Re: Journal Papers vs. Books: The Direct/Indirect Income Trade-off

From: Stevan Harnad <>
Date: Fri, 7 Dec 2001 22:56:30 +0000

On Mon, 12 Nov 2001, Arthur Smith wrote:

> My argument
> was on motivations, which seems to be at the core of any "give-away" vs.
> "non-give-away" distinction no matter what the medium. The fact is that
> the primary goal of essentially all authors is communication, of
> thoughts, ideas, concepts, scenarios, from the author's head, to that of
> the audience. Whether that happens in a journal article, a book, mass
> media, through television specials, musical performances, or whatever.

Yes, but some of us want to give them our thoughts for free (for impact
and uptake), and some of us would like to be paid fee or royalty or
salary income for them.

> famous ones, already established, can pretty much do whatever they want
> - the real question is what do those who are less well known do, how do
> THEY get their important contributions out there and appreciated?

I have often said that refereed research reports are like advertisements
(and access-tolls would be equally absurd and counteproductive in both
cases). Arthur now mentions the early-career stage of a non-give-away
writer. Yes, this too is like an ad, when one is first promoting one's
work and happy to give it away to do so. But the analogy stops there.
The goal is to get to the stage where one can sell it, as the
famous writer does. Most writers never reach that goal, but that doesn't
mean that their motivation doesn't come from trying!

(All this ground has already been covered, more than once, in this
Forum: Please see the earlier items on this thread, and the spin-off
article that came out of it:
Harnad, S., Varian, H. & Parks, R. (2000) Academic publishing in the
online era: What Will Be For-Fee And What Will Be For-Free? Culture
Machine 2 (Online Journal)

> So in essence, my argument is that we can learn a lot from the
> perturbative approach (with, say, trade book publishing as a reference)
> to this "refereed research (journal article)" problem; Stevan claims
> that that's what everybody's doing, and it's wrong because (in physical
> terms), the power series diverges, there's a singularity, and the
> "refereed research" give-away literature is on the other side of the
> singularity and completely different. But I don't see any real proof of
> this - so far things seem to be moving in directions far more parallel
> than not, between book publishing and the refereed literature in the
> online world. The analogies may be even more clear in the music world...

I don't know if the give-away/non-give-away difference is all/none or
a matter of degree, but it's a substantial difference either way. And
the reason both literatures (g and ng) are moving in parallel is very
simple: ng came first, and was the only possibility in the on-paper
medium; on-line, all bets are off. And don't bank too much longer on
the sluggishness with which researchers are finally coming to a
realization of this. They are slow-witted but not wit-less. And music
is even more ng than literature, so it's even less relevant here.

> sh> [...] has no bearing whatsoever on the motivations of refereed research
> sh> paper authors, who always aspire only to get research impact, an indirect
> sh> reward, not text-sales-income, which, far from being a reward, is in
> sh> fact a BLOCKER of the indirect reward, because access tolls block
> sh> access, and whatever blocks access blocks impact.
> sh>
> Does it? One of the two highest impact news sites on the internet is the
> New York Times, even though it (1) has advertising, (2) requires
> registration and probably tracks personal use, and (3) charges for
> access to older articles.

So what? I was talking about research impact (uptake and citation);
the right question to ask is: what would the "impact" of the NYTimes
be WITHOUT 1-2-3? (You seem to be missing the logic of all this...)
Toll-based research in Nature and Science has impact too, but removing
the toll-barriers could only enhance it.

> Even though the first two are not financial
> "tolls" for readers, they should count as something that "blocks access"
> but seems not to "block impact".

You are missing the point...

> Any physics paper we publish that gets
> cited in the New York Times gets many times more attention than just
> putting it up in our own journals (and we do provide at least free
> abstracts for all our papers).

That's a still further matter (and irrelevant): My research gets impact
from a variety of sources. I have no control over whether the NYTimes
covers it, and over whether they charge for toll-free access to their
coverage. I DO have control, however, over whether readers have
toll-free access to the refereed paper itself.

To check your logic, would you please tell me whether you are suggesting
that access-tolls DON'T diminish impact? Perhaps you think they enhance
it? (Don't confuse this causal question with the mere correlation that
there of course is, because of the Gutenberg-era legacy.)

> That, by the way, was part of what I
> meant by "advertising" as applied to refereed research. And even any
> physics paper that gets published in Physical Review Letters will on
> average receive many more serious readers than it would from having been
> posted in any other location (including the arXiv).
> We do have an "access toll", but it seems not to "block
> impact".

False opposition: The same paper that appears in Phys Rev Letters also
appears in arXiv! Freeing online access maximizes impact; I didn't say
that tolls nullify impact (otherwise there could have been no impact
factor at all in the toll-based Gutenberg era!): They just don't allow
it to maximize. (I marvel that I have to state the obvious like

> We also have a free journal (on Accelerators and Beams, funded by
> sponsoring labs) which seems to be doing ok, but despite our best
> efforts at "advertising", just being free doesn't seem to convince
> authors that it has any greater "impact".

Of course not! Who said being free was a sufficient condition for
impact? The sufficient condition is quality. And being free is not a
necessary condition for impact either: But it is IS a necessary
condition for maximizing impact.

> sh> Moreover, the direct rewards from text-sales do not even go to refereed
> sh> research paper authors, but only to their publishers!
> Moreover, the direct expenses for these publications are not even paid
> by authors anymore, but only by their publishers!

And that's fine. As long as there is a market for the on-paper version,
or the deluxe on-line version, let it be paid for as an add-on. But the
vanilla version of the refereed article must also be available on-line
for free. That way the toll-based deterrent can be bypassed where
necessary (and that's the way impact is maximized).

> sh> And in any case (and here we are not speculating or psychologizing),
> sh> most book authors do NOT give their books away, even initially, even
> sh> though most of them never make much money from them (as Hal Varian
> sh> pointed out). (I am not counting here the [possibly much larger
> sh> number of] would-be "authors" who, until the on-line era, could not
> sh> afford the vanity publication that would have been required to make
> sh> their words and thoughts public and permanent: In this exhibitionistic
> sh> age, perhaps everyone thinks their musings are worthy of public notice,
> sh> perhaps even Viva Voce on Reality TV...)
> So DO most authors give their books away, or do they not? If you only
> define authors to include those who sell their books for money, the
> answer is pretty clear... And anyway, DO most scholarly authors give
> their refereed research papers away on their web sites, or do they not?

Most book authors do NOT give away their texts; it's just that most also
don't succeed in selling them either!

And although you are quite right that researchers (with the exception
of some swifter-witted physicists) have been stupefyingly sluggish in
getting around to self-archiving -- and that that slow-wittedness about
what is in their own best interests, and what can be done about it is
an undeniable historical fact so far -- it might not be a good idea to
plan on banking on it for much longer. The token will drop, and then
things may change quite quickly... ("You can fool some of the people
some of the time...")

> sh> And all of this is irrelevant anyway. Because those books that their
> sh> (bona fide) authors DO want to give away gratis in perpetuo, belong on
> sh> the give-away side of the give-away/non-give-away distinction just as
> sh> much as all refereed research papers do -- and the optimal solution for
> sh> them is likewise the same: publicly self-archiving them online!
> Great - so we have some things in common after all! And is this actually
> working for them in the book world to gain impact? Does it work for
> musicians who post free selections at, or via Napster? The
> conclusions I've heard have generally been that the two modes (free
> distribution, and distribution with "tolls") work hand-in-hand to
> enhance impact, despite strong opposition by publishers. My belief is
> that this is true in the refereed research literature as well... but I
> think the analogies are worth a lot more analysis beyond the naive
> economic ones we have heard here.

I think it's so obvious that the disanalogies are the salient ones that
this takes my breath away! Yes, struggling new writers give away
freebies, like self-adverts, but that is all done in the hope that
their work will eventually SELL. Same with musicians. And writers don't
even have a lucrative performance dimension that their give-aways could
be viewed as adverts for (the way freebie CDs might be for live
concerts, though I don't know about that, and suspect it is more the
other way around).

But what on earth is this shadowy "impact" in the case of book writers
and musicians that you are alluding to (apart from potential future
sales)? There are indeed esoteric monograph-writers whose books are
very much like refereed-journal articles, seeking the same kind of
impact, hence g rather than ng; but apart from them, what is this
phantom "impact" you think is the real goal of book-authors?

> sh> [...]
> sh> > most of the barriers
> sh> > to communication have nothing to do with the price.
> >
> sh> Is that so? Tell that to the majority of researchers, at the majority
> sh> of universities and research institutions, who are denied access to the
> sh> majority of the 20,000 refereed journals published annually, because
> sh> their libraries cannot afford the access tolls.
> sh>
> Hmm, I wonder what they do think? "Denied access" is a bit too strong a
> word, since articles are always available via interlibrary loan or
> document delivery.

Ah me, it's the "let them eat cake" response again. So you believe that
if they didn't have the daily free on-line access provided by, say,
arXiv, that all those same daily users worldwide could and would get
toll-based access? Because if I believed that, I would certainly give up
all my efforts to free online access (and would at last understand the
puzzle of the sluggishness!). But I don't believe it for a minute (and
I rather doubt you do either: as an exercise, try recalculating a
day's hits from (nearly
200,000 a day at the US site alone) as interlibrary loans -- in dollars
and in days...)

> But whatever the price of the journal article itself,
> the much greater price generally is the cost of time and attention paid
> by the person reading;

So what?! The time I'm ready to put into it is utterly irrelevant if I
can't afford to access it in the first place!

> every article, every piece of communication, once it
> reaches its audience, demands its attention. If a researcher's time is
> worth $20/hour and he/she has to spend an hour reading and understanding
> a particular paper (since abstracts are generally available free, I
> assume it is full text article reading that we are talking about) then
> that's a $20 "barrier"; far more than the $1.00/article that the library
> may have had to pay for a journal, and probably more than most document
> delivery fees as well.

I can't believe I have to do this, but Arthur apparently needs to have
his logic deconstructed: We will calculate the amount of "work" that
goes into digesting a piece of bread, and then find a formula for
translating that metabolic work into dollars. We then compare the
metabolic price with the purchase price and we find that the former is
higher. -- So let them eat cake! (M. Antoinette)

> A good publisher, a good refereed journal, gains
> that reputation by reducing these reading barriers, reducing the amount
> of time/money wasted in this process, bringing the things that are
> important to the forefront. And deciding which of those services are
> actually worth the "tolls" asked is certainly a role for the library.

As you know (I hope), I have always noted that peer review is essential
and that the true costs of implementing it must be covered. It is,
after all, the peer-reviewed literature to which we are trying to free

But the right answer to your reminder about the necessity of peer review
is not that we are stuck with access-tolls! The access-tolls pay for a
PRODUCT, a text, into which a lot of other add-ons are force-wrapped,
not just the essentials. And as you know, I think we should pay for
those essentials -- and those essentials only -- up-front, if and when
the add-ons no longer cover them -- out of the savings on the add-ons.

Until then, however, we should simply free access by self-archiving.

So let's not have any more of this false opposition between toll-based
quality or else toll-free sludge!

> sh> [...]
> sh> Utterly, utterly irrelevant! All these users' guides and pointers are
> sh> pointless if the paper sought is behind a financial access-barrier the
> sh> would-be user cannot afford to cross! Remove that financial firewall
> sh> and all these pointers (or rather those of them that make any sense)
> sh> are still there to point, but the paper itself is accessible, free for
> sh> all, too!
> Is this some new economic theory, that all you have to do is remove the
> price paid for something, and change nothing else, and everything is
> wonderful?

No. As I have said over and over, self-archiving frees the refereed
paper from toll-based access barriers (online). If and when the
essential peer-reviewing SERVICE (to the author-institution) is freed
from the inessential toll-based PRODUCT charges (the on-paper or
on-line text) and any other optional add-ons (to the
reader-institution), the essentials can be paid for out of the savings
on the essentials (i.e., those former add-ons and options).

> This is the sort of place where analogy with other media is I
> believe useful. Free music sites have had to move to joint agreements
> with recording companies, for legal and financial reasons, so that they
> provide both free and paid content. Online trade book publishing has
> experimented with quite a variety of models - personally I like the
> policy at Baen books, a science fiction publisher, which provides the
> first few chapters of a book for free online, and lets you then pay for
> the rest of the content if you actually want it. But that doesn't seem
> to be a real high-impact solution; there are also the publishers who
> advertise through web site banners and the like, which probably has no
> analogue in refereed journal publishing (but who knows?). The highest
> impact experiment in online book publishing has probably been the online
> customer reviews at A very interesting case, in the context
> of peer-reviewed literature; of course for the most part the content
> there is not actually available for electronic download, you still have
> to get a physical book.

I am astounded by the fact that you keep conflating all these things,
when they are all obviously geared toward sales revenues for the
author! (That's the only thing "impact" means in the ng context.)

> In short, where I believe the future lies is in a mix of free and "toll"
> content, both within the "refereed research (journal article)"
> literature and more broadly; I also believe we can understand this
> future much better by doing a more thorough analysis of our similarities
> and differences with other publication media, and of where each of those
> seems to be going in the online world. I'm not proposing that I have any
> final answers here, but I feel the need to disagree with Harnad's
> dismissal of this line of inquiry.

Well, as I am already saddled with the legacy of researchers'
inexplicable sluggishness about their own interests when it comes
to research impact and self-archiving, I of course can give no
assurances about whether they will see anything at all in this "mix"
model. I alas see nothing at all except miscegenation...

> sh> The superficial similarities are in fact blinding us to the
> sh> profound differences, and their implications. In particular, they
> sh> are slowing our progress to the optimal and inevitable outcome for the
> sh> give-away refereed research literature, which is that it can and should
> sh> and will be self-archived online free for all by its researcher-authors,
> sh> thereby maximizing its visibility, accessibility, navigability and impact,
> sh> to the eternal benefit of research and researchers.
> Just because something is online free does not make it worth reading.

Of course not: Please see above re. necessary vs. sufficient conditions,
and the difference between research impact and potential sales

> Publishers are and will continue to be paid for helping readers cut the
> massive expenditures of their own time that is required to understand
> the literature and read what is worthwhile. Free and for-fee publishing
> are not in irresoluble conflict. Author self-archiving can progress as
> much or as little as the authors themselves want.

It is peer review, and its true cost, that provides the quality-control
and guidance, not the rest of the stuff it is force-wrapped with. (And
of course I have no excuse whatsoever for the sluggishness of
researcher self-archiving to date -- except it's certainly not for want
of MY trying!...)

> sh> The very same option is also open to book-authors, of course, but I
> sh> rather doubt that many will choose to take it. If they are so
> sh> similar, I wonder why?...
> If they are so different, I wonder why "refereed research" authors still
> post so few of their articles online free? We've been at this for 3
> years now in this forum, close to 10 years in others; why is it taking
> so long? I think the similarities will tell us as much or more about
> this issue as the differences...

Ah Arthur, that really hurts! But if it's the similarities that are
behind the sluggishness, then it's the apparent similarities, not real

Stevan Harnad
Received on Fri Dec 07 2001 - 22:57:37 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:19 GMT