Harnad Comments on APA
Interim Internet Publishing Policy

NOTE: The following post by Dr Steven Harnad appeared on the sci.psychology.research USENET newsgroup on 12 December 1996. It is reproduced here with his permission.

From: harnad@phoenix.Princeton.EDU (Stevan Harnad)
Newsgroups: sci.psychology.research
Subject: Re: APA Discourages Internet Publishing
Date: 12 Dec 1996 01:36:37 GMT
Organization: Princeton University
Lines: 375
Approved: ddege@winternet.com
NNTP-Posting-Host: kauri.vuw.ac.nz
Originator: trills@kauri.vuw.ac.nz

From: amead@s.psych.uiuc.edu (Alan D Mead)
Subject: APA Discourages Internet Publishing
Date: 6 Dec 1996 00:01:20 GMT

The latest APA Monitor has a small article on page 15 discouraging researchers from making their work available on the Internet. This seems to some (i.e., beside myself) to be in contrast to recent precedent allowing the current of circulating prepublication copies of manuscripts to colleagues for comment. I have also seen innumerable instances of papers being "published" as technical reports and then submitted for Journal publication; as far as I know, this practice was also permitted.

The publishers of paper journals are (understandably) worried about their revenue flow if people make their work available for free on the Net. But the reality is that there is a huge conflict of interest here: What's best for the publisher is definitely not best for the author (not to mention the reader).

There is only one way that this conflict of interest can be resolved. That resolution is inevitable, and it is optimal: The much-reduced cost per page for producing a peer-reviewed all-electronic journal is much more sensibly paid for by authors rather than by readers. The cost will be written into research grants as part of the expense of disseminating the results. And the work will be available to all readers for free.

If this solution is not obvious to you, then you need to remind yourself of a few facts about the differences between scientific periodical publishing and trade publishing: Trade authors are selling their texts, hence there is no conflict of interest between themselves and their publishers: Both want to charge for access, and to prevent theft of the product. But the authors of scholarly/scientific articles are not selling their texts: No author makes a penny from their articles in APA journals or in any other learned journal. Their reward is elsewhere: in making their findings known to their fellow specialists and in receiving academic credit and advancement for their work.

The present conflict of interest seems to be coming to a head at the prepublication stage: If authors make their papers available on the Web in the form of the unrefereed preprint, then publishers feel there is a risk that readers will no longer consult the refereed, published version in paper. It think this fear is misplaced. When there is so much unmoderated garbage on the Net, it is highly unlikely that readers will want to spend their limited time in winnowing through it all, looking for what's worth reading and can be trusted. So electronic preprints are not really a direct threat to paper journal revenues.

They are definitely an indirect threat, though, in two ways:

First, because of the conflict of interest, it is highly unlikely that authors will comply very long with the ironic directive to remove their preprint from the public eye on the Net at the moment of PUBLICation of the refereed version. They will instead do the natural thing, which is to withdraw the obsolete unrefereed preprint from their electronic archive and swap the refereed, accepted, final draft in its place. Ask yourself what earthly incentive a scholar/scientist might have for blocking access to his findings precisely at the moment that they are meant to be made public in their final, validated form.

Second, the locus classicus of their work will rapidly become the Net: That is where people will find it first, and that is where they will continue to seek it. There is no contest between the power and scope of having the entire literature at your disposal on your desk whenever you want it versus doing it the old way (ordering a reprint, going to the library, interlibrary loan, or personal subscription).

Invoking copyright to try to force authors to keep their refereed work out of the public eye except for those who pay to view it will simply fail: Copyright only works if the author wishes to protect his work from theft. And let everyone be clear about the fact that "theft" by a reader is a very different sort of thing from theft by an author. The theft of a scholarly/scientific article by a reader is a victimless crime insofar as the author is concerned. If you want proof of this, remind yourself that most authors have been willing to pay good-sized sums to purchase reprints that they then mail to reprint-requesters. Consider the Net as one big repository of authors' free reprints.

So "theft" by a reader is not theft from the author in nontrade publication, and hence "violations" of copyright by readers are not violations at all -- from the author's point of view. We will defer the question of whether it is theft from the publisher for a moment. For the time being let us just recall that there is a profound conflict of interest here, a conflict that was born with this new medium.

The other kind of theft -- another author that steals the text and publishes it as his own -- is definitely not a crime the real author wishes to encourage. The Net, it is true, has made it much easier to plagiarise, and authors will be very eager to ensure their priority and to detect any instances of auctorial theft. A proper system of date-stamping and encryption can ensure priority at least as well as paper publication can, and the theft of text, though easier to accomplish, is also easier to detect on the Net. In any case, this second sense of copyright protection is the author's concern, and it is even the concern of the author of a paper preprint. Authors can presumably make up their own minds whether their articles are likely to be so important that someone else may want to steal them and pass them off for their own.

I will now pass into quote/comment mode:

The article says researchers should consider three points before they make their work accessible on the Internet:

1. Some editors may consider works posted on the Internet as previously published (and thus not generally publishable under APA policy).

It is not the journal editor (who is simply a member of the peer author/reader community) who would object to considering manuscripts that have appeared on the Net as unrefereed preprints, it is the publisher. This is again the conflict of interest I spoke of.

In any case, this would be an unenforceable injunction. At best, it can only force authors to make some cosmetic changes to the title and format of the version they make available electronically, so that it is not obvious that it is the same draft as the submitted version. At worst, authors will openly ignore it, forcing editors to become Net-sleaths, trawling for lookalikes of every submitted manuscript.

Even if detected and confronted, the author could rightly claim that until the final draft is accepted by the referee, prior drafts are simply that: prior drafts, which the author is and always has been free to circulate for comments that might help improve the final version.

2. Articles posted on the Internet may be considered public domain and thus could be "incorporated into some else's work and copyrighted."

This is utter nonsense. Eprints are no more "public-domain" than paper preprints. The Net does make it simpler to steal texts, to be sure, but there are powerful ways to establish priority electronically. And it's enough of a job to get one's papers read and understood, let alone worrying that others have nothing better to do than to claim to have written them.

3. Posting published papers could violate the journal's copyright.

It is not at all clear that it would be: If distributing free reprints of one's own work is not a violation of copyright, neither is distributing free eprints. It would be inadvisable for publishers to be too explicit about forbidding free electronic distribution in their copyright agreement lest they make the deep conflict of interest involved here too apparent, and thereby inadvertently hasten the optimal and the inevitable.

In any case, there is no need to treat this question hypothetically: There has already been a test case: Paul Ginsparg's remarkable Physics Eprint Archive at Los Alamos (xxx.lanl.gov). In six years it has already "captured" almost 75% of the physics literature, and not once has anyone dared to invoke copyright. The momentum is simply too large, and the conflict of interest is too obvious. The APS (American Physical Society) is simply trying to come to terms with this, negotiating a cooperative solution whereby it can continue to provide the quality control and validation (through peer review, editing, and formal acceptance), but the locus classicus for both the unrefereed preprint and its refereed, edited successor has clearly become, and will remain, the Electronic Archive.

My prediction is that as libraries begin to cancel subscriptions, the economic model for cost-recovery will shift from the reader-end subscription or site licence or pay-per-view model to the author-end page charges I mentioned earlier. That is the natural way to pay for the dissemination of scientific/scholarly research findings.

And thus the APA's Publications and Communications board has adopted an interim policy to the effect that

1. researchers should not allow Internet access to their work

This is a hopeless injunction: How will it be enforced? And how will authors be persuaded that it is in any way in their own interests? On the contrary, it will highlight the conflict of interest (and hence the optimal and inevitable solution).

2. that researchers who ignore #1 run the risk of having their work "stolen, altered" and "not eligible for submission"

We have six years of successful public archiving of papers in physics without anything like this. Why should psychology be any different?

3. the APA (which generally owns the copyright to your work after it is published) does not allow the full text of any APA-copyrighted work to be placed on the Internet.

The APA will find a way to allow authors to make their papers available for free on the Net. Trust me. They will also find a way to recover the costs and make a fair profit on the essential service they provide, namely, quality control and validation through peer review (and editing).

The issue of publishability is, perhaps, the most problematic as tenure and professional advancement in academic (and to a lesser extent non-academic) positions depends upon a publication record and non- peer-reviewed publication outlets are seldom given appreciable weight by tenure and hiring committees.

Correct. So everyone should be clear on the fact that unrefereed preprints are not refereed publications. Refereeing (and editing) are medium-independent. There will continue to be a hierarchy of journals, based on the rigour of their peer review and the quality of their authors and articles. This too is medium-independent. Promotion-committees will continue to put greater weight on publications in APA journals than on publications in other journals. It is just that the medium in which people will be accessing and retrieving the refereed literature will change.

Does anyone know how a work can become public domain simply by allowing it to be accessible from the Internet? This seems at odds with the spirit of US copyright law and common sense...

It is not just at odds; it is false.

Does anyone know anything more about the motives and rationale of the editors? I have heard some researchers actually accuse the APA of being frightened of the Internet; allegedly the Internet is a threat to the profitability and perhaps the existence of traditional journals. This seems obviously false to me; am I missing something?

It is not obviously false, but it is only a half-truth: Yes, the electronic medium will force paper journal publishers to restructure themselves: They will have to abandon the trade model (which had always sat uneasily with research publication, where the authors were not interested in selling their words but in advertising their validated work put it crudely) and to adopt a different economic model for cost recovery. There will be considerable scaling down, which will reduce the absolute profit, but there will still be a fair profit to be made from providing the service of peer review, validation and editing to the research community.

For additional information, please visit: http://cogsci.soton.ac.uk/~harnad/intpub.html