember reading about someplace, that the millennium will involve everything being owned by and administered from (essentially) the academy).

It’s already being called the Academic Spring (it’s even got a Wikipedia page). I wonder if the final straw, that which gave Timothy Gowers the impetus to make his blog post and the unexpected enthusiasm with which the boycott proceeded, was Elsevier’s support of the Research Works Act which was introduced just the month before.

Most academic computer scientists post electronic copies of their published papers on their personal web-sites, despite prohibitions from many commercial publishers. Does anyone know of any legal actions (or threats of legal actions) against such practices by commercial publishers?

@Bianca re public libraries cancelling JSTOR. That’s sad news. Even with the terrible cuts taking place, it’s hard to imagine that $5000 a year is unaffordable for the whole city of Boston to get access to so many journals. Couldn’t they find a donor to support something like that?

JQ, THAT’s exactly the kind of journal that could move to , say, an Open Journal System installation at a university library, hire someone like me half time or less to run it, provide it free to the public, or maintain your subs scheme for recent issues. The fact that you now have a better deal shows that someone at the commercial publisher has seen the writing on the wall.
The university library where my journal is hosted now runs some 10 or 20 journals , and worldwide, Open Journal System published journals number in the thousands, many of those in the developing world precisely because out here on the periphery publishing has small volumes and crap margins and the big boys aren’t interestd mostly. Open Knowledge Project also makes management software for editing book collections and running conferences.
There are a lot of ex-academics like me, now experienced editors and copy editors who would be happy to run two or three journals with the sort of reviewing and senior board support that journals often have. The costs are not as big as the for-profit publishers would have you believe. They are not doing as much for you as they say.

(Emma basically covered everything I said, and more besides, but I wanted to make very explicit the rejoinder to Christian’s claim that these are two separate issues.)

Rana, that’s interesting. I see it the other way around. I’ve spent too much time chasing articles from old journals to be sanguine about the value of publisher reputations and commitment. If a journal is licensed for open access and printed on paper, people can scan the paper and share the articles whether or not the publisher is still around. If it’s only available electronically, then it’s even easier; maintaining a complete mirror and serving it has tiny costs, easily borne by interested practitioners or even more easily borne by dedicated preservation efforts like the HathiTrust or Internet Archive.

By way of contrast, if a journal is under traditional copyright there is no “official” way to obtain copies if the publisher disappears or doesn’t want archives of past articles to be available electronically. Example: Chemistry & Industry, published by the Society of Chemical Industry in the UK, has decades of pre-electronic material cited by other journals, patents, and the secondary literature. But the Society today does not sell or otherwise offer archival access to their pre-electronic journals. Google has scanned those journals, but the material cannot be read through Google Books because it is still under copyright and the rights-holder has not granted an exception to make it visible. Since the publisher won’t lift a finger or even let others lift fingers to make archives available, the last method of access is physically traveling to a library that retained paper copies or pleading fair use to get an article photocopy through interlibrary loan — but the fair use exception is fairly narrow and under regular attack by rights holders who want even more control. To put it mildly, I think this is at least as real and vexing a problem as ephemeral open access journals whose URLs break after a few years.

Back in the olden days when I was a PhD student I tried to get a copy of a hardcopy journal article on a topic unknown in my country’s university libraries. It was impossible too. I ended up using the newfangled internet to email someone in the US to send me a photocopy. Now the whole run is online.

Myself, I think a journal hosted on a university library server is about as permanent as it gets nowadays. Although my journal is also archived through the Pandora project, Australia’s web archive, so there’s that, too. Matt @ 104 is exactly right.

Matt #104, I’m not disputing the value of things being online; heck, I use online resources quite regularly, and feel grateful at how easy most of them are to obtain. However, that’s not the point I was making.

If I cite a specific article or other publication in my work, that citation is useless if it doesn’t go to something solid and permanent and accurate. That’s my concern, not initial accessibility.

I’ve never had trouble tracking down a published book that was cited (sometimes it’s work, yes, but ultimately a copy does exist somewhere for me to find and read) nor an article published in a venue committed to maintaining some form of permanent unique signifier to use as a citation, whether that’s a permanent, unchanging URL or a page range in a published hardcopy journal.

But an article that was posted once, on a webpage that ceased to exist when its author moved on to another institution? If I can’t cite it using a permanent signifier, it’s worthless as a source, because it means that no one else can find it, confirm that it’s the same source I used, and use it to check the accuracy of my handling of it.

In some ways, sites like that are worse than ephemera, because with ephemera you at least have a single copy to look at, in a named archive, and since you got access to it, someone who cares enough can gain access to it too. A site that’s disappeared, and your only remnants of it are lingering (maybe, if you’re lucky) in a cache site somewhere? That’s cruel to your readers, to send them on a wild goose chase trying to figure out what source you were using.

I’ve done it. I’ve done it far more often than I’d like. It is a real problem, and it’s something that requires thought and commitment on the part of the online editor/host to make sure it doesn’t happen. H-Net does a great job of it. Too many other sites do not. If people are serious about open source replacing the existing structure of journals and articles, they need to think about this, what maintaining permanent unchanging URLs over decades or more of time requires, even as people move on, projects lose funding, technology changes, and so on.

John Q:
It was probably about ten years ago, and I may have misremembered. I thought there would be JSTOR access, but it wasn’t there, or was only available onsite. Maybe it hadn’t ever been there yet. I did definitely see a notice a few years back about cutting access. Maybe rates have changed. That was a long time ago. It isn’t that long since it seemed really great to be able to e-mail catalog searches to myself or copy them to floppy. And as I said, what’s really annoying to me these days is not having access to back issues of The New Republic and New York Review of Books, neither of which the library buys anymore either.

A CD-ROM? You crack me up.

How about a web service that the universities can query at any time for the entire back catalogue, which they can then store locally?

Sneakernet is so 90’s.

Well that’s what DOI is supposed to solve.

bianca steele @110:

I just went onto the BPL website; they list JSTOR as one of the databases to which they maintain a subscription. All that appears to be required for off-site access is a BPL card and a PIN.

JSTOR’s development and operational model were/are different from those of commercial electronic/print publishers like those under discussion here (Elsevier et al). The platform was developed with serious funding by the Mellon Foundation, and was envisioned as a self-supporting ongoing enterprise that would offer a key service to the scholarly world henceforth (electronic back runs of key journals in selected disciplines – originally, ten -, with as permanent a platform as it’s possible to envision). They are flexible about charges/ subscription fees – my husband’s research institute was able to negotiate a modest one-time “installation+start-up fee” plus a reasonable yearly subscription to a select number of journals. In this respect, I think, JSTOR is closer to what some of the commenters above would like to see – reasonably-priced (if not entirely free) access with a pretty solid expectation of permanent access.

If the power goes off, the world ends is the worst case scenario for digital storage. Slow Fires — is the equivalent scenario for traditional library storage. Relatively few books are published on 300-year paper these days, and I think it’s safe to say that no one knows the true extent of digital vulnerabilities — format obsolescence, physical media decay, bitrot, barbarism, etc.

Redundancy seems to be the key, but that’s a matter of whim, not central planning. Does anyone here see in all of this a unified strategy which protects both access and permanence?

Yeah, but Barry, given the reluctance of governmental and non-governmental agencies to bear the cost of migration, can we really rely on a collective of interested folks outside those agencies to do the work of the Lord? While spending over thirty years in library data maintenance, and having the nightmare experience of going through the stacks, bending down the corners of pages in books and watching them crack and crumble in my hands, staring in dismay at racks of half-inch computer tape reels and stacks of 8 and 5-inch floppy disks in our vaults that were no longer mountable on any available reader, I came to wonder. Should the pressure drop in our innovation engines, what will be left?

In the case of cultural artifacts like architecture, I suppose it can be argued that if we have the Pantheon, we don’t really need the Parthenon, or the Temple of Pergamon. When it comes to what’s in books, though, I’m not so sure that we can be so complacent.

Barry Freed:
“We” have been trying to set up directories that separate physical locations from logical descriptions for a while. So far not much more complicated than phone co. directory service or DNS is working AFAIK. You need constant vigilance to keep “local descriptions” from including IP addresses and drive names and the like, for one thing.

1) Physical storage media, like DVDs, degrade over time. There are numerous older media for which no reader works any more, as we have found at Computer History Museum.

2) Computers are really fairly good at automating tasks like copying files to local storage, or at least somewhere other than the original.

3) See WebCitation, a very useful thing for archiving specific files. It takes a little setup, but once you have a bookmark with the right magic value:
javascript:void(location.href=’http://www.webcitation.org/archive?url=’+escape(location.href)+’&email=MY EMAIL’)

all you do is click on that and it queues a retrieval request, emails you when done.
I use an Outlook rule to send them to their own folder.

This doesn’t work for everything, and unlike Brewster’s Wayback, you have to think to do it, but unlike Wayback, you can capture most files immediately. Final fallback is screenshot. I’ve had to use all of these at one time or another.

4) Storage in the cloud is nice … but there is a famous old cautionary tale, “MS Fnd in a Lbry.”

I would like to add to the list the Pensoft journals which are open access journals in the field of zoology, biology, ecology and related fields.

http://www.pensoft.net/journals/

Have to disagree with Rana – I’ve had plenty of problems with deprecated URLs and removed access from big publishers – (E) and (S) spring to mind – they do not have (at least in my field) the interests of researchers at heart.