On Fri, 25 May 2007, Martin J. Osborne wrote:

> Regarding "fair use": is it your understanding that an author can email 
> *the publisher's pdf file* of a published paper to a person who asks for 
> a copy?  (I looked through the list archives, but didn't see an answer.) 
>   My assumption is that she can, given that it is just the electronic 
> analogue of mailing out a xerox of a published paper.
> 
> (Feel free to reply on the list if you think it is of interest.)
> 
> Martin J. Osborne
> Department of Economics
> University of Toronto
> http://www.economics.utoronto.ca

Yes, for simply emailing eprints, it makes no difference whether the author
emails a copy of the publisher's PDF or the author's accepted final draft
(postprint).

It also makes no difference which version is sent via the Institutional
Repository's "Fair Use" Button, in the case of articles that have been
deposited as Closed Access instead of OA because of publisher access
embargoes.

But where it makes a huge difference is in institutional and funder
self-archiving (Green OA) *mandates*. The default version that
should be mandated for deposit is the author's final draft, *not* the
publisher's PDF. The reason is that the author's final draft has far
fewer restrictions imposed on it. (In other words, far more publishers
endorse author self-archiving of the publisher's final draft, and far
more publishers endorse immediate, unembargoed setting of access to the
deposit as OA rather than Closed Access.)

So, if the publisher does happen to fromally endorse immediate,
unembargoed self-archiving of the publisher's proprietary PDF, it's
fine to self-archive that. But the default version that mandates should
specify for all other cases is the publisher's final draft.

By the way, the difference between the publisher's PDF and the author's final
draft means next to nothing to those would-be users who currently have no
access at all. Hence it would be absurd to keep on depriving them of access 
in order to hold out for a difference that makes no difference.

It would in principle be possible to deposit both the author's final
draft *and* the publisher's PDF, the latter always in Closed Access, and,
whenever a user requests an eprint via the Fair Use Button, always to
send the PDF rather than the author's postprint. I would say that at a
point in time when 85% of articles are not being deposited at all, any
which way, and most institutions and funders have yet to adopt deposit
mandates, this would be an example of a needless overcomplication,
discouraging rather than accelerating progress. Both authors and
their institutions and funders do best to forget about depositing the
publisher's PDF at all, except in the specific cases where it has been
endorsed by the publisher for immediate OA (and the author prefers to
do so).

I not only prefer to deposit my final draft, but in addition to depositing
it, I sometimes also deposit postpublication updates and corrections
(clearly tagged as such!) of the published version, which would in
any case supersede the PDF.

    "Optimizing OA Self-Archiving Mandates: 
     What? Where? When? Why? How?"
     http://openaccess.eprints.org/index.php?/archives/136-guid.html

Stevan Harnad