Maximising research access vs. minimizing copy-editing errors

[Stevan Harnad <harnad@ecs.soton.ac.uk>]

On Wed, 5 Jul 2006, Anthony Watkinson wrote:

> I suppose Professor Harnad thinks that if he constantly 
> promulgates the idea (see below) that the only difference 
> between the accepted paper and the final published version is a 
> matter of formatting he will get those not involved in 
> publishing to accept this as a "fact". In fact there is 
> something called "copyediting". There are some publishers who 
> do very little copy-editing or even none at all. However many 
> publishers, especially those who have important journals, do a 
> lot of copy-editing which is not just a matter of house style 
> but can pick up serious errors. The difference between the 
> versions can be significant and this difference is (I 
> understand) being recognised by the current NISO groups working 
> on version. Journal editors of course know this very well too.

The trouble is that Anthony Watkinson and I are addressing two 
completely different problems, hence two completely different 
user populations.

Mr. Watkinson is thinking of the user who has a subscription to 
the journal, with its copy-edited, proofed PDF, and is weighing 
the use of this against the use of the author's final, accepted 
draft -- revised and accepted, but not copy-edited. He is quite 
right that the copy-edited version is to be preferred: I too 
would prefer it, if I had access to it.

But the problem I -- and the OA movement -- are addressing is not 
that one at all. We are concerned with the population of would-be 
users who cannot, today, access the journal version, because it 
is not in one of the journals they or their institutions can 
afford to subscribe to. And the choice *they* are facing is 
access to the author's final, refereed, accepted (but not 
copy-edited) draft, versus no access at all. I very much doubt 
that all those would-be users would be very appreciative of Mr. 
Watkinson's concern to protect them from access to the author's 
final draft on the grounds of potential errors that might arise 
from the lack of copy-editing.

I think Mr. Watkinson may have both the immediate needs of 
researchers and the immediate motivation for Open Access rather 
out of focus and proportion if he imagines that his very 
legitimate scholarly concern to minimize all errors that a 
copy-editor might catch carries any weight at all in the context 
of the overarching research concern that would-be users should 
not continue to be denied access to the final, refereed drafts of 
research findings.

And if Mr. Watkinson is curious about the size and scope of this 
would-be user population, and of the research access problem that 
the OA movement is addressing (compared to the copy-editing 
error-risk problem that he is addressing), a good estimate is 
provided by the 25%-250% higher citation impact of research for 
which the author supplements access to the journal version by 
self-archiving his final draft in his institutional repository. 
That's quite a dramatic difference, but I expect it will prove to 
be even bigger, once we have not only citation data, but also 
usage (download) data comparing self-archived and 
non-self-archived articles (in the same journal and year).

If anyone has any comparative data on the research impact of 
undetected copy-editing errors, I would be very happy to see 
it...

Stevan Harnad
American Scientist Open Access Forum
http://amsci-forum.amsci.org/archives/American-Scientist-Open-Access-Forum.html