From: Sandy Thatcher <[log in to unmask]>
Date: Sat, 11 Feb 2012 08:35:47 -0600

In the special issue of Against the Grain devoted to this topic, "The
Challenges of Bibliographic Control and Scholarly Integrity in an
online World of Multiple Versions of Journal Articles" (April 2011),
which I co-edited with NISO's Todd Carpenter (and to which Stevan
Harnad was a contributor), we dealt with these issues in depth. I
think it is safe to say that there is wide consensus that the Green OA
version is fine for many purposes (including use in classroom
discussions), but that the "version of record" is always best to cite
in any formal writing. One reason among several one could advance for
this practice is that Green OA versions often contain errors in
citations themselves (as a number of volunteer copyeditors discovered
in examining articles posted to Harvard's DASH repository, as I
reported in my essay in the special ATG issue) that are corrected in
copyediting, and thus quoting from OA versions can perpetuate errors.

Sandy Thatcher



At 7:18 PM -0500 2/10/12, LIBLICENSE wrote:

> From: "Kunda, Sue" <[log in to unmask]>
> Date: Fri, 10 Feb 2012 08:49:48 -0800
>
> Yes, I'd like to see more information about this as it's an issue I've
> struggled with as an IR manager for some time now. When we deposit an
> author's post-print into the IR we add a citation based on the
> publisher's version, which goes against what I was taught. It seems to
> me we should create a citation for the actual deposited version. If
> the user then wants to cite the version in the IR as if they've read
> the "version of record" that's up to them (although, personally, I
> would contact the author for the version of record if I decided I was
> going to cite the research in an article and didn't have access to the
> journal).
>
> Sue
>
> ------------------------------------------------------------------------
> Sue Kunda | Digital Scholarship Librarian | Assistant Professor
> Oregon State University - Valley Library |121 The Valley Library |
> Corvallis OR  97331
>
>
> From: LIBLICENSE <[log in to unmask]>
> Subject: Re: Refereed Draft - definitions
> From: "Pikas, Christina K." <[log in to unmask]>
> Date: Thu, 9 Feb 2012 09:49:18 -0500
>
> I was always trained to cite the version consulted but I discovered in
> the Physics community, even if the submitted article cites ArXiv,
> several of the publishers will change the citations over to the
> journal citations in the final editing process. I learned of this from
> a physics blogger (was it Chad Orzel?). It seems a bit sketchy to me
> and it also seems to deflate the citations to ArXiv.  I have not done
> any study to see how prevalent this is.
>
> Christina
>
> ----
> Christina K Pikas
> Librarian
> The Johns Hopkins University Applied Physics Laboratory
> [log in to unmask]
>
> -----Original Message-----
> From: LibLicense-L Discussion Forum
> From: Brian Harrington <[log in to unmask]>
>
> On 02/07/2012 06:48 PM, LIBLICENSE wrote:
>
> From: Stevan Harnad<[log in to unmask]>
> Date: Mon, 6 Feb 2012 22:37:01 -0500
>
> (Contrary to the intuitions of many well-meaning librarians, the
> difference between access denial and open access to the author's
> refereed final draft is the difference between night and day for
> researchers:http://bit.ly/OAnite  It is a great strategic mistake to
> insist on the version-of-record, or to worry that the author's
> refereed final draft is somehow not "good enough". It is infinitely
> preferable to no access: But what is cited is of course always the
> archival version-of-record. The OA version is merely the version
> accessed.
>
> Maybe I'm just pedantic, but shouldn't the version accessed be the
> version cited?  Admittedly, in the humanities fields I'm most familiar
> with, citation is often accompanied by quotation, or at least
> references to page numbers, so the need to consult the
> version-of-record seems self-evident.  But even granting different
> citation practices in other fields, the idea of citing something that
> you haven't actually seen strikes me as going too far.  If the
> refereed final draft is the only version that the researcher needs to
> consult, why not cite it?  This seems especially true if the draft is
> the version that the reader is more likely to have access to.
>
> Brian
>
> Brian Harrington
> Content Development Coordinator
> Project MUSE
> The Johns Hopkins University Press
> [log in to unmask]