by Henry Small
Let me say what a pleasure and honor it is to receive this award. I want to begin by thanking the person without whose support I would not be here tonight, my wife Lois. She has given me the confidence and purpose to move ahead in my work.
Like so many information scientists, I started out in chemistry. After a master's degree, I switched to the history of science and eventually earned a doctorate.
Thus when I finally arrived in the field of information science, I had a lot of extra baggage. This, perhaps, allowed me to view the field from a somewhat different perspective. For example, I saw a database more like a sociologist sees society, as a complex social system. I wanted to know what role information played in scientific discovery and creativity, why and when scientists communicate and how fields of science grow and develop.
Lessons from the History of Science
Actually, the history of science provides many insights into information and communication. One of my favorite examples is the use of anagrams to communicate scientific discoveries, a practice popular in the 17th century. In those days, a scientist might submit a discovery as a coded message. This way he could claim credit for the idea and at the same time not reveal it to his colleagues, the opposite of our modern concept of open publication. Clearly, this was the result of inadequate institutional mechanisms in those days for insuring proper attribution of credit and authorship of ideas, but it also highlights the powerful drive for individual recognition and identity, even as applied to something as lofty as the pursuit of scientific truth.
I began thinking about what I could do with scientific information as a student in the history of science. My major professor also taught a course in chemical information. Seeing a couple of shelves of chemistry journals in his office, I asked with typical graduate student impertinence, "Why do we need the history of science if we have all that stuff?" He proceeded to lecture me, but I saw in my mind's eye a map of the entire scientific literature.
Thomas Kuhn introduced the idea of structure in science in the 1960s. Kuhn's paradigms were snapshots of the structure of science at specific points in time. By flipping through them you might spot a revolution in the making. The problem was that no one knew how to build the camera. In a postscript to his famous book, Kuhn (1970) made note of citation linking as one possible way to get at the problem.
After becoming a certified historian, I spent a couple of years attempting to document the history of nuclear physics at the American Institute of Physics. I soon realized that writing the history of a 20th-century field wasn't easy, as you in ASIS are also finding out in your history project. At AIP I started talking to information scientists. You have two choices, they said: "Build your own database or go work for ISI®." I did the latter.
I saw my first glimpse of the power of citations in the work of Sam Schiminovich (1971) at AIP who was using Mike Kessler's bibliographic coupling (1963).
When I landed a job at ISI, I couldn't wait to get my hands on the database to study the structure and development of science. What I didn't know – but perhaps should have – was that Eugene (Gene) Garfield was interested in the same things. He had made the important connection for me between citation data and the history of science in a report co-authored with the late Irv Sher (Garfield, 1964).
I was very fortunate to become an associate of Gene Garfield. What I learned from Gene is difficult to put into words. First, there is the importance of doing good research; second, of going out on a limb for something you believe in and doing your damnedest. Gene has always been passionate about his projects, in addition to being a great innovator in his field.
As you know I'm a big fan of the ISI database – now dubbed the Web of Science® – especially the way it captures the information behavior of scientists. It comes closer than any other database to reflecting that society of science I was seeking. Without it, my work would have been impossible, and it has been my main inspiration. It is undoubtedly one of the great information innovations of the 20th century. It was laying out proto-information highways long before the advent of the Internet. It has also set international standards for the measurement of science. I want to acknowledge here the role of David Pendlebury and other members of the Contract Research Department at ISI in creating our science indicator products.
It took me only 30 years to realize why citations are so powerful. It has to do with the fact that science is a collective and collaborative process: a lot of independent and creative people around the world, sharing their work through open publication – just like ASIS. The lowly citation facilitates this by allowing one scholar to embed another's work in his or her own, creating an extended, collective argument. Reconstructing these collective arguments is a difficult, but important research task.
Since citations are in many cases appeals to supporting evidence, science evolves tree-like with roots deep into the past. The interweaving of branches from adjacent trees in the academic forest gives strength to the cumulative growth.
The reason the citation index was the ideal database for mapping and measuring science was that, like a relevance judgment, a citation is a purposeful selection of another document. This selection process is complex, but operationally we have an elemental interaction, a binding of two information molecules. The question was: Could we build this up, document by document, until we got a complete picture? The first step for me was co-citation. As I have noted elsewhere (1992), the linguist Bar-Hillel (1959) had suggested a similar approach that he called co-quotation, while at the same time despairing that it would ever work.
The Shoulders of Giants
To explain co-citation, you may have heard Newton's well-known phrase, so brilliantly deconstructed by Robert Merton (1965), "If I have seen further, it's because I've stood on the shoulders of giants." If standing on the shoulder of a giant is a citation, then co-citation is straddling the shoulders of two giants, a pyramid of straddled giants is a specialty, and a pathway through science is playing leapfrog over the giant straddlers. Regrettably, Merton's research did not uncover any historical images of such collective giant straddling.
Soon after I arrived at ISI, I had my second stroke of good fortune in meeting a certain professor at Drexel University by the name of Belver Griffith. Belver and I shared the vision of mapping science with co-citation, and he contributed a great deal to the technique. For example, he introduced the idea of using multidimensional scaling to create a spatial representation of documents. I recently had the pleasure of writing a tribute to Belver on the occasion of his receiving the Derek Price Medal from the journal Scientometrics (Small, 1997).
Belver was also a close friend of the late Derek Price. Derek visited Philadelphia often in the 1970s to salvage whatever scraps of data he could from the cutting room floor, as Derek called it, and also to inspire intellectual mayhem. His modeling of the research front had a major impact on my thinking (Price, 1965).
It was exciting to find that co-citation arrayed highly cited documents so cleanly into specialties, and that, in turn, specialties could join to form a global map. And we did spot a couple of those elusive scientific revolutions. For interpretation I turned to semiotic theory to show how documents stood for ideas. Regarding this work, I plead innocent to being a post-modernist. On the contrary, I believe we were all borrowing from Robert Merton. We have yet to fully appreciate the contributions Merton has made to the varied ways that ideas can be transmitted and transformed on their journey from scholar to scholar.
From philosophy I learned that science was supposed to be somehow unified. If you combine this with Vannevar Bush's idea (1945) of associative information trails, then we ought to be able to navigate from one end of science to the other. E.O. Wilson in his new book on the unity of science (1998) thinks of this as following Ariadne's thread through the labyrinth. I think of it as the circumnavigation of science. What I hope to learn on that journey is how scholars occasionally reach out beyond the confines of their specialties to make connections with knowledge in neighboring fields.
It is no doubt true that our current maps of science are as primitive as the early Ptolemaic maps of the world. But the more we study this large scale, interdisciplinary structure, the better I believe we will understand such fundamental issues as the value of information, the discovery process, the relevance of one piece of information to another and perhaps even the social basis of scientific conviction. This applies to the Internet, too, where the hypertext link is a close relative to the classical reference.
One way to reflect on a career is to recall the problems you thought you were working on, and the solutions you thought you found to them. The older we get the more it seems we focus on what remains to be done rather than what we have accomplished. I am only too aware that the research program I set for myself 30 years ago remains incomplete, but, I hope, ultimately attainable.
The following is an outline, with the benefit of hindsight:
nDefine a fundamental unit of scientific knowledge. For me this was the highly cited document, after initial false starts using indexing terms.
nDefine a quantitative measure of the relationship of these fundamental units to one another. It seemed natural to adopt a measure of co-usage, namely co-citation, which was community dependent.
nAnalyze the structure of these units based on the measure of relationship and interpret their significance. Clustering methods were an obvious choice and it turned out that the aggregates corresponded to research specialties, rather than entire fields or disciplines.
nAnalyze change over time in these aggregates to see if new discoveries and perhaps mini-revolutions, intellectual shifts or other life-cycle phases can be identified. Case studies validated some of these interpretations using the perceptions of participating scientists.
nConstruct the macro-structure of broad disciplines and ultimately science as a whole by successively aggregating the individual specialties. This process led to the creation of maps of science that included a wide range of areas of natural and social science.
nUnderstand the dynamics of science at the macro-structure level; that is, how does science evolve as a whole? This task has proved the most difficult and remains incomplete.
Many issues arise in the pursuit of this research agenda, of which I will only list four:
nimplications of the bibliographic unity of science for the philosophy of science and the debate about the rationality of scientific belief, as posed by the social constructivists;
nimplications for information retrieval, including the possibility of creating document pathways connecting different subject areas, and the understanding of information transitions over space and time;
nimplications for artificial intelligence and scientific discovery where, following Swanson (1987), we seek to make new discoveries by navigating between existing but disconnected literatures; and
nimplications for science policy, where we identify emerging and rapidly changing fields, as well as the cross-fertilization of fields and interdisciplinary dependencies.
The information age is in full swing. The field of information science is experiencing bone-jarring change but at the same time is on the threshold of enormous opportunity. The boundaries of information are also expanding rapidly, in formats, types of users, goals, expectations and sheer volume of data.
I sense that the diversity of interests within ASIS is greater than ever, which is as it should be. However, there is a danger now of a loss of focus or a splintering into narrow interests groups and fiefdoms, just at the moment when we need to learn from each other. To combat this, I favor a more creative mixing of interests, rather than strict adherence to topics and themes, both in the annual meetings and the journal.
My work on the structure of science has led me to believe strongly in the idea of unity in diversity. Yes, we in ASIS do have diverse interests and agendas, but there are pathways that connect us. It is the task of ASIS to blaze these trails and bring us together in new and creative ways.
Bar-Hillel, Y. (1957). "A Logician's Reaction to Recent Theorizing on Information Search Systems." American Documentation, 8(1-4), 103-113.
Bush, V. (1945). "As We May Think." Atlantic Monthly, 176(1), 101-108.
Garfield, E., Sher, I. H., & Torpie, R. J. (1964). The Use of Citation Data in Writing the History of Science. Philadelphia: Institute for Scientific Information.
Kessler, M. M. (1963). "Bibliographic Coupling between Scientific Papers." American Documentation, 14, 10-25.
Kuhn, T.S. (1970). The Structure of Scientific Revolutions, 2nd edition. Chicago: University of Chicago Press, p. 178.
Merton, R. K. (1965). On the Shoulders of Giants. New York: The Free Press.
Price, D. J. D. (1965). "Networks of Scientific Papers." Science, 149, 510-515.
Schiminovich, S. (1971). "Automatic Classification and Retrieval of Documents by Means of a Bibliographic Pattern Discovery Algorithm." Information Storage and Retrieval, 6, 417-435.
Small, H. (1992). "Cogitations on Co-citations." Current Contents, #10, p. 20.
Small, H. (1997). "Comments on Belver C. Griffith, recipient of the 1997 Derek de Solla Price award." Scientometrics, 40(3), 359-362.
Swanson, D. R. (1987). "Two Medical Literatures That Are Logically but Not Bibliographically Connected." Journal of the American Society for Information Science, 38(4), 228-233.
Wilson, E.O. (1998) Consilience: The Unity of Knowledge. New York: Alfred A. Knopf, Chapt. 5.4
Henry Small, director, corporate research, Institute for Scientific Information, is the winner of the 1998 ASIS Award of Merit, the Society's highest honor. Dr. Small was honored for his outstanding contributions to the field of information science and technology and for his groundbreaking work on co-citation as a dynamic measure of the connectedness of scientific literatures. Dr. Small made the following remarks upon accepting his award at the Annual Meeting.