Skip to main content Skip to Search Box
Summary Article: Raymond Bernard Cattell (1905-98) from Key Thinkers in Psychology

People do not display their ‘personalities’ all the time and in every encounter. It follows that attributions of personality traits must take the form of conditionals, such as ‘If criticized, he smiles and thanks the critic’. Logically, traits are dispositions. However, even if one were to subscribe to the principle that each person has a fixed and permanent personality as a bundle of traits, the question of the organization of these conditional attributes can be raised. Is there any reason to think that there are superordinate and subordinate traits? Should the psychology of personality work towards a ‘Linnaean’ catalogue of behavioural dispositions? Cattell seems to have taken for granted that there was a limited range of generic traits of which everyday traits were species. He called them ‘source traits’. They would be revealed by the mathematical technique of factor analysis applied to reveal higher and higher order correlations among subordinate traits at the level of ‘species’. This would be a ‘theory of personality’. Cattell went further and proposed a genetic or biological explanation of personality and other distinguishing attributes of human beings.

Who was Raymond Bernard Cattell?

He was born in Plymouth, England, in 1903. His father was an engineer, working mainly on developments of various kinds of motors for the military. He did very well in school, and took up a county scholarship at Kings College, London, where he read chemistry. In those years London was alive with projects for new ways of living, new educational theories, new views of the place of women in society, and so on. Cattell developed an interest in using scientific research in the interests of social reform. After some practical experience in the progressive school at Dartington Hall, he took up a formal academic training in psychology, taking his PhD at University College, London, in 1929. His supervisor was Charles Spearman, one of the leading statisticians of the era. Young Cattell worked with Spearman on the development of the method of ‘factor analysis’ which was to dominate his research for good or ill for the next 70 years.

In those days, a doctorate was not a necessary condition for appointment to a university lectureship. Cattell had already begun to teach at Exeter University in 1926 before he had completed his PhD. In 1932, he moved to Leicester to the child guidance programme. In 1937, he began his long career in the United States, at first as a research associate with Edward Thorndyke at Columbia, and then at Clark University, where he remained until 1941. In 1944 he joined the Harvard psychology department at the invitation of Gordon Allport. Cattell's first marriage, to Monica Rogers, was dissolved about this time, some blaming his tigerish devotion to work that took precedence over family life. Cattell's conceptual subtlety, though at odds with his statistical methodology, sets him off against the over-simplifications of other trait theorists, such as Hans Eysenck, and seems to reflect the influence of Allport's wide-ranging mind. However, left to himself, he slipped back into a theoretical stance at variance with Allport's point of view, in particular in his adherence to the concept of a source trait.

In 1946, Cattell moved to the University of Illinois at Champagne-Urbana. In the same year he married Alberta Schuetter. The mandatory retirement rules still in force in those days deprived the university of his services in 1973. For the next 25 years he continued to work with the same demonic energy, at first in Colorado. In 1978 he moved to Hawaii, where he taught at the Hawaian School of Professional Psychology.

His last years were marred by a bitter controversy concerning the racist implications of his insistence on a genetic basis for personality. He had been nominated for the award of the Gold Medal of the American Psychological Association in 1997. When this became known, a group of people raised objections to the award, accusing Cattell of expressing racist views in his writings. The controversy became a matter of public concern, and the award was postponed. In the end Cattell himself declined the medal. Whether the bitterness of the dispute and Cattell's sense of being wronged and traduced had effects on his health we will never know. He died on 2 February 1998, very soon after these events.

What did he contribute?

Cattell's researches were driven by method much more than by a theory. His reliance on statistical analyses was absolute. His work is heavily marked by the influence of Spearman, the inventor of one of the most important measures of correlation, and the associated technique of factor analysis.

Cattell published widely, but the book most germane to his work in personality psychology was The Scientific Analysis of Personality (1965). Though, perhaps under the influence of Allport, he included both situations and roles in his conception of the total ‘package’ within which personality played a major part, his main theoretical concept was the ‘trait’. How he interpreted the concept deserves careful examination. He distinguishes three kinds of traits: ‘abilities’, ‘temperamental attributes’, which covered a wide variety of specific responses, and ‘dynamic traits’, involving matters like ‘motives’. Roughly speaking, Cattell's ‘traits’ seem to be tendencies to respond to situations and other people. Dynamic traits come and go with people's shifting motives. For Cattell a psychological type was no more than a collection of relatively permanent traits. He insisted that clarity could only be achieved by paying attention to the details, that is traits that are summed up in ‘type’ talk.

Surface traits are to be explained by ‘source traits’ that ‘operate as an underlying source of observable behaviour’ (Cattell, 1965: 67). They should be distinguished from surface traits which are defined in terms of overt behaviour. For example, Eysenck's extravert/introvert dimension is a distinction in surface traits. This suggestion raises some fundamental issues in methodology. As we will see, Cattell arrives at his catalogue of source traits by a mathematical analysis of the distribution of behavioural descriptions. Source traits are mathematically created groupings of surface traits. Logically they seem to be higher order classificatory categories, like genera in biology. How then could they be the sources of anything? This ontological problem of the nature and standing in psychological reality of source traits was never resolved in Cattell's writings. Had he been an old fashioned positivist this question could never have arisen. According to that account of science, scientific concepts were nothing but ways of referring to clusters of observable behaviours. A source trait, interpreted positivistically, just collected up more specific kinds of behaviour. Cattell wanted such traits to be the sources of specific behaviours. This presumes that they must have some more robust mode of existence in the human being than as mere mathematical artifacts. The most telling criticism of the very idea of source traits came from Cattell's erstwhile colleague, Gordon Allport, as we have seen.

Instead of trying to reduce complex patterns of behaviour to simple relationships between a dependent and an independent variable, Spearman had provided his protégé with a method for analysing the contributions to a certain outcome of a great many variables: multivariate analysis. The principles of this kind of analysis are quite simple. The first step is to correlate every variable with every other. Those which are highly correlated are taken as compound second order variables. Correlations are sought between these to create compound third order variables. The process is repeated until there are no more good correlations, and we have arrived at independent super variables. Each group of second or third order variables is renamed as a ‘factor’, presumed to be ‘responsible’ for the coming into being of manifestations of instances of the lower order members in each group in the appropriate conditions. The analysis does not settle any question as to the psychological or physiological reality of the ‘factors’ since they are merely names for groupings of lower order entities, generated by the mathematical procedure.

Here are some of the surface traits underlying the source trait ‘neuroticism’ (Cattell, 1965: 211), which is itself a name for a cluster of behavioural traits or tendencies, drawn from ‘intelligence’, ‘ego-strength’, ‘dominance’, ‘surgency (extraversion)’, ‘tender-mindedness’, ‘non-conformity’ and so on. ‘Neuroticism’ is the name for the group comprising low ego-strength, high tender-mindedness and high nonconformity. Now identified as a source trait, it begins to play the role of a hypothetical underlying cause of local manifestations of its component traits. Remember that none of these terms describes any actual behavioural tendency or disposition, but a lower order compound variable extracted by analysis of answers to questionnaires. Some of these expressions have made their way into the vernacular.

However, armed with Allport's criticisms, we can easily see that though the names that appear in the list above seem to be psychologically meaningful, they are mere labels. To test this, simply look up the list of actual, observable behavioural dispositions each of these statistical clusters encompasses, and the incoherence becomes evident.

How could someone of Cattell's obvious sharpness of mind fail to realize that the source trait concept had little if any psychological authenticity? If we compare his life experience with that of Allport, a certain narrowness is evident. After he left Harvard he was surrounded by colleagues and students who were very much under his personal influence. At the same time his headlong rush of activities, writing faster than most people can read it was said, precluded the reflective leisure that is one of the conditions for fruitful self-criticism. There is something tragic about a life work built upon sand.

Further Reading
Primary Sources
  • Cattell, R. B. (1965) The Scientific Analysis of Personality. Penguin Harmondsworth.
  • Cattell, R. B. (1979) The Scientific Use of Factor Analysis in Behavioral and Life Sciences. Plenum Books New York.
  • Secondary Source
  • Pervin, L. A.,; Cervone, D.; John, O. P. (2005) Personality: Theory and Research (Ch. 7). Wiley New York and Chichester.
  • Biography
  • Sills, D. L. (1989) R. B. Cattell. In International Encyclopedia of the Social Sciences: Biographical Supplement (Vol. 18). Macmillan New York.
  • © Rom Harré 2006

    Related Credo Articles

    Full text Article Cattell, Raymond Bernard
    Biographical Dictionary of Psychology

    Born : 1905, Hill Top, Staffordshire, England Nat : British Ints : Clinical, educational, evaluation and measurement, experimental,...

    Full text Article Personality
    The Encyclopedia of Applied Animal Behaviour & Welfare

    There is no universally accepted definition of personality, even within the human psychological literature, but the term ‘personality’ in...

    Full text Article Gordon Willard Allport (1897-1967)
    Key Thinkers in Psychology

    The contested concept of ‘personality trait’ has been the focus of both practical studies and theoretical debates for half a century or more. It...

    See more from Credo