In Appreciation: John B. Carroll

Brilliant Contributor

I first met John Bissell (Jack) Carroll in the spring of 1949. He was coming that fall to the Harvard Graduate School of Education as an assistant professor, apparently the choice of two other brilliant statistically-oriented professors in that school, Truman L. Kelley and Phillip J. Rulon. In a slight sense he was replacing me, a one-year full-time instructor just completing his doctorate there and soon leaving Harvard to join the faculty of a Southern university.

Aside from statistics, however, our interests were quite different. Jack’s concerned verbal and linguistic ability. Mine were – temporarily, as it turned out – in animal experimentation. My doctoral dissertation, a study of new area of learning called “partial reinforcement,” involved extensive running of rats on a T-maze. Jack had been greatly influenced by both B. F. Skinner and Louis L. Thurstone; I by Kelley, Rulon, Orval H. Mowrer, Frederick B. Davis, and (at a distance), E. F. Lindquist and Palmer O. Johnson.

It was immediately obvious that Jack was truly brilliant in several research areas. He did good work at Harvard, but the atmosphere of the “Ed School” had changed quickly, as the promising young assistant or associate professors temporarily located there post-war while seeking a more permanent position elsewhere in a department of their specialty. Mowrer, then an associate professor, went to the University of Illinois as a full professor of psychology, devoted mostly to research. Kelley had retired to his beloved California in 1947. Morris Opler, who studied Navaho Indian culture, left for an excellent position in anthropology. Bob Sears arrived at the same time as Jack but didn’t stay long.

Thus, the situation was ripe for Jack to leave for the Educational Testing Service in Princeton, New Jersey, to concentrate on research and development. Later, as his professional interests developed, he moved to the department of psychology at the University of North Carolina. He remained at UNC for the rest of his long career, enriching the field of psychometrics with extremely ingenious and very laborious research on mental abilities. His book, Human Cognitive Abilities: A Survey of Factor-Analytic Studies (Carroll, 1993), was as nearly the definitive treatment of that area as was then possible. Also, the book was a tribute to the power of the personal computer when operated by a master of both the subject matter of intelligence and the technique of that often balky machine.

Not until the last few years of his life, after his wife died, did Jack leave Chapel Hill to live with his only child, his daughter, in Alaska. As was typical of this brilliant psychologist, he quickly adjusted happily to very cold Fairbanks. I can’t recall in my own long life meeting a more pleasant person, not at all the hypercritical sort all too familiar among psychometricians. He had a very even disposition, with the ability to interact productively with professionals far less capable and recognized than he. This was apparent when, within the past year, I had the pleasure of spending considerable time with Jack at two professional meetings. He seemed as sharp mentally as ever, making cogent comments about difficult topics.

Let me conclude this appreciation with an anecdote. Long ago there was an erudite commentator, who spoke about various popular topics and was well received by the media. Once Jack remarked plaintively to me, “I don’t understand why he comes off so much more brilliantly than I. I seem to know as much as he does, but I just don’t talk about it as well.” We can be grateful that Jack didn’t become a media favorite at the sacrifice of his many magnificent contributions. I doubt that this media star of yesterday left nearly the mark that Jack has.

Scientific Role Model

My first acquaintance with Jack was when he came to Minnesota in 1981 to give the Richard M. Elliott Lecture. We began corresponding and seeing each other at professional meetings shortly thereafter. At Minnesota, I learned that Jack was B. F. Skinner’s first PhD student. And, shortly after Skinner recruited him, he began boasting to colleagues that he had just secured the best graduate student in the country. Soon, Jack apparently made a similar impression on his peers: Paul Meehl fondly recalled meeting Jack for the first time at a graduate student gathering. A mutual friend provided the introduction and, following Jack’s departure, turned to Meehl (and said): “Paul, there goes a guy every bit as smart as you, maybe smarter.”

Jack’s career reflects an unfolding of multiple lines of accomplishment, any one of which would constitute an impressive academic career. Many consider Jack, for example, the leading contributor to educational psycholinguistics of the 20th century.

And his 800-page magnum opus, Human Cognitive Abilities: A Survey of Factor-Analytic Studies (Carroll, 1993) is among a handful of the best treatments ever published on individual differences in cognitive abilities; it is certainly the most definitive treatment of their nature and organization. Jack reanalyzed over 460 data sets that he had, over the years, meticulously collected and organized. Before this major 1993 contribution, different models were proposed to characterize and organize human cognitive abilities. Spearman stressed one factor, whereas (at uniform molarity levels) Thurstone preferred seven and Guilford argued for 120 or more dimensions. Jack’s Three-Stratum-Model cast a clarifying light on the dominant, minor, and tiny currents of covariation traveling through the indicators assembled by these and other early pioneers. Furthermore, he illustrated how these nascent formulations, as well as others like Cattell’s and Vernon’s, did not require a complete jettisoning, but rather augmentation, refinement, and structural rearrangement. Undoubtedly, Carroll (1993) will be found on bookshelves and in citations of scholars of human cognitive abilities for decades.

Jack was more than a profoundly talented scientist; he was also a superb role model across multiple fronts. He exemplified how to be courageous and courteous during difficult times. When empirical findings contradicted the contemporary social science point of view, Jack did not shy away from controversy (e.g., his 1995 review of The Mismeasurement of Man). His writing possessed a rare combination of sensitivity and scientific integrity. And, thus, in the 1990s, when misinformation on all sides of the contentious debate about assessing human cognitive abilities peaked, it was not surprising that Psychological Science (1992) and Intelligence (1997) chose Jack to craft opening articles for special treatments of key findings and issues.

A dedicated and skilled mentor, Jack routinely responded to queries from faculty and students in a conscientious, detailed, and exceedingly informative manner. Yet, he could be quite forthcoming when he needed to be. It is easy for me to recall an invited address Jack chaired at APA (2002), where he clearly and firmly told me that my time was up and that it was time to let the discussant speak. He could be effectively forceful while maintaining his charm all at the same time. Jack’s scholarship and personal qualities are why the International Society for Intelligence Research found it fitting to christen their annual award, for the best graduate student paper presentation, the John B. Carroll Award.

Although his curriculum vita is certainly noteworthy in terms of its sheer size, approximately 500 publications, it is not a CV that pales when methodologically sophisticated scientists examine it deeply. His work is rich with psychological insight. Jack’s ideational products are extraordinary blends of analytic-critical and creative-synthetic scholarship. His is the kind of work, the kind of heavy lifting, which would be chosen to impress scientists in other disciplines. In Jack’s methodical and systematic way, personally imperturbable and entirely self-confident, he showed by example how psychology could be practiced as a cumulative discipline. He also showed how models of individual differences reveal consistent patterns of covariation, which reach out and form important external connections, and complement experimental applications, even over decades of cultural change. Without careers like Jack’s, psychology as a scientific discipline would have much less to proudly call its own.

Jack Carroll, known by his colleagues, family, and friends for his profound intellectual gifts, curiosity, optimism, and wit, and his unfailing integrity and sincerity, will be missed deeply by all who had the privilege of knowing and learning from him.

Reference
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge, UK: Cambridge University Press.

Collaborator, Friend, and Teacher

When I joined the L. L. Thurstone Psychometric Laboratory in 1989 as a 25-year-old assistant professor, I was in awe of Jack Carroll’s work and wondered what it might be like to have him as a colleague. From day one there was never an age barrier between us: We immediately started talking and exchanging articles about topics related to exploratory factor analysis, structural equation modeling, the structure of personality, and music (Jack had perfect pitch and was a piano virtuoso). We would visit one another’s offices and I would go to Jack’s house, where his entire basement – his workspace, was packed from floor to ceiling with books and journals.

Early on, Jack taught me how to use the factor analysis program that he created in consultation with L.R. Tucker, and I taught him how to use Windows, a mouse, and the most recent versions of LISREL. To my great joy, Jack was a regular guest lecturer in my exploratory factor analysis graduate seminar. His intent was to talk about his factor analysis program and its special features (compared to standard factor analysis programs on the market); my intent was to allow him to do that, but also to distract him into providing personal stories about factor analysis greats. I’m happy to say that both these intentions were achieved successfully.

In 1998 we conducted some research together on the hierarchical structure of personality in different Five Factor model data sets using the Schmid-Leiman approach, a factor analytic decomposition that permits partitioning of factor variance into components for higher-order factors and their lower-order components (Panter & Carroll, 1999). This method is used frequently when examining the structure of cognitive abilities, but rarely is used to examine personality structure.

At this time Jack was spending most of his time at home, so I would go to his house two or three times a week to work on data sets and talk about findings. I would let myself into the house, chat with his beloved wife, Mary, and then make my way down to find Jack. Amidst bookshelf beside bookshelf and piles upon piles of manuscripts, I would call out: “Jack, are you down here?” and he would cheerfully call out “Over here,” whereupon I would make my way through the academic labyrinth and try to find him at one of the three workstations he set up for research.

We spent hours in that basement working on structural equation modeling analyses and output based on data sets sent to us by personality researchers around the country. I can easily say that I have never worked with a more meticulous collaborator than Jack. When you asked him to read an article (yours or someone else’s) he truly studied it – every word, every number, every reference. Every e-mail you sent him was printed and catalogued using his dot-matrix printer, which, incidentally, he greatly preferred over the laser one I helped him set up. He relied on elaborate labeling schemes for understanding factors and items, and near as I could tell, he would memorize most of the numbers in large analyses. Several times I saw direct evidence of this. In that productive workspace, I learned major life lessons about styles of collaboration, what it means to have true passion for one’s field, and what it means to be a lifelong learner. He mastered current topics in personality structure and in structural equation modeling during the final two decades of his life. Jack was forever open to new ideas and different styles of thinking, and never minded – in fact welcomed – being challenged on data analytic styles and approaches.

In 2002 Jack received the Gold Medal Award for lifetime achievement in the science of psychology from the American Psychological Foundation. The award was presented at the annual American Psychological Association meeting. Jack’s daughter, Melissa Chapin, accompanied him on that trip. (Jack had already moved to Fairbanks, Alaska to be with his daughter after the death of his wife Mary.) It was a wonderful celebration; Jack was absolutely delighted. That next day we cheered for Jack again as part of the Division 5 social hour. Jack was in his element during these celebrations and was truly honored that his lifetime contributions to the field of individual differences was recognized in these meaningful ways.

Reference
Panter, A. T., & Carroll, J. B. (1999, October). Assessing hierarchical structure in the Five-Factor Model. Paper presented at the Society of Multivariate Experimental Psychology, Riverside, California.

The Highest Standard

I first learned of John B. Carroll through his writings, especially his influential first book, The Study of Language (Carroll, 1953). Responding to a request from John W. Gardner, then president of the Carnegie Corporation, Carroll had agreed to survey linguistics and neighboring disciplines in the educational system from K to 16 and beyond. This remarkable book presents his findings and recommendations for the development of a general science of communications.

In the early 1960s, Jack Carroll and I first met. He chaired one committee and I was a member of another that convened simultaneously at the Educational Testing Service in Princeton, New Jersey. Both committees were invited for dinner at the home of ETS president Henry Chauncey and his wife, Laurie. (The home is now Laurie House, adjacent to the ETS Conference Center.) Jack Carroll, an accomplished pianist, perhaps was warned to prepare to entertain after dinner. In any case, he asked me for assistance and we retired to a quiet room to compose a parody on factor analysis. As he played a well-known tune, we sang our composition as a duet. In that convivial environment, it was received with hearty applause. That experience sensitized me to Jack’s unusual talents, creative spirit, and lively sense of humor. I was delighted when in 1974 he and I became faculty colleagues. He was William R. Kenan, Jr. professor and director of the L.L. Thurstone psychometric laboratory, having succeeded me in the latter position.

Jack’s interests in languages began early. At age 13 he became an apprentice to the linguist, Benjamin Whorf, then a student at Yale working with Edward Sapir. Jack had attended a talk by Whorf on the Aztec and Mayan Indians at the Hartford Children’s Museum, after which the two met regularly, to translate texts from the Nahuatl and to decipher Mayan hieroglyphics. In Hartford Public High School, he took courses in Latin and Greek, read French and German, and studied the grammars of Sanskrit and Armenian.

In 1937, Carroll graduated with highest honors in classic studies from Wesleyan University, and attended the Linguistics Institute at the University of Michigan. Sapir was on the faculty. Both Sapir and Whorf advised Jack Carroll not to pursue graduate study in linguistics because career opportunities were so limited; instead, he should study the psychology of language. Consequently, as a graduate student in psychology at the University of Minnesota, he became the first of B. F. Skinner’s PhD degree recipients. His dissertation, “A Factor Analysis of Verbal Abilities,” actually was completed in L. L. Thurstone’s psychometric laboratory at The University of Chicago, where Jack had spent the summer of 1940.

Carroll’s bibliography lists more than 500 entries, including Human Cognitive Abilities (Carroll, 1993), a masterful synthesis of findings from the published literature. He performed factor analyses on 460 sets of data from that literature, using identical procedures on each, enabling him to draw firm conclusions about the structure of intellect.

Carroll’s work spans cognitive psychology, quantitative psychology, language learning and linguistics, and educational theory and practice. The breadth of his influence is reflected by a list of some of the honors bestowed on him: He was a founding member in 1965 of the National Academy of Education. Between 1971 and 2002, he won a dozen major awards for contributions to educational psychology, language learning, linguistics, testing and measurement, and the study of cognitive abilities. In 1998 he was appointed a James McKeen Cattell Fellow by the American Psychological Society “for a career of significant intellectual contributions to the science of psychology in the area of applied psychological research.” In 2002 he won the Gold Medal Award for life achievement in the science of psychology from the American Psychological Foundation.

Jack chose to “retire” in 1982, but only from the University payroll. He was so fully dedicated to research and writing that he largely ignored warning signs about the increasing severity of his diabetes. Once he nearly lost his life when, driving home the two miles from his university office, he lost consciousness and his car struck a telephone pole. After that, he attended more closely to the advice of his physicians.

Throughout his long career, Jack was demanding of high standards in students and co-workers while maintaining the very highest standards in his own work. He served as a model of superb scholarship for all who were fortunate to have contacts with him.

References
Carroll, J. B. (1953). The study of language: A survey of linguistics and related disciplines in America. Cambridge, MA: Harvard University Press.

Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. New York: Cambridge University Press.