Eugene Garfield
Eugene Garfield (September 16, 1925 – February 26, 2017) was an American information scientist, publisher, and entrepreneur best known for founding the Institute for Scientific Information (ISI) in 1960 and pioneering citation indexing as a method for organizing and retrieving scientific literature.[1][2][3] His innovations, including the Science Citation Index (SCI) launched in 1964, transformed scholarly communication by enabling researchers to trace the impact and interconnections of scientific works through cited references rather than traditional subject indexing.[3][4] Born in New York City to a family of Jewish immigrants, Garfield developed an early interest in science and information management during his studies in chemistry.[1] He earned a Bachelor of Science in chemistry from Columbia University in 1949, followed by a Master of Science in library science from the same institution in 1954, and a Ph.D. in structural linguistics from the University of Pennsylvania in 1961.[2][5] His graduate work exposed him to the challenges of indexing complex scientific texts, inspiring his foundational 1955 paper proposing citation indexing as a solution for efficient literature searches and historical analysis of scientific progress.[4] In 1951, while pursuing his library science degree, Garfield joined the Welch Medical Indexing Project at Johns Hopkins University, where he contributed to early developments in medical literature indexing that later influenced systems like Index Medicus and MEDLARS.[6][2] Garfield's career shifted toward entrepreneurship after brief consulting roles in the pharmaceutical industry, leading him to establish Eugene Garfield Associates before founding ISI in Philadelphia.[2] Under his leadership as president until 1992, then chairman until 1993, and later as chairman emeritus, ISI expanded rapidly, introducing key products such as Current Contents in 1957—a weekly alert service for new publications—and the Social Sciences Citation Index (SSCI) in 1973, followed by the Arts & Humanities Citation Index (AHCI) in 1978.[3][2][7] He also developed the Journal Citation Reports (JCR) in 1976, which introduced the influential Journal Impact Factor as a metric for evaluating journal quality based on average citations.[3] These tools laid the groundwork for the field of scientometrics, using quantitative citation data to measure research impact and map the structure of scientific knowledge.[3] In 1986, Garfield launched The Scientist, a news magazine for life scientists that he edited until his death, further extending his influence on scientific discourse.[2] Throughout his career, Garfield authored over 1,000 essays in Current Contents and commented on 4,000 "Citation Classics"—highly cited papers—highlighting patterns in scientific productivity and influence.[6][8] His work earned numerous accolades, including the Award of Merit from the American Society for Information Science in 1975, the Herman Skolnick Award from the American Chemical Society in 1977, and the John Price Wetherill Medal from the Franklin Institute in 1984 for contributions to information science.[9] ISI was acquired by Thomson Reuters in 1992 (later becoming part of Clarivate Analytics), perpetuating Garfield's vision through platforms like the Web of Science.[3] Garfield died in Philadelphia at age 91, leaving a legacy as the "father of citation analysis" that continues to shape global research evaluation and discovery.[10][1]Biography
Early life and education
Eugene Garfield was born Eugene Eli Garfinkle on September 16, 1925, in New York City to a family of second-generation Jewish immigrants from Eastern Europe.[1] His parents had roots in Lithuania, and he grew up in the East Bronx neighborhood, where his early environment reflected the immigrant experiences of many Jewish families in the city during that era.[11][12] In early adulthood, Garfinkle changed his surname to Garfield, adopting the name under which he would become known professionally.[11] He pursued undergraduate studies in chemistry at Columbia University, earning a Bachelor of Science degree in 1949.[13] Garfield then continued at Columbia for graduate work, obtaining a Master of Science in Library Science in 1954, which shifted his focus toward the organization and dissemination of scientific knowledge.[11] Garfield's doctoral studies took him to the University of Pennsylvania, where he earned a PhD in Linguistics in 1961.[13] His dissertation, titled "An Algorithm for Translating Chemical Names to Molecular Formulas," explored computational methods in linguistics applied to chemical nomenclature, bridging his chemistry background with emerging interests in automated information processing.[13] During his academic years, Garfield developed early interests in science communication and information retrieval, influenced conceptually by Vannevar Bush's 1945 essay "As We May Think," which envisioned associative trails for linking knowledge.[14]Personal life
Garfield married his first wife in the early 1950s, with whom he had a son, Stefan; the marriage ended in divorce, after which he became a single parent.[15] He remarried in 1955 to Winifred, his second wife, who predeceased him.[16] In later years, Garfield married Meher, with whom he remained until his death.[5] He was survived by three sons—Stefan, Joshua, and Alexander—a daughter Laura, a stepdaughter Cornelia, two grandchildren, and two great-grandchildren; an earlier daughter, Thea, predeceased him.[17] Garfield resided in the Philadelphia area for much of his adult life, including in Bryn Mawr, Pennsylvania, at the time of his passing.[5] Garfield died on February 26, 2017, in Bryn Mawr at the age of 91, due to complications related to advanced age.[17] His longevity enabled him to witness and guide the evolution of his professional endeavors well into the 1990s.[1] Beyond his career, Garfield maintained a strong personal interest in writing and science communication, authoring over 1,500 essays that reflected his broad perspectives on scientific progress and knowledge dissemination.[18]Professional Career
Founding of the Institute for Scientific Information
Eugene Garfield's vision for improving scientific documentation through citation indexing originated in his seminal 1955 paper, "Citation Indexes for Science: A New Dimension in Documentation through Association of Ideas," published in the journal Science. In this article, Garfield proposed a system that would link scientific works via their citations, enabling researchers to trace intellectual associations and navigate the growing volume of literature more efficiently than traditional subject-based indexes. This idea stemmed from his earlier experiences in information science and built upon concepts explored in his doctoral work, including algorithms for automated indexing.[19][1] In 1960, Garfield founded the Institute for Scientific Information (ISI) in Philadelphia, Pennsylvania, to realize this vision on a commercial scale. The institute was established with a combination of initial funding from a three-year grant awarded by the National Institutes of Health (NIH) to develop a prototype citation index in genetics, alongside Garfield's personal financial contributions, including loans to bootstrap operations. ISI began as a small enterprise, operating from modest facilities and relying on a limited team to handle the labor-intensive tasks of data collection and processing.[3][20] The early years of ISI were marked by significant challenges, including the manual production of indexes, which involved staff meticulously cutting and organizing content from journals, and the need to sustain operations through ancillary services like document delivery to pharmaceutical and research clients. With a focus initially on chemistry and biomedicine—fields where Garfield had prior expertise—the institute bootstrapped its growth by providing practical tools for literature awareness amid limited resources and skepticism from the academic community. To address immediate researcher needs, ISI launched Current Contents in 1962 as a weekly alerting service that reprinted tables of contents from leading journals in these disciplines, offering a low-cost way to stay current without full-text access.[13][21]Development of citation indexing tools
Eugene Garfield's development of citation indexing tools revolutionized scientific literature retrieval by enabling users to trace the impact and connections of research through citations rather than traditional subject headings. Building on his earlier conceptual work, Garfield founded the Institute for Scientific Information (ISI) in 1960 to operationalize these ideas. The cornerstone product, the Science Citation Index (SCI), was first published in 1964 as a quarterly print edition that covered approximately 600 leading journals across the sciences, providing a multidisciplinary view of scholarly connections.[3][22] This initial SCI edition consisted of four main sections: a source index listing articles by author and journal, a citation index mapping cited references to citing articles, a source title index for journal lookup, and a permuted title index that rearranged article titles by rotating keywords to enhance discoverability without relying on controlled vocabularies.[22] The methodology emphasized forward citation tracking, where each article's references were recorded to allow searches from known works to subsequent citing publications, fostering an "association of ideas" approach to navigation.[23] Prior to the SCI, Garfield launched Index Chemicus in July 1960 as ISI's inaugural product, specifically designed to index chemical reactions and substances from the literature by capturing structural formulas and bibliographic details from over 100 core chemistry journals.[24] This tool complemented the broader citation framework by focusing on chemical-specific retrieval, using molecular formula indexes and author listings to support synthesis planning and patent analysis in the pharmaceutical and chemical industries. By 1966, Index Chemicus had expanded its coverage and integration, becoming a vital companion to the emerging SCI for specialized searches.[3] The SCI's scope grew to encompass additional disciplines, with the Social Sciences Citation Index (SSCI) introduced in 1973 to address the need for citation tracking in social sciences, initially covering around 1,400 journals and enabling analysis of interdisciplinary influences like economics and psychology.[3] This expansion applied the same core methodology—source indexing of articles and forward citation mapping—to non-natural sciences, revealing citation patterns across fields previously underserved by indexing tools. In 1978, the Arts & Humanities Citation Index (AHCI) followed, indexing approximately 1,100 journals in arts, humanities, and related areas, which incorporated unique features like cited work indexing for books and non-journal sources to capture the diverse referencing styles in these domains.[3] These extensions maintained the permuted title index and citation linkage principles, allowing users to follow intellectual threads in creative and interpretive scholarship. By the early 1980s, ISI transitioned these tools to online formats through proprietary systems, beginning with magnetic tape distributions in the 1970s and evolving to dial-up access via platforms like BRS/Search, which provided searchable electronic versions of the SCI, SSCI, and AHCI starting around 1974 and expanding significantly in the 1980s.[22] This shift enabled real-time forward and backward citation searching across millions of records, dramatically improving accessibility and speed over print editions while preserving the foundational source indexing and permuted title mechanisms for comprehensive literature discovery.[25]Later career and sale of ISI
In the mid-1980s, Garfield expanded ISI's portfolio by launching The Scientist, a biweekly magazine dedicated to news and commentary in the life sciences, which was distributed free to subscribers and aimed to bridge the gap between scientific research and broader professional discourse.[26][27] By this period, ISI had grown significantly from its earlier years, employing over 500 people and establishing a presence that included international operations to support its global indexing services.[13][28] This expansion was driven by the success of core products like the Science Citation Index, which formed the foundation of ISI's valuation as a leader in scientific information services.[3] Facing challenges including ill health and internal management issues, Garfield and key shareholders sold ISI to the Thomson Corporation in 1992 for an undisclosed sum, marking a pivotal transition that allowed him to step back from day-to-day operations.[28][1] Following the acquisition, Garfield assumed the role of Chairman Emeritus at the newly rebranded Thomson ISI, where he provided advisory input during the integration process.[29] He also continued overseeing aspects of The Scientist and engaged in reflective writing, contributing to the preservation of ISI's historical records.[13] In the 1990s, as Thomson further developed the acquired assets—eventually merging with Reuters in 2008 to form Thomson Reuters—Garfield gradually retired from active involvement, shifting his focus to archival projects and personal essays that documented the evolution of information science.[3][30] This period allowed him to emphasize mentorship and legacy-building, drawing on decades of experience without the demands of executive leadership.[31]Contributions to Information Science
Key publications and essays
Eugene Garfield's seminal 1955 article, "Citation Indexes for Science: A New Dimension in Documentation through Association of Ideas," published in Science, proposed the use of citation indexes as a tool for information retrieval by associating ideas through referenced connections, arguing that traditional subject indexing was insufficient for the growing volume of scientific literature.[19] This work laid the foundational concept for citation-based searching, emphasizing how citations could reveal intellectual pathways and facilitate discovery beyond keyword matching.[32] In his 1961 PhD dissertation at the University of Pennsylvania, titled "An Algorithm for Translating Chemical Names to Molecular Formulas," Garfield developed a systematic method to convert complex chemical nomenclature into empirical formulas, addressing the challenges of parsing organic names for computational processing.[33] The algorithm consists of eight rules for morpheme recognition and substitution, beginning with ignoring locants (numerical prefixes like 1,4-) while retaining parentheses, then replacing name components—such as "eth" for C₂H₅ (2C + 5H) or "nitro" for NO₂ (N + 2O + 1 double bond, DB)—with their atomic and bond values from a dictionary. Ambiguities, like "penta" as a multiplier (5) versus an alkyl group, are resolved by examining subsequent morphemes; plus signs are inserted after morphemes (except multipliers) and adjusted around parentheses before performing multiplications. Hydrogen atoms are finally calculated using the formula H = 2 + 2n_C + n_N - n_X - 2n_DB, where n_C is the number of carbons, n_N nitrogens, n_X halogens, and n_DB double bonds. Examples include translating "methylaminoethane" to C₃H₉N (methyl = CH₃, amino = NH₂, ethane = C₂H₅, adjusted for linkages) and the more complex "N-[3-(diethylamino)propyl]-N-ethyl-2-amino-1,4-butanedioic acid" to C₁₃H₂₆N₂O₄, demonstrating the algorithm's handling of nested substituents and functional groups.[33] Garfield authored over 1,000 essays, initially published weekly in Current Contents from 1962 onward and later compiled into the 18-volume series Essays of an Information Scientist (1977–1993), which explored topics in information retrieval, scientific communication, and the sociology of science.[34] These essays served as reflective commentaries, drawing on his experiences at the Institute for Scientific Information to advocate for innovative documentation practices. Among the key pieces, his 1979 essay "Is Citation Analysis a Legitimate Evaluation Tool?" in Scientometrics defended the use of citations for assessing scientific impact while acknowledging limitations like self-citation and field-specific norms, positioning it as a supplementary rather than sole metric.[35] Garfield also offered reflections on the historiography of science in essays such as "Historiographs, Librarianship, and the History of Science" (1977), where he discussed how citation networks could map the evolution of scientific ideas and integrate librarianship with historical analysis.[36] Beyond these, Garfield contributed chapters to edited volumes on scientometrics, including discussions on citation patterns and knowledge mapping in works like The Use of Citation Data in Writing History of Science (1980), which extended his ideas on algorithmic historiography. He also participated in oral histories, such as the 1997 interview conducted by Robert V. Williams for the Chemical Heritage Foundation, where he recounted the development of his early ideas on information systems and their broader implications.[16]Introduction of the impact factor
In the mid-1960s, Eugene Garfield, in collaboration with Irving H. Sher, developed the journal impact factor as a quantitative metric to evaluate the influence of scientific periodicals, building on his earlier conceptualization of the idea in a 1955 article in Science.[37] This metric was integrated into the Journal Citation Reports (JCR), which Garfield's Institute for Scientific Information (ISI) first published in 1975 as an annual compilation derived from citation data in the Science Citation Index (SCI).[38] The impact factor aimed to provide an objective measure of a journal's citation reception, facilitating informed decisions in academic resource management.[37] The methodology of the impact factor centers on a simple ratio that captures recent citation activity relative to a journal's output. Specifically, for a given year Y, the impact factor is calculated as: \text{Impact Factor} = \frac{\text{Number of citations in year } Y \text{ to items published in } Y-1 \text{ and } Y-2}{\text{Number of citable items published in } Y-1 \text{ and } Y-2} Here, "citable items" typically include substantive articles, reviews, and proceedings papers, excluding editorials, letters, or news items.[37] To illustrate, consider a hypothetical journal Example Science Quarterly. Suppose in 2023 (Y), it receives 150 citations to its papers from 2021 and 2022. In those two years, it published 50 citable items in 2021 and 60 in 2022, totaling 110 citable items. The step-by-step calculation is:- Identify citations: Sum the 150 citations received in 2023 for the 110 items from 2021–2022.
- Identify denominator: Count the 110 citable items published in 2021–2022.
- Divide: $150 \div 110 \approx 1.36.