Eugene Garfield's creation of the Science Citation Index fundamentally changed how scientific research is conducted and evaluated. Before the 1960s, tracking the impact of scholarly work was a laborious and imprecise process; Garfield's index provided the first systematic way to measure a publication's influence through citation analysis. This innovation not only streamlined literature reviews but also established a new metric - the impact factor - that continues to shape academic careers and funding decisions today.
That someone, for most of the twentieth century, was a man named Eugene Garfield. And the reason you can find anything in academic literature today traces back to a question he asked in the early 1950s that nobody else thought to ask: what if we read citations backwards?
The Problem With Subject Headings
To understand why Garfield's question was so disruptive, you have to understand what research libraries looked like before him. Traditional subject indexing depended on a professional cataloguer reading an article and deciding which keywords or descriptors best captured its content. Two different indexers, given the same paper, might assign entirely different subject headings. The system was slow, expensive, and deeply subjective. It also varied from library to library, making broad literature searches a exercise in frustration.
Garfield, a chemist by training who had come to information science through an unexpected path, encountered this problem firsthand. After graduating from Columbia University with a Bachelor of Science in chemistry in 1949 and a degree in Library Science in 1953, he worked as a laboratory assistant at Columbia and found himself indexing all previously synthesized compounds a task that revealed something important to him. He later noted that his inclination toward information was bigger than his inclination toward chemistry.
"He realized that authors themselves had already done a kind of indexing through their citations," according to an analysis of his work at the LIS Academy's informetrics resource. "When Paper A cites Paper B, the author is signaling that the two share intellectual ground. By capturing these citation links, the system replaces an indexer's subjective choices with the collective judgement of thousands of authors."
It was a radical inversion. Instead of asking what a paper is about, Garfield asked what a paper had influenced.
Shepard's and the Legal Analogy
The idea did not come from nowhere. Garfield borrowed it from the legal profession, specifically from Frank Shepard's citation system, which had allowed lawyers since the nineteenth century to track how courts had cited particular legal precedents. Shepardizing the verb lawyers used meant tracing the ripple effects of a single court decision forward through subsequent rulings.
Garfield recognized that the same logic could apply to science. In 1964, when he and the Institute for Scientific Information published the first Science Citation Index, he wrote explicitly that the system was modeled after Shepard's legal citation system. The analogy was not incidental. It was architectural.
"The Science Citation Index was initially modeled after Shepard's legal citation system," Garfield himself wrote in 1979, in his book Citation Indexing Its Theory and Application in Science, Technology, and Humanities. The legal profession had already solved the problem of intellectual lineage. Garfield brought that solution into the laboratory.
The 1955 Paper That Started Everything
Before the index existed in physical form, the idea existed on paper. In 1955, Garfield published a landmark article titled "Citation Indexes for Science: A New Dimension in Documentation through Association of Ideas" in the journal Science. The paper laid out the theoretical case for what he called citation indexing a method of organizing literature based not on subject descriptors but on the citation relationships between documents.
It was, in the context of 1955, a remarkably confident claim. Automated indexing was in its infancy. The idea that a machine could organize knowledge by tracking what scientists did more than what they said about their work required a leap of faith that most information professionals were not ready to take.
But Garfield was not most information professionals. He had been involved in the Welch Medical Library Indexing Project during the mid-1950s, which investigated whether automation could organize and retrieve medical literature without relying on subjective human judgement. That project became the proving ground for his citation ideas. He was not theorizing from a distance; he was building from inside the problem.
The Science paper established the intellectual framework. The Science Citation Index, published nine years later, was the product.
1964: The Index Appears
In 1964, Eugene Garfield and the Institute for Scientific Information ISI, the organization he had founded published the first Science Citation Index in five printed volumes. It indexed 613 journals and 1.4 million citations. The scale was significant. This was not a boutique reference tool. It was a comprehensive attempt to map the citation landscape of mid-twentieth-century science.
Two years later, in 1966, the Science Citation Index became available on magnetic tape, bringing it into the era of automated retrieval. The shift from print to tape was the moment citation indexing stopped being a curiosity and started becoming infrastructure.
According to records at the History of Information, Garfield published an extensive paper alongside the launch entitled "Science Citation Index" A New Dimension in Indexing, also in Science. The title itself was a statement of intent: a new dimension in indexing. He was not offering a better thesaurus or a cleaner card catalog. He was offering a fundamentally different way of organizing knowledge.
From One Index to a Family of Products
The Science Citation Index was the flagship, but it was not the only product. Garfield's vision was systematic. If citations could track the influence of papers, they could also track the influence of journals. This insight led to the Journal Citation Reports and, most famously, the impact factor a metric that calculated how often a journal's articles were cited over a given period.
The impact factor would eventually become one of the most debated numbers in academic publishing, but that came later. In Garfield's original conception, it was a tool for library management a way to make evidence-based decisions about which journals to subscribe to and which to drop. Libraries had always needed a method for evaluating journals they did not yet know well. Citation data offered a quantitative proxy for influence.
Beyond the Science Citation Index and Journal Citation Reports, Garfield also created Current Contents a weekly digest that reproduced the tables of contents of leading journals and Index Chemicus, which tracked new chemical compounds. He founded The Scientist magazine. The Institute for Scientific Information became a commercial enterprise built on the systematic application of citation analysis to scientific literature.
The history of ISI at Clarivate, which acquired the organization, describes it as fostering the scientometric community through research and collaboration a recognition that Garfield's methods had become a field of study in their own right.
Why This Matters for Search and Discovery
Modern readers of WebSearches encounter citation logic every time they use a research database. When Google Scholar ranks a paper partly by how many times it has been cited, that is Garfield's logic in digital form. When Semantic Scholar highlights papers that are semantically related to your search, it is building on the structural insight that Garfield articulated in 1955: citations are signals, not just citations are formalities.
The knowledge graph that powers modern search engines traces intellectual lineage through citation-like relationships between entities. The recommendation engines that suggest related research articles are, in a sense, performing automated Shepardizing on a scale that Garfield could not have imagined.
What Garfield understood, and what took decades for the broader information science community to absorb, was that citations are not merely acknowledgements of intellectual debt. They are votes. Every time a researcher cites a paper, they are implicitly endorsing its relevance to their work. Aggregated across thousands of researchers, those endorsements become a map of intellectual influence that no individual indexer could construct.
The implications extended beyond libraries. Research evaluators use citation metrics to assess the output of institutions and nations. Publishers use impact factors to position their journals. Scientists use citation searches to find not just related work but work that builds on or responds to specific papers. Each of these applications descends from Garfield's 1955 inversion.
What This Means for WebSearches Readers
If you work in search engine optimization, answer engine optimization, or content strategy, understanding Garfield's citation logic is not an academic exercise it is practical infrastructure. The same principles that govern how the Science Citation Index organized literature govern how modern search engines evaluate authority.
When a knowledge panel surfaces certain sources and not others, it is applying a logic of relevance and influence that Garfield helped formalize. When an answer engine cites specific sources in its response, it is performing a function analogous to the Science Citation Index's backward citation tracking finding which sources are most often cited in relation to a given query.
Garfield's insight was that the network of citations contains more usable information than any individual citation. Modern search engines operate on the same principle: the graph of links, mentions, and citations is more intelligent than any single ranking signal. Understanding the origin of that insight helps you read the modern landscape more clearly.
The Man Behind the Index
Eugene Eli Garfield was born in New York City in 1925, the son of second-generation immigrants of Lithuanian Jewish ancestry. He grew up in the East Bronx and attended the University of Colorado and the University of California, Berkeley before completing his undergraduate degree at Columbia. He earned his doctorate in linguistics from the University of Pennsylvania in 1961, with a dissertation that developed an algorithm for translating chemical nomenclature into chemical formulas a fitting bridge between his chemical training and his linguistic interests.
He died in Philadelphia in 2017, at the age of ninety-one. By then, the Institute for Scientific Information had been acquired by Thomson Reuters and later by Clarivate, but the products he created the Web of Science, Journal Citation Reports, the Science Citation Index remained in active use at research institutions worldwide.
The Clarivate blog's retrospective on Garfield's legacy describes him as a pioneer of information science language that, while accurate, somewhat undersells the scope of his contribution. Garfield did not merely pioneer information science. He invented a specific and durable method for making knowledge legible to machines and researchers alike.
The Critique He Did Not Write
Garfield was not unaware of the limitations of his system. The impact factor, in particular, was designed to evaluate journals, not individual papers or researchers a distinction he emphasized repeatedly but that was frequently lost in translation as the metric was adopted by universities and funding bodies for purposes it was never intended to serve.
Citation rates vary dramatically across fields. A paper in mathematics may accumulate few citations over decades, while a paper in biomedicine may accumulate thousands in a single year. Self-citation and citation manipulation are documented phenomena. Coverage biases favor English-language journals. Garfield acknowledged these limitations in his writings, though the systems built on his work did not always carry those caveats forward.
The LIS Academy analysis notes that "the impact factor was never designed to judge individuals" and that "citations do not measure quality" observations that Garfield himself made, but that took decades to filter into policy discussions about research evaluation. In a sense, the success of citation indexing created the conditions for its own misuse: the metrics were so useful for the purposes they were designed for that institutions began applying them to purposes they were not designed for.
The Legacy in Present Tense
Today, the tools that descended from Garfield's work are woven into the fabric of academic research. The Web of Science ISI's flagship database, now maintained by Clarivate remains a primary venue for citation analysis and research evaluation. Journal Citation Reports continue to publish annual impact factor rankings. Citation searches are a standard feature of every major research database.
But the legacy extends further. The citation graph that Garfield built for science has become a template for how digital platforms think about authority and relevance. Google's PageRank algorithm, which evaluates web pages partly by how many other pages link to them, is citation analysis applied to the web. The connection is not coincidental. Larry Page and Sergey Brin, when they developed PageRank at Stanford, were applying to hypertext the same logic that Garfield had applied to scientific literature.
Every time a modern search engine surfaces a result because of its citation count, its link profile, or its position within a knowledge graph, it is operating within a conceptual framework that Garfield helped articulate in 1955. The index that started as five printed volumes has become an invisible layer of the internet.
Where to Read Further
For readers who want to explore Garfield's original vision in his own words, the 1955 Science paper remains the foundational document. Garfield's 1979 book Citation Indexing Its Theory and Application in Science, Technology, and Humanities expands the framework considerably and is cited in several of the sources that document his work. The Clarivate history of the Institute for Scientific Information provides institutional context for how the products evolved. The LIS Academy's analysis offers a clear explanation of how citation indexing actually worked and why it was significant.
Those interested in the broader history of information retrieval will find Garfield's work inseparable from the history of library automation and early computational approaches to knowledge organization. The Science Citation Index of 1964 was not an isolated invention. It was the culmination of a decade of work at the Welch Medical Library and Columbia University, and it pointed forward toward the databases and search engines that now structure how knowledge is accessed globally.
Timeline: Eugene Garfield and Citation Indexing
| Year | Event |
|---|---|
| 1925 | Eugene Garfield born in New York City |
| 1949 | Receives BS in chemistry from Columbia University |
| 1953 | Earns degree in Library Science from Columbia University |
| 1955 | Publishes "Citation Indexes for Science" in Science journal |
| 1961 | Completes PhD in Linguistics at University of Pennsylvania |
| 1964 | First Science Citation Index published in five volumes; indexes 613 journals and 1.4 million citations |
| 1966 | Science Citation Index becomes available on magnetic tape |
| 1979 | Publishes Citation Indexing Its Theory and Application |
| 2017 | Dies in Philadelphia at age 91 |
The index that Garfield built was, at its core, an act of imagination. He looked at the footnotes of scientific papers those dry strings of author names and journal titles that most readers skipped and saw a living network of intellectual relationships. He asked what would happen if you read that network backward, if you traced the influence of an idea forward through time beyond simply documenting its origins. The answer reshaped how knowledge is organized, searched, and evaluated across every field of human inquiry.