Genetic and Environmental Contributions toIndividual Differences: The Three MajorDimensions of PersonalityH. J.Institute of PsychiatryUniversity of LondonABSTRACT This article deals with the contribution of genetic and envi-ronmental factors to individual differences in the three major dimensions ofpersonahty (Psychoticism, Extraversion, and Neuroticism) Twin studies indi-cate, and family studies confirm within limits, the strong genetic determinationof these and many other personality factors, additive genetic variance account-ing for roughly half the total phenotypic variance On the environmental side,shared family environment plays little or no part, all environmental effectsbemg within-family Assortative mating, important in the formation of socialattitudes, has little impact on personahty Dominance may be important forExtraversion Epistasis (emergenesis) may account for the comparative lowvalues of dizygotic (DZ) twins' correlations Evidence for differential hent-ability of traits IS present, but not very strong It is concluded that behavioralgenetics forms a vital part of the psychological understandmg of the causes ofindividual differences m personalityThis article discusses the importance of genetic factors m personality,and It may be useful to begin with a bnef definition of terms Conceptslike "personality" often mean different things to different people, andwithout wishing to prescnbe the correct use of such terms (if there isany such correct use), it may be useful for the reader to know howsuch terms are used within this article Personality in the wider sense(personalityw) may be divided mto two basic categones, cognitive (m-Address correspondence to H J Eysenck, Institute of Psychiatry, University of Lon-don, London, EnglandJoumal cf Personabty 58 1, March 1990 Copynght © 1990 hy Duke University PressCCC 0022-3506/90/$! 50246 Eysencktelligence, mental abilities) and noncognitive (temperament, person-ality m the narrow sense, or personalityN) Cattell (Cattell, Eber, &Tatsuoka, 1970) clearly uses the term personality in the wider sense,because he includes intelligence, social attitudes, and other conceptsin the field covered by his Sixteen Personality Factor (16 PF) ScaleStrelau (1983) uses the term personality in the narrower sense, but con-trasts It with temperament, which he would define as the biological andgenetic basis for personality In this article I follow the usual Amencancustom of using the term personality m the narrower sense, the geneticanalysis of intelligence and ability has been descnbed elsewhere (H JEysenck, 1979, 1982)Textbooks such as that by Hall and Lindzey (1970) suggest that thereIS no substantive area covered by the term "personality", all that thereader is given is eponymous chapters dealing with vanegated theo-nes with little m common and depending more on persuasion thanon empirical demonstration However, in recent years large-scale fac-tor analytic studies have shown convincing agreement on three majordimensions of personality, based on the intercorrelations between lowerlevel traits (Royce & Powell, 1983) These three factors or dimensionshave been vanously labeled by different authors, but m the system I haveonginated (H J Eysenck, 1947, 1952) they are labeled Neuroticismversus Emotional Stabihty (N), Extraversion as opposed to Introversion(E), and Psychoticism as opposed to Super-Ego Control (P)The Eysenck Personahty Questionnaire (EPQ) (H J Eysenck &5 B G Eysenck, 1975) was designed to measure these three dimen-sions, as well as containing a Lie Scale (L) designed to measure dissimu-lation This article is pnmanly concemed with these major dimensionsof personality for two reasons First, more work has been done on thegenetic analysis of P, E, and N than on any other traits or types of per-sonahty, and second, this system has some claim to constitute a buddingparadigm in personality research (H J Eysenck, 1983b) The systemIS firmly based on a biological foundation, and evidence for it can befound m animal studies (rhesus monkeys, rats) where observation ofsocial behavior, breeding studies, and expenmental interventions havedemonstrated similar behavior pattems to P, E, and N (H J Eysenck6 M W Eysenck, 1985) Such evolutionary development stronglyunderpins genetic argumentsGenetic and Environmental Contributions 247Early StudiesThe empincal and quantitative study of the genetics of personality maybe said to have started with the work of Newman, Freeman, and Hol-zinger (1937), they assigned httle importance to genetic influencesIn earlier work (1967) I cnticized their studies m detail In the firstplace, the measures used as tests of personality would not now be re-garded as either reliable or valid, containing for mstance the DowneyWill/Temperament Test, which is essentially based on graphologicalpnnciplesIn the second place, the personality tests used were primarily testsfor adults, but the average age of the group of twins tested was onlyabout 13 years, and must have included children as young as 8, or evenyounger Thus the tests were quite inappropriate for the population inquestionIn the third place, the statistical results obtained for some of the testsgave ccHitradictory conclusions Thus a Neuroticism inventory usedgave an mtraclass correlation of 562 for identical twins, of 371 forfratemal twins, and of 583 for identical twins brought up in separationNote that identical twins who grow up m separation are very slightlymore alike than identical twins brought up together, this suggests avery weak influence of between-family environmental factors, if anyinfluence can be postulated at all The difference between identical andfratemal twins suggests a hentabihty of about 40%, which would begreatly increased if unreliability m the measurmg instrument were cor-rected for It certainly would not seem reasonable to accept the verynegative conclusion of Newman et al's (1937) book, although manytextbooks of personality have done so, without considenng the senousdefects of the studyI have outlined the histoncal development of studies of this kind(H J Eysenck, 1967), most of which have given positive results as faras Extraversion and Neuroticism, in particular, are concemed This arti-cle will only comment on two or three studies before going on to morerecent work using statistical methods of model fitting better equippedto deal with the complexities of the situation The first set of studies(H J Eysenck, 1956, H J Eysenck & Prell, 1951), using twins,