Volume 04 - Number 1 - June 1997
Pages: 1 - 35

FEATURED SCHOLAR INTERVIEW

Barry, H. III. (1997). The historian as Darwinian scientist: Frank J. Sulloway [Interview]. Clio’s Psyche, 4(1), 1, 22-31.

https://doi.org/10.70763/3e98410c45ea98addec555019bbae8eb

The Historian as Darwinian Scientist: Frank J. Sulloway

interview

Abstract

This featured scholar interview, conducted by Herbert Barry III at MIT on April 5, 1997, presents historian of science Frank J. Sulloway on birth order, family dynamics, and the method of psychohistory. Sulloway recounts how his study of Darwin’s acceptance of evolution led him to birth-order research and discusses Born to Rebel and Freud, Biologist of the Mind. He describes a Darwinian model in which siblings adopt strategies to secure parental investment, with birth order a proxy for family niche. He reports multivariate analyses of large historical samples, including scientists, participants in the French Revolution and Reformation, and independent ratings of parent-offspring conflict. He applies the model to Darwin, Luther, Napoleon, and the Reign of Terror, and predicts that Heaven’s Gate members were somewhat more likely to be laterborns. Sulloway contends that historians rarely test hypotheses and calls for scientific history built on falsifiable claims, describing his forthcoming edited volume Testing Theories of History. He is sharply critical of psychobiography that applies a single clinical theory to one life, faults Erik Erikson’s Young Man Luther on evidentiary grounds, and disputes psychoanalytic emphasis on the first years of life. He credits John Bowlby’s attachment research and argues that psychohistory must be consistent with Darwinian theory. He also discusses his training, his mentors Edward O. Wilson and Jerry Kagan, and his MacArthur Fellowship. Barry’s closing note relates Sulloway’s work to Irving D. Harris’s The Promised Seed and to Bossard and Boll’s The Large Family System.

Keywords:

American Association for the Advancement of Science, birth order, Born to Rebel: Birth Order, Family Dynamics, and Creative Lives, Charles Darwin, Frank J. Sulloway, Freud, Biologist of the Mind: Beyond the Psychoanalytic Legend, interview, MacArthur Fellowship, psychohistory, Sigmund Freud, Testing Theories of History

Frank J. Sulloway was born in 1947 in Concord, New Hampshire, and received his PhD in the history of science from Harvard University in 1978. Currently, he is affiliated with MIT as a researcher. Dr. Sulloway has had a remarkable number of research positions and lectureships at leading institutions in America and Europe, although he has never held a professorship. In 1970 he produced, with Mark B. Adams, a series of six films and film guides on various aspects of “Charles Darwin’s Voyage with H.M.S. Beagle” and subsequently published numerous articles on Darwin. Sulloway’s two published books are

Page 1

Freud, Biologist of the Mind: Beyond the Psychoanalytic Legend (New York: Basic Books, 1979) and Born to Rebel: Birth Order, Family Dynamics, and Creative Lives (New York: Pantheon Books, 1996). He is currently preparing, as editor and major contributor, Testing Theories of History to be published by Harvard University Press. Among his numerous honors are the MacArthur Foundation Fellowship Award of over $400,000 (1984-1989), popularly called the “genius award”; his election as a Fellow of the American Association for the Advancement of Science (1989) “for research in the history of biology and the major studies on Darwin and Freud that resulted from it”; the Golden Plate Award; and many fellowships and grants. Herbert Barry (“HB”) interviewed Frank Sulloway (“FJS”) for four hours in his MIT office in Cambridge, Massachusetts, April 5, 1997.

HB: What brought you to the study of birth order?

FJS: I stumbled onto the topic by accident 27 years ago. I was working on the problem of why Charles Darwin became an evolutionist. I had retraced the voyage of the Beagle and was well aware of the crucial data, from the Galápagos Islands, that made Darwin an evolutionist. Soon after his return to England in late 1836, Darwin allowed these Galápagos species to be publicly displayed. As a consequence dozens of top naturalists saw Darwin’s finches, mockingbirds, and tortoises. They became aware that the different islands had different species. Why didn’t they all become evolutionists? I concluded that it was not the evidence that made Darwin an evolutionist, rather it was his willingness to interpret that evidence in a radical light.

I thought this intellectual decision must have something to do with Darwin’s personality. So in 1970 I began studying personality psychology and eventually stumbled onto the research on birth order. At first I was rather skeptical that something that seemed like a biographical footnote could be all that important. But when I compared a sample of early evolutionists with early opponents of evolution, there was a very sizable difference in birth order, shown by a standard statistical test. I obtained a whopping correlation and spent the next 25 years trying to understand why that correlation had emerged.

HB: Which of your works do you view with the most pride?

FJS: I am most proud of Born to Rebel due to the sheer magnitude of the work and the self-sacrifice that went into it. I have existed for two decades as a kind of glorified graduate student without a formal job, living from hand to mouth on competitive grants and fellowships, primarily to write this book. I conducted the research continuously for over two decades, including during the writing of Freud: Biologist of the Mind, on which I spent seven years.

Both books represent my tendency to be extremely thorough in the way I approach research problems and in the way I spend a great deal of time trying to come to the right interpretation. I spent 12 years working on how Darwin came to be an evolutionist before I published anything on that problem. It involved understanding the whole history of what happened to his specimens and tracking down his Galápagos birds in England and measuring each of them. Understanding Freud was a similarly challenging problem in that I felt I had to read most of the scientific and medical literature that he was exposed to in the 1880s and 1890s, as well as the entire corpus of Freud’s works — the published and unpublished psychoanalytic works, including the untranslated works in German. I like labyrinthine kinds of projects.

One of the most satisfying achievements in writing Born to Rebel was to develop a fairly consistent, unitary theory involving almost a dozen factors all combining to shape personality development in a way that is consistent with the goals of childhood understood from a Darwinian point of view. The principal goal of childhood is getting out of childhood alive. What are children trying to do? To maximize parental investment. How do children do that? They develop strategies that implement this goal. All of the variables I ended up identifying as crucial to personality development can be understood in terms of how they lead individuals to adopt the behavioural strategies that they actually do. I was able to use

Page 22

multivariate models to test many of these predictors and thereby explain individual behavior.

HB: How is the reception to Born to Rebel?

FJS: First-born reviewers are three times more likely than laterborns to believe that the theory is wrong (as I have found from surveys). In many cases they have gone out of their way to distort the basic argument. To begin with, the book is not only about birth order. It is about personality development and how family dynamics in various measurable ways influence this process.

My book has definitely polarized its readership. It is not intended, however, to be an evaluative argument about who is a better person. The data tends to show that laterborns are on the side of liberalism — are for freeing the slaves, freedom of speech, freedom of religion, and democracy. Firstborns are generally on the wrong side of those issues. Many first-born readers take this very personally, as if I have something against firstborns, which I don’t. Firstborns excel at many things — they win more Nobel prizes in science, for example — and they are accordingly creative in their own ways, just as laterborns are creative in their own ways.

HB: Do your findings apply only to groups of people or to individuals as well?

FJS: People sometimes say, “Oh, that’s just a statistical generalization, it doesn’t apply to individuals.” Rubbish! If you use multivariate models, you can derive a specific prediction for each individual. Such predictions are particularly meaningful when we are dealing with people such as Darwin who are loaded to the gills with predictors that make them rebels. Darwin’s predicted probability for being an evolutionist, given a dozen or so significant influences, is 94 percent. He’s the fifth of six children; he came from a liberal family; he’s a liberal himself both politically and religiously; he was young at the time he converted to a belief in evolution; he went on a voyage around the world (which is a significant predictor of acceptance of evolution); and he had conflict with his father when he was a child. Virtually everything that makes one a radical is present in his life and therefore makes him more likely than 94 percent of the rest of the population to support evolution. He not only supported evolution but he also developed a novel theory on the subject.

HB: Do you think your book will stimulate hypothesis testing on birth order and family dynamics?

FJS: I would be far more interested in having it stimulate hypothesis testing in general. I am not wedded to any one variable, or even to the role of family dynamics. My book documents the influence in history of many sociological and situational predictors as well.

HB: Does the significance of the variables you considered differ depending upon the situation?

FJS: A tremendously important influence on human behavior is the context of the behavior. That’s why meta-analysis is such a powerful tool for understanding discrepancies in outcomes from one behavioral study to another. The context of the situation dictates the extent to which any particular predictor such as birth order, or social attitudes, or conflict with a parent is relevant to a given behavior as observed in a given event. For example, in England there was hardly a birth-order effect in the reaction to Newton. Everybody respected him and his novel ideas. But in France, where Newton’s theories were controversial, there was a big birth-order effect. That’s an interaction effect. Birth order is a good predictor where the theory is controversial, namely France, but it’s a lousy predictor where the theory isn’t controversial. The greatest challenge in doing historical research is getting the behavior matched to the context. It would not be too hard to do a meta-analysis of 30 major political revolutions — such as the Russian Revolution, the American Revolution, the English Civil War, and the American Civil War — if 30 different scholars would each agree to spend one year working up a sample on one of the 30 events. For a single person to do such a study would take 30 years, so we need more people throwing their hats into the ring in order to achieve a science of human behavior linked with historical context.

HB: What are you working on now?

FJS: I hope in the next year or so to complete an edited volume that I have already done a considerable amount of work on, Testing Theories of History. This volume is an attempt to gather together four or five other scholars in the history of science and in history who have written on the need for testing claims. The general argument is that claims about history are no different from the kind of claims that a

Page 23

paleontologist or evolutionary biologist would make. Evolution is history.

Scientists feel a tremendous need to test their theories but historians feel almost no need to do so. I want to clarify under what circumstances scholars ought to test their claims in history. Some historical claims are not testable, including most biographical claims. But many historical claims, such as generalizations about people, are statistical claims, and thus are testable. The reason that historians rarely test their theories is that they are not trained to test, and they are also not trained to recognize when a test is necessary.

HB: What about your own family, your birth order?

FJS: I’m the third of four boys. My younger brother is nine years younger, a half brother from my father’s second marriage. I grew up as a functional lastborn since I had no younger sibling for nine years. By the way, the important influence in this regard is not biological birth order but functional birth order.

When I was a very young child, in my first year, my mother was hospitalized for about six months. When I was about two and a half she was away for about nine months. One could plausibly argue, based on attachment theory, that having had disrupted parenting probably had a considerable effect on my personality. I have been a fiercely independent person, as witness the fact that I have never had a job. The whole notion of working for somebody else — having somebody literally control my life — is anathema to me. I think that’s related to the fact that my mother was away when I was a child, that I learned a certain degree of independence through separation. I’m also probably more introverted and shy than I would have been had my mother been around on a constant basis.

HB: What was your undergraduate major?

FJS: I was as hybrid an undergraduate, as I am now. I majored in history and science. This field was not the history of science. It was a split major. My science was biology, and my history was American history.

HB: Did you get good statistics training?

FJS: I never took a statistics course either in college or in graduate school. I trained myself to do statistics as an undergraduate by taking a computer course. If you have a database and are interested in doing hypothesis testing, it’s really quite simple to learn the basic rules of statistics.

HB: What is your primary scholarly affiliation today?

FJS: I don’t consider that I have any one topical affiliation. I was trained as a historian of science and have a doctorate in this field. I’ve spent ten years as a postdoc in psychology departments in this country and abroad — two years at Berkeley, six years at Harvard, and two years at University College in London — so I have a broad background in psychology. The only formal teaching I have done was for the Psychology Department at Harvard in a course on Darwin, Freud, and sociobiology. I consider myself both a psychologist and a historian of science.

HB: Have you had any psychoanalytic or psychotherapeutic training or experience?

FJS: I haven’t had any psychoanalytic training at all.

HB: What special training was most helpful in your becoming a scholar capable of using psychosocial research tools with such effectiveness?

FJS: I was very influenced as an undergraduate by having Edward O. Wilson as a teacher. I took his course on evolutionary biology as a junior at Harvard.

HB: In addition to Edward O. Wilson, do any other mentors come to mind?

FJS: Jerry Kagan, a child psychologist of considerable renown, also took me under his wing. The chapter on shyness in Born to Rebel was due largely to Kagan’s own work on that topic. I was interested in showing that the strategies siblings develop based on birth order would be altered by some independent attribute of their psyches that was known to be partly under biological control. Shyness is moderately heritable and has a dozen physiological markers. Although shyness has no direct effect on support of revolutionary ideas, it serves as a powerful moderating influence by entering into interactions with birth order. Shy people are fence sitters so that one finds almost no birth-order effect among them. They try to keep their mouths shut and say the least controversial thing. Shyness makes the lastborn less revolutionary and makes the firstborn more revolutionary (by making them more cautious about expressing a conservative stance). I feel that

Page 24

most of what we want to know about human development is going to involve “interactionist” stories like that. Most interactions are missed without the aid of statistical methods. We tend to see mostly main effects. The only way we will know the full story about human development is to do very intensive, empirical investigations.

HB: Did Erik Erikson, Robert J. Lifton, and the Wellfleet Group have any impact on you?

FJS: No. However, I did have some contact with Erik Erikson when he was writing on Einstein. He wasn’t particularly interested in the work I was doing on birth order, but I was impressed by him as a very warm and interesting person.

HB: What books were important to your development?

FJS: I was certainly fascinated by Darwin’s Autobiography (1958) and Ernest Jones’ biography of Freud, The Life and Work of Sigmund Freud (1953-1957). I’ve always had an interest in with making sense out of people’s lives. Reading the Jones biography and the Fliess letters got me started on my Freud book. Reading Darwin’s autobiography and puzzling about Darwin’s life got me started on Born to Rebel. Why was Darwin such a radical revolutionary? Overall, I was influenced by reading Darwin, far more so than by reading Freud. I was a little disappointed — having invested seven years of my life, and having read the 24 volumes of Freud’s collected works and all his untranslated works in German — to discover how flawed the theory was in terms of one’s ability to apply it in an objective manner. I haven’t used psychoanalysis much in my own work as a historian.

We often think of Darwinism as focusing on innate propensities that are based on genes. But, parallel to the growth of the literature in evolutionary psychology, I came to realize that one could propose a highly environmental form of Darwinian theory. The strategies of siblings are adaptive and are achieved largely through learning rather than through innate propensities. Darwinian selection has allowed us to have large brains so that we can accomplish most adaptations through learning. It was quite exciting, during the last three or four years of writing Born to Rebel, when I began to understand how useful Darwinian theory could be for the understanding of personality development.

There is a fascinating interface going on in research today in the behavioral sciences between personality psychology, behavioral genetics, and evolutionary psychology. From behavioral genetics we have learned that siblings are enormously different and that a great deal of personality is heritable. The heritability for most major dimensions of personality is about 40 percent. Some aspects of the psyche of an individual such as Darwin cannot be explained on the basis of traditional environmentalist accounts. There are aspects of Darwin’s personality — his agreeableness and his sensitivity, for example — that are partly just genetic givens. It’s important for people who are trying to explain behavior not to go zealously overboard in an environmentalist direction. Only about half of the individual differences in personality between siblings is due to environmental influences.

Evolutionary psychology suggests that siblings are very different because they are devising different strategies — their Darwinian interests are different and, therefore, their strategies are different too. These individual strategies emerge as different personality characteristics. Personality development is the whole bailiwick of strategies designed to promote parental favor and to deal effectively with sibling rivals in this competitive situation.

HB: How do you regard the preparation for, and practice of, history today?

FJS: I think there are major problems with how history is done nowadays. It is driven by anecdote and is largely oblivious to the need for hypothesis testing. Although we now live in an age of computers, where accessing and manipulating large amounts of data and doing hypothesis testing are increasingly possible, it is extremely rare for me to ever hear a historian give a talk where a hypothesis is even articulated, much less tested. I’ve heard many debates among historians where it is clear that hypothesis testing would have ended that debate. They just talk back and forth at each other as if a few anecdotes here and there could prove their point. They really do believe that a hundred anecdotes are better than one anecdote, whereas a hundred anecdotes are no better than one to a knowledgeable exponent of hypothesis testing. Anecdotes only suggest hypotheses.

I think there is a major need for a real

Page 25

reformation in how history is taught. This reformation would require having a course early in graduate training, which teaches young historians when they ought to test claims. Formal statistical testing and hypothesis testing are fairly simple to teach. If you send me a graduate student in history, I can teach this person in one week or even a few afternoons everything that he or she will need to know statistically to do his or her project. I can set up a database for them, send them out to collect data, and teach them how to use experts in their field as independent judges or raters. The major problem that historians have is that they are unaware of when they should be testing. Maybe 30-40 percent of all historical problems would lend themselves to testing, but only one tenth of one per cent actually get tested.

Some background in evolutionary psychology is also very important. Darwinism is a general theory that is very useful for understanding animal behavior. Marxism is now out, and psychoanalysis has long been problematical. People who propose a theory of historical behavior should at least be aware of whether it is plausible and consistent in a Darwinian world. If it isn’t, it is probably wrong. A model in this regard is John Bowlby, who came out of a background in psychoanalysis and clawed his way into the world of real science. He recognized the need for hypothesis testing and really boned up on the scientific literature in ethology and evolutionary biology. He began to understand the attachment bond from a Darwinian point of view. The attachment relationship is an adaptive relationship for all young organisms in a perilous world. It keeps them from being scooped up by predators. It really was wrong, as Bowlby emphasized, for psychoanalysts to treat attachment as a form of pathological dependency. Attachment enhances survival, and it goes through its own phases. Disruption of the attachment relationship by an abusive parent, or by parental loss, can be psychologically consequential for the child. A lot of advice based on psychoanalytic theory, being handed out in the 1940s and 1950s, said, “Don’t coddle your children.” This advice was feeding the very problem that it was supposed to cure.

HB: There are some historians who are interested in doing statistical studies and hypothesis testing.

FJS: There are indeed areas in history where hypothesis testing is going on, such as demographic and economic history. One of the problems is that this work gets viewed as “quantitative history.” I don’t like to cast it in this way. What I do is to try to test hypotheses. Some of what I do is quantitative, some isn’t. We don’t say somebody does quantitative biology or quantitative physics. Nobody in the sciences who does hypothesis testing would make a big deal out of the quantitative bit. All scientists want to know is, can a claim be tested? How one actually tests the claim is of far less importance. What we really need is not quantitative historians but scientific historians.

It may sound crazy to have a field of scientific historians, but what’s the alternative? Unscientific historians? Are we going to have people who are oblivious to the normal methods that have characterized the greatest progress in knowledge in the last 500 years? That’s science. What exactly is the scientific method? It’s trying to falsify your claim. Unscientific history occurs when scholars do not try to falsify their claims. Scientific history is just good history. If you’re not doing scientific history, you’re doing bad history.

HB: What was the impact of your getting a MacArthur Fellowship?

FJS: It was very crucial for me. I am the only MacArthur Fellow that I know who used every penny of the fellowship to support my research. I did exactly what they hoped recipients would do. I used it to facilitate full-time research for five years.

HB: Do you plan to make your data files accessible to colleagues?

FJS: My research encompasses databases for the French Revolution; the Reformation; 3,890 scientists; and a few smaller databases, such as American Presidents and U.S. Supreme Court voting. I would have no qualms about sharing some of these databases with scholars who are actively working on related projects. I especially might do so on a reciprocity basis. For example, French historians are in a position to fill out my biographical database on the French Revolution by going through archival data, or parish and genealogical records.

HB: What is the importance of childhood to psychosocial and psychohistorical studies?

FJS: Enormous. Individuals learn a substantial amount of their social attitudes by

Page 26

direct transmission of ideas from their parents. The correlation between social attitudes of offspring and parents is quite high. It’s around .5. However, the same is not true for personality. Individuals learn personality traits as part of the strategy for dealing with family dynamics. The combined influence of social attitudes and personality enable us to explain some aspects of adult behavior, such as voting in favor of a reform movement, or voting Democratic or Republican. It’s a multivariate problem.

HB: Can you identify a range of ages in childhood that is the most influential on subsequent adult behavior?

FJS: I think it’s almost intuitively obvious that the early years are more important than the later years because more of personality is formed during this period. But I would say that psychoanalytic theory has grossly overemphasized the importance of very early childhood, the first year or two when memories are supposed to become “unconscious.” The first ten years are considerably more important than the second ten, and the first five are more important than the second five. But I don’t think the first six months or year are the most important, though it depends entirely on what attributes we are talking about. Attachment behavior, which relates to the capacity for intimacy, probably has an earlier determination — the first year or two seems to be critical. Personality traits, which are a reflection of sibling strategies and family dynamics, are probably formed more during the second through fifth years.

HB: What is the importance of family dynamics and birth order to historians?

FJS: Their implications for historians are considerable because historians have tended to underestimate the importance of within-family diversity. They have assumed, without any evidence, that between-family differences, such as class, religious denomination, or geographic origins, are the influences we should be looking at. But all these influences, when we test them, have relatively little explanatory power. Historians would do well to be more aware of how remarkably diverse siblings are who grow up in the same family and to incorporate this fact into their toolkit of explanatory causes. Birth order is one of the primary mechanisms by which this remarkable diversity arises among siblings. But birth order is not a cause in itself. It is a proxy for strategies that siblings adopt in competing for parental favor, such as the employment of individual differences in physical strength, temperament, and gender.

Most of what’s important in development is not a single event or a trauma. Personality development is a long-term emergent property of interactions with other family members. Attributing aspects of personality to, say, the loss of a parent is generally a mistake. We really need to understand how niches and dynamics in the family change as a result of such events. I’m not saying that the loss of a parent has no influence on a person, but that it’s very easy to grossly overemphasize the influence of this kind of event. If the early loss of a parent can explain huge amounts of subsequent behavior, I would like to see the statistical tests showing it. None of this speculative stuff. A typical psychobiography assertion is “Freud has said that…” and then the author just goes on to weave some hypothetical scenario. That is an irresponsible kind of scholarship and we can do much better. I know, because I’ve tested the influence of parental loss for thousands of historical figures.

HB: How do you define psychohistory?

FJS: To me psychohistory means the application of psychological theories aimed at explaining human behavior in a specific historical context. The psychologist studies people who are alive and can answer questionnaires and be tested in experimental situations. The psychohistorian happens to study people who aren’t alive. The theoretical perspective, and the methods of hypothesis testing, ought to be the same.

HB: Paul Elovitz and some other psychohistorians have been doing studies of recent U.S. Presidents and candidates for President who are alive, interviewing them or people who knew them or grew up with them. The information is of historical importance. In Born to Rebel you make one reference to psychohistory, one which is not friendly. What is your critique of the field?

FJS: I’m not hostile at all to psychohistory as a general discipline. I am hostile to traditions within psychohistory that have taken some theory which they, like some true believer in some religious cult, believe is absolutely true. Such practitioners go back in history and apply this cherished theory to one individual, selecting data willy nilly to fit the theory, and then write some speculative psychobiography of this individual.

Page 27

These psychobiographies don’t come out of academic psychology, where most psychological theories are on a better footing. Instead, they come out of the weakest areas of psychology, which are those based on clinical evidence and involve theories that are most under constant revision because they were never erected on an adequate scientific foundation. I think it’s irresponsible to go back and try to analyze the key to somebody’s personality based on what you think their Oedipus complex was.

A totally different enterprise is doing, judged in terms of tantalizing bits of circumstantial evidence, what I tried to do in Born to Rebel. One begins with a plausible hypothesis, such as one very close to the heart of psychoanalysis — that parent-offspring conflict is relevant to behavior. That’s testable using well-known historical figures. I managed to get independent raters to rate levels of parent-offspring conflict for about a thousand people, which is more than ten times larger than the average study in psychology. Indeed, it turns out that people who have conflict with a parent are more likely to support revolutions. Interestingly, conflict with a parent is far more important for firstborns than it is for laterborns. Firstborns generally have a niche in which they are parent-identified, incurring parental favor by serving as a surrogate parent and by identifying strongly with parents. So if a firstborn has conflict with a parent, that conflict breaks the mechanism. The finding involves an interaction effect. And that tells us that if we are going to apply something remotely like Freud’s notion of Oedipal conflict, its application will be significantly different depending upon the subject’s niche in his or her family. This is the kind of sophisticated theorizing and hypothesis testing that I think we really want in “psychohistory.”

HB: What do you think of Erikson’s psychobiography of Martin Luther and Gandhi?

FJS: I’m not expert on Gandhi; but having read everything on Luther’s childhood, having read Reformation scholars on the problem, and having spent about two years researching my Reformation chapter in Born to Rebel, I know perfectly well that no Reformation scholar takes Erikson’s Young Man Luther (1958) very seriously. The evidence suggests that Luther’s parents were no more harsh than any other parents at this time, and Luther was by no means traumatized by them. I wouldn’t cite Luther as a case of high parental conflict to a level that I know, in my history of science sample, would have made someone a revolutionary. I think Erikson was dead wrong about that.

I have developed a multivariate model of support for the Reformation. Without any information about parental conflict, the model predicts that Luther, a firstborn, will support the Reformation. He comes from Germany, which has the highest support rate of any nation. He comes from the lower clergy, which has the highest support rate of any social group. His parents were lower middle class, which in this rare instance is a predictor of support for change. Given all of these predictors, Luther’s probability of support is 67 percent, just over the midline. All Reformation scholars agree that Luther was one of the more moderate reformers. Had Luther experienced the kind of severe parent-offspring conflict that Erikson claims, Luther ought to have been one of the most radical reformers.

Erikson could get away with the argument of his book because he himself decides what the reader gets to know about Luther’s childhood. How far could Erikson have gotten if he had been limited by the methodological precaution of asking ten or twelve Reformation scholars, as independent raters, where Luther’s conflict with his parents really was on a scale of one to ten. They wouldn’t have agreed with him at all. That’s why he didn’t ask them. If Erickson had wanted to be really honest about it, he should have pointed out that everybody else disagreed with his evidence as well as his conclusions from that evidence. That’s the kind of integrity we need to introduce if psychohistory is going to have more viability as a respected field of inquiry.

HB: We’d like a list of five people in order who you think have made the greatest contribution to psychohistory.

FJS: Counterintuitively, I would list first the book by David E. Stannard, Shrinking History: On Freud and the Failure of Psychohistory (1980), for its antidote value. Anybody reading this book should realize that we should not be practicing psychohistory the way it’s practiced. My second choice is Thomas S. Kuhn’s The Structure of Scientific Revolutions (1962), with the strong caveat that Kuhn in no way attempted to test those claims, although I do think many of the claims he made are in fact testable. Dean Keith Simonton is my third choice. He has tested hundreds of

Page 28

hypotheses about leadership, creativity, and greatness in history. He’s done so in a series of books including Greatness: Who Makes History and Why (1994) and Genius, Creativity, and Leadership: Historiometric Inquiries (1984).

HB: How about the book by Irving D. Harris, The Promised Seed: A Comparative Study of Eminent First and Later Sons (1964)?

FJS: Yes, there’s a book I’ve certainly read. Harris was attempting very informal hypothesis testing. His methods weren’t really rigorous and he used very small datasets. However, one has to give the guy high marks for a book written in 1964, given the level of his competitors. In methodological rigor, Harris is way above the books by Erikson and Freud — I wouldn’t put any of their books on the list. I think the list merits only three or four names at most.

HB: What are your thoughts about the reactions of some people to the coming of the third millennium?

FJS: It’s possible that family dynamics and sibling differences are related to recruitment into millenarian cults that believe the earth is coming to an end. I would predict that the recent Heaven’s Gate cult would recruit people who are religious in general, so that this population is probably slightly biased for firstborns. But it would also tend to recruit out of this population a slightly larger proportion of laterborns who are willing to entertain an oddball set of beliefs that strain credulity. So, my overall prediction is that Heaven’s Gate members were somewhat more likely to be laterborns.

HB: The cult members subordinated themselves to a charismatic leader.

FJS: It depends on the extent to which the cult emphasizes an unconventional belief system, which is a laterborn predilection, versus emphasis on obedience to the authority of a charismatic figure, which is a firstborn trait. It’s a complex question to figure out what is the correct hypothesis in any specific instance.

HB: What are your thoughts on the dynamics of violence in our world? Your book reports that middleborn children are less violent.

FJS: In political and social revolutions firstborns and lastborns tend to be more militant than middleborn children. Firstborns are militant because they are bigger and stronger than their siblings and can effectively use aggression. Lastborns are militant because they tend to be more zealous, more open to experience, and more radically devoted to change. Middle children don’t have the advantage of physical size of the firstborn and they aren’t protected by parents as the lastborn is. They tend therefore to be more diplomatic.

HB: How do you understand the psychology of terrorism?

FJS: This issue is directly related to family niches and sibling strategies. I devoted an entire chapter in Born to Rebel to the psychology of the Reign of Terror during the French Revolution. I tested all the standard hypotheses of French historians on this subject, that the terrorists tended to be urban or young or lower class or predisposed by previous political experience. The two best predictors of support for the Terror were differences in birth order and a tough-minded temperament, with temperament also being predicted by birth order. Younger children, particularly middleborns, are the least tough minded and tend to eschew violence. These sibling differences in attitudes toward the Terror are twice as large as all the differences that French historians have claimed based on social factors like urbanization and class.

HB: Another possible factor is the degree to which the parents are punitive. Lloyd deMause suggests that in previous centuries parents were very punitive and sexually abusive to their children. The violent revolutionaries might have been retaliating against their experiences with brutal parents.

FJS:. It’s a plausible theory. Whether it’s true or not is a matter of doing the empirical testing. I don’t think that we’d find much difference in how loving parents were from the 14th century to the present. Rather, I think that there are enormous variations in how loving or abusive individual parents are in the same population. The between-family differences would swamp the between-century differences. I say this because I have tested the claim for more than a thousand people in my study.

HB: What needs to be done in historical, family dynamics, and birth-order studies in the next decade?

FJS: The single most important class of studies that needs to be done involves all the siblings and parents in the same family. We need

Page 29

to understand more about how, given the niche that the first child actually occupies, what niche the second child then chooses to occupy. That’s different from asking the question, “On average, what is the firstborn’s niche (and associated personality profile)?” This would be asking, “What did the firstborn child actually do, and what did the second child then do in response?” Take the case of Napoleon Bonaparte. He was the second of eight children, with a rather shy elder brother. Napoleon, who was extremely pig-headed and extroverted, was able to usurp the dominant niche in his family. By his late teens he was running the family, which is unusual. If a firstborn does not occupy the typical firstborn niche — that of the dominant alpha male figure — this niche is open for another sibling to occupy.

In the field of history, I’d like to think that historians will cease to worry about whether their discipline is a humanities or a science, which I think is a stupid issue, and will just get on with asking the more interesting questions. The historian’s goal almost always ought to be to have a rank-ordered list of candidates for explanatory power. To do that, one often has to conduct a multivariate analysis. One tests each hypothesis and then rank-orders the effect size of each predictive variable. Then one can say, for example, that the most important predictor of attitudes toward scientific innovation is birth order, the next most important one is social attitudes, then age, personal friendship with the leader, and nationality differences. Some nations are significantly more or less positive toward innovations. This often happens as a result of chauvinism. It was a lot easier for the British to be in favor of Isaac Newton, who was one of them, than it was for the French, because the theory that Newton was attacking was Descartes’ theory which was a French theory.

HB: What else do psychohistorians need to do to gain academic acceptance?

FJS: Let’s take a hypothesis, not a global theory, and let’s go test it in history, with a sufficient number of cases. Our unit of analysis will no longer be individual biographies but rather whole populations. There’s a million problems that can be tackled in this manner. I have thought of going back, for example, and doing a much more elaborate analysis of the antislavery or Abolition movement. We could work up a database of maybe a thousand people, including a control group of famous opponents of Abolition — many black biographies are now in microfilm at libraries like Harvard’s Widener. I’d also love to see a detailed analysis of who went north and who went south in the American Civil War, particularly in the border states where families often split. There are many anecdotal comments about this topic, but nobody has done a really good study of it.

HB: Do you foresee a Darwinian psychohistory?

FJS: Darwinian theory is the one and only general-purpose theory that is capable of explaining the broad features of animal behavior, including that of human beings. This assertion does not mean that psychobiography will end up being nothing but Darwinian. Many lower-level theories need to be included among the basic tools of psychohistory. Still, I am confident that none of these lower-level theories will ever significantly violate Darwinian assumptions. To the extent that the fields of psychobiography and psychohistory wish to become part of modern science, requiring they employ a theory from the behavioral sciences, they must be Darwinian. Born to Rebel offers just such an eclectic, Darwinian approach to the problem of why some people, but not others, endorse radical changes in life.

Interviewer’s Note: Frank Sulloway’s Born to Rebel is an important contribution to our understanding of rebellion and conformity. Human beings make recurrent choices between these two major adaptive strategies. Birth involves separation from the mother, followed by autonomous development and subsequent rebellions against parents, teachers, employers, and other authority figures. Meanwhile, conformity to social demands is required for successful adjustment in relationships.

Sulloway has revitalized the birth order debate initiated in two books a generation ago. More frequent conformity by firstborns and more frequent rebellion by laterborns is consistent with Irving D. Harris, The Promised Seed; A Comparative Study of Eminent First and Later Sons (1964). The author compared famous first sons with famous men who grew up with one or more older brothers in a variety of occupations, including government, warfare, philosophy, and literature. Stronger connection with the parents and therefore incorporation of their ideals caused the first sons to be more idealistic and theoretical.

Page 30

Competition with one or more older, stronger, and more parentally-affiliated brothers caused the later sons to be more pragmatic and observant.

Sulloway described successive children of the same parents as occupying different adaptive niches in accordance with Darwin’s observation that different species occupy different ecological niches. This conception is consistent with an observation by J. H. S. Bossard and E. E. Boll, The Large Family System (1956). Successive children seek unique roles, such as responsible, popular, socially ambitious, or studious. Later children in large families tend to be forced into pathological roles, such as isolate, irresponsible, sickly, or spoiled, because the more desirable roles are already occupied.

The principal new contributions of Born to Rebel are its systematic study of several thousand individuals and its multivariate perspective.

Page 31

About the author

Herbert Barry III

Herbert Barry III, PhD, educated in psychology at Harvard and Yale, taught at the University of Connecticut prior to serving as a research and psychology professor at the University of Pittsburgh for 38 years. He served as President of the International Psychohistorical Association, is a Research Associate of the Psychohistory Forum, and is a proud member of MENSA. He has published 100s of articles with presidents and their birth order among his specialties. Dr. Barry may be contacted at .

How to cite this article

Barry, H. III. (1997). The historian as Darwinian scientist: Frank J. Sulloway [Interview]. Clio’s Psyche, 4(1), 1, 22-31.

https://doi.org/10.70763/3e98410c45ea98addec555019bbae8eb

PDF downloads