An Indian research university has assembled 73 million journal articles (without permission) and is offering the archive for unfettered scientific text-mining

boing
21

I’m Editor in Chief of a society-connected “diamond” open journal. (That means open access and no page charges; this is the model that is growing most rapidly.) I think the reality is somewhere between the common “all the work is done for free by academics” narrative and the “oh but we add so much expensive value” reply from publishers.

I agree that there are real material costs (hosting, DOI fee, etc), but for an online-only journal those are small enough that they can be easily carried by library or association or departmental incidental budget.

The labor expense (on top of the ones like refereeing that people always talk about) is a real cost. My journal is all volunteer, but that limits our size; we would break at 20% more papers accepted. I can barely run the journal and do my day job, if we were not a narrow specialty journal we’d absolutely need a paid staff.

A large chunk of the man-hours is in copyediting, which we do carefully, but honestly the level of copyediting at many commercial STEM journals is abysmal: I see problems all the time in articles in Science which would never pass our review.

Finally, you mention publicity. This is field-dependent; in my small field people know my journal and my board is distinguished, so we have no need for a publicity budget, but I know it can be expensive to buy book display floorspace at the big academic meetings or ad pages in other journals or even on blogs

Scholarly publisher profits (and journal costs) really skyrocketed in the 1980s, possibly thanks to deliberate market manipulation by the likes of Robert Maxwell, but clearly well beyond the actual expense of running the journals. Pirates like Sci-Hub could be forcing a much-needed correction.

8 Likes
22

Oh. Me too.

I do a LOT of peer review. I’ve reviewed text books (at about 50 bucks a chapter), grants (for king and country, i.e. free), high impact journals (free subscription to the print copy for a while. When that stopped, I quit reviewing for them), moderate impact journals (discount on books 99.99 percent of which I don’t want). Editors usually get paid (the compensation varies widely). The production staff (copy editors, technical typist, etc) are compensated.

wazroth is correct: it’s pretty much accepted/tolerated because of tenure and promotion, with a little bit of ego and bragging rights.

Too, there is a LOT of just plain inertia. It’s THE MODEL, so it’s going to be hard to change. Academics are often VERY hidebound.

3 Likes
23

Good on them.

Don’t ‘ask permission’ nor “beg forgiveness” for doing the right thing for everyone.

1 Like
24

I would say conservative. Without signifiers of research quality like publication in highly regarded journals, you end up either relying on letters from bigshots – which perpetuates inequalities in hiring and promotion – or just focusing on things like grant dollars.

3 Likes
25

“Peer review support?” I’ve peer reviewed dozens of papers and never received support of any kind. As a member of editorial boards for academic journals, the most I ever got out of it was an occasional free meal/hotel room if the board met face to face.

I’ve reviewed untold numbers of papers that should never have been sent out for review, making me wonder if the editor was even skimming them first. I think peer review is incredibly important (if flawed at times) but I have no love for scholarly publishers and their insane prices. And don’t get me started on the scam that is textbook publishing :rage:

4 Likes
26

Big Data

2 Likes
27

I think we mostly agree. There are real costs, but the pubishers are making quite healthy profits and a correction is needed. My post was just to address the question of what costs there are/what benefits do they provide at all.

I also emphasize your comment re: issues of scale. Diamond OA/institution-supported works pretty well for a smaller journal. But one of the ones I’m at gets >200 submissions a day, and requires fairly specialized experts to judge them even to find the right reviewers; we have a 3-tiered editorial system with 40+ people involved. As you note “I can barely run the journal and do my day job”…and I am betting your day job - and your science and teaching - suffer as a result. Is that really fair to your career, that of your grad students, the education of your undergrads? (And am I understanding correctly that you do the copyediiting, not a professional staff? Ouch. That’s heroic of you, but I would argue not a great use of someone’s science PhD.)

Again, I absolutely agree that the journals overstepped in what they are charging for the conveniences they provide and that some correction is needed. On the other hand, the current demonization of them arguably also goes a bit too far /ignores relevant facts, including that people are absolutely free to post their own author copies, even after peer review, and even to advertise the fact that they have gotten it accepted in fancy big name journal X…in fact, our journal even encourages them to do that and helps promote each accepted paper, and its open-access version, via twitter.

As for Science and Nature, as I’m sure you know they are both a hot mess for a whole host of reasons.

1 Like
28

data

1 Like
29

Nor does it justify charging $20k per year for a print subscription to Journal of Chromatography A.

2 Likes
30

Never said it did. Read my last paragraph (before the parentheses), please. Also the end of point #3.

31

My issue with the various “gray market” rogue publishing sites is that they are also not free – they have real, largely unrecognized costs in the form of lost productivity and invisible labor. Every time one of our institutional credentials ends up in SciHub’s database, we lose access to IEEE or other major platform for like a week while we resolve the breach. This means lost productivity for researchers, increased interlibrary loan costs while we ILL articles we already have access to, and hours of thankless labor on the part of our e-resources librarians while they negotiate with the publisher to resolve the breach shutdown.

I’m 100% in favor of better models for scholarly publishing, and part of me respects these sites for pushing publishers’ buttons, but publishers are not the ones paying the practical costs. The organizations that are already on the leveraged end of the deal – the universities and libraries – end up bearing the brunt of it.

1 Like
32

Yeah, I’ve basically given up on Nature. Their editors lately seem more interested in publishing articles that make impressive New York Times Science Section stories than ones that have methodological rigor.

1 Like
33

The thing I’m hating about Nature these days is the brand dillution. It seems like they introduce a new title every other month. Nature: Little Polkadot Dresses. Now there is nothing wrong with specialized journals, but by prefixing the title with the word Nature, they make them look more impressive than they are, like you got published in the journal Nature. It is just monitiziing their reputation until it is gone IMHO.

It’s not that editing and distribution are not worth paying for. But at a time when the costs of doing so should be going down; (We’re not paying people to hand set equations, and in this age, there is little reason for a print edition of a specialized journal) the prices charged are going up. Yes, many of the journals are getting bigger and publishing more papers, but I’m not convinced that the quality of the papers is improving, just that the bar is being lowered. IMHO that is because Impact Factor is more highly affected by the ability to publish “superstar” high-citation papers than it is by the ability to sort out and not publish lower quality papers. Since editors and reviewers are imperfect, there is pressure to publish more just in case. Of course the ease of distribution these days also means that editors have to wade through more submissions than used to be the case.

3 Likes
34

#aaronslaw

35

The interesting thing is that the scientific publishers could have done something similar and developed their own cloud services for improved search and analysis if they were such spoiled, lazy parasites.

1 Like
36

Data Cowboy

1 Like
37

I’m surprised a journal like that can find enough good referees for all the papers.

(And am I understanding correctly that you do the copyediiting, not a professional staff? Ouch. That’s heroic of you, but I would argue not a great use of someone’s science PhD.)

It gets easier with time; nowdays I can do a pretty good job very quickly. However, it is has been a real education into the lack of effort some authors put into their submissions. The often-expressed idea that simply writing your paper in LaTeX with a suitable style class constitutes the only significant fraction of a journal’s typesetting effort is uninformed.

Can you elaborate on this? Are you saying that if a publisher learns that someone on your campus has accessed SciHub then you get cut off, or are you saying that if one of your researcher’s papers gets into SciHub then you get cut off? Neither makes any sense, of course, but I’m curious which is the thing you’re seeing? The main issue our librarians have with these databases is that they want us to actually use any resources they pay for in a measurable way, to keep justifying the payment.

Yes; at first I thought these were knockoffs by someone else. The papers in them are often not of high quality.

Yes, many of the journals are getting bigger and publishing more papers, but I’m not convinced that the quality of the papers is improving, just that the bar is being lowered.

In my field it has become harder to publish in the top journals, because they haven’t expanded and the pressure to publish is higher than it ever was (so more competition). Moreover, the venerable journals tend to publish relatively few papers in emergent or interdisciplinary areas, and since that is the direction hiring has moved (often pushed by granting agencies, also by administration against the desires of departments), for many young faculty publishing in these top journals has become nearly impossible.

38

Me too. And my field is smaller – I’ve never heard of editors actually being paid. But maybe they are.

The journal from my society (which I got as part of dues) went from paper to electronic only. Well, that’s OK I guess–except I would usually browse the whole thing because 1) I wanted to keep up with my field, 2) you never know what will be relevant – some new technique, some reversal of accepted science and so on. So I had to log on, go to the title of contents, and download each of them if I wanted to browse them – and indeed print them all out if I wanted to continue my habit of skimming them in the, uh, individual gender-specific reading room.

Not if you’re a US government employee. But so what? Can scientists really sell reprinted articles anywhere?

Yep. But what justifies Elsevier’s profit margin of 37.1%?

2 Likes
39

Editorial board = / = editor in most cases. Peer review support usually means materials and workshops on how to do reviews, not monetary support. If you google the term “peer review support” several examples will pop up. Likewise midrange ‘scientific journal editor salary’ around $50-$60k. That seems congruent with what I’ve seen advertised for editor in chief positions in my field. Whether and how much probably does depend quite a lot on the field, the journal, etc. eLife says they spend 33% of their budget on editor salaries.

Owning the rights: It’s not about selling the reprinted articles, it’s about the freedom to put out your ‘author copy’ without worrying about the journal paywall. I don’t know what the rules are for gov. employee scientists there; I suppose the govt. owns it. As for the IP, which is sell-able, that you have to work out with your institution. Most of the time universities want some kind of cut, or at least credit, but will let you have a pretty large chunk of the royalties or whatever if you sell the IP to a company or start your own. The rationale is that you developed it with their salary and institutional support (lab space, etc.)

Elsevier profit margin: Please re-read my original post. Explicitly said that was not justified.

As I also have said, the publishers certainly have taken things too far. However, a lot of people are making claims that aren’t in line with the facts, such as “editors don’t get paid”, or “all they do is put it behind a paywall”. As scientists, we should do better.

40

If IEEE (or another vendor, but I hear about it most from IEEE) detects that unusual traffic is coming through our subscription, they cut us off and file a breach notice. “Cut us off” in this case means they deny access to the offending IP address – which in this case is usually our proxy server. Then no one off-campus can access IEEE – which is a lot of our users.

The cutoff persists until we can dig into our logs, figure out what user account was accessing the proxy and causing the unusual traffic, block the offending account, notify the user and campus IT that their credentials have been compromised, and file a breach report with the vendor. This is hours of work. A typical breach instance lasts about a week, meaning lots of pissed off users (not to mention pissed off librarians and vendor reps).

Many (but not all) such breaches stem from SciHub and other similar sites. One of SciHub’s MOs is to request your login credentials for subscription resources, then it starts spidering. Once a certain threshold of traffic is detected by the vendor, they start knocking on the library’s door.

Librarians don’t get kickbacks for purchasing database subscriptions or anything. Collection management, especially in subscription resources, is intensely use-driven. At the same time, we have to satisfy the needs and demands of faculty by providing access to the key research and tools. Yes, to that end, we have to “market” our resources in order to communicate to faculty and other users that we have the things they demand and to ensure they get sufficient use to justify the cost. And when a resource gets low use, we cancel it. And then we hear it from the faculty. [1]

And then we have to contend with the double-digit inflation rate on journal and database subscriptions. Trust me, no library wants to be choosing between funding collections increases and improving facilities or giving raises (or to be forced to slash monograph budgets to cover serials increases) but that’s what happens in a lot of cases because of the cost structure of journals and databases.

Librarians want a free (as in speech) and equitable approach to scholarly publishing as much as – probably more than – anyone else in the academy, which is why so many academic libraries have digital scholarship librarians that are experts in open access issues. But SciHub and its ilk make our actual day-to-day lives more difficult and cost us (and our patrons) money, time, and access.

[1] This is not intended as a screed against faculty, but it’s hard to balance the tensions between faculty needs and limited budgets. And one key way it manifests is requests for new resources and complaints when resources get canceled.