Find DOE Collections
"On the Shoulders of Giants – tracing the impacts of
information retrieval systems on science policy"
Dr. Eugene Garfield
Oak Ridge, TN - September 18, 2007
The Office of Scientific and Technical Information (OSTI ) was about five years old when I first entered the field of scientific documentation in 1951. The launch of Nuclear Science Abstracts and OSTI in 1947 coincides with a key event in the history of science information systems –
SLIDE #1 Bernal and Garfield
the Royal Society Conference on Scientific Information held in London (1948) inspired by the polymath John Desmond Bernal about whom I shall say more later.
SLIDE #2 Ralph Shaw
Representing the United States were Ralph Shaw, Mortimer Taube, Watson Davis and
SLIDE #3 J. Murray Luck
J. Murray Luck – the founder of Annual Reviews. Some of the British information scientists were pioneers like Robert Fairthorne and Jason Farradane.
SLIDE #4 Cyril Cleverdon
I met them at the 1957 International conference on Classification in Dorking and then would meet JD Bernal at the International Conference on Scientific Information in Washington D.C. in 1958 – six years after the first Atom Bomb test.
When I asked Sharon Jordan what would be an appropriate topic to discuss at this celebration, she seemed preoccupied with the notion that I could tell you how the field of scientific information systems has affected the world of science. That information is the life blood of scientific discovery, seems axiomatic. I have often repeated that truism but like so many other common sense assumptions – like the virtues of motherhood – it is difficult to prove. There is a comparable belief that is also difficult to prove. I serve on the Board of a non-profit organization called Research!America (www.researchamerica.org ). Its main goal is to educate the public and the members of Congress that medical and scientific research should be expanded -- not only because scientific discovery and curing disease are important to mankind, but also because basic research has positive economic and social impacts. Indeed, Research!America sponsors an annual award to scholars who have demonstrated the economic benefits of biomedical research (http://garfield.library.upenn.edu/researchamerica/award.html )
We have no comparable award for demonstrating the impact of scientific and technical information systems on science progress.
Economists like Edward Mansfield have demonstrated that investments in research provide a 25% return. But that notion is not yet universally accepted or we would not be fighting to maintain our basic research budgets. The first five winners of the Research!America award have each published evidence to support the claim by placing a monetary value on human life, but it is not yet an accepted belief. There are still naysayers who feel we spend too much on research. If not, the research budgets for NIH, NSF, NASA and DOE would be increased automatically every year. In spite of the widespread public belief in the positive impact of research, especially biomedical, the U.S. is spending a lower percentage of its GNP on research than other countries, and less than we did in earlier years. The latest NAS report called this to the attention of the President. The last approved federal budget was slightly in excess of the already reduced amounts. It is relevant to mention here that small countries like Israel and Ireland, with their high investment in innovative research, have become economic power houses.
How can we demonstrate the impact of scientific information systems on the growth of science? This is, of course, a significant problem for science policy analysts. Under the leadership of Alvin Weinberg, the former director of Oak Ridge National Laboratory, the President’s Science Advisory Committee in 1961 issued the so called Weinberg Presidential Commission Report affirming the importance of science information systems.
Members of the Panel on Science Information (page 51) included, among others, Nobelists Eugene Wigner, John Bardeen, Joshua Lederberg. Noteworthy is the recognition given to Francois Kertesz of Oak Ridge National Laboratory who recorded and edited the panel proceedings. To my knowledge no one ever challenged the panel’s recommendations. But the report did not demonstrate a correlation between improved information activities and growth.
Those of us who were trained in chemistry have been taught that it is axiomatic that access to the record of past scientific research is essential. Without efficient information retrieval, chemical research would be slow and even more expensive. Who would want to repeat the research that has been reported over the past 150 years? Yet in 1965, John Martyn of ASLIB in the U.K. reported in the New Scientist that 25 percent of published research is, in fact, unwitting duplication. The implicit value of scientific information activities is reflected in the market place because the worldwide scientific community pays billions each year for access to what was described as the "World Brain" by H.G. Wells in 1938.
SLIDE #5 Paul Otlet & Henri La Fontaine
Today’s Google- and Wikipedia-generation may think that search engines are a new idea, but if you examine the history of encyclopedism you find that as early as 1913 Henri La Fontaine, the co-worker of Paul Otlet (the Belgian documentalist) received the Nobel Peace Prize for their efforts to create a centralized font of knowledge and information, accessible to all humanity.
SLIDE #6 World Brain Holographic etching
Twenty five years ago, as a testimonial to these pioneering efforts, I commissioned the artist-engineer Gabriel Lieberman to create the first holographic etching called the “World Brain”, which can be viewed on the Web and seen in person at the Chemical Heritage Foundation in Philadelphia, a stone’s throw from Independence Hall. (www.garfield.library.upenn.edu/art/isiart/hologram3.jpg)
To establish the impact of scientific information systems on science progress is like discussing the proverbial chicken and egg. Ever since the first scientific journals started in 1665, including Journal des Scavans in Paris and the Philosophical Transactions of the Royal Society in London, it has been axiomatic that research is not completed until its results are made accessible to the entire community. The American Philosophical Society was founded in Philadelphia in 1745 by Benjamin Franklin among others. Its Transactions, started in 1769, is the oldest scholarly journal in America. The exponential growth of the journal literature has been accompanied by enormous growth in the scientific enterprise as decribed in Derek Price’s classic work “Science Since Babylon” “ Little Science, Big Science”.
By the middle of the 19th century, scientists were aware of the need for systems that could make it possible to keep up with the published literature. It was not unusual for scholars to complain that it had become impossible to read everything. Abstracting and indexing services came into being, especially in Germany. Every chemist learns the names of Beilstein and Gmelin.
By the beginning of the 20th century, American science had grown to the point where it recognized the need for access to the worldwide literature, including patents as well as journal literature.
SLIDE #8 Fred Tate and Dale Baker
Last month, Chemical Abstracts Service celebrated its 100th anniversary.
SLIDE # 9 SCI Papers Covered 1955-2006
As I pointed out on Chemical Heritage day last May in Philadelphia, the growth curve for Chemical Abstracts is paralleled by the growth of the Science Citation Index. Derek de Solla Price used similar measures to demonstrate the exponential growth of science.
SLIDE #10 Garfield and Price
While the scientific and library communities in their typically conservative fashion took more than a few decades to accept the novelty of citation searching, it is now taken for granted. Indeed the advent of the internet and search engines like Google has increased the awareness of citation linking and ranking.
Today we are preoccupied with improving the quality of covert intelligence services in our effort to counteract terrorism. This is a modern repetition of our need to improve information gathering techniques. Few of you will have forgotten our shock when Sputnik was launched by the Russians. Our covert and overt intelligence capabilities were challenged then too. Everyone was studying Russian then. In 1964 I and others testified before Congressional Committees on the need for improving our overt information activities.
In 1955, I read a paper at the AAAS meeting entitled the “need for a national science intelligence service”. While we take for granted today’s multi-billion dollar information industry, back in the days of Hubert Humphrey’s Senate Committee and Chicago Congressman Roman Pucinski’s House Committee, there was a great clamor about our failing to recognize the scientific advances made by the Soviet Union which had established their VINITI in 1958. That same year I gave a paper at the 1958 ISCI conference on the need for a unified indexing system which might be compared with trying to get the intelligence services to talk to one another.
SLIDE #11 Garfield in front of log cabin
As a personal note let me recall that in 1960 I changed the name of my young consulting company – Eugene Garfield Associates, Information Engineers – to The Institute for Scientific Information.
SLIDE #12 HFC Certificate
I did this for two reasons… one was to offset the stigma of operating a private commercial for-profit company which published Current Contents. In those days the bias against private enterprise in the information business was real and tangible. While today the National Federation of Abstracting and Indexing Services (NFAIS) accepts
members from both the profit and non-profit world, that was not the case then.
SLIDE #13 Boris Anzlowar (one of five founders of IIA)
That is why I and four other individuals – Bill Knox, Boris Anzlowar, Saul Herner and Jeff Norton established the Information Industry Association. IIA is now called the Software and Information Industry Association.
The second reason I changed the name of my company was the fact that the State of Pennsylvania would not allow us to use the term “information engineers” to describe our activities, since I was not a graduate of an engineering school – this in spite of the fact that a few years later I taught a course in information retrieval at the University of Pennsylvania School of Electrical Engineering to graduate engineers.
When I entered the chemical information world, the importance of science information activities was already well recognized in industry. DuPont had its Scientific Intelligence Department. My first consulting client – Smith Kline & French – had its Scientific Literature Department as did other pharmaceutical companies. It was not uncommon for such departments to command 10 percent of the total R&D budget. And this was paralleled in the U.K. at companies like Imperial Chemical Industries (ICI), in France at Roussel Uclaf, and at Hoechst in Germany. Indeed, the international recognition of the value of scientific information as necessary for sci-tech intelligence was taken for granted. Information pioneers like Joseph Becker, Librarian of the CIA, were known to all of us in the library and information science world. Indeed the CIA library was the first paid subscriber to the Science Citation Index. The second subscriber was the National Library of the Peoples Republic of China in Beijing followed shortly afterwards by the Library of the Soviet Academy of Sciences.
As I recently reported in Dublin at a two-day celebration of the Irish- born polymath John Desmond Bernal, the first quantitative studies of science began with Bernal’s 1939 book “The Social Function of Science” (http://garfield.library.upenn.edu/histcomp/bernal-jd_soc-func-sci/). There were minor uses of bibliometric data reported in the previous decades. However, the “science of science” was initiated by Bernal. The term and field became widely recognized as serious science when Derek deSolla Price published his two books, “Science since Babylon” and “Little Science Big Science” in 1961 and 1963. In the years that followed, his descriptions of the exponential growth of science and other quantitative studies became known as “Scientometrics”.
SLIDE #14 V. V. Nalimov
While the term “Bibliometrics” was coined by Alan Prichard in 1969 it was the Russian polymath V. V. Nalimov who coined the term “Scientometrics” around 1968.
It is widely recognized that the growth of the field of scientometrics is due to the appearance of the Science Citation Index in 1964 which ultimately led to the founding of the journal Scientometrics.
Thanks to the translation of Nalimov’s book by the U.S. Government, it was known to a few Western experts but was not widely disseminated. Without the internet, few people knew the work and its historical roots. However, the advent of the journal Scientometrics in 1978 formalized the transition from the awkward term “science of science”.
SLIDE #15 Bernal
The Society for Social Studies of Science (4S) was established in 1975 and initially it showed an interest in quantitative studies and in Scientometrics, but that has ceased to be the case.
Thomson ISI remains the sponsor of the 4S Bernal Award
SLIDE #16 Merton, Zuckerman, Garfield
whose first few winners included Derek deSolla Price and Robert K. Merton and Thomas Kuhn.
SLIDE # 17 Thomas Kuhn
The advent of Science Citation Index in the sixties, of course had a major impact on the growth of Scientometrics. Without the SCI, scholars would not have had the database needed for conducting studies outside the traditional scientific boundaries of chemistry, physics, etc. The unique inclusion of citation linkages in SCI made it the major working tool of science policy analysts, which facilitated the science indicators movement. The NSF was quick to recognize that bibliometric data provided a measurable basis for conducting bi-annual reports on the state of science in the U.S. Norman Hackerman, who died this year at the age of 96, was Chairman of the National Science Board which issued the Science Indicators for 1976. Here is a quote from that report…
“In 1968 the Congress directed the National Science Board to assess the status and health of science, including such matters as national resources and manpower, in reports to be rendered to the President for submission to the Congress. In 1973, the Board initiated the Science Indicators series, and in 1976 a joint committee of the Congress indicated its continuing interest in this particular series. Science Indicators-1976 is the ninth such annual report and the third in the Science Indicators series. With it, the Board continues its effort to describe quantitatively the condition of science and research in the United States.”
As an extension of my earlier remarks, I thought it would be of some personal interest to say something about Alvin M. Weinberg, the erstwhile Director of the Oak Ridge National Laboratory.
SLIDE #18 Weinberg Report “Science, Government and Information”
Al was Chairman of the PSAC, that is the President’s Science Advisory Committee, which issued the famous Weinberg Report in 1963 entitled “Science, Government and Information.” This document, issued by the GPO, was reprinted in an early 1963 issue of the then new journal called Minerva.
SLIDE #19 Citation Classic “A” Criteria for Scientific Choice
SLIDE #20 Citation Classic “A” ENLARGED
In that same journal issue, Dr. Weinberg published his now classic paper on “Criteria for scientific choice” which went on to become a Citation Classic – posted at http://citationclassics.org with 4,000 Citation Classics Commentaries.
SLIDE #21 Citation Classic “B” – Science and trans science
SLIDE #22 Citation Classic “B” ENLARGED
In this slide and the next you see the Citation Classic commentary by Dr. Weinberg on his paper “Science and Trans-science,” also published in Minerva one decade after the previous citation classic. At the time it was the most-cited paper ever published in Minerva.
SLIDE #23 - Most-cited papers in Minerva
These two classic papers by Weinberg remain in the top 5 papers published in that journal. Reading his comments sixteen years later, one can see that he finally appreciated the difficulty in assigning value in science. It is not a coincidence that his remarks about Polanyi’s democratic republic of Little Science stood in contrast to the “socialistic” republic of Big Science. In that sense, Weinberg was implicitly a disciple of J.D. Bernal, the father of the social studies of science.
I just returned from a Symposium in Dublin, Ireland, celebrating the life of Bernal and used the occasion to show, by visualization techniques, the citational connections between his work in x-ray crystallography and that of his descendents – Rosalind Franklin, Aaron Krug, J.D. Watson, Francis Crick, and Maurice Wilkins. The latter three received the Nobel Prize for the Double Helix Structure of DNA. Bernal also produced five other students who won the Nobel prize.
Back in 1964, Dr. Irving Sher and I demonstrated how citation analysis could help trace the evolution of genetics from Mendel to Watson & Crick.
To conclude, it would have been fun to trace for you the numerous pioneers in the field of scientific information systems but this series of slides will give you a sampling of some of them beginning with James W. Perry whom I met in 1951 and then running through to Calvin Mooers, the inventor of the term “information retrieval”.