Re: Proposed update of BOAI definition of OA: Immediate and Permanent

From: Leslie Chan <chan_at_UTSC.UTORONTO.CA>
Date: Mon, 14 Mar 2005 09:32:48 -0500

On 3/13/05 10:39 PM, "Stevan Harnad" <harnad_at_ECS.SOTON.AC.UK> wrote:

> On Sun, 13 Mar 2005, Leslie Chan wrote:
>
>> I am against updating the BOAI definition of OA because I think the
>> current definition is more than adequate
>
> Leslie, do you think it is more than adequate that (1) 12-month
> delayed-access (NIH Back Access) is being offered in place of immediate
> access (and that if the BOAI definition of OA is not updated, this
> could even be offered in the *name* of OA)

Dear Stevan,
I don't think NIH Back Access is adequate, but I don't think that NIH's
decision was really predicated on BOAI definition of OA so updating the
definition will not change NIH's stance. The problem with NIH's approach is
not simply the extended delay, but the fact that it doesn't make
self-archiving mandatory, rather a recommendation to authors. However this
should not prevent NIH-funded authors from self-archiving, since most of the
journals they publish in are "green" already. So, as you have always pointed
out, the problem is with authors own inaction. The real important question
then is, how would the proposed update of BOAI definition affect author's
self-archiving behaviour. If adding "immediate and permanent" would hasten
author self-archiving, then I would be in favour. But I can't see how.

>(2) that this is being
> used as a pretext for publisher Back-Sliding on Green (e.g., Nature),
> (3) that these policies now risk being copied and cloned, and (4) that
> this this could lock in Back Access in place of (and in the name of)
> Open Access for many years to come?
>
Again, our concerns should be with hastening author self-archiving, not
publishers' PR maneuvers. Publishers will do whatever they do to protect
their revenue streams, and I don't suppose the BOAI definition would matter
much to them. No matter how air-tight the definition, it will always be
subjected to misuse and abuse. It comes with the territory, much like
democracy itself.

> And would you think it adequate if funders were to seek -- and publishers
> to offer -- free access for a fixed period, subsequently withdrawn, in
> the name of OA (as they could, technically, based on the current BOAI
> definition of OA)?

What funders in their right mind would do such a thing, and to what end?


>> the proposed update is needlessly confusing...
>> No one knows what "permanent" means in the digital realm,
>> so why add that level of uncertainly.
>
> Would it be confusing to say instead that free online access must be immediate
> (upon acceptance for publication) and must not subsequently be withdrawn?
> That captures the practical objective without asking for any clairvoyance or
> omniscience on the subject of failsafe digital preservation.

I am weary of of putting prescription ("must be") in the definition. The
current definition is descriptive, not prescriptive. I think you are adding
this out of concern for publishers action. But this put unnecessary burden
on authors, which is the opposite of your intention. What if an author want
to put his old papers on his institutional archive? Since they were not
"immediately" available, they would not be OA by the new definition.

Also, as a faculty member, I have no guarantee that the server on which my
research reside and even our institutional repository will be kept in
perpetuity by our university. In fact I have seen my fair share of servers
come and gone and publications "withdrawn" as a result, through no fault of
the author. In my mind, the articles that are no longer available under such
circumstance are still OA, though unavailable, because it is the author's
intent to make them OA.

>> The current definition is sufficiently clear and sufficiently flexible to
>> allow a broad range of approaches to OA
>
> But the current definition is also sufficiently flexible so that I could
> claim I publish an OA journal (gold) if it makes its articles freely
> accessible online 24 months after publication for one day (only)!

This is the exact opposite of the hypothetical example I provided above.
Clearly there is no intention to provide OA in this case, only false
pretence, and should not be worthy of our concern.

> I could also claim I publish an OA-friendly green journal if I give
> my authors the green light to self-archive their articles for one day
> (only), 24 months after publication.
>
> If we are not ready to allow that an article to which you can have access
> only if you (or your institution) pay an access-toll is either OA or
> "partially OA" (or more OA the lower the access-price), then, by exactly
> the same token, we should not be ready to allow that an article to which
> you can have free access for only one day is either OA or "partially OA"
> (or more OA the sooner or longer it is accessible free).
>
> To allow either would immediately introduce a potential slippery slope that
> would reduce "OA" to an absurdity.

Any publisher that attempts the silly stunt that you described will lose all
credibility, not reduce OA to absurdity.


>> Let's agree that there are multiple flavours of OA (to use John
>> Willinsky's term) and that there are no one-size-fits all solution for all
>> circumstances.
>
> But there *aren't* multiple flavours of OA: There are multiple *roads*
> to OA (mainly the golden road of OA publishing and the green road
> of OA self-archiving, plus combinations and variants thereof).

There are OA articles in plain vanilla ASCII, some in html, some in pdf,
some with richly tagged XML, some with rich metadata, others without. Within
each road there clearly are different variants or flavours, such as
institutional archive and discipline based archive.

> An article that is freely accessible online comes in only one flavour:
> *freely accessible online*. But if it is obvious that an article that
> is freely accessible online only for one day (or only a century after
> publication) is no more OA than an article that is freely accessible to
> me because my institution has paid the access-tolls, then it should be
> obvious why immediacy and permanence need to be part of the definition
> of OA.

Mixing OA articles that are not "permanent" with articles that are
toll-access is not helpful. Again, I can only be certain that my articles
are accessible as long as the server on which they reside is accessible, but
I have no control of the latter. We have to differentiate author's intention
from publisher's intention. Trying to come to a definition that curtails
publishers would unnecessarily demote authors who are providing OA,
particularly those who are not doing so immediately and permanently.

> One-day free access is no more a mere variant "flavor" of OA that merely
> differs from the "flavor" of two-day free access than two-dollar toll
> access is a different flavor of OA from one-dollar toll access: None
> of these are OA, and our definition of OA should make this quite explicit
> by specifying that OA must be immediate and permanent.
>
> This does not imply that $1 toll-access is not preferable to $2 toll-access,
> or 2-day free access to 1-day free access. It just makes it clear none of
> these is OA!
> Stevan Harnad

Toll-access is not OA, that is obvious. Non-immediate and non-permanent
should not make OA non-OA. So I would repeat what I said earlier. If the
expanded definition would indeed hasten author self-archiving, then it is
worthwhile. But I suspect the reverse may be the case, and indeed for
authors for developing countries, the proposed addition is unrealistic.

I just noticed Barbara Kirsop's reply to this tread. I fully concur with her
that we should be concentrating on filling OA archives and encouraging OA
journals instead of debating the "luxury" add-ons.

Leslie
Received on Mon Mar 14 2005 - 14:32:48 GMT