Genes, culture and personality: An empirical approach

866 The present book gives a general account more detailed analyses which have appeared

BOOKREVIEWS of the principal results of the study. in specialist journals.

References

are given for the numerous RICHARD LYNN

L. J. EAVES, H. J. EYSENCK and N. G. MARTIN: Genes,Cultureand Personality: Press (1989). pp. i-vii, l-465.

An Empirical Approach. New York: Academic

This book is an interim report on a collaborative research programme between Eaves, Eysenck and Martin, with considerable help from numerous other colleagues, that has been in place for over 15 years. In a larger sense it is a continuation and elaboration of the research programme on the genetics of personality launched by Eysenck in the early 1950s. Chapter 1 sets the frame of reference for the larger work, justifying the use of a limited number of measures (virtually all developed by Eysenck) and briefly outlining the model fitting and hypothesis testing approach. Chapter 2 outlines the Eysenckian research paradigm in the domain of personality. Chapter 3 presents a succinct overview of the twin method and an informative “meta-analysis” of early twin studies of personality. The interested reader might compare this analysis to an earlier less sophisticated and more empirically driven analysis carried out by Nichols (1978) and not cited by these authors. The two analyses complement each other nicely. Chapter 4 is an all too brief introduction to model fitting. In chapter 5 the authors carry out detailed analyses of the raw data from three major adult twin studies (London, Sweden, Australia) of Extraversion, Neuroticism and Psychoticism as well as an analysis of the National Merit Scholarship twin study based on a younger sample. The bottom line of these analysis is: (a) shared environment is of little significance in the development of personality; (b) genetic and environmental effects may not be the same for males and females for extraversion and neuroticism; (c) there is likely a significant non-additive component of genetic variance for extraversion. In chapter 6 studies of adoptees and extended families are examined and while there are some interesting twists and turns the findings largely support those drawn from twin studies. Chapter 7 is a gem. Developmental psychology would be a very different discipline if every developmental psychologist could be required to read and understand the first three paragraphs of this chapter. Succinctly put, (a) the genotype is fixed, but gene expression is not; (b) change itself may reflect the influence of genes; (c) a phenotypically similar trait in adolescence and adulthood may reflect different genes, An illustration of the latter point is their conclusion, “that a propensity to anti-social behavior in juveniles is distinct, biologically and developmentally, from propsensity to anti-social adult behavior because different genes are responsible for these aspects of behavior at different times” (p. 181). This conclusion is entirely consistent with non-behavior genetic findings on anti-social behavior but, provides a dramatically different explanatory theory. The fact that the authors only had a very modest data set of limited quality to work with illustrates how resistant developmental psychology has been to genetic data and ideas. The authors correctly point out that large longitudinal twin and adoption studies of personality will be needed in order to answer most of the interesting questions regarding the important processes underlying personality development. Studies of nuclear families alone are simply not sufficiently informative. Chapter 8 involves a genetic analysis of individual EPQ items, an issue I find very difficult to get excited about. Chapter 9 deals with the interaction of persons with situations and while brief is very informative. The idea that personality traits may reflect evolved sensitivities (adaptations) to the environment provides an especially interesting perspective on personality measurement and theory. There is, however, to my knowledge, no evidence that personality traits are adaptive in the sense of inclusive fitness. Chapter 10 treats the multivariate analysis of the covariance between traits and chapter 11 deals with the issue of how normal personality symptoms are related to symptoms of psychiatric disorder. The authors conclude, not surprisingly and consistent with Eysenck theorizing on this question, that psychiatric patients are generally more extreme on normal dimensions of personality. They also find very little evidence for major gene effects for psychiatric symptomatology. Symptoms appear to act like items on a personality questionnaire. I found chapters 12 through 16, dealing with social attitudes by far the most fascinating in the book. We have known for a long time that social attitudes are powerful determinants of social interaction and are significantly implicated in marital choice. As usual Eysenck has been publishing in this domain for longer than most of us can remember. We did not know until 1974 when Eaves and Eysenck published a seldom cited paper in Nature that social attitudes are probably importantly influenced by genetic factors. The probable degree of genetic influence is dramatic. These chapters follow-up on that initial work as well as the important paper by Martin, Eaves, Heath, Jardine, Feingold and Eysenck (1986) in the Proceedings of the National Academy of Sciences (7J.S.A.). It is worth noting that almost all the data in the world literature dealing with this topic are owed to this research team. The one striking exception is the work of Sandra Starr on Authoritarianism and she explicitly admits that it was included in her adoption study as a control variable that failed. That is, it was explicitly included in her adoption study to reflect within family environmental influence with little suspicion that it might reflect genetic influence. Not surprisingly Starr and Weinberg encountered difficulty getting their paper published in the peer reviewed scientific literature (Weinberg, personal communication) and it appears in Starr’s book Race, Social Class, and Individual Differences in IQ (1981). I am willing to wager that in the not too distant future this body of work will be considered seminal and will lie directly at the interface between social psychology, sociobiology and behavior genetics. The key issue, discussed at great length in this section of the book is how to incorporate assortative mating into genetic models. Social attitudes, unlike personality traits, demonstrate considerable common family environmental influence. They also, demonstrate extremely high assortative mating coefficients (i.e. in the range of 0.5&0.65). Depending on how mate selection is incorporated in a model considerable variance will be allocated to common family environmental or to genetic influence. The authors painstakingly demonstrate that the process of mate selection for many of these attitudinal variables (religiousness, ethnocentrism, authoritarianism) is far more complex than is apparent on the surface and they wisely do not commit themselves. I find that I am in disagreement with a number of my respected colleagues on my evaluation of this book. I judge it to be a spendid book. My colleagues and I differ primarily in our judgement of how much tolerance we have for data overkill. From the point of view of efficiency of presentation the book could have been trimmed down considerably. From the point of view of introducing the reader to exploratory data analysis (i.e. showing how much fun it is to play with data and how the process of discovery drives a research program) the “excess analysis” is absolutely necessary. It is full of superb examples of how to use model fitting to test hypotheses in behavior genetics. Graduate students and professionals in

867

BOOK REVIEWS

personality research not familiar with the behavior genetic approach to individual differences will quickly come to realize that a model of some sort underlies every attempt at data analysis, whether the investigator has thought it through in detail or not. They will also learn that no data set is adequate to answer all questions, but that a sufficiently rigorous analysis (defined as having a set of ideas regarding the causes of variation followed by the application of the most appropriate model given the ‘theory’ and the limitations of the data) will quickly raise a whole series of new questions. At its best scientific research is programmatic and consists of a process of continuously questioning both current and previous findings by both oneself and others. A successful research program also raises more questions than it answers. This book is a superb example of a successful research program and it will serve as a fertile source of ideas for personality researchers for years to come. THOMAS B~UCHARD

REFERENCES

Eaves, L. J. & Eysenck, H. J. (1974). Genetics and the development of social attitudes. Nature, 249, 288-289. Martin, N. G., Eaves, L. J., Heath, A. C., Jardine, R., Feingold, L. M. & Eysenck, H. J. (1986). Transmission of social attitudes. Proceedings of the National Academy of Science U.S.A., 83, 43644369. Nichols, R. C. (1978). Twin studies of ability, personality, and interests. Homo, 29, 158-173. Starr, S. & Weinberg, R. (198 1). The transmission of Authoritarianism in families: Genetic resemblance in social attitudes. In Starr, S. (Ed.), Race, social class, and individual differences in IQ. Hillsdale, NJ: Erlbaum.

H. CRONIN: The Ant & the Peacock: Altruism & Sexual Selection from Darwin to Today. Cambridge pp. 490 + xv, f17.50.

University

Press (1991).

Today it is commonplace to emphasize that science is not an isolated undertaking that can be divorced from its personal, social or historical setting. Unfortunately, this is usually an excuse for authors to exaggerate the factor that they think has not been sufficiently credited in the past, usually to the detriment of the others. What makes Helena Cronin’s book, The Ant & the Peacock, outstanding is that it takes account of most of the non-directly scientific factors and yet produces a result that is a superb balance of the historical and scientific aspects of Darwin. This is a book that is both an enlightening historical account of the central discoveries of Darwin and the response to them, as well as an up-to-date summary of our best modern insights into them. Furthermore, the two aspects-the historical and the scientific--cannot be divorced, because as Cronin shows Darwin’s two central discoveries-natural and sexual selection-were routinely misinterpreted, misunderstood and misapplied for almost exactly a hundred years after Darwin first put them forward in 1859 and 1872, respectively. As she points out, “the long period when Darwinism was most widely rejected . stretched from soon after Darwin’s death in the 1880s right up until about the 1940s” (p. 37). It is a striking contradiction of the naive ever-on-and-upward picture of scientific progress so often projected by scientists themselves that an authoritative text-book of 1907 could conclude that Darwinism does “not satisfy present-day biologists”-a claim which Cronin corroborates as “undoubtedly true taking all biologists world-wide” (p. 49). She continues, “Darwinism, . . typically, was condemned for being unscientific” and “was stigmatised as speculative, untestable, inexact and-worst of all, for this put it entirely beyond the pale of science-teleological” (pp. 49-50). Two issues in particular contributed to Darwin’s difficulties, and these are the principal themes of the two parts of the book, vividly portrayed in the title, with the ant corresponding to the riddle of altruism, and the peacock to the issue of sexual selection. Altruism appeared to contradict natural selection because of the way that term was interpreted, both by Specerian Social Darwinists, and even by the co-discoverer of evolution by natural selection, Alfred Wallace. Here the problem was one of seeing natural selection as concerned with the individual organism, and particularly with its survival and success-or ‘fitness’. This made selfishness self-explanatory, but self-sacrifice a seeming anomaly because it promoted the ‘fitness’ of others. Almost exactly a century after the publication of The Origin of Species W. D. Hamilton was to make a discovery that finally solved the riddle of the altruism of the social insects. He showed that such selfless sacrifice of individuals to the apparent good of the social whole was not an example of natural selection at the level of the group or species as many had concluded after Darwin, but of the fact that selection selects ultimately at the level of the individual gene. He showed that a gene for altruistic self-sacrifice in one individual could be selected if its effect on the ultimate reproductive success of identical copies of itself in others was greater than the cost to itself of such self-sacrifice in the first place (Hamilton, 1964). Cronin gives an excellent account of how genetic advances made this ultimate triumph of Darwin’s individualistic concept of selection-what today we would call ‘the selfish gene’ viewinevitable. Sexual selection, writes John Maynard Smith in the Foreword to the book, “was largely ignored in discussions of evolution for a hundred years” (p. ix). Helena Cronin gives by far the best and most detailed account to date of why. She shows that, for the best part of that century, sexual selection was wrongly seen as “flying in the face of natural selection” (pp. 114-l 17) and that here Darwin’s most formidable adversary was none other than Wallace-more ‘Darwinian than Darwin’ in the view of many contemporaries of both. Indeed, even today one often finds natural and sexual selection treated as if they were different principles-often with Hamilton’s ‘kin selection’ added as if it were a third. In reality, Hamilton’s solution to the problem of insect altruism is based on the same insight that explains why natural and sexual selection are not fundamentally different principles. This is the surprisingly modern realization that ultimately natural selection is nothing other than the differential reproductive success of individual genes, with what Darwin called ‘sexual selection’ often being critical for sexually-reproducing species. Seen as the biodegradable packing for the individual genes they contain, peacocks with their elaborate tails are just as explicable as sterile insect workers if we consider the possibility that, like any naturally-evolved trait, such tails and such sterility promote the ultimate reproductive success of the potentially immortal genes which produce them.