A Football Player in a Maximum Security Prison

I was conceived towards the end of the World War II when the mathematician John von Neumann and other very bright European immigrants to North America were creating game theory, the first modern computer, and the atomic bomb.

Over the years I learned that physical force was an instrument used by the “bad” as well as “good” guys to achieve their aims. At the time of my birth, other European immigrants to North America were making major contributions to the study of children’s development, including the development and prevention of aggressive behavior.

The work of two of these immigrants would eventually be central to my early professional education: Fritz Redl, a student of Anna Freud in Vienna, was creating Pioneer House in Detroit to “treat” aggressive children (Redl & Wineman, 1951), while Erik H. Erikson was gathering information to write Childhood and Society (Erikson, 1950).

At the end of the war my father returned to the main occupation he had held since he was eighteen years old: football player. I thus spent my childhood with a father who was celebrated in the newspapers, the sport magazines, and on radio for his dazzling punt returns and accurate placement kicks.

I started organizing football games with cousins and friends before I could read. My success in creating research teams for longitudinal and experimental studies on the development and prevention of aggression may be due to the long experience I had in getting friends together to play American football, a game where each player has a very specific role in achieving the team’s success.

During my academic career I learned that it is less painful to be hit by an opponent when you are running with the football towards his goal line than when you receive an aggressive anonymous review for a grant proposal, or a paper submitted to a scientific journal! During a football game you know who hit you and you learn how to run away from that person on the next play or in the next game. To survive in Academia, you must learn to dodge unknown aggressors!

I naturally majored in physical education (University of Ottawa, Canada), and eventually did a MSc degree in educational psychology at the University of Montreal (Canada). Before doing my PhD in educational psychology at the University of London in Great Britain, I did clinical work in a residential center for juvenile delinquents and in a maximum security prison for mentally ill offenders.

Clinical psychology at the University of London was dominated by Hans Eysenck, whom my psychodynamically oriented professors at the University of Montreal referred to as the devil incarnate, because of his behavioral orientation. I perceived the clash between psychoanalysis and behaviorism similarly to the clash between Catholicism and Protestantism.

Having liberated myself from my Catholic education during my adolescence, at the University of London I liberated myself from psychoanalytic theory, mainly by reading Eysenck’s Crime and Personality (1964). I first reacted to his ideas with the criticisms of my previous psychoanalytically oriented professors in Montreal. But I rapidly realized that I understood the logic of his arguments much better.

This did not prevent me from having friends who were being trained at Anna Freud’s Hampstead clinic. Interestingly, some twenty years later I was asked by Peter Fonagy, who had become director of the Hampstead clinic, to be on their International Advisory Board!

The second major influence during my PhD was developmental psychology. Professor Lindley had been one of the initiators of the series of birth cohort longitudinal studies which started in the early 1950s in Belgium, England, France, Sweden, and Switzerland.

A few months after I started my PhD, I was given a Times newspaper book review (Who Becomes Delinquent? West & Farrington, 1973) by my upstairs neighbor, a Canadian postdoctoral student in cancer genetics, Alan Bernstein, who eventually became President of the Canadian Institutes of Health Research. He knew that my thesis was on the treatment of juvenile delinquents and wondered if I had heard of the study described in the Times newspaper article. I had not! And I was far from realizing that this study and its authors would have a profound impact on my career.

Indeed, my PhD research on the effects of residential treatment for delinquent boys made me realize the importance of understanding the early development of antisocial behavior, and the Cambridge Study in Delinquent Development (Farrington, 2003) is still one of the classic models.

The third major influence was my co-supervisor, the statistician Daisy Penfold, whom I met every month in her office at the end of the afternoon, so that we could have a glass of sherry and talk about the challenges of gardening in a cold country like Canada.

I attended a few of her lectures on statistics, but her main influence was through her questions during our meetings. They forced me to spend large amounts of time reading on statistical analyses to find answers to her questions. I knew I had earned my PhD when I presented her my statistical analyses strategy for longitudinal data based on Joachim Wohlwill’s (1973) work for developmental stages. She said: I have never heard of this type of analyses, but it makes sense.

My principal supervisor, Robert Andry, had published a book on the parental pathologies of juvenile delinquents, and thus helped me think about families’ impact on juvenile delinquency. The main thrust of my PhD research was to assess the impact of treatment centers’ social climate in six residential institutions on the behavior of juvenile delinquents over an eight-month period.

There were two other important influences during my PhD. While reviewing the literature on the treatment and prevention of juvenile delinquency I found a fascinating book that that had been published twenty years earlier and that I had never seen cited.

It was the first report sponsored by the World Health Organization (WHO) after its creation in 1948: “Psychiatric aspects of juvenile delinquency” was written by the Swiss child psychiatrist Lucien Bovet (1951). The aim of the report was to describe the best practices to deal with the juvenile delinquency problem that many countries faced after the World War II.

Bovet emphasized Thorsten Sellin’s comments on the lack of observed effects of the Cambridge–Somerville experiment which had started in the early 1930 (Powers & Witmer, 1951). I could not imagine, at the time, that I would eventually become a close friend of Joan McCord who spent her career evaluating the very long-term effects of the Cambridge-Somerville study.

One of Sellin’s comments was that the boys were probably too old (mean: 10.5 years) when the preventive intervention started. Bovet concluded: “Any rational prophylaxis must, therefore, attack the basic disorders of which delinquency is a sign. Its form and scope will be that of a vast mental health campaign” (p. 45). He even went on to add: “Treatment of the nondelinquent little girl or teen-age girl might perhaps have been the most efficacious prophylaxis for the delinquency which a few years later will break out in her sons” (p. 46).

To this day I am convinced that if Bovet’s 1951 report had been taken seriously the science of preventing chronic antisocial behavior during childhood and adolescence would be much more advanced. Unfortunately, he could not promote his book because he died with his wife in a car accident not long after its publication, and hardly anyone referred to the first ever WHO monograph, not even the latest report on violence by the WHO (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002)!

At the end of my PhD, in the summer of 1976, I went back to the University of Montreal as an assistant professor. I was told that in four years I would be asked to replace the head of the department who had reluctantly accepted the job for one term, and that I needed to quickly gain experience by taking charge of the undergraduate and graduate programs. Naively, I thought that it was normal to do this administrative work, teach, and start a funded research program in a department created only five years earlier with psychodynamically oriented clinicians who did not appreciate quantitative research!

A Kindergarten Cohort Study to Understand Behavior Problems during Adolescence

In 1980, the largest school board of Montreal asked me to help them assess the behavior problems of kindergarten children. It did not take me long to realize that this was a wonderful opportunity to start a longitudinal study with a large sample of kindergarten males from low socioeconomic areas of Montreal and trace the development of aggressive behavior.

I was clearly influenced here by West and Farrington’s (1973) study of delinquency development in London. I eventually decided to randomly choose a group of these disruptive kindergarten boys for a prevention experiment and invited international investigators to help plan this study. The enthusiastic support of Joan McCord, David Farrington, Lea Pulkkinen, and Rolf Loeber was extremely encouraging. We eventually published an edited volume on early prevention of antisocial behavior (McCord & Tremblay, 1992).

Rolf Loeber convinced me to focus on physical aggression, while Lea Pulkkinen helped me sort out the different personality approaches to longitudinal research. She also played an important role in my association with the International Society for the Study of Behavior Development.

David Farrington was convinced that we needed longitudinal-experimental studies such as the one I started and systematically invited me to take part in different committees, including the US National Institute of Justice-MacArthur Foundation planning committee for the study that eventually became the Project on Human Development in Chicago Neighborhoods.

The two years of intensive work on this committee introduced me to the world of American Criminology. I had much pleasure in working with Al Blumstein, Al Reiss, Felton Earls, Lee Robins, Rob Sampson, Denise Kandel, Del Elliott, Joan McCord, Malcolm Klein, David Rowe, David Hawkins, Terry Thornberry, Lloyd Ohlin, Michael Tonry, James Q. Wilson, and many others. Interestingly, David Farrington and I were the only non-American members, and we initiated the joint writing of a chapter between England and Canada by sending each other fax copies!

The Montreal Longitudinal-Experimental study is now close to its fifth decade and has shown significant long-term beneficial impacts on substance use, school achievement, criminal behavior, and earnings (Boisjoli, Vitaro, Lacourse, Barker, & Tremblay, 2007; Algan et al., Reference Algan, Beasley, Côté, Park, Tremblay and Vitaro2022).

In 1995, I was invited to participate in the creation of the National Consortium on Violence Research (NCOVR) funded by the US National Science Foundation and National institute of Justice. This led to an extremely productive and stimulating collaboration with Daniel Nagin of Carnegie-Mellon University to describe the different developmental trajectories of antisocial behavior. The initial results (Nagin & Tremblay, Reference Nagin and Tremblay1999) clearly showed that a large majority of boys from the poorest inner-city areas of Montreal were using physical aggression less frequently as they grew older. Only a very small group of boys (4%) did not show the declining trend. When interviewed at fifteen and seventeen years of age, they were the boys who reported the highest frequency of physical violence, and they were the ones most frequently found guilty of infractions before eighteen years of age.

These results were showing that there was a serious problem with “social learning theory” concerning the development of aggression from school entry to adulthood. Frequency of physical aggression was not increasing while they were increasingly exposed to aggression from the media and their peers (Tremblay, 2000, 2006).

The developmental trajectories were also indicating that the increasing arrest rate for physical aggressions during adolescence appears to be due to a change in adults’ reactions to youths’ aggressions with age. Because puberty leads to an increase in height, weight, strength, and cognitive skills, it also leads to a dramatic change in the negative consequences of physical aggression. Adults are much more fearful of receiving aggressive behavior from a sixteen-year-old than from a ten-year-old. This leads the police and judicial system to start arresting and convicting individuals who have been physically aggressive with others since early childhood (Broidy et al., 2003).

When I reviewed the literature on physical aggression during early childhood in the early 1990s, I realized that there were very few studies, and essentially no longitudinal studies tracing the development of physical aggression from infancy to adolescence. One of the factors that limited the study of aggression in young children was the idea that the physical aggressions of children were not “true” aggressions because young children could not intend to behave aggressively (Kagan, 1974).

I was now convinced that Aristotle was wiser than most modern developmental criminologists, psychiatrists, or psychologists when he wrote that any investigation of causes needed to start at the beginning. So, following his advice, we finally started to study large cohorts of children from birth to understand how and why children learn not to use physical aggression.

It took close to ten years to find the resources to start these studies, and it took all my power of conviction to get my colleagues to accept the inclusion of physical aggression items in the parent rating questionnaires before children were two years of age. Their argument was that children are not aggressive before that age, and parents would be shocked to be asked if their child had physically aggressed against someone.

Our results eventually showed that from nine to forty-two months of age the frequency of physical aggressions increases substantially and then starts decreasing (Côté, Vaillancourt, LeBlanc, Nagin, & Tremblay, 2006; Tremblay et al., 1999; Tremblay et al., 2004; Girard et al., 2019). These studies show that the downward trend in frequency of physical aggression started before three years of age and continues in adolescence. Frequency of anger outbursts and physical aggression increases rapidly from the first year after birth to approximately the third, and then the frequency decreases.

Early Biopsychosocial Developmental Mechanisms Leading to Chronic Physical Aggression and Other Mental Health Problems

Much of the work we have done with our longitudinal studies was “simply” to describe development from birth to adulthood. Surprisingly, reviewers of papers we sent to developmental psychology journals over the years often criticized these papers by saying that they were “only descriptive,” as if taking five, ten, fifteen, and twenty years to describe intra-individual change in behavior was not crucial for the science of human development.

We need to have a precise idea of the long-term developmental trajectories of human behavior to identify the mechanisms that drive development. However, we did not limit our work to describing development. We also made numerous attempts to use longitudinal data to identify potential biopsychosocial mechanisms which could be used to prevent or correct the development of chronic aggression (e.g., Barker et al., 2007; Booij et al., 2010; Côté et al., 2007; Fontaine et al., 2008; Van Bokhoven et al., 2005).

We are studying these biopsychosocial mechanisms with two strategies that are more appropriate than longitudinal studies of singletons: twin studies and experimental preventive studies.

The Future: Qué Sera Sera, Whatever Will Be Will Be …

I envy those who will study human development over the next century. The field will certainly make leaps and bounds. However, I consider myself lucky to have lived at a time when the opposition between social learning of aggression and genetics of aggression led, slowly but surely, to some form of integration of these perspectives with epigenetics. I was very fortunate to take part in the unforeseen leap to intergenerational environmental effects on gene expression (Tremblay & Szyf, Reference Tremblay and Szyf2010). I expect and hope for a major acceleration of our understanding of lifespan biopsychosocial mechanisms in the next few decades.