Open Access Now: On the Virtues of Not Over-Reaching, Needlessly

Sandy Thatcher (President, American Association of University Presses) wrote in liblicense:

"If peer-reviewed papers are allowed to be posted on IRs and personal web sites, how is this not open access? My understanding is that this is what Stevan Harnad calls Green OA."

Yes indeed, peer-reviewed papers, being deposited and made freely accessible in an IR, is indeed Open Access (OA) -- 100% full-blooded OA (and there's only one OA: Green OA self-archiving and Gold OA publishing are just two different ways to provide OA).

I think I can sort out the confusion (and disagreement):

The original BOAI definition went a little overboard in "defining" OA.

(I am co-culpable, since I was one of the co-drafters and co-signers, but I confess I was inattentive to these details at the time. If I had been thinking more carefully, I could have anticipated the consequences and dissented from the definition before it was co-signed. As it happened, I dissented later.)

The original definition of OA did not just require that the full-text be accessible free for all on the web; it also required author-licensed user re-use and re-publication rights (presumably both online and on-paper).

If I had not been too addle-brained to realize it at the time, it would have been obvious that this could not possibly be the definition of OA, if we were not going to price ourselves right off the market: We had (rightly) claimed that what had made OA possible was the new medium (the Net and the Web), which had for the first time made it possible for the authors of peer-reviewed journal articles (all of them, always, author give-aways, written only for usage and impact, not access-toll income) to give away their give-away work big-time, at last, as they had always wanted to do (and had already been doing for decades, in a much more limited way, in mailing reprints to requesters).

That fundamental new fact would have been turned into just an empty epiphenomenal totem if our punch-line had been that on-paper [i.e., print] re-publication rights are part of OA! For one could just as well have claimed that in the paper era (and no one would have listened, rightly).

Third-party paper re-publication rights have absolutely nothing to do with OA -- directly. OA is an online matter.

Of course, once Green OA prevails, and all peer-reviewed papers are accessible for free, worldwide, 24/7, online, then on-paper re-publication rights will simply become moot, because there will be no use for them!

And that's why I was so obtuse about the original BOAI definition of OA: Because although it was premature to talk about paper re-publication rights, those would obviously come with the OA territory, eventually; so (I must have mused, foolishly) we might as well make them part of our definition of OA from the outset.

Well, no, that was a mistake, and is now one of the (many) reasons the onset of the "outset" is taking so long to set in! Because here we are, over-reaching, needlessly, for rights we don't need (and that will in any case eventually come with the OA territory of their own accord) at the cost of continuing to delay OA by demanding more than we need.

The same is true about the other aspects of this unnecessary and counterproductive over-reaching: Just about all of what some of us are insisting on demanding formally and legalistically already comes with the free online (i.e., OA) territory (and immediately, today, not later, as re-publishing rights will, after OA prevails):

We don't need to license the right to download, store, print-off and (locally) data-mine OA content. That's already part of what it means for a text to be freely accessible online.

Nor do we need to insist on the right to re-publish the text in course-packs for teaching purposes: Distributing just the URL has exactly the same effect!

Re-publishing by a third party (neither the publisher, nor the author, nor the author's employer) is another matter, however, whether online or on paper.

Simple online re-publishing is vacuous, because, again, harvesting the URLs and linking to the OA version in the IR is virtually the same thing (and will hence dissolve this barrier completely soon enough, with no need of advance formalities or legalities).

Third-party re-publishing on-paper (print) will have to wait its turn; it is definitely not part of OA itself -- but, as noted, to the extent that there is any use left for it at all, it will eventually follow too, once all peer-reviewed research is OA, for purely functional reasons.

Finally, Google can and will immediately machine-harvest and data-mine OA content, just as it harvests all other freely accessible online content. It will also reverse-index it and "re-publish" it in bits, online. Other data-mined database providers may not get away with that initially -- but again, once we reach 100% OA this barrier too will dissolve soon thereafter, for purely functional reasons. It is not, however, part of OA itself; and insisting on it now, pre-emptively (and unnecessarily) will simply delay OA, by demanding more than we need -- because more is always harder to get than less.

And (to finish this rant) this is why I have urgently urged a small but critical change in the Harvard-style copyright-retention mandates (which, because they ask for more than necessary, need to add an opt-out clause, which means they end up getting less than necessary): We don't need to mandate copyright retention at this time. It's wonderful and highly desirable to have it if/when you can have it, but it is not necessary for OA. And OA itself will soon bring on all the rest.

All that's necessary for OA is free online access. Indeed -- and I'm afraid I risk getting Sandy's hackles up with this one again, so I will avoid using the controversial phrase "Fair Use" -- mandating immediate full-text deposit, without opt-out, with the option of making access to the full text OA wherever already endorsed by the journal (62%) and "almost OA" (i.e., Closed Access plus the semi-automatic "email eprint request") for the rest will be sufficient to get us to 100% OA not too long thereafter.

Holding out instead for copyright-retention, with an author opt-out option, in order to try to ensure getting 100% OA from the outset, is simply a way to ensure getting nowhere near 100% OA-plus-almost-OA from the outset, because of author opt-outs.

Summary: OA is free online access, and the necessary and sufficient condition for reliably reaching 100% OA soon after the mandate is implemented is to mandate immediate deposit with no opt-out. There is no need for licensed re-use and re-publication rights; copyright retention should just be described as a highly recommended option (which is what a "mandate with an opt-out" amounts to anyway). All the rest will happen of its own accord after 100% immediate-deposit is achieved. (The only obstacle to 100% OA all along has been keystrokes, nothing but keystrokes, for which all that is needed is a keystroke mandate.)

Stevan Harnad
American Scientist Open Access Forum