[At the bottom of D. Goodman's posting is S. Harnad's reply.]
Stevan Harnad wrote:
> But as to the kind of server on which it is self-archived: If it is
> accessible webwide (rather than just to the university's internal
> users) this is indeed a full-blooded green light! (The distinctions
> between "individual" and "institutional" servers are empty and
> incoherent nonsense and should be ignored and omitted from the
> SHERPA/Romeo list of conditions.)]
The distinction between an individual server and an institutional one
is real, if you are concerned about availability tomorrow, as well as
today. A server maintained by an individual (or even a research group)
depends entirely upon the individual for stability. The individual may
move universities -- the papers will still be available, but they may
take some finding. The individual may retire or die -- someone may keep
up his server for a while, but not necessarily. The individual may move
from the academic community and return to industry and no longer care
about his papers. Many individuals may prove weak links in attempts to
maintain consistent metadata or OAI harvesting.
A server maintained by even a university department is more
stable. Departments have administrators, and most science departments
have information technology staff. Departments may merge or split, but
the continuity is maintained. Of course this is still not a stable as a
university server or a disciplinary society based server or a national
library based server, but it's a different order of magnitude.
The fact that publishers make these distinctions indicates that they are
aware of this. Any system relying upon individual scientists is fragile,
and requires a stable parallel system, such as those that publishers
produce. LOCKSS may develop into such a system, for example.
David Goodman
-------------
Stevan Harnad's reply to David Goodman::
On Tue, 6 Apr 2004, David Goodman wrote:
> Stevan Harnad wrote:
>
> > But as to the kind of server on which it is self-archived: If it is
> > accessible webwide (rather than just to the university's internal
> > users) this is indeed a full-blooded green light! (The distinctions
> > between "individual" and "institutional" servers are empty and
> > incoherent nonsense and should be ignored and omitted from the
> > SHERPA/Romeo list of conditions.)]
>
> The distinction between an individual server and an institutional one
> is real, if you are concerned about availability tomorrow, as well as
> today.
The SHERPA/Romeo list is about publishers' self-archiving policies. It
is not about server-technology or data-preservation. It is not part of
publisher self-archiving policy to stipulate how well self-archived
material is to be maintained! To repeat: The distinction between an
individual and an institutional server -- from the publisher's standpoint
-- is empty and incoherent nonsense. I quote from the sample institutional
self-archiving policy in the Eprints Handbook:
"Copyright agreements may state that eprints can be archived on your
personal homepage. As far as publishers are concerned, the EPrint
Archive is a part of the Department's infrastructure for your personal
homepage."
http://software.eprints.org/handbook/departments.php
In other words, the distinction between an individual and an institutional
server is a difference in how one's institution sectors and labels a
disk! I would urge that we not spend our time on such pseudo-issues
while there is so much OA still to be provided! Nor should we conflate
it with preservation issues (and pseudo-issues) that have absolutely
nothing to do with the SHERPA/Romeo listing, which is merely cataloguing
which publishers (or journals) have or have not given the green light
to author self-archiving (of the postprint, the preprint, or both) so
far.
> A server maintained by an individual (or even a research group)
> depends entirely upon the individual for stability. The individual may
> move universities -- the papers will still be available, but they may
> take some finding. The individual may retire or die -- someone may keep
> up his server for a while, but not necessarily. The individual may move
> from the academic community and return to industry and no longer care
> about his papers. Many individuals may prove weak links in attempts to
> maintain consistent metadata or OAI harvesting.
All true, and all utterly irrelevant to the question of which publishers
do or don't give their green light to author self-archiving! Please
see the many past AmSci threads
http://makeashorterlink.com/?Q594324F7
on the preservation non-problem for the OA (Open Access) *version* of a TA
(Toll Access) journal article self-archived by its author -- as long as TA
journals continue to be the primary locus of articles! The self-archived
version is just an extra, a bonus, provided for those would-be users whose
institutions cannot afford the toll-access to the primary (TA) version.
The problem, today, for these (still mostly nonexistent) OA versions of
TA articles is not preserving them but providing them! OA versions are
supplements, not substitutes.
> A server maintained by even a university department is more
> stable. Departments have administrators, and most science departments
> have information technology staff. Departments may merge or split, but
> the continuity is maintained. Of course this is still not a stable as a
> university server or a disciplinary society based server or a national
> library based server, but it's a different order of magnitude.
All worthy considerations, insofar as the maintenance of university
servers is concerned, but completely irrelevant to the issue at hand.
> The fact that publishers make these distinctions indicates that they are
> aware of this. Any system relying upon individual scientists is fragile,
> and requires a stable parallel system, such as those that publishers
> produce. LOCKSS may develop into such a system, for example.
There may be many things of which people (publishers or otherwise)
are aware or unaware. But what we are discussing here is merely
which publishers have or have not already given their green light
to self-archiving. I may be aware that females are more reliable
self-archivers than males, but it cannot be my official policy -- or
if it is, it is merely asking to be laughed at and ignored -- that only
females (or only males) may self-archive!
Stevan Harnad
Received on Tue Apr 06 2004 - 21:25:09 BST