Bernard Naylor, who entered the correspondence with a paper he wrote for another forum, now offers extended remarks that take up the issues of the whole series. His new contribution views the journals publishing system holistically and takes up issues such as prestige, pressure to publish, conservatism of authors and publishers, and the prognosis for acceptance of electronic publications by all the players in the current academic information chain.
From: harnad@clarity.princeton.edu
Date: Tue Aug 9 17:12:03 1994
To: gs@reagan.ai.mit.edu (UNESCO List EJ LIB), serialst@uvmvm.bitnet (Lib Serials list), vpiej-l@vtvm1.bitnet (Pub-EJournals)
Subject: On Trade vs. Esoteric Publication
The following remarks by Bernard Naylor, Director, University of Southampton Library, are followed by comments from Stevan Harnad.
From: "B.Naylor" B.Naylor@soton.ac.uk
Date: Thu, 4 Aug 94 17:38:16 BST
A SMALL CONTRIBUTION TO THE SUBVERSIVE DISCUSSION
Bernard Naylor
Director, University of Southampton Library
- Having quite a large department to run, I've only been able to keep one ear so far on the progress of your interesting discussions. I'm now snatching a few moments to make a comment or two which I hope you will find constructive.
2.1 The model which divides scholarly communication into "trade-scholarly" and "esoteric" is conceptually tidy, but I'm not convinced it takes full account of the actual state of affairs. While we are locked (for the time being) into the commercial and paper mode, I believe that journal articles can actually be described as lying at different places on a long and seamless spectrum rather than falling neatly into two groups. Often, the actual place where a journal is located on the spectrum is hard to define, and it may change with time as an article becomes hot or loses heat. Hence we have to imagine that there will be confusion in an indeterminate, but probably large, central area of the spectrum if the model to which you are working is to emerge. This will be while "journals" (or, more precisely, the communications they contain) follow through the process of being assigned to one of the two groups. And even then, any given article may show a tendency to migrate from the group to which it has become assigned.
2.2 The journals of highest prestige (which are most often the ones with the fewest apprehensions about the present financial situation for journals) will continue to be the magnets for people who think their work is of the highest quality and deserves the widest scholarly attention. Hence, the division of scholarly communication into two sets is likely to be seen as having connotations about quality of content. (I.e. "publish trade-scholarly" equals higher quality and relevance. "Publish esoterically" equals lower quality and relevance.)
2.3 Scholars here are already under heavy pressure from their departments to publish in highly-rated outlets and I cannot see that pressure easily letting up. It can have very serious financial consequences in the UK for a department if its members don't achieve that, because the process of awarding them money on account of their research makes that an important criterion. I cannot imagine any scholar readily saying to his/her head of department: "Actually, my work is of relatively low scholarly quality and relevance so I do not propose to try to get it published in any of the journals which our profession ranks highest" - even though, in practice, that may be where it will end up.
2.4 Money is also relevant here. I doubt whether a discussion about the altruism of the members of academe would get us very far. I just think that any scholar who could get some money for publication -- as well as transmitting ideas -- is quite likely to be attracted by that possibility. The attraction of that will become even stronger if the author perceives someone else (e.g. an information broker) making money out of something the scholar has made available free. It could be argued that the activities of the information broker should be stopped, but I think that would break down because for some people information brokers provide a very useful service, enabling them to use their own time in other, more productive ways.
2.5 I therefore think that the possibility I raise in my "Future of the Journal" paper, namely that the electronic (and esoteric) sector could feature for some time as the sink for the material of lesser quality, could quite likely emerge.
3.1 I've found the debate about where the costs of publication really lie very interesting. It may turn out to be true that the "ordinary" publishers are exaggerating the costs which the network will not substitute, by a factor of about three (their "over 70 per cent" as against your "less than 25 per cent"). If it does, I shall be a bit surprised since most of them depend, for earning their living, on being more or less right abut that kind of thing.
3.2 I think one reason for their caution about electronic substitution is because they find speculation about alternative cost-recovery models very complex and difficult. But one of the two main purposes of substitution is to address a perceived economic problem -- the other being to introduce one further industry to the IT revolution. The road from yesterday contains enough litter from high-tech disaster projects to suggest some caution, at least, on the economic/costing front. Remember; the conclusions reached are for testing in settings where red faces are the least of the penalties for being proved wrong.
4.1 On the question of technology, I am reasonably confident that I have not come across any technological problems which I don't think can be solved. There are still important questions as to: how soon? and by whom? and at what cost? At present, the "information superhighway" itself does not exist internationally in a form which could cope with the information traffic currently carried (not very efficiently) by print on paper (the "fat pipe" under the Atlantic at 1.5 Mbps is surely hopelessly inadequate?) and the other problems I referred to in my paper are also with us and have to be resolved if an electronic solution is to operate satisfactorily. Print on paper is currently carrying a colossal amount of information into all sorts of unlikely academic places.
4.2 An important additional consideration on this point, in my view, is what I might call "the mental preparation of the sector." The academic sector, in my view, has been right to point to the serious mistakes made in some industries where technological advance has been implanted without adequate preparation of the work force, in the form of reorientation of work processes and training of working people. We are now seriously contemplating the most dramatic change in the working habits of scholars for some centuries, but I look in vain for convincing signs that our sector appreciates this and is collectively bending its mind to preparing for the consequences. There is far too much "Throw the technology at them and they'll get on with it". A change of this magnitude is going to cause confusion and muddle anyway for a time, but we ought to be doing far more to minimise that. Our present approach, in my view, is asking for trouble. One of the biggest drags on the introduction of advanced technology solutions lies in the attitude of the people for whom we consider them to be intended. It's crucial that a much better effort is directed to changing that. When I had to cancel some physics journals a few months ago, our physicists gave me the impression that I was bringing the roof down on their department, and Paul Ginsparg wasn't mentioned once in their remonstrations -- though I've no doubt they're well aware of what he's doing.
4.3 I recently asked a group of about twenty chief university librarians (that is, about 20 per cent of the total UK cohort) "by what date do you expect your present number of current subscriptions to print-on-paper journals will have been reduced by 80 per cent." The extremes were "1995" and "2050" with 2010 by far the favourite prediction (2010 was mine as well -- honestly). (I said I thought that might be a way, for some of them, of saying "after I've retired".) There is some belief out there that the present system will collapse and change suddenly quite soon but it's not very widespread -- though it may be right. Otherwise, it implies that Southampton University Library must shed 300 print journals every year from now till 2010. At present, that does not look very likely and this means the process will have to accelerate substantially (and I expect it will) later in the period.
4.4 Re-reading some of what you and Andrew have written makes me wonder whether we are really that far apart. My job is to cope with today's reality (which is largely, like it or not, both papyrocentric and commercial) and try to anticipate the next few years' (say, five or at most ten) changes reasonably intelligently. You and Andrew appear to be speculating about a revolution which I am satisfied will come -- though I seriously doubt your present economic assumptions -- and I sometimes get the impression that Andrew might not disagree strongly with my speculation and the speculation of my librarian colleagues as to the time scale. I'm less sure, Stevan, about your views as to likely time scales.
5.1 One of the points I'm trying to underline in pointing out that journal publishing is an industry is that it will not sit quietly by and let itself be subverted. We must assume that, if a concerted plan emerges to cut major traditional publishers out from the knowledge communication business, they will fight very strongly for their "share of the action." Commercial publishers will have revenue streams to defend, not least in the interest of the people they employ and of their shareholders, and even learned societies and academic presses could face massive upheaval if revenue-earning titles (which are also circulated gratis to some as a privilege of membership) appear to be in danger of complete substitution, and they will fight to prevent that.
5.2 One of the main points that publishers are likely to raise in this country, and probably in Western Europe generally, will be to question (to put it no more strongly) the propriety of academic institutions using public money (all UK universities bar one very small one are funded by the taxpayer) in order to drive a viable industry (as they see it) to the wall. Perhaps the government will be happy to see the publishers as resembling the toll bridge owners of past centuries whose bizarre privileges were bought out or set aside with the growth of the road network. But the mood in the UK, even among universities themselves, is to press for more and more activities to go down the "charging" road. Our government believes passionately in markets -- I think it claims it got its predilection for them from the United States! -- and I can see it saying: "You must allow commercial academic publishers onto the academic superhighway on terms which allow them to compete fairly for the survival of their role". I don't think this necessarily means that it should always be carried out in the same way as now -- but they must be given a reasonable chance to adapt. I'm genuinely surprised that you appear to think that it will be politically acceptable in the United States to make most scholarly information a non-tradeable commodity. I don't think it will be here.
- Maybe these few remarks have irrevocably confirmed my typecast for you! They are among the reasons why I think evolution is a better mode than revolution and why I think that the economic problems which libraries and others face in the scholarly communication field are unlikely to be resolved by an agreement that it should all be for free in an electronic setting.
Comments by S. Harnad:
2.1 The model which divides scholarly communication into
"trade-scholarly" and "esoteric" is conceptually tidy, but I'm not
convinced it takes full account of the actual state of affairs... I
believe that journal articles can actually be described as lying at
different places on a long and seamless spectrum rather than falling
neatly into two groups... there will be confusion in an indeterminate,
but probably large, central area of the spectrum...
There is indeed a continuum from trade (scholarly) publishing to esoteric (scholarly) publishing, but the lion's share of the kind of periodical publication that I and most publishing scholars and scientists are concerned with falls quite safely within the unequivocally esoteric region of that continuum. The gray area is not at issue, nor is it an issue.
At the root of our disagreement is your identification of the factors you apparently think determine a journal's place in the continuum. You think they have to do with the prestige of the journal, whereas I think they are much bigger factors than that, and that they eclipse the relatively trivial differences between high and low prestige esoteric journals (they're ALL esoteric, compared to journals whose individual articles actually have a nontrivial readership size). More below.
2.2 The journals of highest prestige (which are most often the ones with
the fewest apprehensions about the present financial situation for
journals) will continue to be the magnets for people who think their
work is of the highest quality and deserves the widest scholarly
attention. Hence, the division of scholarly communication into two sets
is likely to be seen as having connotations about quality of content.
(I.e. "publish trade-scholarly" equals higher quality and relevance.
"Publish esoterically" equals lower quality and relevance.)
Here is the root of our disagreement. Of course the highest prestige journals are the ones with the fewest financial worries (because they are the last ones likely to be cut from library budgets), and, by definition, they are also the ones that authors and readers value the most highly. But this has NOTHING to do with the trade/esoteric continuum! The high prestige journals, like their lowlier cousins are ALL esoteric if one takes the proper measure of esotericism. This proper measure is NOT:
(1) how much a journal costs,
(2) how many libraries subscribe to the journal, or
(3) how likely libraries are to drop the journal.
Nor is it:
(4) how eager authors are to publish in the journal, or
(5) how heavily publications in the journal weigh with promotions committee.
It is not even:
(6) how much weight readers assign to articles in that journal,
(7) how many individual subscribers there are to that journal, or even
(8) how many readers browse that journal.
We are closer, but not quite there yet, with:
(9) how many readers READ a particular article in that journal
and even closer with
(10) how many readers CITE that article.
But even with the last two measures, the relevant comparison is not between the more and less prestigious journals in a given field (it is a foregone conclusion that prestige will correlate positively with (9) and (10)). What one must look at is the average readership per article in relation to the true cost of producing that article. An even more dramatic way to depict it would be as the RATIO OF THE PER-PAGE READERSHIP TO THE PER-PAGE COST. That ratio may differ a little if one compares high- and low-prestige journals within a given field, to be sure, but if one compares it with the ratio for pages that really DO have a market -- either popular scientific and general intellectual periodicals or the magazine market in general (with or without adjusting for the contribution of advertising revenues, wherever they kick in), then the true locus of most of scholarly/scientific publication along the trade-esoteric continuum will, I suggest, become plainly and unequivocally apparent. And once we scale up to this broader sample of the continuum, the relatively trivial differences between high and low prestige journals will be altogether eclipsed.
Let's be more specific. Though it's risky to resort to figures from hearsay (and that is all I must confess I have so far), I am confident enough in what I am about to point out that even if I am wrong by one or two orders of magnitude, the upshot is the same: The average published scientific article has fewer than 10 readers and no citers; I'll bet the same is true for the average piece of scholarship in the humanities.
You aren't speaking of average work? Alright, let's move up to the top end of the Gaussian distribution and consider just the top 5% of the articles in a given field: How much higher do you think those figures are likely to be for them? (And don't forget that this excludes 95% of what is published, and is reckoned on a per-article, not a per-journal basis: not even the most prestigious journals produce exclusively, or primarily, winners.)
So what are the figures for a winner likely to be? Twenty readers, two hundred, two thousand? Suppose it's two thousand. We all know that in paper those rare articles that generate a huge demand are supplied mostly in the form of preprints and reprints, because a journal certainly cannot calibrate its print run by banking on occasional individual articles. (Journals print issues and volumes, whereas "citation classics" are relatively rare individual papers, exceptional by definition.)
But citation classics and the best-sellers among the separata are not what the economies of scholarly publishing are based on. Even if we restrict ourselves to the single journal in each academic specialty that is by consensus the most prestigious one, the per-page readership ratios for most of its pages will be off-scale compared to the periodical literature that actually has a market.
Why has this never come up before? Because no matter how absurd the per-page ratios were in scholarly periodical publication, there was nothing anyone could do about it (and they certainly weren't going to give up the reporting and reading of scholarly research), because the economies of paper were such that the only way to recover the true costs of making the research available at all was by levying reader-access charges. Yet of course it was never the (on average less than 10) readers who actually paid the costs per page; it was the vast infrastructure (mainly University libraries) that was set up to SUBSIDIZE the minuscule demand there really was for any particular page of this esoteric (sic, I now state without trepidation) corpus. The nonreaders (all of us) subsidized the readers (each of us) of any particular article.
Well that is simply no longer necessary; or, rather, the subsidy can now be set up in a much more sensible way, matching the true demand structure and the nature of the service provided by the publisher: not by continuing to treat a no-market commodity as if it were a viable trade item and benightedly trying to sell it to the vanishingly small number of scholars who may ever want to see it, but by charging the much lower true per-page costs of esoteric publication at the point where it is the most sensible to charge them: At the point of access to the peer community's eyes and minds. That much more modest per-page cost will be the price of the service that esoteric publishers perform for AUTHORS (and their institutions and research support agencies) in making their work available to their fellow-specialists, globally, in perpetuum, and, of course, for free. And we will all be the better off for it.
2.3 Scholars here are already under heavy pressure from their
departments to publish in highly-rated outlets and I cannot see that
pressure easily letting up .... I cannot imagine any scholar readily
saying to his/her head of department: "Actually, my work is of relatively
low scholarly quality and relevance so I do not propose to try to get it
published in any of the journals which our profession ranks highest" -
even though, in practice, that may be where it will end up.
But this point is only pertinent if one accepts your assumption that high-prestige journal = paper (= trade) and low-prestige journal = electronic (= esoteric).
Whereas, as I have tried to argue, I think that assumption is incorrect. Currently ALL journals, low prestige and high, are paper. The trade/esoteric dichotomy, as I have tried to argue above, has nothing to do with prestige. And it will continue to have nothing to do with it as journals become electronic; prestige will continue to depend on the rigor of the peer review and the quality of the authors and submissions, not on the medium.
It is true that today's high-prestige paper journals are likely to be the last to be cancelled by libraries for economic reasons, and it may even be true that this irrelevant side-factor will affect the initial conditions among electronic journals (the first new ones, and the first ones to migrate, will not be the high-prestige ones), and that will of course be regrettable, and will retard the inevitable, for reasons that are NOT to scholarship's advantage. But that still has nothing to do with the trade/esoteric continuum, and hence does not provide a rational basis for drawing conclusions about the applicability and appropriateness of the trade model to no-market papers, whether high prestige or low.
2.4 Money is also relevant here. I doubt whether a discussion about the
altruism of the members of academe would get us very far. I just think
that any scholar who could get some money for publications -- as well as
transmitting ideas -- is quite likely to be attracted by that possibility.
The attraction will become even stronger if the scholar perceives someone
else (e.g. an information broker) making money out of something the scholar
has made available free. It could be argued that the activities of the
information broker should be stopped, but I think that would break down
because for some people information brokers provide a very useful
service, enabling them to use their own time in other, more productive
ways.
This passage is rather complicated, and again involves some assumptions and contingencies that I think are erroneous: First, one of the marks of the esotericity of most of scholarly and scientific periodical publication is that the author does NOT make a penny from the sale of his text, and does not, and never has expected to. (It is an instant signal that one is in the trade rather than the esoteric region of the continuum if this is not true, and the author expects and does receive royalties for those pages. This happens in popular and general-audience scholarly/scientific writing. OF COURSE most of us would jump at the opportunity to make a few bucks from publishing our words, but how often do we get a chance to do that?) Trade publishing is medium-independent. When there is a potential paying readership, one can and should charge, whether on paper or on the airwaves. It's just that this contingency is absent in the region of the continuum I am concerned with; no author is making money there on paper, and no author will make money there on the Net either.
So whereas I agree that scholars are not altruists -- if there were money for them to make from the sale of their words (and if this were not too much at odd with their scholarly mission, if any) -- then they would certainly be happy to collect it; but the fact is that for the overwhelming majority of the scholarly/scientific corpus there IS no money to be made from the sale of their words -- at least not money for THEM (the author/scholars). Money is being made, to be sure, but it is being made exclusively by the paper periodical publisher, not the author.
Which brings me to the second (in my view incorrect) assumption: To show why this assumption is incorrect, I must first re-introduce what I have come to refer to as the "Faustian Bargain" that esoteric authors have reluctantly entered into with paper publishers -- and let me stress that in this metaphor it is PAPER that is the devil, not paper publishers, for they too are victims of the tyranny of the true costs and technology of that unfortunate medium. This bargain is ONLY Faustian in the case of esoteric publication -- publication in which the market for most papers is virtually zero, the author does not make a penny, and the sole motivation is to reach the eyes and minds of one's peers and posterity with one's findings. To have to treat that special transaction on the same model as the quite normal and un-Faustian bargain between a paper publisher and an author who makes a living by selling words is very close to absurd, yet for centuries there was no choice: The only hope an esoteric author had of reaching the tiny potential non-market of peers was to allow the publisher to charge for access to the writing, even though the author would not make a penny, because that was the only way that the expenses and a fair return could be paid for the true and sizeable per-page costs of the technology and logistics of paper periodical production and distribution.
To repeat: The Faustian bargain was reluctantly accepted, despite containing an essential internal conflict of interest between the esoteric author's desire to reach as many interested peers as possible and the publisher's need to restrict access with a price-tag to defray the substantial per-page costs and a fair return for investment and effort, BECAUSE THERE WAS NO OTHER CHOICE.
So if the first incorrect assumption was that esoteric authors can and do make money from selling their articles, the second one was that their Faustian symbiotic relationship with the paper publisher would somehow carry over to his "information broker" counterpart in the electronic medium. But what are we imagining here? The true per-page costs (if my estimate is right) are now down to less than a quarter of what they were in paper, so there is no longer a Faustian dependence on a technology whose sizeable costs need to be recovered by blocking access to esoteric work that already has virtually no market; the author, the author's institution, library, scholarly/scientific societies and research (publication) grants can with a little perestroika EASILY collaborate to subsidize life-long published page quotas as needed, thereby allowing access to be free for all (as it always should have been).
So who is this "information broker"? The editor of the electronic journal and the editorial staff (their services are already reckoned as making up most of the remaining < 25% per-page costs)? The (unpaid) referees? The author's own word-processing budget? The copy-editors/proof-readers (the rest of the < 25%)? Or the classifiers and maintainers of the electronic archive (these are currently called "librarians" and they do not normally get a cut from the sale of the author's text in any case).
So what is it that the "information broker" is doing that allows the author to get on with scholarly life instead of being a jack of all trades? In paper, this was clear: The author did not have to spend time printing, disseminating and archiving the work for all; the publisher did it for that author (but unfortunately had to charge admission in exchange). Where is the counterpart of this in the electronic medium that would warrant reincarnating the Faustian bargain yet again, now that there is clearly no need for it?
2.5 I therefore think that the possibility I raise in my "Future of the
Journal" paper, namely that the electronic (and esoteric) sector could
feature for some time as the sink for the material of lesser quality,
could quite likely emerge.
Even within the paper medium itself, new journals always have to struggle initially to establish a niche (and many fail, or fail to attain a high level of prestige). This is true in spades when it is not just a new journal that is at issue, but a new medium. So the prediction that the Net will carry material of lower quality than paper initially is quite a safe one to make (I myself have described the Net as a "Global Graffiti Board for Trivial Pursuit" till not that long ago), but this has nothing at all to do either with esotericity or the nature of the medium itself. The initial disparity is simply due to history, demographics and initial conditions. Paper currently holds virtually all the scholarly cards (and the hearts and habits of the major card-players) and the Net is a haven for the young and unscholarly or not-yet-scholarly.
But don't bank on it. The house of cards is also poised to collapse. All that's needed is something that will overcome the primarily MENTAL obstacles that exist currently: mainly superstitious habits and beliefs about correlations between paper and quality, correlations that are there, to be sure, but for arbitrary historical reasons rather than functional ones -- indeed, as I am trying to suggest, these reasons are becoming increasingly DYSfunctional, and THAT will be what finally makes the paper house of cards collapse.
3.1 I've found the debate about where the costs of publication really
lie very interesting. It may turn out to be true that the "ordinary"
publishers are exaggerating the costs which the network will not
substitute, by a factor of about three (their "over 70 per cent," as
against your "less than 25 per cent"). If it does, I shall be a bit
surprised since most of them depend, for earning their living, on being
more or less right about that kind of thing.
Indeed. But the PREMISE of such livelihood-preserving calculations is that one must continue to earn one's living essentially the same way. (As I suggested, those who have calculated the true per-page costs as 75% rather than 25%, as I do, have only been reckoning what electronic processing will save from a system that is designed to produce PAPER pages, as they do; let them redo it for a system designed solely to produce electronic pages.)
If I am right, publishers will have to be prepared to do a major restructuring (and reconceptualizing) of their role in esoteric scholarly publishing or else this anomalous portion of the symbiotic author/publisher system will simply break off and start anew, as an autonomous form of publication. The key factor is that the Faustian Bargain is no longer necessary; the true per-page costs (and a fair return) for esoteric publication can be recovered on a page-subsidy model. There is no longer any need to charge admission to a show that virtually no one wants to see.
3.2 I think one reason for their caution about electronic substitution
is because they find speculation about alternative cost-recovery models
very complex and difficult.
Indeed; but that does not make those models wrong, or impossible. Nor does it support your suggestion that publishers must be doing the calculations correctly because that's what they do for a living...
But there are (and there inevitably will be) publishers who DO understand the arithmetic (MIT Press seems to be one such publisher), and it is my hope that their efforts will eventually lead in a direction that serves the interests of all, without Faustian conflict, and without heroic efforts to preserve the unstable and far-from-optimal status quo.
But one of the two main purposes of substitution is to address a
perceived economic problem - the other being to introduce one further
industry to the IT revolution. The road from yesterday contains enough
litter from high-tech disaster projects to suggest some caution, at least, on
the economic/costing front. Remember; the conclusions reached are for
testing in settings where red faces are the least of the penalties for being
proved wrong.
By all means. But I don't see esoteric scholarly/scientific publishing as a potential cash cow for any big IT ventures (perhaps for popular and wide-spectrum scholarly/scientific publishing, but not for the esoteric region of the continuum). This no-market form of publishing is essential to us all -- it's what keeps human learned inquiry going. But the trade model just does not fit it (and never did).
Besides, most of us are talking about shrinking and restructuring rather than big, risky investments; that's mostly what the migration from paper to the Net entails (see Andrew Odlyzko's paper).
At present, the "information superhighway"
itself does not exist internationally in a form which could cope with
the information traffic currently carried (not very efficiently) by
print on paper (the "fat pipe" under the Atlantic at 1.5 Mbps is surely
hopelessly inadequate?)... Print on paper is currently
carrying a colossal amount of information into all sorts of unlikely
academic places.
This is an empirical question, and not in my domain of expertise, but I have heard that, once we decide to make a commitment to carrying the (scholarly/scientific) information that way, the Net will be ABUNDANTLY able to bear the weight. In any case, it is not clear to me who should be worrying about capacity problems these days, when people are passing vast quantities of, for example, porno-graphics, freely back and forth on the Net. Metaphors, of course, settle nothing, but my own image of the full corpus of esoteric periodical publishing (see Andrew Odlyzko's essay for some sample figures) as the flea on the tail of the dog, insofar as the Net's carrying capacity is concerned, especially in the future, when more and more of the rest of the traffic will be commercial and paid for (because it DOES have a market).
Humanity will be better served by granting that flea a free ride in perpetuum, rather than treating it as if it were another commercial traveller.
It's odd, by the way, that the "Net Capacity" and "Net Highway Toll" alarms are so often sounded by individuals who are no better informed than you or I are, but in whose interests it would be if there WERE an awkward capacity limitation or a prohibitive toll...
4.2 An important additional consideration on this point, in my view, is
what I might call "the mental preparation of the sector." The academic
sector, in my view, has been right to point to the serious mistakes
made in some industries where technological advance has been implanted
without adequate preparation of the work force, in the form of
reorientation of work processes and training of working people. We are
now seriously contemplating the most dramatic change in the working
habits of scholars for some centuries, but I look in vain for
convincing signs that our sector appreciates this and is collectively
bending its mind to preparing for the consequences. There is far too
much "Throw the technology at them and they'll get on with it". A
change of this magnitude is going to cause confusion and muddle anyway
for a time, but we ought to be doing far more to minimise that. Our
present approach, in my view, is asking for trouble. One of the biggest
drags on the introduction of advanced technology solutions lies in the
attitude of the people for whom we consider them to be intended. It's
crucial that a much better effort is directed to changing that.
I agree with this completely. Authors, readers, librarians, etc. do need to be prepared, informed, etc. But that's not an argument for the status quo, or even for slowing down (the pace of the move toward electronic journals is already excruciatingly slow, to my tastes: 99.9999% of the esoteric scholarly periodical literature is still in paper, after all; hence there is no precipitous hurtling toward an unknown doom going on here!). Preparing the populace should be undertaken pari passu with the migration itself, but certainly not prior to it, or instead of it...
When I had to cancel some physics journals a few months ago, our
physicists gave me the impression that I was bringing the roof down on
their department, and Paul Ginsparg wasn't mentioned once in their
remonstrations - though I've no doubt they're well aware of what he's
doing.
This is not surprising, and I don't think it is evidence of anything more than that no one has a rational command yet over what is happening and what is about to happen. It is because of this irrational factor that I never make predictions about WHEN it will all happen; I simply affirm THAT it will happen (and the sooner the better)...
There is some belief out there that the present system will collapse
and change suddenly quite soon but it's not very widespread - though it
may be right. Otherwise, it implies that Southampton University Library
must shed 300 print journals every year from now till 2010. At present,
that does not look very likely and this means the process will have to
accelerate substantially (and I expect it will) later in the period.
I suspect it will be a critical-mass effect: After a slow linear phase, consisting mostly of new electronic journals rather than migrations of established paper journals (though you are probably right that, unfortunately, the weaker paper journals make take to the skies first) as well as regressive "hybrid" projects (paper journals offering a double deal: paper plus electronic version, but to subscribers only, aimed eventually at developing an electronic-only subscriber base), a critical mass of free-access refereed electronic journals will form and demonstrate that they can be as rigorous as refereed paper journals, but with the added advantage of (1) universal searchability and access as well as (2) interactivity (peer commentary). That -- with the help of subversive projects like Paul Ginsparg's HEP Archive and Public Preprint/Reprint Archives created by authors -- will trigger a relatively rapid and dramatic restructuring that will end with most or all of the esoteric periodical corpus airborne.
4.4 Re-reading some of what you and Andrew have written makes me wonder
whether we are really that far apart. My job is to cope with today's
reality (which is largely, like it or not, both papyrocentric and
commercial) and try to anticipate the next few years' (say, five or at
most ten) changes reasonably intelligently. You and Andrew appear to be
speculating about a revolution which I am satisfied will come - though
I seriously doubt your present economic assumptions - and I sometimes
get the impression that Andrew might not disagree strongly with my
speculation and the speculation of my librarian colleagues as to the
time scale. I'm less sure, Stevan, about your views as to likely time
scales.
Ah, I refrain from committing myself to numbers when it comes to time scale. I stick to (1) editing and adapting PSYCOLOQUY as a model and to (2) trying to describe what is actually happening, what is possible, and what is rational in talks and papers. Chronology must fend for itself. (But if it's my 'druthers you're inquiring about, it can't happen too soon for me.)
5.1 One of the points I'm trying to underline in pointing out that
journal publishing is an industry is that it will not sit quietly by
and let itself be subverted. We must assume that, if a concerted plan
emerges to cut major traditional publishers out from the knowledge
communication business, they will fight very strongly for their "share
of the action." Commercial publishers will have revenue streams to
defend, not least in the interest of the people they employ and of
their shareholders, and even learned societies and academic presses
could face massive upheaval if revenue-earning titles (which are also
circulated gratis to some as a privilege of membership) appear to be in
danger of complete substitution, and they will fight to prevent that.
No doubt. But what will eventually prevail, I hope, is what is in the best interests of scholars/scientists and Learned Inquiry itself. As most of our intellectual wares (99.9999%, as I said, and add more 9's for the retrospective literature) are currently on the paper flotilla, it is in ALL of our interests to ensure that that flotilla does not sink prematurely. I believe a benign solution is possible to effect an orderly transition to the skies, one that will be fair to all; publishers simply need to be flexible and innovative, and not be tempted to adopt the short-sighted strategy of filibustering in favor of some version or other of the status quo. It just won't fly.
5.2 One of the main points that publishers are likely to raise in this
country, and probably in Western Europe generally, will be to question
(to put it no more strongly) the propriety of academic institutions
using public money (all UK universities bar one very small one are
funded by the taxpayer) in order to drive a viable industry (as they
see it) to the wall. Perhaps the government will be happy to see the
publishers as resembling the toll bridge owners of past centuries whose
bizarre privileges were bought out or set aside with the growth of the
road network. But the mood in the UK, even among universities
themselves, is to press for more and more activities to go down the
"charging" road.
You will not be surprised, perhaps, that this scenario evokes little empathy. I hope the motivation on all sides will be more constructive than this. Esoteric scholarly publishing is motivated by something far more important to us all than the money to be made from selling its texts.
Our government believes passionately in markets - I think it claims it
got its predilection for them from the United States! - and I can see it
saying: "You must allow commercial academic publishers onto the
academic superhighway on terms which allow them to compete fairly
for the survival of their role." I don't think this necessarily means that
it should always be carried out in the same way as now -- but they must
be given a reasonable chance to adapt.
This makes it sound as if the only problem is access to the Internet: But there WILL be a lot of toll-way traffic on the Internet. That is absolutely irrelevant to the issue under discussion here. Unless publishers are planning to re-tool themselves as telecommunications companies, they are not the ones for whom those bells would toll in any case! (This is a red herring, just as the capacity argument is.)
Paul Ginsparg's HEPnet currently gets 35,000 "hits" per day -- 35,000 physicists the world over retrieving articles. It is simplistic to conceptualize Net use as a finite resource (dramatic increases in Net capacity can be gotten for relatively small investments in money and material; and it makes little sense to tax the number of bits received or the time spent receiving them in an interdigitating network with varying transmission times, especially if the Net is far from saturation), so it is a mistake to imagine a toll on each "hit."
But suppose things did go in that direction: If the Net were privatised (and apparently it is about to be), AND if the Universities chose to pass on to their user communities the 10% increase in costs this would entail over their current flat connection costs, this still would not be reckoned on a "per hit" basis, but as a flat rate. Users would pay their own flat rates for unlimited sending and retrieval, and the rest would depend on WHAT they were retrieving: If it was a commercial newspaper or magazine, they would be prepared to pay a further toll, as they do now (and the writers of the material, and their publishers) would make a fair profit from that toll.
But what if it was a scholarly article that only ten people would ever want to read, and from which the author would never make a penny? Even on this commercial tollway model there is no way to make it rational to squeeze a further fee out of the would-be esoteric reader. It's a foregone conclusion that the author, eager to be read, would happily spring in advance for ten complimentary tickets for those few who will ever want to see the show. In short, the trade model makes no sense for esoteric publication EVEN ON A COMPLETELY COMMERCIALIZED NET, and there is still no money for the publisher to make for the sale of words no one wants to buy!
The real service that esoteric publishing provides is to authors, their institutions and their research funding sources: The "product" they provide is not the author's words for those who want to buy them but the means of access to the eyes and minds of the author's fellow-specialists. Hence that's the natural place to seek to recoup the true expenses of providing that service. I hope it is becoming evident by now how hopelessly Procrustean a trade model is for a product/service/market like this.
I'm genuinely surprised that you appear to think that it will be
politically acceptable in the United States to make most scholarly
information a non-tradeable commodity. I don't think it will be here.
The United States has little to do with this. The question is: Why and for whom do scholars publish? The answer is radically different from the answer to the question of why an author who makes a living by writing publishes. The esoteric market will simply reflect this, once it has been released from its Faustian bonds to the commercial market. "Trade" there will still be, but it will be the selling of the services (25% of paper costs per page) of the esoteric publisher to the esoteric AUTHOR and his institutions rather than to the esoteric READER and his institutions, as it was under the Faustian model. There is a "subsidy" in both cases: Institutions provided it through their library subscriptions the old way; the new way, much less expensive, the (25%) subsidies will be up front. And apart from the Institutions having to pay much less for the availability of the esoteric scholarly corpus, individual scholar/readers will be the greatest beneficiaries, now able to search, browse and read one another's work in the Virtual Library to their hearts and minds content, never having to worry about paying a toll to go where virtually no one cares to go anyway.
What possible objection (or role) could the United States have in something like that?
- Maybe these few remarks have irrevocably confirmed my typecast for
you! They are among the reasons why I think evolution is a better mode
than revolution and why I think that the economic problems which
libraries and others face in the scholarly communication field are
unlikely to be resolved by an agreement that it should all be for free
in an electronic setting. -- Bernard Naylor
For the record, I'm for evolution rather than revolution too, and I have my fingers crossed for a peaceful, fair and orderly evolution rather than a revolution that produces casualties to anyone. And I'm also for paying the true costs of refereed electronic periodical publishing (they are low, but they are not zero); I am simply opposed to having them paid (and/or surcharged) on a completely inappropriate trade/subscription model (subsidized by University Libraries). Everyone's interests would be better reflected and served if they were paid an author/subsidy model.
Stevan Harnad
Date: Wed, 10 Aug 94 11:45:34 -0600
From: Paul Ginsparg 505-667-7353 >
Subject: Re: On Trade vs. Esoteric Publication
stevan, (i know i promised e-mail blackout, but i found myself with a bunch of networkphilic physicists so it became reasonable to restructure the local environment with a lan and internet feed...)
just a few comments on the most recent (as usual from my pragmatic and non-visionary approach -- in current commercial parlance "just do it"):
From: "B.Naylor" B.Naylor@soton.ac.uk
Date: Thu, 4 Aug 94 17:38:16 BST
4.1 On the question of technology, I am reasonably confident that I
have not come across any technological problems which I don't think can
be solved. There are still important questions as to: how soon? and by
whom? and at what cost? At present, the "information superhighway"
itself does not exist internationally in a form which could cope with
the information traffic currently carried (not very efficiently) by
print on paper (the "fat pipe" under the Atlantic at 1.5 Mbps is surely
hopelessly inadequate?) and the other problems I referred to in my
paper are also with us and have to be resolved if an electronic
solution is to operate satisfactorily. Print on paper is currently
carrying a colossal amount of information into all sorts of unlikely
academic places.
i have always been perplexed by what goes into other people's calculation of bandwidth. finally i realized that they are talking about transmitting uncompressed bitmaps: an 8.5" x 11" page scanned at 300 dpi is roughly a Mb. but remember that all the white space compresses extremely well so this is already a grotesque overestimate. moreover in an earlier segment of this thread, we pointed out that the "scan and shred" technique was backward-looking, just adding additional expense with network distribution an ad-hoc afterthought. the savings are enormous if one avoids the paper stage entirely, instead retaining the electronic form the documents typically already possess.
(for comparison a page of ascii is 3-4kb, and less than half that after compression -- so we typically gain a factor of 500 over the raw bitmap. using a sophisticated page markup language such as compressed postscript [or preferably its successor [Adobe's pdf] with full font and graphics capability, the savings remain dramatic.) yes it will be a major task to render all the currently existing "print on paper" to network distributable form if that is desired, but we need to point out repeatedly that the issue of costs under discussion centers on a medium that is text from inception to final distribution. even the "fat pipe" as a bottleneck is a red herring in any event, one can always run mirrors (as i run in italy and japan) to ensure adequate bandwidth -- that way only a single transfer across the weakest link is necessary.
and speaking of "unlikely academic places", i am currently organizing a physics summer school in the french alps (les houches, next to chamonix -- i had expected to be out of network range but found an old hp 715 here and together with some other stone age implements forged a 64 kbit/s internet link -- say goodbye to minitel) and the students are from all over the world. the most common comment regarding the physics archives is how much they have already improved the situation in developing and former eastern-block countries, where the "colossal volume of print on paper" does not penetrate due to cost and other issues. at lanl, the systems i'm running still consume less than .01% (i.e. .0001) of the lanl.gov backbone capacity so we really do realize stevan's "flea on tail" metaphor.
When I had to cancel some physics journals a few months ago, our physicists
gave me the impression that I was bringing the roof down on their
department, and Paul Ginsparg wasn't mentioned once in their
remonstrations - though I've no doubt they're well aware of what he's
doing.
well it is still premature. note that the current physics archives branched out from high energy particle physics and do not yet cover close to the whole of physics. this was primarily due to lack of resources (i.e. zero) at my end, a situation recently rectified, and when i get back to the u.s. in mid sept there will be a dramatic horizontal expansion. (but even so it may be a couple more years before you can cut physics journals without complaint.)
4.3 I recently asked a group of about twenty chief university
librarians (that is, about 20 per cent of the total UK cohort) "by what
date do you expect your present number of current subscriptions to
print-on-paper journals will have been reduced by 80 per cent". The
extremes were "1995" and "2050" with 2010 by far the favourite
prediction (2010 was mine as well - honestly). (I said I thought that
might be a way, for some of them, of saying "after I've retired".)
your colleagues are incorrect. the driving force will not only be economics but the enhanced functionality of the electronic medium. there are many things that the new medium supports (see, for example, http://xxx.lanl.gov/hypertex/ ), including the overall fluid nature (on-line annotations, continuously graded refereeing, automated hyperlinks to distributed resources including non-text based applications, etc., etc.) that simply have no analog in print. it will be more or less like moving from radio to television -- radio remains for those things for which it's better optimized, but the majority of new material will move to the new medium.
5.1 One of the points I'm trying to underline in pointing out that
journal publishing is an industry is that it will not sit quietly by
and let itself be subverted. We must assume that, if a concerted plan
emerges to cut major traditional publishers out from the knowledge
communication business, they will fight very strongly for their "share
of the action". Commercial publishers will have revenue streams to
defend, not least in the interest of the people they employ and of
their shareholders, and even learned societies and academic presses
could face massive upheaval if revenue-earning titles (which are also
circulated gratis to some as a privilege of membership) appear to be in
danger of complete substitution, and they will fight to prevent that.
this will proceed quickly. the subversion is that their bottom line will be removed.
formerly we were at their mercy because we needed their production and distribution facilities. now we can outdo them on both counts, at dramatically reduced cost. at the same time, we expose how little intellectual added value they provide in general (i mean the validation and identification of significant research which can only come from within the community.) sure they could remain in the game if they were willing to scale down to the more efficient operation enabled by the fully electronic medium, but the bottom line will not be there for them and they will have little ability to compete with a streamlined operation organized by researchers in alliance with their research libraries (and perhaps non-profit professional societies).
5.2 One of the main points that publishers are likely to raise in this
country, and probably in Western Europe generally, will be to question
(to put it no more strongly) the propriety of academic institutions
using public money (all UK universities bar one very small one are
funded by the taxpayer) in order to drive a viable industry (as they
see it) to the wall.
but why is it currently viewed as appropriate to use gov't funds to sponsor this same "viable" industry in the form of overhead on grants that eventually makes its way to research libraries for transfer to them???
re:
Date: Tue, 9 Aug 94 17:12:04 EDT
From: "Stevan Harnad" harnad@princeton.edu
Paul Ginsparg's HEPnet currently gets 35,000 "hits" per day -- 35,000
physicists the world over retrieving articles.
one small clarification: actually it is just over 20,000 users (and more than just physicists since it has branched out into other fields including computational linguistics and economics). the 35,000 "hits" per day include all variety of searches, etc., not each an article retrieval.
Paul Ginsparg
From: amo@research.att.com (Andrew Odlyzko)
Date: Sun, 14 Aug 94 06:57 EDT
To: Bernard Naylor B.Naylor@soton.ac.uk
Subject: Balance Point and the economics of ejournals
Bernard,
Thank you very much for your comments, and please excuse the delay in responding to them, but I was away on a trip when they arrived. In the meantime, Stevan Harnad and Paul Ginsparg have responded to your message. I agree with what they say, and have only a few minor comments to add.
- Inadequacies of present networks:
You are right that the present 1.5 Mbps pipe over (or under) the Atlantic is inadequate for full-scale scholarly communication. It can carry about 5 TB (tera-bytes) in a year. In my article I estimated that just the mathematical literature alone requires about 1 TB to store (with fax compression). However, network speeds are increasing at dramatic rates, and soon we will have an adequate infrastructure in place.
As an aside, we can obtain much higher quality of material and lower communication burdens by converting the old documents into TeX, say. This is not as hard as it might seem. In mathematics there are about 20 M pages of printed material. At the rates that skilled typists in the US command, it would cost $ 200-400M to typeset them into TeX. By going to the Third World we could lower this to the $ 50-100 M range. (Obviously there aren't enough skilled typists in the Third World or even in the industrialized world to do this quickly, but the conversion could be done over 5-10 years, which would offer opportunities to train the necessary labor force.) For comparison, the existing mathematics print journals cost about $ 200 M per year. Thus a fraction of the annual cost of today's system would suffice to convert all the literature to a modern format. I expect similar estimates apply to other fields. The trouble would be in organizing this conversion effort.
- Time scales:
Here are some comments on what you wrote:
4.3 I recently asked a group of about twenty chief
university librarians (that is, about 20 per cent of the
total UK cohort) "by what date do you expect your present
number of current subscriptions to print-on-paper journals
will have been reduced by 80 per cent". The extremes were
"1995" and "2050" with 2010 by far the favourite prediction
(2010 was mine as well - honestly). (I said I thought that
might be a way, for some of them, of saying "after I've
retired".) There is some belief out there that the present
system will collapse and change suddenly quite soon but
it's not very widespread - though it may be right.
Otherwise, it implies that Southampton University Library
must shed 300 print journals every year from now till
2010. At present, that does not look very likely and this
means the process will have to accelerate substantially
(and I expect it will) later in the period.
I am rather surprised that so many of these librarians picked 2010 as the date for a major change. I would have expected them to be much more conservative.
I agree fully with you and Stevan that the drop in paper journal subscriptions will be very nonlinear. The most dramatic part of the drop is likely to occur between 2000 and 2010. I would be surprised if it occurred this decade, since networks and computer capacities are not adequate yet, and there is tremendous inertia in the system. On the other hand, it's hard for me to imagine print journals surviving more than 15 years in large numbers, as by 2010 the world will be fully "wired."
4.4 Re-reading some of what you and Andrew have written
makes me wonder whether we are really that far apart. My
job is to cope with today's reality (which is largely, like
it or not, both papyrocentric and commercial) and try to
anticipate the next few years' (say, five or at most ten)
changes reasonably intelligently. You and Andrew appear to
be speculating about a revolution which I am satisfied will
come -- though I seriously doubt your present economic
assumptions -- and I sometimes get the impression that
Andrew might not disagree strongly with my speculation and
the speculation of my librarian colleagues as to the time
scale. I'm less sure, Stevan, about your views as to
likely time scales.
You are right, we are not that far off in our opinions. My essay looked at the future about two decades from now, when all the novel features should be in place. Its aim was to show people what will be available then, and why the present system is bound to collapse. It dealt hardly at all with how to get there from here. That is a thornier issue, and one I now have to devote some thought to, as I am on some committees that are supposed to make recommendations for near-term actions.
The transition to electronic publishing is likely to be turbulent, and I do not wish to make it sound too easy. As just one example, it is likely that various administrators will seize on the projections of low-cost electronic publishing of scholarly journals and decide that this will allow them to save the bulk of their libraries' costs. However, while electronic journals can easily eliminate or at least drastically lower journal subscription costs, these costs are at most one third of the total cost of running research libraries. Since conversion of libraries to digital formats is going to be a long process, immediate savings are likely to be considerably smaller than such administrators might hope for.
Best regards,
Andrew Odlyzko
Forward to Chapter XV
Backward to Chapter XIII
Share
Twitter
Facebook
YouTube
Flickr
Assessment Blog
Policy Blog
E-mail
RSS