Ok, I knew I would draw some fire for that last message.
Strangely, it doesn't seem to have made it onto the
web archive (though Stevan quoted it extensively) - I think I've
been spending too much time on this debate - is this a nefarious plot?!
I should mention again that I don't speak for my employer
(the American Physical Society) and when I get carried away with
a point I may not even be speaking very well for myself. Still,
I think I said a couple of things worth defending, perhaps by
clarifying a bit.
What got me into trouble was when I wrote:
>as> are scholarly authors really so different from "trade authors"?
to which Stevan Harnad responded:
>
>I will try to show the fallacy in each of Arthur's examples, below. It
>is easy to state in advance what he keeps doing, inadvertently: He
>keeps presupposing the reader-end, trade-model as an unquestioned,
>silent premise, and then he (unsurprisingly) keeps deriving it as a
>conclusion: [...]
drawing sharp distinctions between the motivations of trade
and scholarly authors. I agree that the prospect of direct
payment for your words, whether royalty-based, salary-based, or
whatever, is an important distinction - but it is not the ONLY
reason why "trade" authors write, and the other motivations, which
are primary motivations for "beginning" trade authors or journalists,
seem be very similar to those of scholarly authors. Harnad argues
that the prospect of future reward through a "fat contract" makes
a big difference, but really how is that so different from the prospect
of future reward to a scientist through grants, tenure, or prizes
based on a significant publication track record? Yes the future reward
is based on selling your words, but is not publication to please
a grant funding agency also "selling your words"?
>There is a stark dissociation here, between the two types of author and
>the revenue paid for their words, a dissociation that the superficial
>similarities -- there had to be SOME, after all, they are both authors,
>both publish, etc. -- invoked here cannot obscure.
Exactly, both are authors, both publish - that simple fact gives
them a lot in common, and I don't believe it is just superficial.
Harnad sees black and white, but I see a wide spectrum between scholarly
and trade publication, with no sharp boundaries. And yes the fact that
scholarly authors can write books that sell as trade literature
is relevant - in some fields this is the primary written communication
means, rather than journal publication. What happens to this argument
when books become available online too - probably only a few years
away, since publishers like Springer are already testing the waters?
>(Now here comes a more complicated one; if I succeed in making
>the fallacy evident here, it should advance us considerably:)
>
>The prestige of the refereed journal is a red herring too, at least in
>the analogy with the prestige of the trade journalist's employer: The
>real meaning of the refereed journal's prestige is that the work has
>earned certification as being of the quality that merits closer
>attention -- in terms of being read, cited and built upon!
>
Trade journal articles are not meant to be read or cited? The
articles are chosen and carefully placed by editors based on some
kind of quality criteria - the imperative is to make a trade journal
interesting enough that readers keep coming, and paying, for it.
Prestige = readership (quality and quantity) for both. Is this really
so different? And think about trade books by scholarly authors
intended for, say, the graduate student level or above - are they
not carefully critiqued before publication, and intended to be read,
cited, and built upon?
>WHO BENEFITS FROM GATE TOLLS?
>
>as> In my view, scholarly authors have much in common with all "trade"
>as> authors - all authors want their work presented in a manner that
>as> will ensure a wide readership and particularly recognition from
>as> their peers.
>
>Correct. And what ensures peer recognition among refereed authors is
>refereeing: Do all trade authors want that too?
Do scholarly authors actually like critical referee reports? Trade
authors go through a critical editorial process that always modifies
and often improves the text. Nonfiction trade authors often solicit
commentaries from other authors to vett their manuscripts and later
to use as jacket blurbs. After books are published, critics publish
their commentaries. In scholarly publishing we have perhaps LESS review
than in many areas of trade publication.
>[...] S/SL/PPV is not the only way
>to make an honest living in publishing! There's also page charges.
And advertising. Some scholarly journals advertise, and could possibly
support themselves with neither page charges nor fees to readers.
It might be something to investigate further, but I doubt the
industries that cater to scholars have enough marketing funds to
support all of peer review. Just thought I'd bring it up.
> [...] And what
>ensures the widest readership is free access for all on the Web: Do all
>trade authors want that too?
No - I think free access on the Web does NOT ensure the widest readership.
It ensures the widest POTENTIAL readership, certainly. But the only
way you'll actually get people reading things is to somehow point them
there, and how do they get to your little web site, if they've never
heard of you and can't think of any of the handful of keywords that
brings your site to the top of the AltaVista heap? Assuming they
even attempt to use AltaVista to search for scholarly articles?
[xxx is a separate case, which we can discuss later.]
That was the crux of my argument: a publisher helps readers and
authors meet. That concept is central to publication. Publication in
a Physical Review journal means first that your paper (selected through
a rigorous review process) is accessible
by a variety of mechanisms (online links, searchability, tables
of contents, author indexes) within the grouping of articles that
comprises the journal, and just as importantly that your article
will be accessible through INSPEC, Science Citations, and similar
cross-journal search tools. Soon your article will be accessible by
direct or indirect links from the reference sections of papers in other
publishers' journals. The publisher attempts to create as many
reasons as possible for readers to WANT to read your article. Because
that's what the publisher gets paid for - under S/SL/PPV that is.
So the publisher and author actually have similar goals - they
want to create a need among as many readers as possible to see the
article in question. And if too many readers are frustrated by
S/SL/PPV complications then the publisher loses out as well as the author.
Now, if the S/SL/PPV model goes away and we have page charges supporting
publication, the motivation to the publisher to catch reader interest
is indirect (authors will presumably pay more for a journal that can
catch more readers) and it is likely that actual readership
of an article will decrease, even if potential readership increases by
providing the articles for free. It seems a paradox, I know, but I
believe it is deeply rooted in the economics of the problem.
Anyway, that's my argument for S/SL/PPV over page charges. Take it
or leave it...
On tide-over funds - I would like to see a more detailed model
or action plan for this. I thought at first you said only online-only
journals would qualify, but now we seem to be expanding the scope.
Would it be centralized or decentralized? Organized by field, or
geographical location? Lots of details missing, and some of them
may matter if it is actually to be done.
The details of the page charge scheme also seem to be somewhat
missing here. Does it come directly out of authors' grants? Is it
mandated in some way by granting agencies? Can we write down
a specific criterion that a granting agency could legislate, eg.
Any page charges for publication in free-to-reader online-only
journals must be paid in full through the supplementary
publication award (limited or unlimited?).
And how should journals handle unfunded authors? Make up the difference
by charging funded authors more? That could lead to large inter-national
discrepancies. Could there be different page charge levels depending on
what services were provided to the author? Wouldn't there be an incentive
among publishers to keep hiking page charges if neither authors nor
libraries directly feel the pain?
On xxx:
> sh> The optimality [of free/online-only] to the reader is obvious.
>
>as> That's the kind of statement referees are supposed to catch...!
>
>Let them look at xxx's access statistics.
>
>http://xxx.lanl.gov/cgi-bin/show_weekdays_graph
>
> [...]
>But Arthur, what are you saying? That all the xxx authors above are
>just being irrational?
I wasn't talking about xxx in these arguments. I think it's important to
distinguish the run-of-the-mill individual or even departmental
web server from xxx, which is an important and active site with a large
steady readership (even if many of them are lost pornography-seekers :-)
For a good fraction of physics (20-40 percent of basic physics, probably
around 10 percent of "all" physics, somewhere around 1 % of scholarly
publishing) xxx is a reliable source of the latest papers, and has
become a de-facto publisher, archiver and abstracter/indexer, minus of
course traditional peer review. It is working in this manner for those
fields probably for a variety of peculiar reasons that may be hard to
duplicate for other fields - nevertheless it is modestly continuing to
expand and gain acceptance. Certainly it's a good thing for those
fields of physics to have an excellent forum like this for rapid
communication - but it also presents a threat to any S/SL/PPV journal
as you frequently point out - among a community with historically
little support for page charges that means anything like current
peer review mechanisms is under siege, and I worry. But not
too much yet...
On "backsliding":
>I think you are imagining a transition scenario from Nirvana back to
>S/SL/PPV (i.e., from the optimal/inevitable back to now) that would
>require things to flow uphill instead of down.
Well that's the question, is it uphill or downhill? It seems to
me the slope depends mostly on the culture of the scholarly community,
which is a hard thing to change. Right now, even if it seems like
it shouldn't be, the reality is that readers (library committees)
have a nonzero but limited budget out of which they select which of
competing scholarly journals to pay for access to. The vast majority of
authors continue to seek publication in a traditional journal without
putting anything free on the web, in some cases because prestigious
journals forbid it.
Maybe a tabular "picture" will help clarify what we're saying.
Assuming it's just author and reader having to pay for peer review,
call "A" what the authors (in aggregate) pay to the publisher [or whoever
is coordinating the peer review], and "R" what the readers (in aggregate)
pay. We'll constrain the total T ( = "A+R") for the moment, agreeing that
it could be reduced considerably from the current total. I'm
discounting the fact that "T" could be a bit lower under some models
due to differences in implementation costs - I don't think these costs
are significant in the long run.
Payment
Model Authors Readers
1 A > T R < 0 A possible advertising model
2 A = T R = 0 Page-charge supported, free-to-reader
3 0 < A < T 0 < R < T Combined S/SL/PPV and page charges
4 A = 0 R = T Current S/SL/PPV model (no page charges)
5 A < 0 R > T Royalty model (trade publication).
Harnad claims that only a mix of (2) and (5) is stable together (Model
(1) hasn't been considered at all here). My claim is that the
whole range, from (1) to (5), could easily be present in the scholarly
publishing mix, and the final stable distribution will depend on
field (what kind of authors and readers you have), funding structure
(what kinds of publications charges authors are funded for), and
culture (are authors willing to pay page charges for free distribution,
are they willing to refuse royalties up to a certain level for
restrictions on distribution, where do readers go to look up information).
I also claim that the principle effect of online journals is, through
both communication efficiencies and automation, to decrease considerably
the total cost T. Electronic distribution also makes it easy to
increase the number of readers sharing in the burden R, without any
substantial increase in T. The net effect is to decrease the costs
both to authors and readers, but more substantially to readers, and
equilibrium economics would suggest the result is a shift downward in this
table.
Harnad's argument is that, through mandated changes in the funding
structure and education of authors about the superiority of model 2,
the distribution can be changed significantly, and shift upward
rather than down. It's possible, but I don't see it being easy,
and certainly not "inevitable"!
Arthur (apsmith_at_aps.org)
Received on Tue Aug 25 1998 - 19:17:43 BST