Make your own free website on

This is one of a series of EXCERPTS from older articles put online by John Ray as a public service. The articles concerned are in general otherwise available only by special request to a University or other major library.


Proceedings of the National Academy of Sciences, USA Vol. 83. pp. 4364-4368, June 1986

Note: As this article uses very sophisticated methods of statistical analysis it might as well be in Greek as far as the layman is concerned. I have therefore highlighted in red the principal findings. The authors found that heredity alone could explain all their findings about transmission of conservatism. For a fuller interpretation of the implications of the article, see here

Transmission of social attitudes


ABSTRACT Data gathered in Australia and England on the social attitudes of spouses and twins are largely consistent with a genetic model for family resemblance in social attitudes. There is substantial assortative mating and little evidence of vertical cultural inheritance.

The facility with which humans learn and their great investment in mate selection, parental care, and education make the human species a model system for the study of cultural inheritance. Until comparatively recently, however, genetic models for family resemblance such as those devised by Fisher (1) were superior to cultural models because the former were quantitative and led naturally to statistical estimation and hypothesis testing. The emphasis of theoretical analysis has changed over the last 10 years, with the formulation of many quantitative models for the contribution of cultural inheritance to individual differences and family resemblance (2-8). Such models have explored vertical transmission between parent and child, horizontal transmission between siblings and one-to-many oblique transmission between teacher and students......

Over the last 15 years, data on very large samples of monozygotic and dizygotic twins and spouses have been collected in the attempt to provide more powerful resolution of the basic elements of cultural and biological inheritance. In the case of personality measures, mate selection is virtually random and the resemblance between relatives is almost entirely genetic in origin (9-12). There is little evidence that cultural inheritance contributes to individual differences in personality in the populations studied so far.

Social attitudes present a marked contrast to personality measures. They too show substantial family resemblance (8, 13, 14), but the similarity between mates for social attitudes is also considerable (14. 15). Secular changes in attitudes are so rapid (16) that frequent revision of test instruments is necessary. On the face of it, such findings lend support to purely cultural model for family resemblance. However, studies of attitudes so far have not tested the assumption that vertical transmission is cultural. They have not addressed the alternative hypothesis that individuals are influenced by their genotypes in their acquisition of particular opinions from the range current in a given society (17). Insofar as latent genetic factors influence the individual's preference for particular attitudes, vertical transmission will have a genetic component and purely cultural models will be inappropriate.

The Samples

In two separate studies, social attitudes questionnaires were mailed to twins enrolled on the Australian National Health and Medical Research Council Twin Registry and on the Institute of Psychiatry Twin Register. London, England. Both registries comprise volunteers and there is a marked excess of female monozygotic twins in both studies. Zygosity was determined by querying similarity in childhood and confusion of one twin for the other by parents, friends, and teachers. and it has been validated by blood-typing in subsets of both sample (18, 19). The Australian sample (Table 1) obtained responses from 3810 of 5967 pairs (64%) to whom questionnaires were originally mailed. The British sample 1 Table 21 generated 825 complete pairs, a pairwise response rate of =50%. The Australian twin sample was supplemented by 103 pairs of spouses studied by Feather (13), and the British twin sample was supplemented by 562 spouse pairs ascertained from the London area.

Test Instruments

The Australian study employed a 50-item version of the Wilson-Patterson conservatism scale (refs. 20 and 21; Table 1), which comprises a series of one-word items (e.g., "censorship") to which the subjects rate their agreement by circling "Yes," "?," or "No." The test yields a "conservatism" score by weighting odd-numbered items "+1" and even-numbered items "-1." A subsample completed the scale twice at an average interval of 3 months and test-retest reliability was 0.86 in 64 females and 0.92 in 32 males.

The British study of twins and spouses employed a Public Opinion Inventory comprising 40 frequently encountered statements relating to such issues as religion, sex, treatment of criminals, and nationalism (22). Respondents rate their agreement with each item on a 5-point scale. The instrument was scored for two factors: (i) radicalism, which describes the "left vs. right" dimension in British politics, and (ii) toughmindedness, exemplified by approval of capital and corporal punishment. In the entire British sample (n = 2774), the correlation between these two factors was 0.17.

Data Summary

..... The path model in Fig. 1 represents genetic and cultural components of vertical transmission in nuclear families in the presence of phenotypic assortative mating. The model assumes transmission of additive genetic differences (A) and a direct cultural effect of parental phenotype (P) on the phenotype of offspring. The regression of phenotype on genotype is h in males and h' in females: the partial regression of offspring phenotype on parental phenotype is b in males and b' in females. The parameter b, therefore, embodies vertical cultural inheritance in the model. The phenotypic correlation between mates is p. The paths from phenotype of parent to additive genetic effect (x, and w') can be expressed as functions of h, h', b, and b' at equilibrium under cultural transmission and assortative mating.....

The method of nonlinear weighted least squares was applied to the z transforms of the observed correlations to recover estimates of the parameters that correspond closely to maximum-likelihood values (23, 24). Since the z's are normally distributed and independent, the sum of weighted residuals is approximately distributed as Chi-squared for n - p degrees of freedom, where n is the number of observed correlations and p is the number of free parameters in the model. The residual Chi-squared may be used as a guide to the goodness-of-fit of the model and Chi-squared values for certain alternative hypotheses may be compared to justify reducing the general model to a more parsimonious form.

The results of the model fitting are given for the Australian data in Table 3 and for the British data in Table 4. Models in which h [heredity] is set to zero in both sexes give a very poor fit to the data in both studies (see models 1 and 2) since the associated residual Chi-squared values are large. On the other hand, models that leave out cultural inheritance (b = b' = O) give an extremely good fit when allowance is made for assortative mating (models 3 and 4). The Australian sample gives no evidence of heterogeneity over sexes in the contribution of genetic and environmental factors since models 4 and 6 fit no better than models 3 and 5. Because the spousal correlation mu = 0.675 used in this analysis is from another study (13), is based on relatively small numbers (n = 103), and has not been age-corrected, we repeated the analysis specifying a value of mu = 0.40. Even with this conservative estimate of the spousal correlation, the conclusions of model fitting were unaltered and the estimate of b = 0.07 was trivial and nonsignificant.

For the two factors scored in the British study, the best fit is obtained when different genetic contributions to variance are allowed in males and females (model 4). A greater proportion of variance is genetic in males than in females for radicalism and in females than in males for toughmindedness.


We are aware of the many criticisms of the twin method and of the responses others have made to these (25). For example, it is alleged that monozygotic twins see each other more frequently than dizygotic twins and that this greater frequency of contact is reflected in greater monozygotic similarity in attitudes. Since the correlation between reported frequency of contact and absolute intrapair difference in conservatism scores is -0.08 in female and -0.14 in male twin pairs in the Australian sample, any such effect must be trivial, even if the cause of such covariation is in the direction asserted.....

Back to List of excerpted references

Go to Index page for this site

Go to John Ray's "Tongue Tied" blog (Backup here or here)
Go to John Ray's "Dissecting Leftism" blog (Backup here or here)
Go to John Ray's "Australian Politics" blog (Backup here or here)
Go to John Ray's "Gun Watch" blog (Backup here or here)
Go to John Ray's "Education Watch" blog (Backup here or here)
Go to John Ray's "Socialized Medicine" blog (Backup here or here)
Go to John Ray's "Political Correctness Watch" blog (Backup here or here)
Go to John Ray's "Greenie Watch" blog (Backup here or here)
Go to John Ray's "Food & Health Skeptic" blog (Backup here or here)
Go to John Ray's "Eye on Britain" blog (Backup here or here)
Go to John Ray's "Leftists as Elitists" blog (Not now regularly updated -- Backup here or here)
Go to John Ray's "Marx & Engels in their own words" blog (Not now regularly updated -- Backup here or here)
Go to John Ray's "A scripture blog" (Not now regularly updated -- Backup here or here)
Go to John Ray's recipe blog (Not now regularly updated -- Backup here or here)
Go to John Ray's "Some memoirs" (Occasionally updated -- Backup here)

Go to John Ray's Main academic menu
Go to Menu of recent writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page (Backup here)
Go to Selected pictures from John Ray's blogs (Backup here)
Go to Another picture page (Best with broadband)