Re: Savings from Converting to On-Line-Only: 30%- or 70%+ ?

From: Stevan Harnad <harnad_at_COGSCI.SOTON.AC.UK>
Date: Sat, 29 Aug 1998 09:01:34 -0400

On Fri, 28 Aug 1998 Arthur Smith <apsmith_at_APS.ORG> wrote:

 as> There are of course many other electronic-only journals, but they
 as> "own" the content in the traditional manner, and do the distribution
 as> themselves.

 sh> Correct. But if free, they have nothing to lose (and a good deal to
 sh> gain, and save) by reconfiguring as an xxx overlay.

as> Then why have they not done so? People generally do things that
as> are in their economic self-interest.

Give it a bit more time before drawing any strong conclusions about
what is and is not possible or likely. There is still a great deal of
confusion and naivete about (as contributions to this discussion -- not
yours -- illustrate). Moreover, there is a powerful status quo in place
-- powerful both in terms of the generations and lifetimes of habit
forged in its image and in terms of the huge financial interests (few
of them as concerned about the learned community's best interests as
the APS is) vested in it.

This creates a great deal of inertia and even resistance to what will
still prove, in the light of hindsight, to to have been the obvious
path toward the optimal and the inevitable. Even simple, obvious things
do not yet look simple or obvious to most people at the moment.

Wait a bit. Even a forum like this could have a disproportionate
effect in setting things in motion in a more sensible direction.

But to address your question ("Why not yet, then?") more directly:

New, online-only journals who have gotten costs down to below 30% today
still can't demand page charges to recover them yet because of the
prevailing culture (owing to the status quo). Page charges would simply
kill off these fragile new journals.

That culture has to be changed, and will be, by the pressure of
subscription cancellations brought to bear on publishers as a direct
result of the use of xxx and of authors' free home archives by readers.
Only when this pressure is felt by all will page charges be seen as the
appropriate cost-recovery mechanism; until then, tide-over subsidies
for free online-only journals are the way.

Together, free public eprint archives and free online-only
journals will be the experience and practise that changes the
culture, but, alas, this cannot happen overnight.

as> I brought up ATMP because it uses EXACTLY the model you propose for
as> publication, except that it has somehow NOT been able to run itself
as> off author page charges. Why not? You dismiss this, but I think it is
as> well worth investigating as a practical example of what can be done,
as> technically, politically and economically.

This is simply a consequence of the fact that such journals are
fledging in a world that is still uniformly accustomed to doing things
in a completely different way. Moreover, the stigma from the
insult/injury factor currently associated with page-charges that I have
mentioned several times -- that page charges in the paper era, levied
on TOP of S/SL/PPV, and in exchange for nothing more than paper and its
firewalls blocking free access, were an abomination! -- hardly helps
speed things along, even though it is, ironically, a purely paper

But this bad press for page-charges in exchange for toll-gated paper
will dissipate when the (negative) connection between page-charges and
a free literature for all begins to be appreciated (through actual
practise and growing dependence on xxx and the home archives).

 sh> The model I proposed was (1) author-end page charges to support the
 sh> cost of (2) refereed journals free for all with (3) xxx as the public sh> mode of access. Overlaying (3) looks like the simplest and most
 sh> efficient way to achieve (1) and (2). But that is not what you are
 sh> questioning; you are questioning whether we could/should achieve (1)
 sh> and (2) at all.

as> Exactly. Why should we? What's wrong with (1) no author page charges,
as> (2) modest charges to access quality refereed journals with (3)
as> journal publishers or third parties supplying the distribution
as> mechanisms? It is not only workable (since it is what is being
as> done now) but it can bring all the cost savings and added value
as> from on-line conversion to bear at a pace that no volunteer effort
as> can match.

The answer is simple: Because page charges are access barriers,
toll-gates, fire-walls. They would mean I cannot just link to xxx and
pick up everything and anything. I must go through a financial obstacle
course of S/SL/PPV (assuming that publishers' licensing packages can be
made to interdigitate into a seamless literature at all). Maybe I will
be lucky enough to be at an institution that can afford to license the
entire corpus; probably not. But with free public archiving by authors
of all their work, there are no maybe's: It's gathered into xxx and
cheap search and linking will take care of the rest.

The optimality to the reader is obvious. Now ask yourself why AUTHORS
should wish it otherwise? Are financial firewalls that gate access to
their work any advantage to them? Is there any earthly reason I should
not archive all my work publicly, in my home server and xxx?

So the incentives are all there, for the authors and the readers. And
that's the whole learned community: Who else is there?

That is the point to remind everyone: But the literature is not
COMPLETELY free; there is still the cost of quality control (peer
review, editing). Now how would you prefer to pay that? Up front, out
of library savings, research grants, and university publication
budgets, or out of S/SL/PPV. But note that with the former, you keep
the free, seamless access for all; with the latter you are back to the
tolls and restrictions of S/SL/PPV.

You are posing this question from the point of view of the present,
paper-bred S/SL/PPV status quo. You ask the reader (without mentioning
the fire-wall problem): wouldn't it be simpler to keep the present
system, with cost-reduction, rather than to move to page-charges, with
all their bad associations?

This is a Trojan Horse. It would make us forever hostage to
counterproductive toll-gates when, for the same money, rechanneled in a
more sensible way, we can free the literature for once and for all, for
one and all.

 as> What's so special about a truly electronic-only journal? It merely
 as> eliminates one of the production and distribution pieces of the
 as> process.

 sh> You are
 sh> presupposing the answer to the 70/30 question, based on the
 sh> way you are presupposing that it should all be done!

as> No you are presupposing to understand something you have no practical
as> working example of, while I am basing my comments on what I have seen
as> and my experience with a number of online journals (both online-only
as> and not).

I am the founder and editor of one of the first free, refereed
online-only journals, Psycoloquy, subsidised by the American
Psychological Association, since 1990.

It is still tiny, not comparable with APS journals, or with the much,
much bigger, busier and hence more costly (to implement) paper journal
I have edited for over 20 years (BBS), but I have no doubt that
its day will come...

So my experience (and that of the founders of the other brave new
e-only journals) is bottom-up. Yours is top-down, from the status
quo. We'll eventually meet somewhere, but our future course will, I
hope, be set from the ground.

as> I fully accept the "30" answer -
as> sure we could cut costs by a factor
as> three through online-only production (given various things that will
as> probably take at least a decade to implement). I have no argument withas> that. My argument is with supporting it with author page charges and
as> making everything freely accessible to readers, an approach which I
as> think will either destroy peer review entirely or will lead to the
as> demise of non-profit publishers while the for-profits increase their
as> monopoly power.

I have tried to give the most critical argument for it: to do away
with access barriers. If we agree that we are talking about the
same much-reduced costs, then I am just saying S/SL/PPV is simply
the wrong way to recover them in this new medium, because, by
definition, toll-gates mean access blockage, and this special
category of authors (unlike trade authors) does not share the toll
receipts, condone them, or in any way benefit from them. In paper,
they had no choice, because costs were so high that there was no way to
make the literature free.

But in this new medium there is a way. It's not the way the trade
literature will or should take; but it is certainly the optimal and
inevitable way for this anomalous little subset (relatively speaking)
of the human written corpus, one in which, perversely, authors have
been giving away their texts, wanting nothing but readers in return,
all along.

(This is the point, in the endless back-and-forth cycle with advocates
of the status quo, where quality control is usually introduced as if it
were somehow at odds with the author's wish to give away his work to
all would-be readers: A red herring. It is precisely the cost of that
quality control whose optimal mode of recovery we are here discussing:
should it be via the existing fire-walled route of S/SL/PPV or the
seamless one made possible by page-charges?)

 as> The majority of users of electronic journals print out a copy of
 as> articles they are interested in - does that make every electronic
 as> journal not electronic-only?

 sh> Do you really think the 70,000 daily hits on xxx are all (or mostly)
 sh> printer downloads?

as> I have no idea what they are. What's a "hit" anyway? I was talking
as> about the articles readers are actually interested in enough to take
as> home and read in the bathtub.

Now that sounds circular, but it does give me the cue to quote the
passage about this "bathtub" argument that has just been squeezed
out of my forthcoming Nature article by length limits (Sept. 10:
"On-Line Journals and Financial Fire-Walls"), as a idea at what
most of those "hits" amount to:

    The (2) less-than-optimality of even the most advanced current
    screen-reading will no doubt continue to be invoked as grounds for
    holding the new medium at arm's length. What this tireless plaint
    fails to reflect on is what proportion of our intercourse with the
    periodical literature really calls for the _kind_ of reading
    (bath-, bed-, beach-based) that is usually envisioned in this
    context, as opposed to the desk-based searching, skimming,
    spot-checking, citation-tracing, and active cut-pasting and
    quote/commenting that is the mainstay of Learned Inquiry. Besides,
    there's still the desk-side printer for whenever a huggable copy is
    needed; and long after the printer is obsolete, there will still be
    the option of micro-thin "virtual papers" that can simulate any
    papyrocentric feature (optimal or counter-optimal) after which
    habit hankers, from portability, pliability, and thumbability to
    even the musty smell of shelf-worn vellum.

 as> Ahh, here is the key sentence: "once most of the literature is
 as> on-line and free"...

 sh> But that's been the premise and the conclusion all along: that
 sh> that is the optimal and inevitable condition, and the question is,
 sh> how to get there as soon as possible? The status quo (which you
 sh> appear to want to modify relatively little) is certainly not the
 sh> condition on which conditional statements like the above one are
 sh> based.

as> I'm happy to modify the status quo, but I prefer evolution to
as> revolution. I fully expect in the distant future ALL of the
as> literature to be on-line
as> and inexpensive. But probably not free. Does it make such a big
as> difference, cheap vs. free?

Yes. Vide supra. It's the difference between the toll-free corpus for
all that we could have, and the toll-ridden one for some that we have

 as> Even if the current linear growth rate continued
 as> it would be several decades before even 50% coverage of physics was
 as> reached.

 sh> How many papers do you think are published monthly in all of physics? sh> The slope (45 degrees) and intercept (now over 2000 papers monthly)
 sh> from 0 after 7 x 12 months is known for xxx:

as> Yes, it's been seven years. 25,000 papers in one year is a lot, more
as> than the Physical Review handles. But the subject coverage is very
as> sparse, concentrated almost exclusively in high energy theory,
as> astrophysical theory, and some parts of condensed matter theory. In
as> basic physics there should be at least a factor of 3 or 4 more than
as> what is there, and including areas of applied physics gives another
as> factor of 3 or 4. Covering all of science means at least a couple of
as> extra orders of magnitude, and the growth has not been sufficient in
as> recent years to reach that lofty goal in less than a century. There
as> could certainly be an "inflection point" in the offing, but wouldn't
as> it have happened by now? There has been little in the way of
as> technology changes in the last 4 or so years to make it that
as> much easier for other areas of physics or science to join the
as> bandwagon. Burgeoning systems generally show exponential,
as> not linear, growth.

You are right to ask why it has not happened yet, as the technical
means are indeed in place. I think the answer is human habit hysteresis
(the overwhelming tendency to keep doing things the old way).

This has obviously been overcome (lightning-fast!) in the fields you
mention. How long will it take in other fields? Perhaps making xxx more
WORD-friendly will help. (We are doing this with CogPrints, in
collaboration with xxx, which can then subsume CogPrints, covering
Psychology, Neuroscience, Linguistics, and large portions of Biology,
Computer Sciences and Philosophy.)

Discussion and even some polemics may help too...

as> But it could still happen. If authors en masse suddenly decided to
as> post to xxx, what then? I grant your assumption, and then what?

as> 1. Peer reviewed journal subscriptions crash

Not if we can negotiate one-time subsidies to tide them over the
transition to page-charges.

as> 2. For-profits simply raise their prices (and make out with their
as> 3-year contracts) while non-profits scramble.

You left out the driver of the S/SL/PPV crash: Library cancellations,
driven by the user preference for the free literature. That will hit
all S/SL/PPV publishers, for-profit and non- alike. In the face of
cancellations, raising S/SL/PPV just accelerates the process.

as> 3. Non-profits try to introduce author page charges. For-profits keep
as> theirs at zero. Authors flock to for-profits for publication (the
as> recognition imperative is still important) and non-profit submissions
as> crash.

Both for's and non's will be squeezed by the drop in S/SL/PPV
revenue caused by the preference for free use, so they will be
equally driven toward page-charges. The winners will be the ones
who get stable transition strategies in place sooner, and then

as> 4. Non-profits turn to government funding. For-profits cry foul and
as> unfair competition.

Foul? It is government funding of research and universities that
is subsidizing the texts they get from authors for free already,
and then subsidising the buy-back through S/SL/PPV! If anyone has
any complaints about that, let them talk to me...

Refereed journal AUTHORING is a non-profit industry!

as> 5. The non-profits fold or are bought out by for-profit publishers.

The pressure exerted by growing free use against S/SL/PPV revenues
will be felt equally by both sectors. The non-profit publishers
are better placed to make the transition smoothly as their
constituency is the authors and readers themselves.

as> 6. The for-profits notice that xxx also is government subsidized...

There are so many ways I could answer that. Let me leave it to
other participants in this discussion...

as> Is my scenario any less likely than yours?

All depends on the strength of habit hysteresis: If habits and
the status quo prevail, yours is more likely. If reason and
the interests of the learned community and learned inquiry itself
have a chance, then my scenario is both optimal and inevitable.

Stevan Harnad
Received on Tue Aug 25 1998 - 19:17:43 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:24 GMT