A Short History of the Big Five (Five-Factor Model)

The standard account of the Big Five's origin starts with an idea usually traced to Francis Galton, who speculated in the 1880s that the most important differences between people would, over enough time, get encoded as single words in ordinary language. That idea, now called the lexical hypothesis, makes a testable prediction: catalogue every personality-describing word in a dictionary, and the structure hiding inside that list should reflect the structure of personality itself.

Psychologists Gordon Allport and Henry Odbert put the prediction to work in 1936, combing an unabridged English dictionary and pulling out close to 18,000 words used to describe a person's traits and behavior. Their list was raw material, not an answer: far too large and unstructured to describe anyone's personality on its own. The next decades of research are largely the story of different scientists trying to compress that list down to its real underlying dimensions.

In the 1940s, psychologist Raymond Cattell took a working subset of the Allport-Odbert terms and applied factor analysis, a statistical method for finding clusters of ratings that tend to move together. Cattell's own path led him to a model with sixteen factors, published as the 16 Personality Factor Questionnaire (16PF), considerably more granular than the five-factor structure that later researchers would settle on.

Cattell's correlation data turned out to matter more than his sixteen-factor conclusion. When other researchers later re-examined the same kind of data with different methods, a simpler five-factor pattern kept surfacing underneath it, the first hint that sixteen dimensions might be more than the data actually supported.

Working on a U.S. Air Force research contract, Ernest Tupes and Raymond Christal re-analyzed personality-rating data across several samples, including data descended from Cattell's own work. In a 1961 technical report, they reported that five factors recurred consistently across the samples they examined — one of the earliest clear statements of what would become the Big Five structure.

Their report was written for a military research audience and didn't reach wide academic circulation right away. In hindsight, though, it's commonly treated as the moment “five” first emerged clearly from the data, well before the structure had the widely recognized name it carries today.

Two years later, psychologist Warren Norman independently replicated a closely matching five-factor structure using peer-rating data of his own. A single finding is a data point; a second, independent team reaching the same structure is a pattern. Norman's replication is often cited as the point where the five-factor result stopped looking like one research team's idiosyncratic finding and started looking like something durable.

It still took years of further replication, across different samples, rating methods, and eventually languages, before the five-factor structure became the field's default description of personality rather than one contender among several.

Lewis Goldberg, working in the same lexical tradition as Allport, Cattell, and Norman before him, is widely credited with coining the term “the Big Five” in the early 1980s. By his own account the name was meant descriptively rather than dramatically: “big” because each of the five factors is extremely broad, covering a great deal of narrower trait vocabulary underneath it, not because the model was claiming special importance.

Goldberg spent much of his career arguing for the structure's robustness and building tools other researchers could use to test it, including the International Personality Item Pool (IPIP), a public-domain bank of research items released for scientists to use and adapt freely. That project is a useful landmark in the model's history (worth naming here descriptively) and is separate from this site's own test: every one of our 50 statements was written originally for this instrument.

While the lexical tradition was counting words, Paul Costa and Robert McCrae approached personality from the questionnaire tradition instead, building structured inventories rather than analyzing dictionaries. Their early work centered on three factors — Neuroticism, Extraversion, and Openness, which is where the name NEO comes from — before later research led them to add Conscientiousness and Agreeableness, arriving at the same five-dimension structure the lexical researchers had found by a completely different route.

Costa and McCrae generally preferred the term “Five-Factor Model” (FFM) over “Big Five,” and their instruments — the NEO Personality Inventory and its successors — became widely used research measures built on the model. Today the two names are mostly used interchangeably, though FFM still signals the questionnaire tradition specifically for researchers who care about the distinction.

Here's the detail that gives the Big Five much of its credibility: lexical researchers counting trait-words and questionnaire researchers building structured scales were doing genuinely different work, using different data and different assumptions, yet both converged on the same five dimensions. That kind of convergence, repeated across many samples, several languages, and several decades since, is a large part of why the Big Five remains a leading framework in personality research today.

None of that makes the model a finished, closed case. The structure has held up well, though not perfectly, across cultures and instruments, and researchers still debate finer points — for instance, Colin DeYoung and colleagues have proposed broader factors that may sit above the five, a reminder that this is an active area of research rather than settled history. For the fuller discussion of how strong the evidence actually is and where its honest limits sit, see the science behind the Big Five.

“Big Five” and “Five-Factor Model” now describe the same five traits, and writing outside specialist research commonly treats the two names as interchangeable — this site uses “Big Five” as its primary name for that reason. For the plain-English tour of what the model actually measures, start with What Is the Big Five?, or go straight to the traits hub to meet each of the five traits this history eventually produced.