Publisher Proxy Deposit Is A Potential Trojan Horse: I

Ann Okerson: "If your publishing organization is providing for your authors the service of deposit of their articles according to various mandates, particularly NIH (beginning on 4/7) could you kindly describe the nature or extent of these services"

Paul Gherman: "At Vanderbilt, our Medical Library has been doing significant work contacting publishers to find out what their policy and procedures are. One discovery is that some of them intend to charge authors between $900 and $3,000 to submit articles to NIH. Some will allow for early posting, if the fee is paid."

I suggest not colluding with publishers offering to "Let us do the [mandated] deposit for you".

The reason is simple, if we take the moment to think it through:

(1) The OA movement's goal is to provide Open Access (OA) to 100% of the world's peer-reviewed research article output.

(2) The goal of [some] publishers is to preserve the status quo -- or the closest approximation to it -- for as long as possible, at all costs.

(3) The providers of both the research and the peer review, in all disciplines, are the employees of the universities (and research institutions) worldwide (and the fundees of the funded research).

(4) The only way to cover all of OA's target space is for all research output, from all universities worldwide, funded and unfunded, across all disciplines, to be made OA.

(5) OA self-archiving mandates from universities and funders (39 so far, worldwide) can ensure that all research is made OA.

(6) Some funders (e.g. NIH) -- but no universities (e.g. Harvard) -- have mandated direct university-external self-archiving (in PubMed Central) instead of direct university-internal self-archiving (and subsequent central harvesting). This was an unnecessary strategic mistake.

(7) University-external, subject-based self-archiving does not scale up to cover all of OA output space: it is divergent, divisive, arbitrary, incoherent and unnecessary.

(8) The way to scale up systematically to capture all of OA output is for both funders and universities to mandate that all research output, in all disciplines, from all universities worldwide, funded and unfunded, should be self-archived in each university's own Institutional Repository (IR). (The deposits, or their metadata, can then be externally harvested into whatever subject-based, disciplinary, or multidisciplinary central collections we may desire.)

(9) The universities are the research providers; the universities are the co-beneficiaries of showcasing, archiving, auditing, assessing and maximizing the visibility, usage and impact of (all) their own research output, funded and unfunded, across all disciplines.

(10) The universities are also in the natural and optimal position to monitor and reward their own employees' compliance with both university and funder self-archiving mandates.

(11) It would hence systematically undermine the scaling and convergence of OA self-archiving mandates onto university IRs to transfer responsibility for compliance to an external party -- the publisher as their employees' proxy self-archiver -- depositing in arbitrary and divergent external repositories.

(12) Universities and funders should universally mandate self-archiving directly in each author's own university's IR; they should say "no, thank you" to offers of proxy self-archiving on behalf of their employees from publishers. External collections can then be harvested, as desired, from the IRs that will then cover 100% of OA output.

[Similar considerations, but on a much lesser scale, militate against the strategy of universities out-sourcing the creation and management of their IRs and self-archiving policies to external contractors: accounting, archiving, record-keeping and asset management should surely be kept under direct local control by universities. There's nothing so complicated or daunting about self-archiving and IRs as to require resorting to an external service. (More tentatively, I am also sceptical that library proxy self-archiving rather than direct author self-archiving is a wise choice in the long run -- though it is definitely a useful option as a start-up supplement, if coupled with a mandate, and has been successfully implemented in several cases, including QUT and CERN.)]

Stevan Harnad
American Scientist Open Access Forum