Open Access Archivangelism

Guenther Eysenbach suggests that what should be mandated (by institututions and funders) is not Green OA self-archiving but Gold OA publishing, because it would be more efficient and less costly. I explain below why neither of these is the case now (but that Green OA mandates may indeed pave the way for an eventual transition to Gold OA that will be less costly and more efficient).

Parallel Publishing?
    (1) There is no parallel publishing with Green OA self-archiving.
    (2) Publishing (whether Gold OA or non-OA) is publishing.
    (3) Green OA is access provision (to published articles), not publishing (of already-published articles).
    (4) A posted, unpublished document is not listed in an academic CV as "published," and a published article that is also posted (or emailed, or photocopied) is not listed as two publications.
    (5) Users cite the published article, not the access-source; where available, they also list the URL(s) for access purposes.

Society pays twice?
    (1) Who pays twice for what?
    (2) Tax-payers pay to fund research.
    (3) Institutions pay to subscribe to the journals in which the research is published.
    (4) No institution can afford to subscribe to all (refereed) journals; most can afford only a few.
    (5) So OA is needed so that all users can access all articles.
    (6) Green OA self-archiving supplements what is currently accessible to subscribers, by making it accessible to all would-be users, webwide.
    (7) So who is paying for what, twice, with Green OA?
    (8) Even with Gold OA today, no one would be paying twice.
    (9) But hybrid "open choice" (optional gold) publishers would be paid twice -- once from institutional subscription money, and once from research grant money.
    (10) If and when universal Green OA causes universal subscription cancellations, all journals can downsize to Gold OA, paid for by redirecting part of the windfall institutional cancellation savings rather than by redirecting scarce research funds from research.

Mandate gold OA and "pay once"?
    (1) No one is paying twice.
    (2) Institutions and funders cannot mandate publishers' choice of cost-recovery model.
    (3) Institutions and funders cannot mandate authors' choice of journals.
    (4) Until and unless subscriptions are cancelled, Gold OA requires extra funds, usually diverted from research.
    (5) Green OA can provide immediate 100% OA.
    (6) It just needs to be mandated by institutions and funders.

Pay for Gold OA by "Topping up" research grants?
    (1) Topping up from what funds?
    (2) Research is already underfunded and research funds are scarce.
    (3) Redirecting research funds to pay needlessly for Gold OA publishing today just makes research funds scarcer, needlessly, because providing Green OA costs nothing, and subscriptions are still paying the costs of publishing.

Canada's CIHR Mandate without "infrastructure"?
    (1) What infrastucture?
    (2) Researchers whose institutions already have an OAI-compliant Institutional Repository (IR) can deposit there.
    (3) For researchers that do not yet have an IR there are many back-up central repositories in which they can deposit, such as DEPOT, CogPrints, or Arxiv, all OAI-interoperable.
    (4) What is lacking is deposits, not repositories to deposit in -- and the mandates will cure that.
    (5) There is no need for a "Canada PubMed Central" as a locus of direct deposit.
    (6) Central repositories can harvest from Institutional Repositories, through OAI interoperability.

Self-archiving is not free?
    (1) Not free to whom?
    (2) And to whom does it cost what?
    (3) Institutions create repositories (for many reasons: record-keeping, performance assessment, access-provision, visibility).
    (4) The cost to the institution per paper deposited is a few dollars.
    (5) The cost to the researcher per paper deposited is a few keystrokes.

Can subscription savings pay for repository costs?
    (1) Institutional repository costs per paper deposited are negligibly small.
    (2) Central repository costs are up to those who think central repositories are needed, over and above OAI-interoperable institutional repositories.
    (3) None of this has anything to do with publication costs or Gold OA fees.
    (4) Today Gold OA fees per paper are not negligibly small, and they divert scarce funds from research.
    (5) If and when Green OA causes subscriptions to become unsustainable, journals will cut costs by abandoning the paper edition and PDF-generation, offloading access-provision and archiving onto the distributed institutional repository network, downsizing to peer review alone, and converting to the Gold OA cost-recovery model -- paid for, per paper, by institutions' annual windfall savings from having cancelled journal subscriptions.

How much does it cost to run a Repository?
    (1) Per deposited paper, next to nothing.
    (2) And institutions derive many benefits from their IRs, having nothing to do with journal subscription costs one way or the other.

Stevan Harnad
American Scientist Open Access Forum

Stan Katz has blogged a summary of the OA session at the Association of American University Presses (AAUP) Meeting in Montreal in the the Chronicle of Higher Education Review. Here are six little clarifications:

(1) SK: "Open Access [OA] [means that] all scholarly articles ought to be mounted on free public access websites maintained by their universities.

" This brief definition is fine as a 1st cut.

(2) SK: "[T]he obligation [i.e., recent mandates like Harvard's] to “publish” by mounting articles on free websites will make it impossible for nonprofit presses and learned societies to sustain themselves."

The mandates are for authors to mount (i.e., self-archive) the final, refereed draft of their published article, not to "publish" it by mounting it. This self-archived "Green OA" draft is a supplement, provided for all users whose institutions cannot afford access to the publisher's version. It is not itself another publication.

The publisher might be nonprofit or commercial; that is not the relevant question. The relevant question is whether or not the supplementary OA draft causes cancellations, rendering journal subscriptions unsustainable for covering costs. So far, in the few fields where OA self-archiving has been taking place the longest (15 years) and most extensively (100%) -- e.g. high-energy physics, published by the American Physical Society (APS), the Institute of Physics (IOP) and Reed-Elsevier -- the publishers report that they find no detectable subscription cancellations associated with self-archiving:

However, if and when OA self-archiving ever does cause catastrophic cancellations, making subscriptions unsustainable, then, and only then, journals can (a) offload all their former access-provision and archiving functions, along with their costs, onto the distributed network of institutional repositories, (b) downsize to peer review alone, and (c) convert to the "Gold OA" publication-cost-recovery model, charging the author-institution, per outgoing paper, for peer review and certification instead of charging the user-institution, per incoming journal, for access. The institutions will (on the very same hypothesis, of catastrophic cancellations) have more than enough annual windfall subscription cancellation savings out of which to cover those charges.

(3) SK: "Harnad’s suggestion is that the universities transfer the payments they are currently making to their academic presses to subsidize peer review and archiving of their faculty scholarly output."

No, my suggestion is only that universities should mandate self-archiving. Then, if and when the resulting universal OA should ever make subscriptions become unsustainable, the universities' subscription savings will, by the same token, be freed to pay for the university's peer-review costs. That is no subsidy: It is direct payment for a service, out of the very same funds formerly used to purchase a product.

(4) SK: "[U]niversities increasingly expect their presses to be self-sustaining economically, and are unlikely to put up the necessary funding"

Extra funding by whom for whom for for what? While universities are subscribing to journals, that pays for peer review. If and when journal subscriptions collapse, the university savings will pay for the peer review. (And again this has nothing to do with university presses in particular.)

(5) SK: "[N]ot all scholars (and nonprofit publishers) are connected to universities. Who will subvene their publications?"

Unaffiliated scholars are rare enough, and the per-paper costs of peer review alone are low enough, so that a small surcharge on the charges of the affiliated authors (the vast majority) will take care of these outliers.

And as noted, neither whether publishers are nonprofit or commercial publishers, nor whether they are university or learned-society publishers, is relevant to any of this.

(6) SK: "[OA has] different... implications...for the humanities and social sciences"

OA pertains to refereed journal articles publication in all disciplines. Humanities and social sciences are not exceptions in any way. All research, in all scholarly and scientific disciplines, benefits from maximizing its uptake, usage, applications and impact by eliminating the access-barriers that have been made obsolete and unnecessary by the advent of the PostGutenberg Galaxy.

Stevan Harnad
American Scientist Open Access Forum