I stumbled into graduate school having no idea what I wanted to study and ended up specializing in a field that didn’t exist.

It’s hard to imagine a time when there wasn’t much interest in studying normative adolescent development. But such was the case in 1974, when I began my graduate studies at Cornell. Had I not been randomly assigned to work as John Hill’s research assistant, my career would have been very different.

How I Became Interested in Adolescence

I never intended to become an academic psychologist. At Vassar I was a premed psychology major. Applying to graduate school was a backup plan suggested by my apartment-mate, lifelong friend, and collaborator, Jay Belsky, in case I wasn’t accepted into medical school. Because I was certain I’d be admitted, however, I didn’t spend any time investigating psychology graduate programs. By January of my senior year, I had been rejected from all but one of the schools I’d applied to, so I had no choice but to put the backup plan in place. Jay had his heart set on Cornell’s program in Human Development and Family Studies, and with little time to spare and no guidance from anyone else, I hastily applied there too. I wanted to become a family therapist, not even pausing to consider that Cornell’s program didn’t have a clinical program.

When people ask me how I got interested in adolescence, the answer is that it was pure chance. The absence of a plan to study developmental psychology made it easy to accept whatever assistantship I was assigned, with whomever I was assigned to. Absent any interest in becoming an academic psychologist, all the potential advisors in Cornell’s department were fungible. John, I figured, was as suitable as anyone else. Plus, he was the department chair.

When people ask me why I stayed interested in adolescence, I have a much longer and, I hope, more satisfying explanation. Studying adolescence has enabled me to merge three interests that have driven my research over the years: the impact of the family on mental health, the contributions of biological forces to psychological development, and the application of scientific knowledge to social policy.

My Brief Flirtation with Family Therapy

As I’ve explained, my interest in familial influences on mental health was already in place before I entered graduate school. The fact that Cornell offered no clinical training didn’t stop me from obtaining it, though. During my second year of grad school, I paid a visit to a community mental health clinic serving children and families, where I met the lead clinician, Dan Matusiewicz, who had trained with Salvador Minuchin, at the time one of family therapy’s iconic figures. Dan was happy to take me on as an apprentice. I spent one semester reading books and papers he recommended, talking about his cases, and observing some of his sessions.

Soon thereafter he suggested that I take on a client or two. I was petrified. My first case involved a marital couple who fought about their sex life. I had to stifle a laugh when the husband said his wife didn’t like it when he “kissed her cupcakes,” which should have tipped me off that I wasn’t well-suited for clinical work. My second involved a family with a defiant and delinquent teenager who routinely stayed out until one or two in the morning, hanging around with peers who kept similar hours. I tried to teach his mother and father how to be better parents. Each week, I would meet with my mentor, discuss my cases, and plan what I might do next. I don’t think I harmed either client, but I doubt I helped them.

My friends wondered how a twenty-three-year-old who had never been married or raised a child could possibly know how to help couples or families. “You don’t need to be a chicken to know that hens lay eggs,” I would say. Several years after I finished grad school and joined the faculty at University of California Irvine, I applied for licensure (at the time, research and teaching hours counted toward licensing requirements, and California had no specific license for clinicians). After passing a standardized written test, I sat for an oral exam. The examiners passed me but said that I needed to restrict my practice to families with teenagers.

I took on a few clients and soon concluded that I wasn’t cut out for clinical work. Clients’ inability or refusal to take my advice, which was my ridiculously naive approach to therapy, frustrated me. (“If you won’t listen to me,” I told one set of parents after three sessions, “I can’t help you.”) I once heard the eminent scholar of family relationships and one of my role models, Mavis Hetherington, explain that, despite her clinical training, she never became a therapist because she was “incapable of unconditional positive regard.” Me too. Unconditional positive regard is a bad trait for a scientist, where healthy skepticism is necessary. I continued to study and write about adolescents and families, but I took down my shingle. I realized that there is a world of difference between being steeped in the literature and applying this knowledge to treat individuals struggling with marital or parental difficulties. Maybe you do need to be a chicken, after all. I did continue to study and write about families for the rest of my career, though, including several books on parenting for the general public, where I could give advice without knowing whether it was followed.

The Climate at Cornell

My first year in graduate school, 1974, was an important year for both John Hill and Urie Bronfenbrenner, who would later become a second mentor. Bronfenbrenner had just published an editorial in which he wrote, “[P]olicymakers should look to us, not only for truth, but, we must modestly confess, for wisdom as well. In short, social policy needs science.” This proposition has guided my entire career. And because I believe that doing policy-relevant research comes with an obligation to report the findings of one’s research to the general public, I have made it a point to author or co-author books for general audiences. Urie’s work in the mid seventies, on the ecology of human development, also got me interested in studying the context in which young people develop, including the family, peer group, school, workplace, neighborhood, and culture.

John had just written a draft of a white paper for the National Institute of Education outlining a comprehensive framework for the study of adolescence. As I’ve explained, at that time the field of normative adolescent development simply didn’t exist. True, there had been seminal works by Anna Freud, on detachment; Erik Erikson, on the identity crisis; and Jean Piaget, on formal operational thinking, all published in 1958. But empirical support for the major propositions of each of these viewpoints was lacking (and would never really materialize). Healthy adolescents, it turns out, don’t detach themselves from their parents. The identity crisis is hardly universal and is more likely to take place during young adulthood than adolescence, if at all. And there is little support for the notion that formal operational thinking characterizes a coherent stage of cognitive development that all adolescents experience. Indeed, today, it’s hard to find research on detachment, the adolescent identity crisis, or formal operations.

No Grand Theories of Adolescence

The failure to confirm much of what Freud, Erikson, and Piaget wrote about adolescence is one reason theory has played such a small role in the development of the field. Once these approaches lost their influence, by the early 1980s, no grand theories of adolescence stepped in, with the possible exception of evolutionary developmental theory, which has grown in its influence recently. That perspective has its proponents (myself included) but remains controversial, in part because of its connection to G. Stanley Hall, the earliest writer to discuss adolescence as a developmental period. Despite his many brilliant insights into the nature of adolescence, some of Hall’s ideas have been dismissed as anachronistic, racist, or wacky, but most of his writing was amazingly prescient.

The general absence today of grand theories of adolescence also inheres in its origins as a largely clinical enterprise. Before 1970, much of what had been written about young people was derived from studies of psychologically troubled populations. As empirical research on more representative community samples began to appear in the literature in the early 1970s, it became increasingly clear that that the portrayal of adolescence as a period of “normative disturbance,” as Freud had written, was largely incorrect. Nevertheless, the stereotype of adolescence as a time of storm and stress led scholars to turn their attention to the study of individual differences rather than to search for developmental universals in order to understand why some teenagers were troubled and others were not. Unlike the study of development in infancy, which focuses mainly on what all babies have in common, the modern study of adolescence has been mainly about how people differ. One early exception to this was seminal work by my friend and collaborator, Ellen Greenberger, on the development of psychosocial maturity, which is as insightful today as it was when it was initially published in the mid 1970s, and which has had an enormous influence on my own work.

Finally, the general absence of theory in research on adolescence is a product of the long history of research on adolescence concentrating on emotional and behavioral problems, like depression, delinquency, and substance abuse. This has its origins in the clinical perspective on adolescence that dominated most of the twentieth century but has continued because the major funding for research on adolescence has come from the National Institutes of Health (NIH). Most of this research has been designed to inform prevention and intervention, rather than to test theoretical propositions.

A Tale of Two Serendipities

The framework John Hill articulated in his white paper was based on the idea that there were just three universal features of adolescence: puberty, abstract thinking, and social status redefinition; that these changes played out in contexts that varied across SES, ethnicity, culture, and historical time; and that the impact of these changes on psychological development was therefore contextually conditioned.

Because of my interest in family therapy, I devoured studies of family interaction during my first semester of graduate school. One afternoon, I came across an article by Theodore Jacob, which reported the results of a cross-sectional study comparing family interaction in preadolescence and adolescence. I wondered, if family interaction changes during the transition into adolescence, what was driving it? Hill’s model suggested where to look: puberty, abstract thinking, and status redefinition.

By the end of my first year, I had designed a short-term longitudinal study with three observation points separated by six-month intervals, in which family interaction would be measured in triads of firstborn early adolescent boys and their parents during the transition into adolescence. But rather than using only age or time as the primary independent variable, I also used puberty, abstract thinking, and the status change from elementary to secondary school.

The results were clear: pubertal development, not abstract thinking, changing schools, or age, drove the process of intrafamilial transformation. As boys approached the apex of puberty, they began to interrupt their mothers more frequently, and conflict between them increased. I later replicated these results in a sample of both girls and boys, but two serendipitous events changed how I ended up thinking about the findings.

The first was a conversation with Jay, during a visit to Penn StateFootnote ¹, where he was teaching. He and I were discussing the reasons puberty might provoke parent-adolescent conflict. At one point, Jay asked, “How do you know that puberty is causing the family conflict, rather than the reverse?” I said that this seemed like a crazy idea, but because my ongoing study of families was both longitudinal and sufficiently powered, I could look at both causal directions. I ran some analyses when I returned home.

Much to my astonishment, I found that, at least among families with girls, not only did pubertal maturation predict family conflict, but family conflict also accelerated pubertal maturation, whereas parent-adolescent closeness delayed it. Over the years, the fact that the quality of the parent-adolescent bond affects the timing of pubertal maturation in girls has been replicated many times in many different samples.

The second serendipitous event was also the result of a conversation, this time with John. He had called to see if I was able to pinch-hit for him on a panel discussing social relationships in human and nonhuman primates. The featured speaker was the eminent primatologist, Jane Goodall.Footnote ² Knowing nothing about adolescence in other primates, I made an appointment to discuss my study with Chris Coe, who was the director of the Wisconsin Primate Center, where Harry Harlow had done his groundbreaking work on mother-infant attachment. Chris said that my finding was hardly surprising, and that a similar phenomenon had been long observed in some species of monkeys. Housing a juvenile monkey in close proximity to their mother, which maintained their emotional closeness, suppressed pubertal maturation. Once the dyad was separated, pubertal maturation soon commenced.

I wondered about the evolutionary significance of this pattern. Perhaps the squabbling that took place in human early adolescents’ families was an atavism that evolved to drive adolescents out of the family unit in order to reproduce. If conflict between mothers and daughters results in earlier puberty, it allows for more menstrual cycles over the course of daughter’s lifetime and, as a consequence, a larger number of offspring.

Around this time, Jay had been thinking about evolutionary approaches to socialization, which argued that having a large number of offspring is part of a reproductive strategy that values the quantity of one’s offspring over the quality of the child-rearing one can provide, which is important when animals live in high-risk environments in which many infants do not survive. In collaboration with Pat Draper, Jay and I co-authored a theoretical article in which we linked contextual stress with the child’s exposure to harsh parenting, which in turn led to accelerated puberty and higher fecundity. These propositions have been confirmed in many empirical studies.

Policy-Relevant Research

I noted earlier that my relationship with Urie Bronfenbrenner sparked my commitment to doing policy-relevant research. Some developmental scientists study what they study to test theoretically derived hypotheses; I’m not one of them. My work has mainly been driven by an interest in contributing to debates about social policy. And my first foray into this world grew out of a conversation with Ellen Greenberger, who had done the psychosocial maturity research I mentioned earlier. She hired me out of graduate school to join the faculty of the Program in Social Ecology at the University of California, Irvine (UCI).

My ending up at UCI was yet another serendipitous occurrence. In the spring of 1977, I was close to finishing my dissertation and had no job in sight. Ellen was looking for someone to teach courses in developmental psychology to cover for the unexpected departure of a faculty member. Desperate to hire someone quickly, she called the department chairs of the top developmental programs to ask if they had any students who were about to finish and were seeking employment. Luckily for me, one of her calls was to John Hill.

A few months later, I was driving across the country to Southern California to take a job as a one-year replacement. During that year a tenure-track position opened up. I interviewed for it, got the job, and stayed at Irvine for the next six years, where I was granted tenure shortly before being recruited away by the University of Wisconsin.

After I got to UCI, Ellen and I had an auspicious lunch meeting, during which she asked my opinion about her daughter’s interest in taking an after-school job. I thought it was a good idea, but Ellen disagreed. On the way back to the office, we stopped at the library to see if we could find any studies of the impact of school-year employment on adolescent development and well-being. None existed. We wrote a successful grant proposal to the National Institute of Education and launched the Adolescent Work Project.

I once again had the good fortune of being wrong. Contrary to widespread belief that having a job built character, we found that when students worked more than twenty hours a week, working took its toll on academic performance and mental health. These counterintuitive findings received a good deal of press attention and influenced some states to change their child labor laws to restrict the number of hours students were able to work when school was in session.

Our study was the first in a series of policy-relevant studies I worked on over the next forty-five years. It also introduced me to someone whose telephone call about twenty years later would completely change my career. But I’m getting ahead of myself.

Beyond the Classroom

Serendipity struck again after I joined the Wisconsin faculty in 1983. I was doing the study of parent-adolescent relationships mentioned earlier when I received a call from Fred Newmann, an education professor who had just received a large grant to start the National Center on Effective Secondary Schools. The Center had plans to do research on pedagogy and school administration, but they were seeking someone to add a study that looked at influences on adolescent achievement outside the classroom; and the Center would provide the funding. I called my Wisconsin colleague Brad Brown, an expert on adolescent peer relationships, and invited him to collaborate. I’d handle familial influences and after-school jobs, and Brad would handle peers and extracurriculars.

I asked a graduate student to review the literature on familial influences on school achievement, and she came across an unpublished paper by Sandy Dornbusch, a Stanford sociologist who was studying parental influences on school performance. I called Sandy, whom I had never met, and invited him to collaborate with Brad and me. We pooled our funding and designed and launched a study of 10,000 students in nine different high schools. Over the next ten years we would publish many articles on how parents, peers, extracurriculars, and part-time jobs influenced student achievement and engagement, and a book arguing that reforming classroom practices was inadequate if these outside influences weren’t taken into account. Like the earlier study of student employment, this study received considerable attention in the popular press.

The MacArthur Foundation Changes My Life

I left Wisconsin in 1988 to move to Temple University. During the mid 1990s I became involved with two MacArthur Foundation Research Networks, one as a consultant and one as a member. Each of these experiences had a profound formative impact on my career. A paper that my then student Beth Cauffman and I wrote for the foundation’s Network on Mental Health and the Law, on maturity of judgment, drew from and built on Ellen’s model of psychosocial maturity, and ultimately became a blueprint for my later work on adolescents’ criminal culpability. And my membership in the foundation’s Network on Psychopathology and Development stimulated and nurtured my interest in adolescent brain development. But my involvement in those two networks was just the beginning of a long and productive relationship with the foundation.

I had gotten to know many people at MacArthur through the work I did for each of these networks, but this activity also reconnected me with Laurie Garduque, who had directed the Adolescent Work Project at UCI in the late 1970s, and who later became a program officer at the foundation. One day, completely out of the blue, Laurie called and invited me to direct a new network on Adolescent Development and Juvenile Justice. MacArthur had grown concerned that the American justice system had lost sight of the fact that adolescents and adults were different in ways that warranted differential treatment under the law. Laurie wanted the network to be directed by a developmental psychologist interested in adolescence. Our network, which included psychologists, criminologists, legal scholars, and legal practitioners, conducted studies on juveniles’ competence to stand trial, desistance from crime, and criminal culpability.

Doing Policy-Relevant Research

Many scientists who are interested in social policy incorrectly believe that the best way to influence policy is to do good basic science, put it in the hands of policymakers, and let the lawmakers take it from there. One of the most important lessons I learned from the juvenile justice network is that if one wants to do research that impacts social policy, the research questions need to be formulated from the outset to inform answers to specific policy questions. Through my collaboration with the legal practitioners on our network, I came to see that it is far better to first understand the questions policymakers want answered – for instance, at what age are juveniles competent to stand trial – and use one’s scientific training to design research that is both rigorous and relevant. At our network meetings, scientists would often propose interesting studies, grounded in theory and past research, only to be told by the practitioners that no matter what the study found, it would not have an impact on policy because of practical or political considerations that would create insurmountable obstacles to change. So when students often ask what they need to do to prepare to do policy-relevant research, I tell them to get the best possible research training but to spend enough time familiarizing themselves with the policy arena in which they want to work before designing any studies.

In 2003, my network colleague Elizabeth Scott, who taught law at Columbia University, and I published an article titled “Less Guilty By Reason of Adolescence,” in which we explained why deficiencies in adolescents’ judgment and decision-making made them less responsible than adults for their behavior, and therefore less criminally culpable. We argued that this inherent immaturity of judgment should ban the use of the death penalty in cases involving juvenile offenders (at the time, the death penalty was legal in many states if the offender was sixteen or older) because it violated the Constitution’s Eighth Amendment’s prohibition against cruel and unusual punishment. It obviously was cruel, we said, to punish someone so harshly for an act he or she didn’t have complete control over.

When we learned in 2004 that the US Supreme Court (SCOTUS) was going to hear arguments in a case challenging the constitutionality of the juvenile death penalty, we rushed to get our article into the hands of as many of the court’s law clerks as possible. I was asked by the American Psychological Association (APA) to help draft portions of an amicus brief to the court explaining how science shows that adolescents are inherently less responsible than adults. In 2005, by a five to four majority, the Supreme Court, citing our article and the APA brief, abolished the juvenile death penalty. Several subsequent SCOTUS rulings used our work to ban or restrict the use of life without the possibility of parole in cases involving juvenile offenders.

I spent much of my time as director of the MacArthur network giving talks to professional organizations of defense attorneys, judges, and state officials. As the science of adolescent brain development grew and gained popular attention, as evidenced by newspaper stories, magazine covers, and New Yorker cartoons, I began to get questions from legal practitioners about the significance of the neuroscience for the treatment of adolescents under the law. Fortunately, I had developed a preliminary understanding of adolescent brain science through my earlier collaboration with members of the MacArthur Network on Psychopathology and Development, and although I didn’t know much, I knew enough to know what I didn’t know.

I took a sabbatical from Temple in 2007 and spent the year steeped in the burgeoning literature on adolescent brain development. This gave me enough of an understanding to explain to non-neuroscientists how immaturity in adolescent brain development contributed not only to juvenile offending but to risky behavior more generally. Since then, my work has focused on the application of adolescent brain science to matters of legal, social, and public health policy.

Concluding Thoughts

When I’ve taught graduate seminars in developmental psychology, I’ve employed different organizational schemes. But my favorite approach is to structure the syllabus around individual scholars, rather than overall aspects of development. I select a dozen or so well-known researchers and assign five articles each week from one person’s career, beginning with their first publication and sampling chronologically from their bibliography, concluding with one of their most recent publications. Our class discussions focus not only on the person’s work, but on the evolution of their thinking, and how careers and ideas evolve over time.

I’m not sure how I would characterize my own career if I were the subject of such an analysis. So much of it was shaped by serendipity that it’s not easy to see much coherence or continuity. Indeed, the most important turning points in my intellectual development have been triggered by unforeseen occurrences that I had little to do with: rejection from medical school and following Jay to Cornell; having John assigned as my doctoral advisor; being hired by Ellen and asked whether I thought her daughter should take an after-school job; being asked by Jay whether I might have reversed cause and effect in my interpretation of the puberty/family conflict findings; having to pinch-hit at the last minute for John at a symposium with Jane Goodall; being offered unexpected funding to study noninstructional influences on high school achievement; getting calls from the MacArthur Foundation to be involved in, and ultimately direct, a research network; and being asked repeatedly about adolescent brain development when I gave talks to attorneys and judges.

Like the field of adolescence, over time my research, which started with my interest in clinical practice, came to be more influenced by biology, brain development, the study of context, and social policy than I could ever have imagined. And in those fifty years, so has the study of adolescent development.