“The plural of anecdote is not data.”
I first encountered this saying during one of the more humbling weeks of my professional life. I had just finished presenting what I thought was a compelling argument to a roomful of skeptical analysts — an argument built almost entirely on striking stories I had collected from interviews. A colleague pulled me aside afterward, said nothing for a moment, then quietly wrote seven words on a sticky note and pressed it into my palm. The note read: The plural of anecdote is not data. I laughed it off in the moment, but that night I sat with it differently, turning it over like a stone with unexpected weight. It wasn’t cruel — it was clarifying. That single sentence reframed everything I thought I understood about evidence, persuasion, and the seductive trap of a good story told often enough.
That experience sent me down a long rabbit hole. Where did this phrase actually come from? Who first said it, and why does it still cut so cleanly decades later? The answer, as it turns out, is more complicated — and more interesting — than most people expect.
The Quote and Its Deceptive Simplicity
On the surface, the saying seems almost too obvious to need stating. Of course a collection of stories isn’t the same as systematic evidence. Yet researchers, journalists, policymakers, and everyday people fall into exactly this trap constantly. The phrase lands with such force precisely because it names a mistake we all recognize — in others, and eventually in ourselves.
The saying also carries a quiet grammatical joke. Data is technically the plural of the Latin datum, meaning a single given fact. Anecdote pluralizes conventionally into anecdotes. So the sentence works on two levels simultaneously: it makes an epistemological claim, and it does so with a wink toward the mechanics of language itself. That layered quality helps explain why it has stuck around.
Additionally, the phrase travels easily across disciplines. Scientists use it to police methodology. Lawyers invoke it to challenge testimony. Journalists cite it to defend the difference between feature writing and data reporting. Few short phrases carry that kind of cross-disciplinary utility.
The Earliest Known Appearance
Tracing this quote to its origin requires patience. Many widely repeated attributions point to famous economists or psychologists, but the documentary record tells a more specific story.
The earliest verified appearance comes from a 1982 journal article. Kernaghan was a Canadian scholar working in public administration — not exactly the celebrity intellectual most people imagine coining a phrase this punchy. Yet the evidence points clearly in his direction, at least provisionally.
This matters because the quote frequently gets attributed to far more famous names. Roger Brinner, a prominent economist, appears in many online citations. So does the Nobel laureate economist Paul Samuelson. Neither attribution holds up under scrutiny when you follow the documentary trail carefully.
The provisional nature of this attribution deserves emphasis. Kernaghan and Kuruvilla may have crafted the phrase themselves, or they may have borrowed it from an oral tradition that simply wasn’t captured in print earlier. Academic publishing in 1982 left enormous amounts of spoken intellectual culture undocumented. Therefore, an earlier origin remains genuinely possible.
Paul Samuelson’s Related Observation
Before the 1982 article, economist Paul Samuelson made a thematically parallel point that deserves attention. Writing in Newsweek in July 1967, Samuelson stated:
“Anecdotes do not constitute social science. They must be documented by some notion of the frequency of their occurrence.”
This is not the same phrase. However, it expresses the same intellectual conviction with similar directness. Samuelson’s version is wordier and more explanatory — it lacks the aphoristic snap of the seven-word version. Nevertheless, it shows that this idea was circulating in serious intellectual discourse at least fifteen years before the phrase crystallized.
Samuelson was one of the most influential economists of the twentieth century. His willingness to make this argument in a popular magazine column — rather than a peer-reviewed journal — suggests he considered it an idea worth broadcasting widely. That instinct proved correct.
Additionally, Samuelson’s framing points toward something important: the phrase didn’t emerge from nowhere. It crystallized a concern that serious thinkers had been wrestling with for decades, if not longer.
Irwin S. Bernstein and the Scientific Context
The phrase gained significant traction in academic circles through a 1988 commentary by Irwin S. Bernstein, a professor of psychology and primatologist. Writing in the journal Behavioral and Brain Sciences, Bernstein deployed the saying with real argumentative force:
“Relying on anecdotes, no matter how numerous, fails to specify any value of the independent variable, or the intentional state of an animal. The plural of ‘anecdote’ is not ‘data.'”
His context was a heated debate about whether primates demonstrated intentional deception — a question that generated many compelling anecdotes but, Bernstein argued, very little controlled data.
His point cut deep. Researchers studying animal behavior had accumulated hundreds of striking stories about clever primates apparently deceiving rivals or humans. Each story felt convincing. Together, they felt overwhelming. But Bernstein insisted that volume of anecdote doesn’t substitute for measurement of variables. The sheer accumulation of stories, however vivid, cannot tell you how often a behavior occurs, under what conditions, or why.
This is the core epistemological problem the phrase identifies. Furthermore, Bernstein’s framing made the phrase memorable to a generation of scientists who encountered it during their training.
The Phrase Spreads Across Disciplines
Through the late 1980s and 1990s, the saying appeared in an impressive range of contexts. Each new usage reinforced its cross-disciplinary authority.
In 1989, The Christian Century published a variant phrasing with an anonymous attribution: “research is not the plural of anecdote.” This flip construction is interesting — it emphasizes what research is rather than what data isn’t. The anonymous attribution signals something important: by 1989, the phrase had already detached from any specific author and entered the broader intellectual commons.
Meanwhile, legal scholarship picked up the phrase as well. In 1995, Edith Greene published a book review in The Justice System Journal that used it to challenge a popular critique of the American jury system:
“I have no doubt that some juries make egregious errors, that some jurors are hopelessly prejudiced, and that some verdicts are indeed outrageous. But this collection of anecdotes, and Adler’s take on them, does not convince me that the jury system is in trouble. The plural of anecdote is not data.”
Greene’s usage demonstrates the phrase’s rhetorical versatility. She wasn’t dismissing the individual stories as false. She was insisting that their accumulation doesn’t automatically produce a systemic conclusion.
Also in 1995, historian Sally Scully of San Francisco State University used the phrase in the Journal of Social History with a subtle qualification: Scully’s version acknowledged the phrase’s validity while pushing back on its limits — a reminder that the saying, like all good aphorisms, contains tension worth exploring.
The Antithetical Twin: “The Plural of Anecdote Is Data”
Here is where the history gets genuinely strange. Running parallel to this phrase is its exact opposite: The plural of anecdote is data. This version argues that accumulated personal experience does constitute a form of evidence — particularly in qualitative social research, oral history, and community-based knowledge production.
The two sayings aren’t simply contradictory. They reflect a genuine and unresolved tension in how different disciplines think about knowledge. Quantitative researchers tend to embrace the negated version. Qualitative researchers, ethnographers, and oral historians often find more truth in the affirmative form. Both camps have legitimate points.
Additionally, the existence of both versions helps explain why the attribution has always been murky. If the two phrases evolved together — perhaps in academic hallways and seminar rooms rather than published papers — then neither version has a single clear inventor. They emerged from a conversation, not a monologue.
Why the Misattributions Happen
Despite the documented 1982 origin, the phrase regularly gets credited to economists like Roger Brinner or even to figures like Frank Kotsonis. These attributions circulate confidently across the internet, repeated without verification.
The pattern makes psychological sense. A sharp, memorable phrase feels like it should belong to a famous person. When we encounter it without context, we unconsciously attach it to the most intellectually prominent figure we associate with the idea. Economists get credited because the phrase sounds like something an economist would say. Psychologists get credited for the same reason.
Ironically, this process is itself a kind of anecdotal reasoning — we assume authorship based on plausibility and familiarity rather than documented evidence. The phrase about anecdotes gets misattributed through exactly the cognitive process it warns against.
The Phrase in Modern Usage
Today, the saying appears everywhere from peer-reviewed methodology sections to Twitter arguments about public health policy. Its modern resonance has only intensified as social media platforms have made it easier than ever to share compelling personal stories at scale.
The phrase now functions almost as a methodological shorthand — a way of quickly flagging the difference between illustrative evidence and systematic evidence. Journalists use it to distinguish feature writing from investigative data analysis. Doctors use it to explain why case reports don’t override clinical trials. Policy analysts use it to push back against arguments built entirely from constituent stories.
However, the phrase also attracts legitimate criticism. Source Some scholars argue that it gets weaponized to dismiss lived experience, particularly the experiences of marginalized communities whose realities rarely appear in official datasets. This tension is real and worth taking seriously. The phrase doesn’t mean that stories are worthless — it means they require a different kind of rigor before they can support systematic conclusions.
Sally Scully’s 1995 formulation captured this nuance well. She acknowledged the phrase’s validity while insisting that individual biography still catches what statistics miss. Both things are true simultaneously.
What Kenneth Kernaghan Actually Contributed
Kenneth Kernaghan deserves more credit than he typically receives. Source As a scholar of Canadian public administration, he worked in a field that constantly navigates the tension between policy anecdote and policy evidence. His 1982 article addressed selection problems in public personnel management — a context where anecdotal thinking about individual cases can easily distort systemic conclusions.
In that setting, the phrase wasn’t a clever bon mot. It was a methodological warning with real stakes. Personnel decisions made on the basis of compelling individual stories, rather than systematic data, produce biased and inconsistent outcomes. Kernaghan was making a practical argument, not just a philosophical one.
This origin gives the phrase a grounding that its later, more abstract usages sometimes obscure. It began as a tool for improving institutional decision-making. That practical root helps explain why it has remained useful across so many different fields.
The Lasting Power of Seven Words
What makes this phrase so durable? Several things work together. First, it’s grammatically clean — seven words, one idea, no hedging. Second, it exploits the Latin roots of data in a way that feels slightly learned without being pretentious. Third, it names a mistake that intelligent, well-meaning people make constantly. Therefore, it never runs out of occasions for use.
Furthermore, the phrase respects its audience. It doesn’t say stories are bad. It doesn’t say personal experience is worthless. It simply draws a line between two different kinds of knowing — and insists that crossing that line requires work, not just accumulation.
That’s a sophisticated epistemological point compressed into a form anyone can remember. Source That compression is the craft achievement, whoever first managed it.
Conclusion
The phrase the plural of anecdote is not data traces most reliably to Kenneth Kernaghan and P.K. Kuruvilla’s 1982 work in Canadian public administration, with a thematically related observation from Paul Samuelson appearing as early as 1967. By 1988, Irwin Bernstein had given it scientific currency in the pages of Behavioral and Brain Sciences. Through the 1990s, it spread into legal scholarship, social history, and popular discourse. Today it functions as one of the most widely invoked methodological reminders in English-speaking intellectual life.
Additionally, its antithetical twin — the plural of anecdote is data — reminds us that the conversation is still open. Stories matter. Systematic evidence also matters. The real skill lies in knowing which kind of knowing a given question actually requires.
That sticky note my colleague handed me years ago didn’t mean my stories were wrong. It meant I hadn’t yet done the harder work of turning them into something that could bear the weight I was asking them to carry. That distinction, it turns out, is worth seven words — and apparently worth several decades of debate.