Science in the Twilight Zone -- Rebuttal

First, the scientific societies have kept silent on library funding for decades. Is this intentional? No one could imagine that they are not aware of the atmosphere of crisis and triage. I have puzzled over this. The advantages societies enjoy, such as tax exemptions, tax-deductible donations, bequests, dues, special postal rates, grants and other subsidies, and political recognition as voices of the research community, are meant to serve society. Their general objectives, the advancement and diffusion of knowledge, proclaim that their mission is broader than their business interests. Is this true? Writing in The Scientist I addressed the American Institute of Physics [AIP] and American Physical Society [APS]:

"I challenge AIP and APS to explain why they have not sought a solution to the library crisis that would benefit everyone." (Henderson 1998)

The reply, from AIP executive director Marc Brodsky, stated, "AIP has taken the routes available to it to help hard-pressed libraries...." (ibid.) In short, he focused on pricing and business competition rather than explain the 25-year old unarticulated (therefore secret) policy of silence that impacts present and future members, students, and the sponsors of research.

There is more such evidence:

• A letter from the head of the American Chemical Society [ACS] department of government relations and science policy informed me, "ACS has decided not to be an active participant in this discussion ... and will be back in touch with you if the Society decides to take an active interest in the library funding issue." (July 1, 1993)
• The head of the American Mathematical Society expressed a similar position, with no explanation, by telephone. (July 8, 1993)
• Government agencies toe a parallel line. For instance, an assistant secretary of the Department of Education advised that a rigorous regulation of indirect costs, "could unnecessarily interfere with decisionmaking ... [and] be viewed as an inappropriate intrusion." (Dec. 30, 1992) I see this sort of "academic freedom" protecting mediocrity rather than excellence.
• The head of the National Science Board responded to my inquiry by detailing its interagency activities, proclaiming leadership, and avoiding any word regarding the libraries used by its academic researchers. (March 25, 1996) I see this as government permitting its contractors to provide $15 billion services with inadequate quality assurance.

Why have scientific societies withheld using their influence to support library funding, to reform indirect library costs, and to solve the library crisis? What happened to assessment and accreditation? Why have they kept their reasoning secret? Why have government agencies not intervened?

The societies appear to be preoccupied with an often bitter competition for academic library business and for papers (Henderson 1998; Henderson 1997). Perhaps they harbor some hope that academic libraries and authors will give them a monopoly over all research in an apocalyptic shakeout over price-per-kiloword. Perhaps they believe an economically discouraging market may stop young talent from straying and starting new journals, financed by commercial interests, that siphon important papers away from association editors. This sort of anticompetitive solution would sacrifice customers - libraries, researchers, and research sponsors - principles, and progress.

In short, an "honest assessment" is unlikely until the societies explain themselves with candor.

Second, some of the problems of the associations started with their decision, not their inability, to start new journals that reflect the "twigging" of the tree of knowledge. No one can blame commercial publishers for catering to unsupported interests of researchers. ACS announced a moratorium on new journals 25 years ago (Ballhausen 1973). Other associations followed a similar path. American Physical Society, for instance, held Physical Review at 4 parts for decades, growing it to 80,000 pages after nearly doubling its price in 1970 by subdividing it (Bulletin of the American Physical Society 1971, 700-702). Only in 1993 did it spin off a 7,300 page "interdisciplinary" 5th part. The catalogs of AIP claim a fourth of the world physics literature via only 80 journals. This strategy forces authors to crowd into a handful of journals with a loyal library following rather than risk a 320-page startup.

Third, crowding papers expands journals' size, delivering more pages each year. Science increases about 5% annually in constant dollar funding, personnel and journal articles (a stable trend observed for over 300 years). The figures quoted by Dean Mobley that I find in the U.S. Periodicals Price Index cover U.S. associations but not foreign commercial publishers. (American Libraries. May 1996) Foreign journals also suffer from the added boost to prices by the considerable devaluation of the dollar. Demands by libraries for electronic editions have introduced new costs and, obviously, increases in prices. It is reasonable that journal prices outrun the Department of Labor's Consumer Price Index; they have unrelated purposes and bases.

Fourth, the library crisis itself forces sharp price increases. Journal (and monograph) prices are propelled upward by sales reduced by libraries' failure to expand at the same rate as research. This malaise cut across the entire spectrum of arts and sciences, affecting monographs as well as journals, evidenced early on by the 1979 National Enquiry on Scholarly Communication and more recently a 1997 conference in Washington DC. (Johns Hopkins University Press. 1979; {http://web.archive.org/web/19980513072318/http://www.arl.org/newsltr/195/osc-sch.html}) D. W. King et al. observed that a loss or gain of sales has a substantial effect on the relationship of income to cost, particularly where average fixed costs represent more than 70% of the total (King 1981). The impact of every cancellation on price is substantial, as I pointed out in further detail in Serials Librarian (Henderson 1992). Average sales of scientific and technical monographs fell from 3437 to 1894 copies between 1960 and 1975 according to figures compiled by King Research. (Statistical Indicators of Scientific and Technical Communication. NTIS PB-278279) University presses indicate their average printing is around 600 copies now. Journals suffered similar losses.

Eventually ACS was compelled by its members (not by librarians) to launch a new journal. On a visit to ACS headquarters in 1981 I learned it was to start its first new journal in more than a decade in the narrow but "hot" niche of organometallics. A commercial publisher had taken it over, drawing many papers. I recall their surprise a year or so later that libraries were unwilling to subscribe while many members placed orders at deep discounts.

Such discounts drew fire in the 1970s, when librarians first faced huge budget cuts. It may have precipitated the custom of blaming publishers for library impoverishment. Ultimately differential pricing was the basis of an Internal Revenue Service [IRS] action against ACS. IRS alleged that member discounting was "inurement," putting ACS's nonprofit status into jeopardy for two years. (Chemical & Engineering News. Nov. 3, 1980. p. 8) My impression is that ACS member discounts increased from 75% in 1981 to 90% or so.

Fifth, many nonprofit publishers enjoy tax-related and other subsidies enumerated above. They also demand authors' payments to support production costs. Under idealistic rules for Federal grants, publication may not be contingent on payment of "page charges;" and, the author's institution is not obliged to pay. The latter part of the rule sent many associations into the red in 1968, when universities, running scared at Lyndon Johnson's approaching retirement, decided to hold on to their cash. (Physics Today. Dec. 1968. p. 126-127) As a result of this and other concerns, associations have been reducing their dependence on page charges and offsetting lost income by raising library rates. (Physics Today. July 1986. p. 51-58)

Dean Mobley suggests that increasing numbers of research universities competing for limited grants contributed to the shrinkage of research dollars per university. This may be true, depending on which universities you are talking about. The top 30 out of 100 spend half of the total research dollars. (Science & Engineering Indicators. Appendix Table 5-4) There are also imbalances in growth. I counted 71 U.S. university members of ARL as of 1970, the watershed period. Now I count 106 - about 1.5 times more. U.S. academic R&D (constant dollars) increased by a factor of 2.5, enlarging both the dollar pie and its pieces. World science output increased more than 3 times. These indicators suggest to me that the library should grow faster than the university if it is to absorb and disseminate the work product of research. Universities held library growth far below this goal. The average "Research University I" (by Carnegie Classification standards) now contributes <1% of this world research literature. In order to support its programs, its off-campus users, and its future, it must purchase and maintain the remaining >99% via its library and other avenues of communication.

This leads me to repeat Vannevar Bush's emphatic dictum, "If the colleges, universities, and research institutes are to meet the rapidly increasing demands of industry and Government for new scientific knowledge, their basic research should be strengthened by use of public funds." (Bush 1945)

I believe that reasonable financial support for libraries' growth through Federal science policy would be there if universities and societies supported it. Science policy does recognize libraries as legitimate overhead spending. It fails to follow this to a reasonable conclusion, as I pointed out in my original article. The answer to my question, "are science libraries related to science?" may be "nevermore" - and a source of considerable grief to researchers, faculty, students, publishers, and librarians. And to taxpayers who are led to believe that U.S. science is based on rational methodology and principles of excellence. The fate of libraries was explained some time ago by University of Massachusetts librarian, Richard Talbot. (Bowker Annual. 1984 p. 74-82) He pointed out that libraries got less money than they need because they were forced to compete with the programs they are supposed to support. I would add that libraries also compete with an overly powerful bureaucracy.

Administrators distanced themselves - as Dean Mobley points out - by calling the serials crisis a "library problem." While blaming publishers for the sins of provosts, they foster other myths to promote their own interests. One of these is the claim of penury. In contrast to the fables believed by chemistry librarian David Flaxbart and Dean Mobley, the growth of academic R&D in the late 1970s and 1980s was healthy (though less robust than the 1960s). The growth of libraries, comparing constant dollar spending, was not. (Table 1) Professing financial necessity, universities promote government and foundation grants, donations, bequests, weakening of authors' exclusive rights, and, "relief from the growing costs of providing access to an ever-expanding volume of scholarly output" (Policy Perspectives 1998). They are not entirely reticent about library funding. Their premise is that they should not be held responsible for excellent collections, even if it means shortchanging students, researchers, instructors, and the taxpayers who finance research and student loans. They seem to say they "can't afford decent libraries."

What do we see? Universities choose to invest billions of dollars in stocks, bonds, and real estate instead of library collections and at the expense of knowledge. Many retained surpluses ten times what they doled to their beleaguered libraries. A dozen U.S. schools like Duke, Princeton, and Chicago cut library spending in 1996-97, according to Association of Research Libraries, while reporting excess revenues ranging up to 268 million dollars on IRS form 990. Others, like Case Western, Columbia, Cornell, Emory, Harvard, MIT, Notre Dame, Pennsylvania, Princeton, Rice, Stanford, Tulane, Vanderbilt, Washington (MO), and Yale unnecessarily held spending far below the six percent minimum recommended by experts. (Table 2) The total "profit" reported by all U.S. research and doctoral universities exceeded $3 billion in 1992-93 - more than enough to restore decent collections. (Table 3) This desertion of integrity makes the misposting of grant overhead (while "forced" to cut science library subscriptions) seem to me like a Three Stooges episode. What happened to the university's mission of research, education, and public service?

Another myth suggests that commercial publishers emerged recently, and took advantage of some incapacity of associations. Springer, Wiley, Williams & Wilkins, Lippincott, Mosby, and many others have been around for a long time, longer than most U.S. association publishers. Walter J. Johnson, who co-founded Academic Press with Kurt Jacoby after escaping the Nazis, considered himself a fifth-generation publisher. Commercial publishers have a long history, starting with the for-profit founding of the Philosophical Transactions of the Royal Society by Henry Oldenberg in 1665. Macmillan published Physical Review over 100 years ago.

The image of these publishers as "carpetbaggers" goes along with the myth that all good science is now done at American universities. Actually, two foreign papers appear for every paper authored in the U.S. according to the National Science Board. (Science & Engineering Indicators. 1998) The hapless situation is that many foreign papers are not found in U.S. libraries and are poorly disseminated. Foreign vendors are now major suppliers of photocopies. Do the journals cost too much? How do you figure? The winners of the 1987 Nobel Prize in physics, for instance, reported research stimulated by a paper in Revue de Chimie Minerale - a journal not to be found in the library of the U.S. center for superconductivity research, the University of Houston. The breakthrough was done by scientists in Switzerland. Foreign papers are not well cited or co-authored by U.S. researchers. (Science & Engineering Indicators) I believe the decimation of library collections contributes heavily to this deficit of insularity and loss of progress.

The myth of the commercial publisher as "profiteer" and "rapacious" is baseless and probably libelous. A "profiteer" is typically someone who sells scarcities, such as food, fuel, or safe passage illegally during wartime or disaster. Science information is not scarce; it is, if anything, in an economic surplus. I doubt that a "profiteer" can trade in "excess" goods as claimed, particularly when the customers deliberately created the economic imbalance. (I would also point out that if I am a "lobbyist," as portrayed by Flaxbart, I am a very poor one having no office in Washington DC; I would rather be described as an "advocate of dissemination.")

The Association of Research Libraries embarrassed some and offended others with such accusations based on spurious findings drawn from unavailable data. (Report of the ARL Serials Prices Project. 1989) Stanford economists R. Noll and W. E. Steinmuller gently reproached the study, "Several previous studies have attempted to calculate publisher profit margins. This line of inquiry remains problematic, even when all potential sources of increasing journal costs are taken into account." They added "... we discovered that all existing sources of data were inadequate for a reliable empirical study of the causes of price increases..." (Noll and Steinmuller 1992). The illegitimacy of the ARL claims certainly annoyed me, with no immediate interest except in fairness. The victims turned the other cheek. Nobless oblige. The damage done to the reputations of publishers and researchers, nonetheless, lingers on.

Commercial publishers, like the four houses singled out by ARL, do foster niche journals which are expensive. They invest in new specialties which societies would subordinate (or suffocate, according to some) in their omnibus products. They finance journals sponsored by societies or informal groups ("invisible colleges") of scientists. My experience indicates that groups of scientists are more likely to approach publishers with proposals than the other way 'round. Specialized publications have small circulations, by definition. They pay for translations and editorial work. They provide interested readers with a browsable focus on topical content. Commercial publishers also pay taxes. They maintain offices all over the world. They extend discounts to intermediary vendors. Many readers who favor niche journals over omnibus publications suggest that the price per "article of interest" makes them more valuable to the researcher than the behemoths that range over an entire discipline. The latter may be cheaper per-kiloword, but irrelevant material is just so much dry sawdust in the meat loaf. Such extraneous bulk forces every researcher to consult very expensive A&I services in order to find anything.

Report of the ARL Serials Prices Project also shamelessly attacked researchers, using the phrase, "excessive publication." I see the reasoning that tenure requirements exacerbate the library crisis as thoughtless and unbecoming an organization with "research libraries" in its name. Here is why: Derek de Solla Price's Law indicates that 25 per cent of scientists account for 75 percent of all papers (Price 1963). The majority of authors publish very little, perhaps one or two papers in a lifetime. Why would the prolific segment, who probably achieve tenure early and keep generating papers anyway, be concerned with the so-called need to "publish or perish?" Moreover, as spending in academic research increases (about 7 times in constant dollars since 1960), the work product must also increase. The ratio of authors to papers is relatively constant and is defined by Lotka's Law of Productivity - also noted by Price (Price 1961).

Professor Rob Kirby praises the "xxx" information exchange as a potential solution to researchers' needs. It also resonates with the desire of administrators for "relief from the growing costs of providing access to an ever-expanding volume of scholarly output". (Policy Perspectives. ibid) The "xxx" solution is, of course, too good to be true. National Institutes of Health and Atomic Energy Commission both tried similar programs in the 1960s, emphasizing dissemination of preprints. They quickly collapsed (Abelson 1966). The longevity of "xxx" at Los Alamos National Lab is a testament to the energy of Paul Ginsparg, who persuaded the government to go into the document delivery business in competition with the private sector. Its most attractive feature is also its Achilles' Heel: "xxx" supplies for free at taxpayer expense reprints that others sell. Our society favors market-oriented solutions over government welfare, particularly where the private sector is fully capable of providing an equal service and "welfare clients" are wealthy and located largely outside the United States. The "xxx" project will probably become politically detestable as the mountain of unrefereed content archived on government servers grows like all science, exponentially, and universities continue to pile up billion dollar profits.

More important, the "xxx" solution is no journal and it is no library - not even close. Journals offer an aim and scope, a point of view, and often considerable information and service beyond authors' reports. Libraries provide reference works, browsable collections of journals, monographs, and personal guidance as well as resources too expensive and too broad to interest an individual, a department, or a laboratory. We are a very long way from replacing a good science library staffed by experienced professionals with an electronic box stashed under a desk -- although that solution appears to have its supporters.

Flaxbart suggests that scientists' natural conservatism is responsible for the tedious migration of research to all-electronic media. I believe that their diffidence is also the child of hostile economic conditions and the sister of industry consolidation. The investor must be able to forecast a return. More experimental investment in new media and faster migration to new technology would be encouraged by robust libraries. This was demonstrated in the 1960s, when bolder publishers first adopted author-produced cold type, photo-offset printing, and cataloging-in-publication. As it is, much of today's development depends on ephemeral grant money, government lobbying, and public relations concerns. Basic questions of obsolescence and archival quality remain unanswered. Why would anyone consider an outlet for formal publication when it might not be around in three years time?

My complaint focuses more on the preemption of resources based on presumptions that fail to address the productivity of research. In spite of passionate appeals by faculty and librarians, administrators cut library growth. I watched for 20 years as "access not ownership" displaced the Alexandrine ideal by executive fiat. They now have the brashness to ask for a specific demonstration of library value. (See T. Saracevic and P.B. Kantor. Journal of the American Society for Information Science. 48,6 p. 527-562. 1997) No research supported the sudden paring of library funding in 1970. A late effort aimed at excusing library cuts, known as the Pittsburgh Study was thoroughly repudiated. (For example: C. Borkowski and M. J. N. MacLeod. Library Acquisitions: Practice and Theory. 3 p. 125-151. 1979)

As a boon to science and scholarship, the promise of resource sharing clearly backfired. It decimated collections, forced prices skyward, narrowed database coverage, and put some publishers in the position of asking authors to pay them as they might a vanity press. It made "red-tape-and-wait" interlibrary borrowing equal to collections and drove many researchers to travel more. It discouraged publishers from producing reviews and references essential to improving dissemination. As a miraculous way to save money and increase resources, the "new paradigm" of "access not ownership" belongs next to atomic energy in the potter's field of false prophets.

If the current system of scientific publication is indeed broken, I believe it is the folks in the administration building who broke it. Economist David J. Brown suggests this is because they have no interest in progress (Brown 1996). Operating with unchecked power, they ignore all pleading - and certainly all logical arguments - from faculty and library specialists. Not until the societies, the government agencies, and the accreditation boards make this their official issue can I see science libraries and librarians restored to science.

Mr. Flaxbart demands better quality control of research. I agree. Improved quality requires better peer review. Referees who first pass on grants and then on publication must be fully informed. Instead, thanks in good part to the deterioration of libraries, they are as knowledgeable as (and perhaps worse than) the authors they judge. Without an up-to-date collection, they cannot even check citations. Referees are commonly given two weeks to pass on a submission. The latest survey indicates average ILL/DD performance of 97 ARL libraries was beyond such deadlines! [{http://www.arl.org/newsltr/195/illdds.html}] How are referees supposed to spot duplication, plagiarism, and error if they lack the essential resource? Undisseminated findings are the main obstacle to quality and productivity. Five years after publication, as news of discoveries and revised theories filters up through reviews, informal exchanges, and other communications, reviewers have more accurate opinions (Garvey 1979). Of course, five years after publication is too late. To fix the system we need more dissemination, not less; better collections, not more resource sharing theories. Science policy is the key, as I described in some detail in Society (Henderson 1998a).

This analysis of impact on the effectiveness of research is missing from all the calculations modeling "access not ownership." These estimate library costs of borrowing versus purchase by means of use studies. They ignore the consequential cost of "drivel" produced when information is unevaluated or unavailable to authors and referees. Preliminary studies suggest that half to three quarters of published research is "drivel." If the payload "per ton" of journal articles in terms of useful and reliable new knowledge was 1000 lb. 30 years ago, it is probably 500 lb. now and dropping. One day, productive research "per ton" may be measured in ounces. If academic research spending is $15 billion today, then $11 billion or more may produce "drivel," as characterized by Mr. Flaxbart, through lack of information.

How does the $2 billion spent on libraries at Research Universities look in the context of $11 billion spent to produce "drivel?"

A more fundamental problem with the technology-based alternatives proposed by Kirby and others is their focus on documents, rather than knowledge-oriented solutions. This criticism was raised by an elite presidential panel when the volume of science was about one-sixth the present. (President's Science Advisory Committee, 1963) Its recommendations are worth repeating:

"The technical community must recognize that handling of technical information is a worthy and integral part of science.... We shall cope with the information explosion, in the long run, only if some scientists are prepared to commit themselves deeply to the job of sifting, reviewing, and synthesizing information; i.e., to handling information with sophistication and meaning, not merely mechanically. Such scientists must create new science, not just shuffle documents: their activities of reviewing, writing books, criticizing, and synthesizing are as much a part of science as is traditional research."

It came as no surprise that the House Science Committee produced a policy document that ignored Speaker Gingrich's request to address the problems of bureaucrats and coherence (U.S. Congress 1998). Words like "libraries," "librarians," and "synthesis" are not to be found. Nor is there any indication of how to deal with the growth of scientific information! It casts its vision with familiar goals but mistakes public relations for basic tools of research:

Science, driven by the pursuit of knowledge, and technology, the outgrowth of ingenuity, will fuel our economy, foster advances in medical research, and ensure our ability to defend ourselves against ever more technologically-advanced foes. Science offers us an additional benefit. It can provide every citizen-not only the scientists who are engaged in it-with information necessary to make informed decisions as voters, consumers and policy makers. For the scientific enterprise to endure, however, stronger ties between this enterprise and the American people must be forged. Finally, our position as the world's most powerful nation brings opportunities as well as responsibilities that science and its pursuit can, and should, address.

I wonder how the House Committee believes science can provide us "with information necessary ..." without the library research that sifts and distills the massive literature into coherent digests of discovery and the libraries that organize and disseminate it.

The unsigned "Publish or Perish" white paper -sponsored by AAU, ARL, ACS, et al. - readily admits to the imbalance of research and library growth. Apparently confident that faculty will not soil themselves with administrative detail they offer a dare:

"If the answer is instead a bigger library budget, then the faculty must be central to the process of deciding which budgets are to get smaller." (Policy Perspectives. ibid.)

My response to this challenge is threefold.

• First, end the shell game of library overhead payments. Science policy must support libraries in proportion to their use by and the needs of scientists off-campus and on. Although these libraries are owned by universities, they serve as a national resource. The research community and the taxpayer has an interest in their quality. Policy must respond to the growth of world literature and the potential impact of devaluation of the dollar.
• Second, universities must invest more in libraries and less in real estate and financial markets. The library collection is the physical asset that separates research universities from trade schools. I don't believe scientists can attract serious science grants on the basis of trade school resources.
• Third, cut administrative spending across the board. Middle managers, vice-presidents, marketing, real estate speculation, the mail services, social events, and administrative travel would top my hit list. I would eliminate recent superfluous innovations like Yale's "vice president in charge of New Haven." I also would aggressively pursue more resource sharing and outsourcing of bookkeeping, accounting, payroll, and purchasing.

To sum up, I asked whether science libraries are a part of science. In response I got a productive question about the role of the societies spiced with the demonology of mammon and the miracle of the government trough. The answer must be that organizations seemingly anointed to a higher purpose will never see the land of milk and honey as long as they secretly sacrifice principle at the altar of the golden calf. Universities could spend more on their libraries, but they don't care about the contribution of library dissemination to the university mission as long as students and grants keep coming. Science is about information, both as ingredient and as work-product. When the library is perceived as a cost center rather than an asset essential to effective research and education, we are in trouble. Investments in research cannot be harvested unless they are thoughtfully sown and fully processed. The full range of academic scholarship is also at risk, as journals push monographs aside on the basis of research grant dominance. Societies of scientists that were chartered to address this very question also appear to favor commercial priorities. Associations of publishers are populated by these scientific societies, producing a certain ambivalence. Perhaps it is up to associations of librarians to defend the official status of science libraries in science - while they still exist.

References

Ballhausen, C.J. 1973. Too many chemical journals. Chemical and Engineering News. December 10, 1973.

Brown, David J. 1996. Electronic Publishing and Libraries: Planning for the Impact and Growth to 2003. Bowker-Saur, London, New Jersey.

Bush, Vannevar. 1945. Science--the Endless Frontier. U.S. Government Printing Office, Washington, D.C.

Garvey, W.D. 1979. Communication: The Essence of Science. Pergamon, Oxford.

Henderson, Albert. 1992. Forecasting changes in periodicals prices. Serials Librarian. 21(4): 33-43.

________. 1997. The ethics of the library crisis and the first amendment. Publishing Research Quarterly. 13(4): 31-40.

________. 1998. Decade-long legal battle focused on journal cost, impact. The Scientist. 12(2): 7-8. [Online]. Available: {http://www.the-scientist.com/?articles.view/articleNo/18779/title/Decade-Long-Legal-Battle-Focused-On-Journal-Cost--Impact/} [December 15, 1998]

________. 1998a. The incoherence of the new science policy. Society. 35(6): 38-43.

King, D.W. 1981. Scientific Journals in the United States: Their Production, Use, and Economics. Stroudsburg, Pa.: Hutchinson Ross Publishing Company.

Noll, R. and Steinmuller, W.E. 1992. Serials Review. Spring/Summer 1992: 32-37.

Policy Perspectives. 1998. 7(4): March 1998. [Online.] Available: {http://www.thelearningalliance.info/Docs/Jun2003/DOC-2003Jun13.1055537929.pdf. [December 15, 1998].

Price, Derek de Solla. 1961. Science since Babylon. Yale University Press, New Haven.

________. 1963. Little Science, Big Science. Columbia University Press, New York.

U.S. Congress. House of Representatives. Committee on Science. 1998. Unlocking Our Future. Toward a New National Science Policy. Sept. 24, 1998. [Online]. Available: {http://www.access.gpo.gov/congress/house/science/cp105-b/science105b.pdf} [December 16, 1998].

TABLES

Table 1.
Growth of U.S. Academic R&D and 41 ARL Libraries in constant dollars in 5-year periods 1970-75 through 1990-1995.

Table 2.
Private universities' profits (surplus revenue) 1997 & library spending 1996-1997.

Table 3.
Revenues, expenditures and profits (surplus revenues) for research and doctoral universities 1994-5.