RE: Versions

["Evan Owens" <epo@ithaka.org>]

A story about versioning:

A month after the first issue of the Astrophysical Journal Letters was
published on the web in the summer of 1995, an author wrote in saying,
here's a revised, improved version of my paper, please take the old one
down and post this one instead.  The reaction at the University of Chicago
Press, the journal office, and the American Astronomical Society was
horror: this has been published, it is the scientific record, we can't
change it.  Over the years, however, I have wondered if perhaps the author
didn't have it right after all. Perhaps that author saw a benefit of the
new medium that we, in replicating the conventional print practice, did
not see as clearly.

I understand the requirement to preserve the integrity of the scientific
record.  I wonder though if there isn't an opportunity for technology
(and publishers) to add value by managing the revision and update
process in such a way that the record is preserved and also gracefully
updated, annotated, extended in ways that are useful to authors and
readers.  

I suspect that because the technology of web and electronic publishing
allows modifications, the pressure to allow such modifications will only
grow over time.  The authors know that web sites can be updated; the best
response will be to have a system for managing those changes and revisions
in a way that protects the record and serves the community interest.  If
the formal publication process doesn't meet this need, it may well appear
in the informal channels in a way that is unmanageable in the long term.

I admit to a certain self-interest here: one of the critical issues in the
archiving of electronic jouranls is understanding how publishers manage
corrections to their content and web sites. There are already multiple
(sometimes conflicting) versions in the formal publication process (HTML,
PDF, SGML/XML); the thought of authors creating additional versions in
MS-Word and releasing those into the community sends shivers up my
archival spine, as it were.  Ignoring for the moment the problems with
MS-Word as an archival file format, the opportunities for textual
divergence are frightening to contemplate.

Evan Owens 

-----Original Message-----
From: Anthony Watkinson [mailto:anthony.watkinson@btopenworld.com] 
Sent: Tuesday, June 15, 2004 5:39 PM
Subject: Versions

I have a question for him and for this list. For Professor Harnad and his
disciplines the authentic form of the message of the scientist is the
postprint (as he has named it) - the form of the message that has
(finally) been accepted for publication. This is the so-called vanilla
form. BMC publish this vanilla form. PLOS and the great majority of other
refereed journals go through another iteration before publishing. It is
called "copy-editing" - an unfortunate term because somehow it makes one
thinks of correcting references and putting in commas. Of course
correcting references is necessary to make linking possible. Putting in
commas irons out ambiguities. There are also (as I am sure PLOS staff)
would agree major work necessary to get the paper in final form after the
peer review has done its work. PLOS would not have gone for this extra
expence if they did not find that their community wanted this extra work.

My question for Jan and others is - what is the authentic form of the
minutes when there are two versions with some sort of external
creditation. There is the version in the institutional repository,
accepted by those who run those repositories, and there is the "published"
version. I would suggest that authors see the published version as the
authentic version (though I have no evidence for that other than what the
authors and editors say to me and to other publishers). This is clearly
not a question for BMC because the versions are identical but what do PLOS
and its supporters think?

Anthony Watkinson