[open-science] Norms and flexible understandings question

Wed Feb 11 16:48:11 UTC 2009

2009/2/10 Heather Morrison <heatherm at eln.bc.ca>:
> The idea of norms rather than licenses does have some appeal.

Just to be clear: I take this to mean vis-a-vis stuff like share-alike
(I think we are all agreed now that we need a 'license' whatever you
do -- i.e. some formal statement regarding the material (even if that
is PD dedication))

> Question:  can a normative approach accomodate things like flexible
> understandings of what is permitted, and allow for evolution of such
> understandings?

While I understand the attractiveness of such flexibility I am
concerned it starts to blur into fuzziness. One of the major reasons
for making stuff open is to remove the permissions barrier: I can use
stuff without having to go and consult with its producer.

Once norms become fuzzy either we start requiring significant efforts
from users or we end up with the norms being ignored -- and neither
outcome seems attractive.

> Flexible understandings = stuff like friendly licenses.  To go back
> to Dorothea's example, if one researcher shares data freely while a
> second researcher cannot share for privacy reasons, there are
> probably a few understandings that the first researcher would think
> of as friendly to sharealike.  For example, if the second researcher
> shares her data privately with the first researcher, and publishes
> her results OA immediately on acceptance for publication, the first
> researcher may well be satisfied that this is Sharealike.

I think this example is an interesting one (and I believe John
Wilbanks has raised it before in discussions -- though not in this
thread).

I take your point here about flexibility but I do think both ways: in
the example you give the user of open data (i.e. the first researcher)
is having to do quite a lot of thinking regarding what they are
supposed to do to comply with the 'spirit' of the norms. In a world in
which data sharing (we hope) is going to happen *a lot* I think this
could get a bit burdensome.

On a point more directly related to the original example we should
also be clear about what share-alike means here.

If the first researcher is just using the data 'internally' to do
calculations and then publishing a paper using those calculations
share-alike isn't going to come into it: they are not producing *and*
distributing a derivative 'database' (of course attribution still
'binds' ...). Thus, sharealike will only come into it if you are
making a 'derived' database and publicly distributing *that*.

Regards,

Rufus