Re: Journal publishing and author self-archiving: Peaceful Co-Existence and Fruitful Collaboration

[Stevan Harnad <harnad@ecs.soton.ac.uk>]

On Mon, 18 Dec 2006, Sandy Thatcher wrote:

> I'm afraid I don't share your "serene confidence that there are 
> plenty of available OA hosts, big and small, ready to take on 
> the implementation of peer review for migrating established 
> journal titles and ed-boards, scaled down to OA publishing."

That's fine. It's all speculation anyway, on both of our sides: 
speculation that self-archiving will or won't lead to 
cancellations, and if so, speculation about when, and how much; 
and speculation that, if much and sudden, current publishers will 
or won't jettison their titles rather than downsize; and 
speculation that, if jettisoned, there will or won't be OA 
publishers happy to take over the titles.

What's sure, because already tested and demonstrated, is that 
self-archiving is highly beneficial to research and readily 
feasible, right now, through mandated self-archiving. Hence 
self-archiving can and should and will be mandated at this time. 
The data-free speculation and counterspeculation about its 
possible eventual effects on publishing has been going on for 
over 10 years now, so the data-based practical step is already 
well overdue.

One point, though, is a point of logic rather than of 
hypothetical conjecture: In your reasoning about your 
hypothetical scenario that you consider the most probable one 
(catastrophic cancellations, abandonment of journals by their 
non-OA publishers, and failure of the abandoned journals to 
migrate to OA publishers because OA costs could not be met and 
there were not enough would-be OA publishers able or willing to 
meet the demand) you have inadvertently conflated two very 
different factors: One is the current cost to universities of 
hosting their journals' editors' offices, and the other is the OA 
publication cost to universities for their own research article 
output.

These are two entirely different things. Journal hosting costs 
have nothing to do with OA, or OA publishing. Whatever journal 
hosting universities are doing today, in the non-OA era, for 
non-OA journals, while paying journal subscriptions for whatever 
journals they subscribe to, the only change in the OA era, if 
subscriptions were indeed all cancelled suddenly, as you 
hypothesize, would be (1) sudden, substantial windfall savings 
for universities, and (2) sudden, substantially lower publishing 
costs for journals (because, ex hypothesi, they no longer sell 
texts, paper or online, but only perform peer review).

Those lower publishing costs would (again, ex hypothesi) be paid 
in the form of OA publishing charges, for each university's 
article output, out of each university's subscription savings. 
This has nothing at all to do with a university's journal hosting 
costs!

(Perhaps what you were doing was conflating the university as a 
journal subscriber, the university as a research article-provider 
[with its associated OA publishing costs] and the university as a 
potential OA publisher itself!  None of this, except possibly the 
last, has anything to do with the free resources many 
universities currently provide for hosting the journals -- OA or 
[mostly] non-OA -- of publishers other than themselves!)

> Partly I don't [share your confidence in successful migration 
> if mandated self-archiving were to induce sudden cancellation] 
> because I think, to work most efficiently, there needs to be 
> more structure to the system than self-archiving or IRs 
> themselves can provide, even with pretty good federated 
> searching.

There is more structure. Indeed the *only* requisite structure: 
each of those hypothetically migrating journal titles has an 
established editorial board, referees, authors and readers; they 
all migrate with the title. So does the implementation of peer 
review, which is the same for all journals: only the contents and 
the peer-review quality standards differ. All that structure 
remains with the migrating title.

Search and federation have nothing to do with it (and will be 
incomparably more powerful in the OA era).

> The editors of single journals would need to find a way to join 
> together with editors of other journals in their disciplines, 
> or related disciplines, so as to form a group of journals that 
> could serve a whole discipline, or special area of interest, 
> well.

I think these are hypotheses again, and actually I disagree: a 
field is best served by mostly independent journals: No need to 
amalgamate except when it helps increase efficiency and cut 
costs. (And cost-saving amalgamation has nothign to do with 
federated search!)

> That is typically what scholarly societies have done, and maybe 
> some of them could take over the journals abandoned by large 
> STM publishers-if they don't continue to feel just as 
> threatened by OA as the commercial publishers do!

What OA publishers would take on after the hypothetical collapse 
of user-institution subscriptions as the means of covering costs 
is the OA publishing cost-recovery model: author-institution 
publication charges.

There would be more than enough to cover these much-reduced costs 
for peer-review service-provision only, out of the (hypothesised) 
windfall institutional subscription savings.

> An ideal structure would be something like what CIAO and 
> AnthroSource represent, respectively, for International 
> Relations and Anthropology in the social sciences, which 
> encompass not only journals but also monographs, working 
> papers, conference proceedings, and grey literature. As 
> director of a press that worked with our library and SPARC to 
> help set up such a structure for another social science 
> discipline, rural sociology, I can tell you that this is no 
> trivial or inexpensive task!

And most of it not necessary for hypothetically migrating 
journal-titles only. The author-institution publication charges 
will cover the much-reduced costs. (You are here combining 
journal-specific OA and OA self-archiving matters with completely 
independent non-OA matters.)

> Sudden change is very difficult to plan for, and my worry is 
> that if such a scenario were to happen, no really adequate 
> structures would be in place save for a few like the ones I've 
> mentioned to provide for an organized environment of knowledge.

But we are not talking about actual sudden change, but about 
hypothetical sudden change, compounded by hypothetical 
unwillingness of OA publishers to take over migrating titles. 
None of those contingencies are based on any evidence at all: 
just speculation.

Sandy, what I think research needs now is actual OA, through the 
self-archiving mandates that have already been demonstrated to 
generate OA and its benefits to research. Hypothesizing and 
counter-hypothesizing alas does not solve research's immediate 
access and impact problem.

I actually emerged (temporarily) from a long-standing, 
self-imposed moratorium on speculation about hypothetical 
after-effects of self-archiving, in honor of your accession to 
the AAUP presidency, Sandy! All of these conjectures and 
counter-conjectures have been made before, many, many, many times 
across the long years. I deliberately stopped engaging in these 
speculations because I noticed that they were slowing the 
progress of OA: people were speculating instead of 
self-archiving.

Now that we have, I think, aired our respective conjectures, I 
think we can agree that there is no way to know which ones are 
correct, and that all we can do, and advocate, is what we think 
is most probable and useful, on the evidence available.

> Possibly, yes, some individual editors would immediately try to 
> keep their journals going by setting up their own 
> self-publishing OA operations.

No, I don't mean editors, self-publishing (though that is a 
possibility). Editors usually prefer to just edit -- i.e., select 
referees, adjudicate referee reports, decide what revisions need 
to be done, and when they have been successfully done, and to 
accept or reject articles accordingly. They don't want to handle 
the administrative part -- contacting authors and referees, 
reminding, tracking deadlines, etc. The publisher needs to 
arrange for that, and, as you mention, there can sometimes be 
economies of scale from doing this for multiple journals at once.

> But who would pay for the editorial support services that the 
> major STM publishers now provide?

The author-institution OA publishing charges, paid for their own 
article output, (out of their own user-institution subscription 
savings).

> Departmental budgets can be stretched only so far, and these 
> might be tapped already for supporting their own authors 
> publishing in other OA journals.

Sandy, I am afraid you seem to be double-counting here...

> (This is part of the "free rider" problem that university 
> presses have long suffered from, because they do not publish 
> for their own university faculty primarily but provide a 
> service to the system as a whole. Universities like to fund 
> their own faculty first, their presses second, and the same 
> would likely be true for editorial offices of journals.)

We are not talking here, particularly, about university 
publishing. We are talking about journal publishing, university 
article output, and how to cover the costs of publication in the 
OA era if/when there is a sudden collapse of subscriptions, and a 
migration of suddenly abandoned journal titles.

An OA journal "virtual" editorial office that only provides peer 
review, does not generate text, either hard copy or digital, does 
not store, does not provide access, etc. is a much lighter 
proposition than the traditional journal office...

> Academic editors would need to spend more of their time doing 
> the kind of work that professional publishing staff now do, at 
> a cost to the university that would overall be greater (because 
> faculty are paid, generally, better than professional 
> publishing staff).

Not at all. That is what the OA publishing charges would be 
paying for (the administration of peer review); and in the 
virtual world, there is no reason those administrative functions 
should be performed at the editor's university unless so desired. 
An OA publisher could do it centrally for multiple journals, as 
BMC, Hindawi and PLoS do (except scaled down a good deal more, to 
peer-review alone).

> Universities would do well to start creating these structures 
> now, but I don't see that as likely to happen because most 
> administrators, I suspect, share your view of gradual change 
> and will think there is plenty of time to prepare.

It's not universities that need to scope this out in advance, but 
publishers (current ones and OA publishers). (You are conflating 
the role of the university as a journal subscriber, as a research 
article provider and peer-review service consumer, and as a 
potential OA publisher!)

> Sure, library funds once used for purchasing STM journals could 
> be diverted, but this is not so straightforward a process as 
> you seem to assume, as many libraries now share the burden of 
> subscription payments whereas I suspect that the distribution 
> of editorial offices will be more highly concentrated in the 
> most-research intensive universities where the leading scholars 
> reside-and I can't see Ball State contributing its savings from 
> library subscriptions to supporting Yale faculty's editorial 
> offices!

Sandy, you have the Escher drawing misperceived! The redirection 
of an institution's windfall subscription savings (on the 
hypothesis of catastrophic cancellations) is toward paying the 
publication charges for that same institution's outgoing research 
article output (for peer review and certification). It is not 
redirected in order to subsidise OA journals hosted by that 
university. OA journal costs are to be covered out of the 
publication charges, relying no more (nor less) on university 
charity (hosting) than before.

You are assuming that the OA publisher to which titles -- 
hypothetically released by their prior non-OA publishers because 
of hypothetical catastrophic cancellations and unwillingness to 
downsize to a more modest OA niche -- will migrate will be 
universities:

Why? OA publishers will be publishers; some of them may be 
university publishers, but that is a different operation (and 
budget) from a university qua university. And others will be 
learned societies, others will be commercial, and still others 
will be independent non-profits. Nothing to do with redirecting 
university budgets or even university windfall cancellation 
savings.

> We at Penn State are doing our small bit by serving as a test 
> site for the DPubs "open source" software that is designed to 
> provide a platform for managing the editorial and production 
> processes not only for journals but also for conference 
> proceedings and, ultimately, edited volumes and monographs in 
> electronic form. But there should be many other efforts like 
> this going on if we are to avoid a very messy transition period 
> if my hypothesised scenario of sudden change comes true.

There are; and there will be, if need be.

Best wishes,  Stevan