My memories of Paul Kline

Paul Kline was born in 1937, London, and died on September 25th 1999 (at Bath, UK) from pneumonia, which was associated with very recently diagnosed acute leukaemia. It was sudden, unexpected, and a desperately sad loss for both for his family, his friends, and colleagues. He was aged just 62. He became an academic psychologist almost by accident, training as an educational psychologist following a spell teaching classics and then moving on to read for his PhD at Manchester. This work combined the two elements – psychological assessment and Freudian theory – that fascinated him for the rest of his life, and which led to two editions of his standard work Fact and Fantasy in Freudian Theory. His classics training also served to provide the basis for his views on academia, lecturing, and studentship. Paul had great respect for academia, but, very much as a vocation where one is dedicated to the elucidation, greater understanding, and the eventual imparting of new knowledge to students and other academics alike. As had been said of him many times, Paul was an "old-style" academic, one who aspired to excellence solely via intellectual pursuits and research. In his latter years, this frequently brought him into dispute with institutional administrations, which now tend to put greater value on administrative and financial prudence than academic excellence. A more formal description of his considerable contribution to the area of individual differences can be found in Barrett and Cooper (in press, Personality and Individual Differences). However, in keeping with Paul’s own rebellious nature, I have chosen to depart from the usual format, and instead, to write a testament to Paul and his work from my own, idiosyncratic, viewpoint. For a few minutes, take a walk with me down memory lane!

My first recollection of Paul Kline was at a department "freshers meet the staff" day, back in 1975, at the University of Exeter, where I had enrolled as an undergraduate student. Because I was about 10 years older than most of the incoming students, and because I generally dislike these or any kind of formal gathering, I was stuck on a back wall like a lonely drawing pin! I had noticed this small, bearded, diminutive man with a pipe stuck in his mouth, who was also looking as uncomfortable with the proceedings as myself. He came over and asked"what’s your name then?" No social etiquette or small talk, no sophisticated academic banter or bonhomie, just a direct question that was fired at me more as a verbal challenge than a mild query! I told him, and asked for his in turn – then found out that suddenly I was talking with one of the three members of that great triumvirate of Cattell, Eysenck, and Kline. It was instant "chemistry" between us. He smoked a pipe, Jamaican and Havana cigars (when he could afford them!), hated formality, was spontaneous, possessed a wicked sense of humour, and was utterly outrageous in some of his comments and caricatures. Whether he or I knew it at the time, that 15 minute or so "meeting" was to lay the foundation of our initial student-mentor and later working relationship for the next 20+ years.

During my 2nd and 3rd years as an undergraduate, I was able to take tutorial semesters with Paul Kline, on personality, psychometrics, and general individual differences. It is here that our relationship really began to take hold – and the serious pipe and cigar smoking began! A gentle and almost permanent blue haze would hang over his end of the corridor – walking into his room was a sensory experience all of its own! But, how I remember his intellectual energy and creativity, and the real excitement and enthusiasm of those special days with Paul. He was incorrigible – his sense of humour was just crazy – the smallest stimulus could sometimes set him off (and me soon following!) on a chain of zany and ludicrous extrapolation that sometimes had me collapsed in pain with laughter. His lectures were just marvellous. No notes, no slides, no OHP, just a little structure – then lashings of bravura, personal anecdotes, key concepts, and facts – but not in any particular order! Taking notes in Paul’s lectures was pointless. You just sat there and let his intellect and enthusiasm absorb and wash over you for an hour. Then you went to the library or textbook to finish up the details. University academia at its finest.

Following my first degree, I enrolled with Paul Kline as my PhD supervisor. Now we really began working together – on the psychometrics of some key personality questionnaires. It is at this stage that I really began to appreciate the sheer depth of knowledge and critical thinking faculty possessed by Paul. By this time (1979), he had already collaborated with Raymond Cattell on the book The Scientific Analysis of Personality and Motivation (Cattell and Kline, 1977) and had just written the book Psychometrics and Psychology (Kline, 1979). Over the next three years, we ploughed through analysis after analysis of the EPQ and 16PF – publishing a series of papers that culminated both in the successful defence of my PhD, and in his substantive monograph The Factors in Personality Questionnaires Among Normal Subjects (Kline and Barrett, 1993).

By this time, I had the mainframe computer at Exeter on a computational hair-trigger. I only had to press a few keys, and streams of psychometric analyses would pour forth like the output from a demented numerical savant. I remember both Paul and I sitting down in late 1981, and like two drunks who had been totally sloshed for 3 years, suddenly sobering up and realising where we were and what we were meant to be doing. Yes, we had produced analyses and valuable insights for the field – but had we actually achieved more knowledge of the construct of human personality? The slick automation and almost mechanical nature of the many procedures of which we were now capable suddenly served to show us the futility of continuing any further with this kind of research. It was clear to us back in 1981 that there was little point in proceeding any further with questionnaire design and analysis, if we truly wished to gain greater understanding of the constituents of personality.

I left Paul and Exeter in 1982, and went to work with Hans Eysenck in 1983. Paul and I kept in loose contact over the years until about 1993, when suddenly I ended up collaborating on Paul’s 1994 paper Studies with the PPQ and the five factor model of personality (Kline and Barrett, 1994). The Professional Personality Questionnaire was an attempt to generate a 5-factor instrument, similar to Costa and McCrae’s NEO. It was commissioned by an entrepeneur, for graduate personnel selection. However, it was one thing to produce factors for pragmatic purposes, quite another to claim that these factors were somehow fundamental. It seemed to both of us that the field had failed to learn a lesson that we had learned so well back in the early1980s. Factor analysis of questionnaire items is not the way to elucidate causal understanding of personality constructs. A point that Hans Eysenck also made again and again. Further, Cattell in 1957 had already demonstrated the 5–factor model – as part of his 2nd order factor solution to the 16PF. It baffled Paul (and me) as to how, almost 30 years later, the same familiar solution could somehow be re-presented as "new knowledge". Still, it was, and is, and hence Paul’s analytical response in this paper. During 1994/5, we also worked on the major analysis of the Concept 5.2 occupational personality questionnaire, a paper for which we were awarded the 1996/7 citation of research excellence by Anbar Electronic Intelligence (Barrett, Kline, Paltiel, and Eysenck, 1996). A fact that amused Paul no end – especially when he received his certificate!

However, 1996 is the year that Paul’s work and thinking was transformed. He had been asked to review a paper for the British Journal of Psychology by Joel Michell – something to do with measurement and science. Evidently, other referees had indicated that it should be rejected – Paul was asked for an opinion. He recommended that the paper be published. That one decision has had profound ramifications that even Paul did not realise at the time. I remember meeting up with him on my return from a spell in New Zealand, at the July 1997 Spearman conference in Plymouth, UK. If ever a conference served to underline the complete and utter stagnation of an area, this was it (apart from Art Jensen’s offering). Paul was mortified. I was wondering if time could really stand still for so many years without anyone else noticing that nothing was changing. We both decided then and there that if we were going to do anything anymore, it would be to try and work at the theory, processes, and measurement of the constituents of intelligence; the same for personality. He had mentioned the Michell article – and that there seemed to be something of import in it. I had not seen it as it did not appear in print until September – so was unable to comment in any sensible way. Then the Michell (1997) paper appeared, with Paul’s and other peer reviews. The world metaphorically rocked on its axis for a few hours.

To say that it changed both Paul’s and my own perceptions of psychometrics and psychological measurement is no exaggeration. The clock was turned back for both of us – it was as if we were back in the heydays of Exeter – blazing a new trail once again – but now armed with considerable experience, expertise, and Joel Michell to guide us! The reason why both Paul and I avoided all substantive empirical work during this phase was that we both realised we needed to think – and (for me at least) plunge into physics, math, and measurement theory. Michell’s paper was, and still is, a devastating paper in its exposition of the constituent properties of a quantitative science, and the poverty of quantitative scientific measurement within psychology as a whole. The initial result of a year’s thinking about the implications of all this for psychometrics culminated in what I consider to be Paul’s finest book – The New Psychometrics (Kline, 1998, reviewed by myself in October's issue of Personality and Individual Differences, 1999 - a copy is available here). His deduction concerning the "unit problem" is fundamental. During the latter part of 1998 and early 1999, Paul and I were fortunate enough to share the stage as keynote speakers at two BPS "millenium" conferences. We were now jointly working away on the empirical and theoretical implications of Paul’s New Psychometrics – and beginning to make some real inroads into the problems. Before he gave his paper in the second conference, Paul had said to me "I’ll soften them up, you finish them off". We had real fun at these conferences, and I think the audience really did appreciate that Paul, in his own inimitable way, was imparting something that was very special indeed. He spoke as always with passion, integrity, acumen, and humour.

The final problem we were working on over the summer of 1999 was concerned with how to generate substantive theory and an empirical program of research that would embody the essence of Paul’s New Psychometrics. This had stumped both of us for months! But then, a solution had suddenly come to me. I rang Paul to let him know that we were in business – and to say that I was going to visit him in the next few weeks to lay out the whole series of arguments and propositions as I saw them, and work face to face until we had jointly hammered out the final theory and strategy. Things went quiet for a month – then I heard from Paul’s wife, Penny – he had been taken to hospital. Within a week, he had died. Just 62 years of age. Dreadful.

However, his work lives on. I was fortunate enough to assist his publishers (Free Association Books) with the proof-reading of his forthcoming new book on introductory psychometrics A Psychometric Primer. He had finished the manuscript just weeks before he died. So, I was able to appreciate that Paul could integrate the essence of the arguments concerning his New Psychometrics into an introductory textbook. Even more fortunate is that Paul’s new revision of his Handbook of Psychological Testing has just appeared (Kline, 2000). I managed to purchase a copy and have just read it. This is a book, in my opinion, that defines the essence of what psychometrics and psychometric testing is all about. For me, it is the natural successor to Nunnally (for those readers who know the seminal importance of Nunnally’s 2nd edition Psychometric Theory (1978)). This book provides the logic, fundamental thinking, and essential psychometrics that is required to truly appreciate both the benefits and limitations of psychometrics – as a form of measurement of psychological attributes. Again, he has successfully woven the conceptual structure and arguments of the New Psychometrics into the subject matter, with especial regard to applied practitioners.

If I were asked to indicate where Paul had made his major contributions, I would offer his earlier work on the intellectual and empirical examination of psychoanalytic theory. Then, his work on the psychometrics and factor analysis of human abilities and personality (including the studies with Colin Cooper on the empirical elucidation of perceptual defence mechanisms). The string of introductory books that accompanied his explorations are still some of the best books written for students in the area of individual differences. Paul’s psychometrics was what I like to call the "thinking person’s approach" to psychometrics. He was not particularly interested in the minutiae of psychometric procedures or the mathematics, but rather, in the substantive issue of how these procedures could assist a researcher to empirically investigate psychological constructs.

However, I cannot but regard his current work on the New Psychometrics as the pinnacle of his achievements. This is not the already obsolete "New Rules of Measurement" promulgated by Embretson and others (1996) working in the area of item response theory, but a fundamental reorientation and reconceptualisation of all current psychometrics procedures within a conceptual framework of a quantitative science. I know from personal experience that the level of thinking and logical analysis that has been required by Paul to define such an approach dwarfed that required for his earlier work in psychometrics. In essence, Paul has been instrumental in creating the foundation for a paradigm shift in the area of psychological measurement. The tragedy is that my friend is not with me to share in the great adventure that he has begun.

I hope that for a brief moment of time I have been able to make Paul "come alive" so to speak – especially for those readers who might have heard Paul speak, met him, been one of his students, or was a friend or close family. I miss him, his intellect, his kindness, and his wicked sense of humour.


Barrett, P., Kline, P., Paltiel, L., and Eysenck, H. J. (1996). An evaluation of the psychometric properties of the concept 5.2 Occupational Personality Questionnaire. Journal of Occupational and Organizational Psychology, 69, 1-19

Cattell, R.B. and Kline, P. (1977). The Scientific Analysis of Personality and Motivation. London: Academic Press.

Embretson, S. (1996). The New Rules of Measurement. Psychological Assessment, 8, 4, 341-349

Kline, P. (1972). Fact and Fantasy in Freudian Theory. London: Methuen.

Kline, P. (1981). Fact and Fantasy in Freudian Theory (2nd Edition). London: Methuen

Kline, P. (1979). Psychometrics and Psychology. London: Academic Press.

Kline, P., and Barrett, P. (1983). The factors in personality questionnaires among normal subjects. Advances in Behaviour Research and Therapy, 5, 141-202

Kline, P., and Barrett, P. (1994). Studies with the PPQ and the five factor model of personality. European Review of Applied Psychology, 44, 1, 35-42

Kline, P. (1988). The New Psychometrics: Science, Psychology, and Measurement. London: Routledge.

Kline, P. (2000). Handbook of Psychological Testing (2nd. Edition). London: Routledge.

Kline, P. (2001). A Psychometric Primer. London: Free Association Books.

Michell, J. (1997). Quantitative science and the definition of measurement in Psychology. British Journal of Psychology, 88, 3, 355-383

Nunnally, J.C. (1978). Psychometric Theory, 2nd Edition. New York: McGraw-Hill