Re: The July 6-7 NYAM "Freedom of Information" Meeting

From: Mark Doyle <doyle_at_APS.ORG>
Date: Tue, 11 Jul 2000 14:31:16 -0400


> From:
> Date: 2000-07-11 09:57:23 -0400

> I think you did not quite get the point I was trying to make. Just to set the
> record straight for those who did not attend:

Hmmm, in reading your e-mail here, I don't think I missed the point at all :^).

> 1. I made it very clear in the beginning of my talk that I was speaking 'a titre
> personnel', and not on behalf of any organisation.

I don't think I implied anything else. Michael Jacobson may have
implied such a thing though (and it was certainly how you came across
despite your disclaiimer - after all, you closed with an invitation
for PMC "to join us [CrossRef]".)

> 2. My main point in comparing CrossRef and PubMedCentral was in terms of how to
> reach the, what I believe(d) to be, common objectives of both intiatives: a.
> "Every research worker should have access to the refereed scientific (journal)
> literature of his/her choice/level' b. "All facilities of the Internet should be
> available to him/her, so as to ensure maximal efficiency and effectiveness of
> his/her research efforts".

There is no way that a) is fullfilled by CrossRef and I don't believe
CrossRef addresses this issue at all. This is what I find disingenuous
about your line of reasoning.

[reordering your text....]

> All CrossRef does is lead the end users to the Website of the article of their
> choice: it is up to the publisher of the journal in which that article is
> contained to determined the terms of access, i.e., for free or for fee.

Exactly! And that is why CrossRef has nothing to say about a) while
PubMedCentral is conceived precisely to address a). There can be no
more complete access than if it were all free.

> As the
> core (biomedical) literature gets subscribed to (or licensed) by practically all
> academic libraries, access is for free for most academic end users.

First of all, it was pretty clear that from Varmus's talk that not
all researchers have access. But more importantly is at what cost does
this access come? The reality of the serials crisis can't be denied.
Every day librarians make choices about what access to cut off. You
may say this doesn't happen to the core of the literature, but I would
argue that then a very large percentage of the literature is non-core
and it is truly extravagant to make the non-core material available
in the traditional channels (perhaps as part of large consortia deals
- with evey bottle of Glenlivet, you have to take a case of Mad Dog
20/20). I also believe that the core literature isn't being made
available to all researchers who could benefit from the access.

PubMedCentral is an attempt to make this entire problem go away by
changing the model. The real question is whether this forces the baby
out with the bath water. Can high quality QC and archiving be done
outside of the subscription model? This is an experimental question
and Varmus was very persuasive about arguing for the need for

> Those
> biomedical publishers who choose to make their journal files available for
> free, and who, as a result, would typically participate in the PubMedCentral
> initiative, can do so within CrossRef as well. Moreover, their articles do not
> only remain interlinked with those that are not free for all, they will, in
> future, also be linked with the non-primary literature (such as textbooks,
> (major) reference works, (numerical) databases etc.), a part of the literature
> that nobody has (yet) suggested should be given away for free.

This is conveniently ignoring the fact that a journal who wants to
make their content available for free needs a channel to make it
available. Without something like PubMedCentral they have to pay for
the electronic infrastructure to distribute the journal. With PMC,
they can focus just on providing QC. Certainly CrossRef isn't such a

> My argument was (and is), that it will be a long time (if ever) before the PMC
> initiative reaches a critical size of , say, 20% of the biomedical articles
> published.

Probably true, but mostly because of entrenched publisher interests.
But even the APS wouldn't be able to immediately comply with a request
to put all of Phys. Rev. contents into something like PMC - we have
no other model for paying for peer-review or for converting
information to long-term archival formats. What the existence of PMC
does, however, is allow one to explore how to move beyond the current
subscription model to provide alternative ways to pay for these
necessities. It is clear that the new electronic way of doing things
enables one to decouple distribution from peer-review. Furthermore,
such a decoupling means that the costs of distribution can be greatly
reduced (cf. Ginsparg). The main distinction between the
publisher-centric view and the Varmus/Ginspargian is precisely this

> Until that time, research workers are deprived of even the most
> elementary 'facilities of the Internet' (such as reference linking) for a
> critical part of their literature. Meanwhile, the CrossRef initiative, which
> assumes that publishers keep their articles on their own Websites and only
> requires them to make available their articles' metadata in a 'central
> facility', has quickly reached more than such a critical size, thus allowing
> reasearch workers access to a large (virtual) file within the next few months.

I am not saying that CrossRef isn't needed (though I do think
CrossRef's goals can be met in other ways without such heavy
"mainframe" centralization, but that is another argument). I am saying
that CrossRef in and of itself does nothing to address point a) and
since a) is the raison d'etre for PubMedCentral, it isn't reasonable
to compare the two as if one can substitute for the other (and this
doesn't even talk about non-journal literature that PMC would like to
include). To imply that PMC (or the length of the timescale for PMC
achieving criticality) is "depriving" researchers of anything is
really over the top. There is no doubt that CrossRef will help
publishers link articles together and researchers will thus benefit.
But this doesn't expand access or redistribute costs for granting
access (the costs for distribution would also be greatly reduced in
the centralized PMC model compared to the present system). Nor, at
least for the biomed literature, does CrossRef come close to providing
the kind of tool that PubMed itself is (and this does provide an
alternative to CrossRef for linking together the biomed literature -
the strength of CrossRef is that it extends beyond just the biomed

> The conclusion is, therefore, that the PubMedCentral solution (i.e. hosting the
> journals centrally) for reaching the above mentioned objectives IS CONTAINED IN
> the CrossRef solution, thus making PMC superfluous.

I am sorry, but this just doesn't follow for the reasons stated above.
CrossRef is a convenience - it does nothing to improve true access.

> As an added thought (not
> expressed at the meeting): PMC seems to competing more with Highwire than anyone
> else - a number of the PMC journals are on Highwire - why have them both places?

Good question. Highwire happens to be a rather expensive way to
distribute journals. I'll leave it to the readers to draw the inference.


Mark Doyle
Manager, Product Development
The American Physical Society
Received on Mon Jan 24 2000 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:47 GMT