The core idea of the lexical hypothesis
The lexical hypothesis provides the theoretical foundation for an approach that takes natural language as the starting point for understanding human personality. Its central claim is that the individual differences which matter in human social life come to be noticed by the people who speak a language and are eventually encoded in that language as words. The inference, in other words, is that the more important a personality characteristic is to human beings, the more words should exist for expressing it. The hypothesis can be traced back as far as Francis Galton in 1884, but the first person to carry it out as systematic research was the pair of Gordon Allport and Henry Odbert in 1936. They extracted some 4,500 words describing personality from Webster's New International Dictionary, and that list became the point of departure for all the Big Five research that followed.
From Allport and Odbert to Goldberg
The roughly 4,500 words Allport and Odbert extracted were far too many to submit to factor analysis as they stood. During the 1940s Raymond Cattell organized them into about 171 trait clusters and then used factor analysis to extract sixteen factors, published as the 16PF. Cattell's sixteen factors, however, proved difficult to replicate, and in the 1960s Ernest Tupes and Raymond Christal reanalyzed his data and reported a five-factor structure. Their finding attracted little attention at the time, but in the 1980s Lewis Goldberg conducted lexical research of his own, confirmed the same five-factor structure, and gave it the name Big Five. Goldberg's contribution lay in systematizing the approach based on the lexical hypothesis and in encouraging comparable research in languages other than English. Lexical studies have since been carried out in many languages, among them German, Dutch, Italian, Japanese, and Chinese, and structures close to the five factors have been obtained repeatedly. Some languages, however, have yielded six- or seven-factor solutions, so it cannot be said that the same five factors emerge everywhere.
The significance of the lexical hypothesis, and criticisms of it
The greatest achievement of the lexical hypothesis is that it offered a way of deriving the structure of personality inductively from the collective knowledge of humanity, meaning language, rather than from the hypotheses of a particular theorist. This made it possible to build a data-driven model of personality that does not depend on any researcher's theoretical preferences. Criticisms nevertheless exist. First, there may be important personality characteristics that are not encoded in language at all, sexual orientation and certain cognitive styles being possible examples. Second, research confined to adjectives overlooks aspects of personality that are expressed through verbs and nouns. Third, since language is a cultural product, characteristics that a particular culture does not treat as important are unlikely to be reflected richly in its vocabulary. The HEXACO model of Ashton and Lee, with its six factors, is an example of what widening the scope of lexical research can yield: it added an honesty-humility factor that the Big Five does not capture. One reason a model derived from the lexical hypothesis reads so easily is that it was assembled out of everyday words in the first place. When a description of your own personality, or your partner's, strikes you as sounding about right, that is because the words in it grew out of the vocabulary people have always used in sizing one another up.