Friends and colleagues share memories of Donald W. Fiske, who died in April at 86.
A Eulogy for Donald W. Fiske
Donald W. Fiske was an exceptional individual in many respects. He and wife Barbara moved to Hyde Park, IL and he joined the faculty at the University of Chicago in 1948, where he remained his entire career. Don’s research promoted rigorous methods to make psychology a true science, and a generation of scholars felt his influence.
Don co-authored an early article with Donald Campbell that provided a quantitative approach for measuring differences between people. This analysis separated the distorted information gathered in questionnaires, interviews and other research methods from the information that was a truer reading of a person’s actual characteristics. It became a classic in the field and is still required reading for many students in psychology.
This paper, entitled “Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix,” appeared in Psychological Bulletin in 1959, and is the most cited article – over 2,000 citations, twice as many as any other article – in the publication’s 100-year history.
The paper addresses a fundamental problem in scholarship – how to interpret psychological measurements. When scientists seek to study an abstract characteristic of an individual, such as intelligence or extroversion, they need measures or instruments of that characteristic. Among Don’s contributions was the establishment of a rigorous means of evaluating whether measures were valid and what procedures were needed to make them valid.
Although Fiske often investigated personality traits and measures, the methods he developed were tremendously influential because they could be applied widely in the field of psychology. Fiske pointed out ways in which the concepts in the field were imprecise, and he provided a means of increasing their precision. As a result, psychology stands today on a much stronger scientific foundation.
Fiske was the author, co-author, or editor of 11 books dealing with measurement and other methodological issues in psychology. One of his books was “Measuring the Concepts of Personality” (1971), which traced the core issues of his early research.
Fiske continued his work by studying the use of observer ratings, including people’s self-observations. He argued that researchers should avoid interpreting self-observations as accurate measures of the traits they were describing. His books, “Face-to-Face Interactions: Research, Methods, and Theory” (1977, co-authored with Starkey Duncan), “Strategies for Personality Research” (1978), and “Interaction Structure and Strategy” (1985, also co-authored with Duncan), established more reliable methods of measuring personality traits. He advocated using fine-grained nonverbal behavior and concrete indicators so that judgments about personality could be made reliably.
In later work, he examined how psychology and the other social sciences could become more rigorous. That research was pursued in his book “Metatheory in Social Science” (1986, co-edited with Richard Shweder). His contributions dramatically advanced how behavioral and other social scientists think about the abstract meaning of their concrete observations and measures. In fact, he was elected President of the Society for Multivariate Experimental Psychology in 1968 as a direct result of these contributions.
Despite his many books, publications and paradigm shifting research on methods and measurement, Don’s proudest contributions to the social sciences were, of course, his children Susan and Alan.
In the Donald W. Fiske Distinguished Lecture Series – a living tribute to a great contributor to science and society – Susan was the inaugural speaker, and the pride Don felt that day was apparent to all. Don took an active part in the lecture series each year and continued to be a constructive influence on the department until his death.
Don will not only be remembered for his many contributions to psychology and to the University of Chicago, but also for his quick wit, keen mind, and generous spirit. We will miss him dearly.
A Brave Intellect Remembered
In reflecting on almost 20 years of collaborating with Don Fiske, one of his most striking characteristics was that he was able to tolerate me, and in good grace. I remember many research decisions arising in our studies on which I was advocating a certain position vigorously, and, I thought, irrefutably. He would sit quietly, and when I was finally done, he would respond with a carefully detailed position with which a reasonable person could only agree. I would find myself somewhat chagrined that my position was less irrefutable than I had thought, but nonetheless caught in the powers of his experience, expertise, and tact.
I initially approached Don because I wanted to study individual differences in face-to-face interaction. It seemed intuitively compelling that people have recognizable interaction styles. The idea, thoroughly in line with the general consensus among investigators of “nonverbal communication” of that time, was that we could describe both interaction process itself and individual differences within interaction through counts and timings of various actions by participants, such as gazing, smiling, gesturing, and speech.
We undertook a large study in which these counts and timings were summarized over entire interactions (Duncan & Fiske, 1977). Included were several different types of variables based on these summary counts and timings. Using these “simple rate” variables, we carried out extensive correlational analyses and analyses of variance. In due course we began to interpret the results.
One prominent flashbulb memory I have from working with Don occurred while we were discussing various interpretations of a set of interesting results. We simultaneously came to the same realization, and, looking at each other without speaking, each of us knew that the other had the same idea. It had become apparent that it was impossible to interpret the results using the simple-rate variables in the study. The problem was that a fundamental property of interaction is that it is composed of closely related sequences of actions. That is, each participant’s actions were contingent, at least in part, on the actions of the partner. When information on interaction sequences is not included in the data, the results become uninterpretable.
The resulting Duncan and Fiske monograph reported this and several other studies (1977). In our final chapter, we included the following statement, originally italicized for emphasis: “We therefore recommend that investigators of the conduct of face-to-face interaction abandon [studies using simple-rate variables], replacing them with studies based on analysis of action sequences.”
Although not warmly received, the position has never, to my knowledge, been disputed in principle and may have contributed to the rapid demise of the large “nonverbal communication” area of study in social psychology. As a constructive alternative to simple-rate variables, we proposed a new type of variable based on interaction sequences. These “action-sequence” variables were subsequently used in a set of studies reported in a second monograph (Duncan, Fiske, Denny, Kanki, & Mokros, 1985).
As we continued our studies in face-to-face interaction, Don applied these developments to his principle area of expertise – personality measurement. Once again, he arrived at conclusions that were sharply critical of long-standing practices in that area. These conclusions became the subject of vigorous debate among investigators. Though unpopular, Don’s contributions resulted in fundamental reconsideration of research practices in personality measurement.
To me, Don represents the quintessential psychological investigator; he is meticulous, broadly read, methodologically authoritative, sharply analytical, systematic, uncompromising, and intellectually courageous. He simply did not flinch from the intellectual positions to which his analyses drew him. Perhaps underlying his many substantive contributions, the model of a psychological investigator he provided us all is his most enduring bequest.
Duncan, S D, Jr, & Fiske, D W. (1977). Face-to-face inter action: Research, methods, and theory. p.315. Hillsdale, NJ: Lawrence Erlbaum Associates.
Duncan, S D, Jr, Fiske, D W, Denny, R, Kanki, B G, & Mokros, H B. (1985). Interaction structure and strategy. New York: Cambridge University Press.
Metatheorizing with Don Fiske
Donald Fiske was a proud and committed “positivist;” indeed, he was one of the purest, most significant positivists the discipline of psychology had in the second half of the twentieth century. By positivism I mean an approach to the production of knowledge that seeks to ground all claims about what the world is like in what the anthropologist Ernest Gellner (another positivist) called “terminal facts,” and in deductive logic that draws a sharp distinction between objectivity and subjectivity, placing the latter, interpretative things in the hands of the humanists. Don Fiske always drew a sharp distinction between the sciences and the humanities. He loved the humanities but located them on the opposite side of a continuum that ran laterally from the positive sciences to the interpretive arts.
On his end of the continuum the words used in ordinary syntax presented an epistemological problem, because words have multiple meanings, and outside of their local context their meanings can be quite indeterminate, making agreement on application hard to achieve. Often these meanings are too inscrutable to serve as a useful language for science, he believed. The language philosopher Paul Ziff says, in familiar terms, “A cheetah can run faster than a man.” What claim about the world is actually being made? A three-legged sick cheetah is a cheetah. Does the sentence mean that a three-legged sick cheetah can run faster than a man? And how long is the race? Is the sentence false if there are many men who can beat any cheetah foolish enough to enter a marathon? Does the sentence merely mean that some cheetahs can beat some men under some circumstances, with the implication that some men can beat some cheetahs under some circumstances? You can see the problem. Don Fiske believed the problem was deep and that ordinary verbiage got in the way of developing a universal, context-free language for a positive science.
He also thought big things were much harder to scientifically understand than small things, which is why he preferred to break things down into the smallest possible units. For his positive science he preferred to study things that could be seen rather than unseen – no transcendental entities. Reliable observation of atomic units with a minimum reliance on the semantics of everyday life was his notion of how to build a universal psychological science. He was in favor of the study of behavioral acts and very skeptical about the study of unseen global subjective states, such as latent personality traits. His doubts extended to the study of motives and emotions. I can easily imagine him saying something to the effect of: “Stick to terminal facts if you want to do science, or else turn to the humanities, which is a noble but different calling.”
Don was also my colleague and friend at the University of Chicago, and invariably the issue of how science ought to be done was a topic of conversation and debate during lunch and dinner conversations. I knew his work before I arrived at the University of Chicago in 1973, because the anthropologist Roy D’Andrade had brought Don Campbell’s and Don Fiske’s astonishing “multitrait-multimethod matrix” article to my attention while I was writing a PhD thesis on the influence of culture and semantic structures on personality assessment (Campbell and Fiske,1959). That paper by the two Dons may well be one the most significant methodological pieces ever written in the social sciences, and its challenge – tied to the problem of method specificity or “method variance” – has yet to be met. To my surprise and delight, Don knew all about my thesis work when I arrived at Chicago, and, as I later found out, had been supportive of my appointment. He was one of the first members of the faculty to invite me out to lunch. That was the beginning of our friendship and of many debates about knowledge production in the social sciences.
Those conversations went on for nearly three decades, but surely the highpoint was our collaboration in the early 1980s when we co-organized a conference on “Potentialities for Knowledge in Social Science,” held at the University of Chicago on September 11-14, 1983. Don was displeased with the state of knowledge in several areas of psychology, and he read widely in the social science crisis literature, systematically indexing and classifying every type of argument about whether the idea of a social science made sense, and what type of science it might be.
I learned what it means to be organized and systematic by watching him build his taxonomy of critiques. Here are the kinds of questions he had on his mind: How come there is a lack of convergence over time in theories and concepts in the social sciences? How come social science generalizations, when they are valid, are so restricted in scope and bound to particular methods and populations studied? How come social science’s findings and formulations seem so remote from practical social policy issues? Was Rene Descartes right that meanings, intentions, ideas, values and emotions cannot be studied scientifically?
That 1983 conference, which Don organized with some modest help from this friend, was designed to address such questions, and was published under the title Metatheory in Social Science: Pluralisms and Subjectivities (Fiske and Shweder, 1986). That book contains one of the clearest statements of Don Fiske’s version of positivism (Fiske, 1986). The meeting was notable, in part, because Don succeeded at gathering together so many of his peers, including Campbell, Paul Meehl, Lee Cronbach and others who had helped shaped the scientific canons of American psychology in the post-World War II era, and now had somewhat differing views of the “faith.” Ken Gergen was also there, and Paul Secord too, and scholars from several disciplines, including one physical scientist.
The meeting was also notable because it was so disputatious, in part because Don had managed to invite the full variety of positivists, post-positivists, and anti-positivists. But he was eager to hear those disparate voices and to assess the state of science. He attempted to produce progress in an uneasy discipline by first reading every critical book and essay he could find, and then inviting and organizing a contentious debate. As I look back and think fondly about Don and his projects, the “Metatheory” enterprise stands out in my mind as his monument to the future of a scientific psychology.
Campbell, D T & Fiske, D W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin. 56:81-105.
Fiske, D W (1986). Specificity of method and knowledge in social science. In D W Fiske and R A Shweder (eds), Metatheory in Social Science: Pluralisms and Subjectivities. Chicago: University of Chicago Press.
Fiske, D W & Shweder, R A. (eds). (1986). Metatheory in Social Science: Pluralisms and Subjectivities. Chicago: University of Chicago Press.
The ‘Children’ of Donald W. Fiske
As one of Donald Fiske’s graduate students, it took me years to accept his implicit invitation to think of him as “Don.” To us, he was “Mr. Fiske” (in the Chicago tradition) in person and “DWF” on the many notes and written comments that he left on our desks in his research suite. He was a formal man, and his manner did not invite familiarity. However, part of this formality was planned with regard to his relationships. On the day I defended the dissertation that I had developed under his watchful eye, he left me a complimentary note with only a few suggested revisions, and he signed it “Don.” I had passed an important threshold in our relationship – only my years of automatic training stood in the way of acknowledging the change.
In contrast to his reserved personal style, Donald Fiske’s intellectual style was open and very personal. For much of his mid- and late-career, he reflected on how personality psychologists like himself seemed to have taken a blind alley in science. He would invite his students into his office to think about totally new approaches to personality research. He encouraged us to throw away the measures, scales and approaches that he and his post-WWII cohorts had worked on for over two decades. This could have been a depressing activity, but Don Fiske was not morose or brooding. Instead, he was excited to move on, to look around at other sciences for inspiration, and to consider new data and methodology.
Only later did I appreciate how different he was from psychologists who like having their names associated with specific theories, scales, or paradigms. There are no Fiske inventories or statistical coefficients, although he was more than qualified to develop such tools. Instead, he sought – and taught his students to seek – a deeper truth. He wanted to understand personality as a system of behavior in the real world, and he was deeply suspicious of operational definitions that seemed to make this task too simple. Most strikingly, he deconstructed his own previous contributions in the same even-handed way that he took apart the contributions of other personality researchers of his day.
As he reflected on his work in the festschrift that Susan Fiske and I organized for his 75th birthday, he wrote that he thought of himself as a skeptic, rather than a pessimist. Some of those whose work he criticized might not agree, but Don expected this diversity of opinion. His self-characterization rang true for me, because he was at once critical and respectful, interested in new ideas and procedures while remaining cognizant of their limitations. He advanced personality research using a dialectic process, first building the best set of methods he could, and then turning to see whether these best methods produced results that met his high expectations. They rarely did.
Having the opportunity to study and think with Don Fiske was one of the great fortuitous events in my career. He equipped me with an approach to science that carried well beyond personality research. His appreciation for statistical methods as a means, but not an end, in psychology helped me set an academic course that has been satisfying, and hopefully fruitful. Remarkably, many of the scientific themes he emphasized in my training thirty years ago are still highly relevant to my substantive work on stress, coping, and distress. As I collect daily self-reports, reports of family members, and information about objective events, I feel myself walking around the fundamental behavioral questions, examining them from this side and that, in the Don Fiske way. I do my best in passing on this critical perspective to my own students, who are Don Fiske’s grand-students, and as such will pass the enlightened torch.