[Journal-notes] Re: Publish OA if you can - self-archive either way
harnad at ecs.soton.ac.uk
Sun Oct 16 16:27:27 EST 2005
On Sun, 16 Oct 2005, Jan Velterop wrote:
> Dear Stevan,
> Facts, yes. Evidence, however, constitutes facts in support of a
> conclusion. The facts presented do perhaps support, but not the conclusions
> bandied about. Neither your conclusions that self-archiving doesn't harm
*Hasn't* harmed journals to date (after 14 years). That's what the
*evidence* is (or isn't). The rest is predictions and speculations.
> and that OA publishing threatens to slow the adoption of OA,
Who on earth said that?
What I said (and say again) is that equating OA exclusively or even
primarily with OA publishing slows the provision of OA.
> nor Sally's conclusion that OA publishing is not economically viable, are
> supported by facts. The facts presented only support the conclusion that so
> far self-archiving has not materially harmed journals in physics and that
> new OA journals are not likely to be profitable in record time.
With that I agree. (Compare carefully with how you phrased this in the passages
above, with which I cannot agree.)
> No predictive quality, just retrospective. One has to be careful to regard
> facts as predictive evidence. When Louis XVI of France concluded that the
> French people were incapable of killing their king, he saw the fact that
> they had never before killed their king as evidence. The facts in the
> basket below the guillotine strongly supported the alternative conclusion.
> (Puts "content is king" in perspective, doesn't it?).
We are not talking here about a one-off event like killing or not
killing a king. (This sort of argument, when used in philosophy of
science courses, is usually used as a basis for Humean scepticism about
the justifiability of making any predictions at all based on evidence to
date, because the evidence can always prove wrong, as when the chicken --
who thinks, on the basis of all evidence to date, that there will always
be a tomorrow for her -- makes her fatally fallacious extrapolation.)
But I assume Jan is not trying to make a (self-defeating) Humean argument
here against all evidence-based inferences and extrapolations. And a
negative effect on subscriptions is not a sudden, all-or-none, one-off
thing, like a beheading. We have had 14 years of continuous evidence
that there is no correlation between self-archiving and subscription
decline. Things might change, of course; but things might change with
respect to countless other patterns we take to be established until
there is evidence to the contrary.
> Another fact. Some ALPSP societies fear that their journals will be harmed
> by self-archiving mandates. Their fear is not irrational if one accepts
> that there are laws and patterns in economics.
If 14 years of self-archiving (in some physics areas reaching 100% OA
a number of years ago) have not confirmed those "laws and patterns in
economics," then I would conclude that those laws may not be laws after
all (or not pertinent to the anomalous case of the research literature,
written by its authors only for research impact, not royalty revenue,
and given away (sic) to publishers and reprint-requesters alike).
And the only "pattern" here so far is the observed empirical pattern of no
> But even if their fear were
> irrational, I always understood that psychologists are generally of the
> view that bombarding sufferers with facts is not a good way of dealing with
> their fears. Especially not if those facts can easily and rationally be
> dismissed as irrelevant.
Easily, yes; but rationally? I thought the premise here was "even if their fear
> Therefore, it is important to persuade those
> societies that there are ways of regaining control and so allaying their
For you, Jan, the objective may be to allay publishers' fears, even if
irrational. I have my hands full with the irrationality of researchers,
who are so slow to do what is in their *own* best interests (and partly
out of irrational fear of publishers!)...
> By offering authors the choice of publishing with OA, for instance.
Commendable, and desirable. But far, far too slow. And cannot be
accelerated to 100% by a mandate -- whereas OA self-archiving can.
> And by persuading funders that mandating OA can be done when regarding
> publishing as a part of research -- not apart from it -- and thus regarding
> the cost of publishing as integral to the cost of doing research (the
> Wellcome Trust has taken a truly leading stance in this regard).
Wellcome has mandated *OA self-archiving*, not OA publishing. It cannot
mandate OA publishing. No one but the publisher (or the publisher's
subsidiser, if any) can mandate that.
> You don't answer the question 'why not OA'. Just repeating the fact that we
> don't have widespread OA yet and that it has been the question for the last
> 10 years, is not answering why.
The primary reason we don't have 100% OA is because only 15% of authors
are self-archiving, even though 100% could. A secondary reason is that
only about 7% are publishing in OA journals, but that is because only
about 7% of journals are OA journals.
And the solution is to mandate OA self-archiving, which will generate
100% OA. And then there will no longer the question "Why not OA?" because
100% OA will be there.
> Let me attempt an answer. In my view it's a
> combination of -- but not limited to -- the following:
> 1. There is little to suggest that the ego-system we call Academia works
> logically or rationally (would the impact factor be invested with the power
> it has, for instance, if Academia did?);
Publication bean-counting in the pure publish-or-perish days was an
extremely weak 1st approximation to the measuring and rewarding of
research productivity and impact. Then came journal impact factors, a
slightly better approximation. Now we are in the era of measuring and
rewarding authors' own citation impact, not just the average citation
impact of the journals they publish in.
What is irrational is to be rewarding citations while not maximizing
them! But I still have some faith that the facts -- about OA's power
to increase citation impact -- will prevail, if not soon enough with
researchers themselves, then with their rewarders -- their institutions
and funders, the ones in a position to mandate that they maximise their
citation impacts, by self-archiving.
> 2. There are not enough journals yet that offer OA publishing, especially
> not enough with an impact factor;
We agree on that, but there are two ways to go: Try to increase the
number of OA journals, or try to increase the number of articles
self-archived. With increasing the number of OA journals, you are
up against publishers, who have all kinds of fears, and whose main
allegiance is not to research impact but to protecting their revenue
streams from risk. With self-archiving, you are up against researchers,
who also have irrational fears, but whose main allegiance *is* their
research impact. And they can be mandated by their own employers and
funders to maximise the very thing they are already being rewarded for:
their research impact (by self-archiving).
> 3. The practice of incorporating the cost of publishing into the cost of
> doing research is not yet widespread enough;
That's because most publishers are still interested in recovering their
costs the way most of them do now. That's why 93% of them are non-OA
publishers. I commend your efforts to try to entice them by getting
research funders to agree to cover OA publishing costs, but I don't for
a microsecond believe that that's going to bring us much closer to 100%
OA, whereas OA self-archiving can, and will, once mandated.
> 4. Where the practice exists, it is not always made clear enough to
Don't be surprised that researchers (even if they purport to desire OA)
are sluggish in doing what is in their own best interests. They don't
self-archive nearly enough; they don't even publish in OA journals (when
they can) nearly enough.
> 5. There are not enough repositories for self-archiving;
We can always use more repositories, but it is *certain* that that is
not the real problem, because the many repositories we *do* have are
nearly empty! What's missing is the self-archiving mandate (to fill them).
> 6. Where there are, researchers are not always convinced that self-
> archiving doesn't harm journals (after all, many of them also accept that
> there are laws in economics);
I highly doubt that the lion's share of the 85% self-archiving that is *not*
being done today is not being done because researchers, like publishers, are
worrying about the fate of their journals! Most self-archiving is not being done
because of pure inertia: researchers are sluggish, set in their ways, already
have more than enough to do, and (as they have *told* us, in the Swan & Brown
surveys) they won't self-archive till their employers and/or funders mandate it
-- but then 95% of them will do it, 14% grumblingly, 81% willingly. And the five
institutions that already have self-archiving mandates (CERN, Southampton,
Minho, QUT, and soon Zurich) are demonstrating that mandates work,
as the surveys indicated.
Let us not forget that "publish or perish" is itself a mandate, without
which the sluggish researchers would not be giving the bean-counters *any*
beans to count...
> 7. Last but not least: OA advocates have lost their united approach which
> was so wonderfully captured in "Publish with OA if you can - self-archive
> if you can't".
The unity was never there: I was the only one who kept flashing that
unified message (in AmSci) from 2001-2003, while you and just about
everyone else except Peter Suber were conducting a unilateral "gold
rush" in which OA was equated exclusively with OA publishing and the only
mention of archiving was as a form of storage of OA journal content! (Have
a look at the Bethesda Statement and the original Berlin Declaration.)
But the unified approach is still there: See the latest Berlin Declaration:
The only difference is that the unified statement (and the right one)
is "Publish OA whenever you can -- but self-archive either way"...
> You asked me some specific questions:
> >Are you, Jan, interested in (1) making common cause with the
> >publishers who are trying to fend off self-archiving mandates on
> >the grounds
> >of the worries, with no evidence to support them, or are you
> >interesting in
> >(2) encouraging self-archiving mandates, for the sake of reaching
> >100% OA as soon
> >as possible (after all these years)?
> I hope you will allow me to ignore the bush-ite nature of this binary
> question ("if you're not for me you must be against me"), and answer it
> with the option you don't give me, option (3): I am in favour of mandating
> OA, by publishing with OA where possible and self-archiving where it isn't.
Since mandating OA publishing is not an option, a straightforward, unhedged way
of putting it is the way the last Berlin Declaration meeting put it:
"In order to implement the Berlin Declaration institutions should
implement a policy to:
1. require their researchers to deposit a copy of all their
published articles in an open access repository
2. encourage their researchers to publish their research articles
in open access journals where a suitable journal exists (and
provide the support to enable that to happen)."
Notice that it is necessarily the self-archiving (only) that is being
*mandated*, whereas the OA publishing is being encouraged (and supported).
> Your second question:
> >Which side are you on, Jan? *For* RCUK mandating
> >self-archiving, or against it?
> Is already answered, but for the sake of avoidance of any doubt: I am in
> favour of RCUK mandating that their grantees provide OA, by publishing with
> OA where possible and self-archiving where it isn't.
Although it's a bit fuzzy on what is (and can be) mandated and what
cannot, and although there needs to be self-archiving in *both* cases
(when publishing in either the 7% OA journals or the 93% non-OA journals),
that's close enough.
I'll remind you, though, the next time you say something that seems to
contradict this apparent commitment: "The best defense against the dangers
[sic] of self-archiving is pro-actively offering open access publishing."
I assume you would not take my commitment to OA publishing very seriously
if I said "The best defense against the dangers [sic] of OA publishing is
More information about the Jrnlnote