Re: Definition of Open Access

From: Stevan Harnad <>
Date: Fri, 30 Jun 2006 15:05:40 +0100

On Fri, 30 Jun 2006, Matthew Cockerill wrote:

> Don't forget ramblers:

Good one!

> Different people mean different things by open access.
> e.g. For some, rights of reuse are a key part of the open access concept.
> For others they are not. I would say that neither group is wrong,
> they are just using the term differently.

So far, fine, and very ecumenical. But you agree, don't you, that if
people are advocating, say, nonsmoking, for people's health, it makes
a rather big difference whether by "nonsmoking" they mean "not having
an ignited cigarette at all" (NS1) vs "not inhaling what they puff from
their own ignited cigarette" (NS2)! The consequences, say, for a campaign
for a public smoking ban would be rather radical if the proponents of
NS2 argued that an NS2-ban would suffice (and that the rest is just a
democratic definitional matter: "different stokes for different folks"!).

I would not just say they were using the same term differently, but that the
DS2 folks were simply *wrong* (or, alternatively, missing the point)
about what nonsmoking is and means.

> I guess you could always footnote it as "open access" in the BOAI
> sense. But still, as with the American constitution, the same wording
> can be understood very differently by different people, so you're still
> not out of the semantic thicket. Which is why I would say that rather
> than discussing the meaning of the term "open access" and what is and
> isn't required for it, I think it's better to focus on practical goals.

But practical goals hinge on what we mean by OA! My practical goal
(since well before the term "OA" was coined or defined) has been to
make the full texts of all journal articles accessible toll-free, to
anyone and everyone, on the web, immediately and permanently. To reach
that goal, all that has to be done is -- to make the full texts of all
journal articles accessible toll-free, to anyone and everyone, on the web,
immediately and permanently (and that's all I ever meant by "OA").

Our disagreement, Matt (as well as my disagreement with your predecessor,
Jan Velterop) has always been about whether there is any need for or
benefit from adding the extra constraint of specifying reuse/republication
rights as part of what it means to provide OA. I have nothing against
specifying and providing reuse/republication rights, over and above making
the full texts of all journal articles accessible toll-free, to anyone
and everyone, on the web, immediately and permanently. I just object to
calling that part of the definition of OA, because by asking for more,
it risks getting less.

But I full acknowledge that I am on thin ice here, historically and
formally, because the (in my view) arrant definition of OA -- specifying
that reuse/republication rights are part of what OA means -- is *in fact
the original BOAI definition of OA* -- and "immediately and permanently"
is not! So I am in fact objecting to the very definition that I co-drafted
and signed off on!

But we need to admit and try to correct our mistakes.

I think it was a mistake to over- (and under-) constrain our definition of
OA in that way, and I have often expressed the wish that we would correct
it, "officially," so to speak, because although the BOAI definition
was very important and historically the first, it was not infallible,
and these were later-discovered flaws.

To see why, please note that if stipulating reuse/republication rights
were indeed part of what is meant by OA, then the OA self-archiving
mandates of the FRPAA, RCUK, and EC would not really be *OA*
self-archiving mandates!

The other side-effect of being too insouciant about the meaning of OA is that
according to the variant definition of OA used by some, OA just means BOAI-2
(OA publishing): BOAI-1 (OA self-archiving) is not OA at all! And, worst of
all, the BOAI definition (which includes the clause about reuse/republication
rights) is in fact in contradiction with its *own* definition of OA, if it tries
to simultaneously define OA as providing reuse/republication rights *and* to
define BOAI-1 (self-archiving) as providing OA!

Fortunately or unfortunately, these are subtleties far too fine-grained to
actually be noticed and acted upon by either pro- or anti-OA camps. The
level of detail and logical rigor and attention has been a good deal
coarser and more superficial than that. But the definitional problem
does keep rearing its head now and again, making mischief every time...

Stevan Harnad
Received on Fri Jun 30 2006 - 16:03:17 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:24 GMT