[Open-access] Collections of Libre material

[Peter Murray-Rust]

On Mon, Feb 13, 2012 at 6:40 PM, Douglas Carnall
<dougie.carnall at gmail.com>wrote:

> >> Especially for scientists access to complete articles and data
> >> is compulsory, but I guess that for "laymen" illustrative pictures and
> >> abstracts would be sufficient.
>
> >I always get nervous when I see this sort of scientist/layman
> >distinction, and I think we should work to eradicate such a boundary
> >as much as possible.  (I was a layman myself until a few years ago,
> >and would have hated to be fed a watered-down version of research
> >while an elite priesthood of scientists got the Real Stuff.
>
> I'd like to reinforce this point. As a translator and editor I very
> often deal with unfamiliar topics and need to get up to speed quickly
> with the language and jargon typical in a field. It is a major
> frustration in my work that the most authoritative work is locked up
> behind paywalls. Typically I need to briefly access one key term in a
> handful of articles to understand how it is used in the field. As the
> prevailing rate for technical translation is around $0.12-0.20/word,
> accessing 3 or 4 articles at $30 each to check a single term is
> completely unfeasible. But that would be the best way to ensure high
> quality. I find paywalls vexing precisely because dumbed down
> popularizations are useless to me.
>

This is a brilliant example of how people don't realise the different uses
to which articles can be put.  What percentage of a domain do translators
need? For example if we got 10% of all papers is that likely to be enough.

Another similar requirement is my own field of computational linguistics.
To train machines to interpret text you need a marked up corpus. For that
you absolutely have to have BOAI material - reading free through a paywall
is useless. It needs to be redistributable

>
> The point more generally is that neither the author nor the publisher
> can possibly conceive of all the potential ways that a scholarly work
> might be useful when it is freely available. If the scholarly
> literature could be treated as one vast linguistic corpus, I am sure
> that interesting developments in scientific communication,
> terminology, and translation would follow, for example.
>

Exactly right. I think we should try to collect as many different types of
example as possible.

And of course translation is a commercial activity so CC-NC is useless.

P.

>
>
-- 
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-access/attachments/20120213/dafd6960/attachment.html>