Re: Nature 10 September on Public Archiving

From: Mark Doyle <doyle_at_APS.ORG>
Date: Wed, 16 Sep 1998 15:42:26 -0400


On Wed, 16 Sep 1998, "Albert Henderson <>" wrote:

> Savings at the production level are one thing. Real savings are another.

> I think one should take into account the cost of this transition to
> research and its underwriters. There has been a tremendous investment
> in new technology and systems over recent decades. (No one has any idea
> how much money has been invested in technology that obsolesces every 3
> years or so.) It still, as Arthur Smith writes above, is no more than
> "close." Outside the U. S. -- where 2/3 science articles originate --
> they are not so automated and enthusiastic.

You seem to be ignoring why people invest in the new technology. Perhaps it
is because they find it immensely useful in getting their work done?
Researchers invest in technology for many reasons and in the past electronic
publishing hasn't been a primary one. A side benefit has been that they
find they are getting increased access to information and better
communications with their colleagues in distant places. The cost of
technology for accessing xxx is extremely minimal, less than subscribing to
a single year many journals. Increasing coverage at xxx makes the return on
such a minimal investment even higher. Establishing and maintaining a print
library that contains (not to mention grows with) the same information as
xxx has little, if any, return on investment in this way.

> "Forcing/educating authors into better practices" is more than a decade
> away.

Nonsense. We do it all the time. We get REVTeX from well over 60% of our
submitters. We are improving authoring tools to help authors improve their
practices. For those that lag, we provide services. The issue here is
whether such services should be covered by subscriptions. From where I sit,
the answer is no, especially if there is a low-cost investment that a
researcher can make that will not only help them publish in traditional
journal, but also communicate with new colleagues and access repositories
like xxx.

> In spite of the buzz we hear constantly, the effect of ejournals to
> date in terms of citation studies has been minimal. (S.P. Harter. 1998.
> Scholarly Communication and Electronic Journals. JASIS: Journal of the
> American Society for Information Science. 49(6):507-516).

Irrelevant as Harnad already pointed out.

> Unlike the electronic-only journal, most publications must run two
> distributions. Many readers lack the infrastructure required to
> participate in epublishing.

A decreasing number and probably not a majority... The readers that don't
have the infrastructure, only have minimal access to the traditional
literature anyway. The barriers for getting such an infrastructure are
falling. For instance, it is much cheaper now to provide phone service in a
developing nation than it was to develop it in the US. You don't have to
replicate all the telephone poles and wires and switches (which represent an
enormous investment). Instead, you either provide well-placed microwave
towers or use a satellite at a fraction of the cost of the US

When I worked on xxx at Los Alamos, we regularly received e-mail from users
all over the world who extolled their increased access to the literature.

> Associations like AAAS, with members who
> prefer their paper copy, are particularly hard pressed to cut out
> print.

'Prefer' is not 'require'. Anyway, no one is saying print has to be cut
out. The issue is whether print or electronic is the primary deliverable.
Current publishing is heavily slanted towards print production so electronic
production is fit in as an ad hoc added cost. But this isn't the optimal
solution. Having two distribution methods does not require two production
processes. Rather, you can quite effectively use a single production process
geared towards electronic distribution with printing as a single extra step
for which publishers can charge users who want it.

> So epublishing while appearing to provide economies actually
> becomes an extra burden that demands extra investments in technology,
> human resources, etc.

Only because of the way it has evolved as an ad hoc add-on to the current
print-oriented processes. But, yes, of course there are investments
required, but these investments, if done correctly, can have extremely high

> There are more pressing problems flowing from the poor productivity of
> research. Even Newt Gingrich complains about the poor dissemination and
> synthesis of scientific results. The taxpayer is not getting his/her
> money's worth. The researcher is insulated from important information.
> Going electronic offers no solution to this.

How does increasing library subscriptions to paper journals increase
productivity? Having more paper in your library doesn't make it any easier
to find the articles that are relevant to your research. Going electronic
means more effective searching, faster browsing, automated filtering with
notifications, more open communications, better access, more time for doing

> British Library consultant David J Brown indicates the new media will
> not survive if the economic market is not large enough. (ELECTRONIC
> PUBLISHING AND LIBRARIES. London: Bowker-Saur 1996)

So? That doesn't mean that the market isn't large enough. I think it is far
more likely that publishers who don't evolve to take advantage of the new
media won't survive. If the market changes, it is up to the publishers to
respond to the change in the market, even if that includes streamlining
production so that expenses are more in line with what the market will

> He also points out
> that library growth has not been sufficient to absorb the growth of
> research. I would add, nor is library growth able to disseminate what
> it cannot absorb. In my eyes this is due to universities diverting
> library finances to support administrative growth.

> A solution to the problems of dissemination would be easily derived by
> financial support of library growth that would keep up with research.

A sub-optimal solution you mean. Another solution is to change the process
so that library growth isn't required. Such solutions exist.

Mark Doyle
APS Research and Development
Received on Tue Aug 25 1998 - 19:17:43 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:27 GMT