COMMENT OBITUARY Eugene Garfield (1925–2017) Inventor of the .

think you’re making history, Gene!” at conferences, making prototype indexes and So said Nobel laureate and molecular sending proposal after proposal to the US Pat- biologist Joshua Lederberg to his friend ent Office, the National Science Foundation Eugene“ I Garfield in 1962. They were building and the National Institutes of Health. Funding the (SCI), now the finally became available after 1957, when the Analytics , with launch of the Soviet Union’s Sputnik satellite long-sought grants from US funding agen- unleashed panic in the United States about the cies. Today, we cannot imagine research information crisis in science. Funders wanted without indexes that reveal how articles are ways to evaluate their effectiveness. Lederberg ANALYTICS OF CLARIVATE COURTESY cited. Garfield enabled an entire field: sciento- and Garfield joined forces to build an auto- metrics, the quantitative study of science and mated citation index across science. technology. Nonetheless, for many years, the SCI Garfield died on 26 February. We met in made a loss, supported by profits from Cur- 1992, when I was writing a history of the rent Contents and other ISI services. Neither index. That was a few months before he scientists nor librarians saw much use for sold the Institute for Scientific Information these expensive books (a ten-year set could (ISI), the company he had founded (initially cost US$25,000) with their long lists of code named Documation) in 1956 in Philadel- in small print. The exception was the commu- phia, Pennsylvania, to Thomson Reuters. nity of historians and sociologists of science. He stayed on as chairman emeritus, a bomb For example, Derek de Solla Price, a science of energy, still coming up with ideas for historian at Yale University in New Haven, applying citation indexes. Connecticut, and sociologist Robert Merton at Garfield also launched The Scientist In 1951, he landed a job at the Welch Medical immediately saw the SCI — a monthly magazine for life scientists Library at in Balti- as an instrument for analysing the dynamics — together with indexes in the social sci- more, Maryland, where almost all information and structure of science, and each developed ences and humanities, and services that services of the National Library of Medicine theories about citations in research. alert researchers to new relevant publica- were born. He explored new ways to deal with Since the early 1970s, the SCI’s influence tions. The ISI’s flagship product was Current the exploding medical literature, which was has extended. Quantitative analyses of output Contents, which compiles the tables of con- outpacing the capacity of human indexing, and citations have been used to evaluate fund- tents for recent scientific journals. He built and developed machine methods for search- ing programmes, research groups, individu- a host of services to summarize, filter, index ing and cataloguing using punch cards. als and nations. This use increased markedly and classify articles. His tools allowed scien- In 1953, the library organized what it billed after the Journal was marketed tists to learn how publications were used in as the First Symposium on Machine Methods in the SCI Journal Citation Reports starting in later research and to find related ones — an in Scientific Documentation. This introduced 1975 (the impact factor had been computed ability now so crucial that it is hard to imag- Garfield to Shepard’s Citations, a system for for selected journals in the SCI from the early ine that it had to be invented. Garfield was legal citations invented in 1873 that tracked 1960s). Garfield came to see the impact fac- also a prolific letter-writer; he developed his how US court cases cited earlier ones. It was tor as a mixed blessing, “like nuclear energy”. best ideas in communication with scientists, a radical departure from subject indexing, Although he felt that citation indexing and the scholars, policymakers and technical experts. which then dominated thinking in science. impact factor could be remedies for the limi- This correspondence and his more than 1,000 Garfield contacted William Adair, a former tations of peer review, he was uncomfortable published essays are gold mines for historians. vice-president at Shepard’s who had expertise with their misuse as performance indicators. Garfield was born on 16 September 1925 in citation indexing, to see whether comput- Garfield was fascinated by art. The former into a family of second-generation Jewish ers could be applied to the problem. So began ISI building, designed by architects Denise immigrants living in ’s East a mutual education — Garfield learnt about Scott Brown and Robert Venturi, housed Bronx. He and his sisters were raised by their citation indexing and taught Adair about the an impressive collection, including striking mother and her family, a mix of left-wing scientific literature. While working, Garfield murals by Huichol artists from Mexico. labour activists and entrepreneurs. Garfield did a master’s degree in library and informa- Garfield’s enthusiasm was not the picked up traits from both. tion science at Columbia University in 1954, bookkeeper’s but the visionary’s. He saw in In 1949, he graduated from Columbia and obtained a PhD in structural linguistics his creations a better science for society and University in New York as a chemistry major. at the University of Pennsylvania in Philadel- the ideal of a unified body of knowledge When he went to work there as a laboratory phia in 1961. By 1955, Garfield had developed accessible to all. ■ assistant, he discovered that he was not good the concept of a scientific citation index and at the bench. He indexed a closet full of introduced it to readers of the journal Science Paul Wouters is professor of previously synthesized compounds so that (E. Garfield Science 122, 108–111; 1955). and director of the Centre for Science and he did not have to remake them. It would become one of his most highly cited Technology Studies at Leiden University, This exercise convinced Garfield that he articles, yet the response at the time was luke- the Netherlands. was more information scientist than chemist. warm. He campaigned relentlessly, presenting e-mail: [email protected]

492 | | VOL 543 | 23 MARCH©20 201717 Mac millan Publishers Li mited, part of Spri nger Nature. All ri ghts reserved.