[open-science] Should scientific text be put in the public domain rather than licensed with CC-BY?

[Thomas Kluyver]

On 12 January 2011 17:46, Michael Nielsen <mn at michaelnielsen.org> wrote:

> To play Devil's Advocate: if my database has been produced by data mining
> other people's texts, at what threshold level of use of their texts do I
> start to need to attribute?  How should my software decide?
>

Well, clearly that's open to discussion, but software could naively decide
along the lines of "Am I reproducing a chunk of >n consecutive words", and
no doubt you could improve on that algorithm. But if you are going to
reproduce, say, two paragraphs of written work, you should attribute it
anyway, as per the community norms others have discussed. I don't think
having the community norms backed up by legal measures creates extra
problems for legitimate use.


> - Some unwanted potential outcomes (people are free to rip your work off
>> without
>> giving you any credit)
>>
>
> I don't think this is likely.  My understanding is that much US Government
> research is released into the public domain (indeed, some of mine was, when
> I worked at Los Alamos National Lab).  But I've never heard anyone complain
> that this causes problems with plagiarism.  I'm sure there are cases
> scientists have sued one another before over this kind of thing, but I've
> never personally witnessed an instance.  On the other hand, I've personally
> seen dozens or hundreds of instances where community pressure has been
> brought to bear on issues around atttribution.
>

I agree that it's fairly unlikely, but it is a potential downside, and I'm
still not convinced that there are upsides.

To those discussing the problems of multiple authorship: What about the way
Wikipedia deals with it? According to its terms of service (
http://wikimediafoundation.org/wiki/Terms_of_Use), by submitting, you
license your contribution CC-BY-SA and GFDL, and you agree what attribution
will be considered sufficient (essentially a link to the page). Revision
history is, of course, recorded, so it's possible to see what each person
did, rather than a simple list of names. In fact, they do offer a list of
authors names as an option for attribution, but for obvious reasons, that's
probably almost never used.

Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-science/attachments/20110112/b6bf10ee/attachment-0001.html>