From Galton to Bouchard: Nature/nurture, twins, correlations and controversy

 When you have an estimated IQ of 200 and Charles Darwin in your gene pool, it is probably not surprising that you might say the following:

I have no patience with the hypothesis occasionally expressed, and often implied….that babies are born pretty much alike, and that the sole agencies in creating differences between boy and boy and man and man, are steady applications and moral effort.  It is in the most unqualified manner that I object to pretensions of natural equality.

These were the words of Francis Galton in 1869 in his book Hereditary Genius.  He was literally (at least to the best of anyone’s knowledge) the person who coined the terms Nature versus Nature.  His abilities and accomplishments were so wide ranging that followers can admire almost any Galton that interests them–the mathematical/statistical wonder, the explorer/geographer, fingerprint expert, meteorologist, psychometrist, or pioneer in the study of human nature.  He was the half-cousin of Darwin, sharing their grandfather, Erasmus Darwin.

An amazing amount of what he did laid the groundwork for modern behavioral genetics.  He invented the correlation, the standard deviation and bivariate normal distribution–all essential to later measures of traits and abilities.   He was the first to develop psychometrics ( questionnaires and other measurement techniques) and his fascination with individual differences  was the beginning of the modern study of both intelligence and temperament.  A particularly interesting link is his establishment of the “Lexical Hypothesis“.  The idea behind this is that those qualities that are most important in human nature will become encoded in language.  Therefore, you can use those descriptive words to establish all possible examples of human traits, combining them into categories (like agreeableness).  Most intriguing is the fact that this is the basic approach that was used in creating today’s most respected temperament test–the Five Factor Model or Big Five.

Not surprisingly, for a man of his abilities, he was fascinated with the question as to whether outstanding abilities were hereditary.  He began by looking first at biographies of famous men and their families, and very soon concluded that as you went from father to son, and sibling to sibling, and compared that with first and second degree cousins, the number of exceptional relatives of an exceptional person declined.  He took this a little further by sending out a questionnaire to all members of the Royal Society in England asking about birth order, occupation, reasons for interest in science, etc.  This further strengthened his hereditary hypothesis and sent him on to study twins in what we would call today twins reared apart and twins reared together.

And so, family studies, questionnaires, twin studies and the statistics that he developed to analyze these, all led him to the belief that strong hereditary forces were at work, with nature being far stronger than nurture.  Unfortunately for his posthumous reputation he then went, in what was to him, a logical step further.  He coined the term ‘eugenics’ and wrote of his conclusions in a further book in 1883,  suggesting that society should encourage more children from  eminent families with monetary incentives and establish some sort of index of family merit.

These ideas, right or wrong, later fed into all the horrors of the Nazi race-based eugenics movement, including the death camp experiments on twins carried out by Joseph Mengele.  Without question this was a large force in bringing on the almost total return to a nurture philosophy in the later 20th century.  No doubt it is also the reason today that Charles Darwin is the good cousin and Francis Galton, for all his accomplishments is the dark cousin.  Evolution–good! Heredity–bad?

Nevertheless, Galton lives on in all his many contributions, but especially in our statistics books and in the modern study of temperament.  Today, researchers who have done the most to carry his ideas forward, at least in the area of temperament, are Thomas Bouchard and his colleagues who have done remarkable work in the University of Minnesota’s twin studies, and Costa, McCrae and colleagues who developed the Five Factor Model.  These creations have worked hand in hand to give us a sturdy picture of human temperament from a genetic point of view.

There are outstanding twin studies being done all over the country, and all of these studies examine many things in addition to temperament,  but the University of Minnesota has done remarkable work on genetics in twins, and the relationship to temperament scores on the Big Five.  In a study on twins reared apart, they gathered 100 pairs of twins (59 identical and 47 fraternal) and administered among other instruments, the Big Five questionnaire.  These scores were compared with 522 identicals reared together and 408 fraternals reared together.  Selection of twins was done very carefully with a great deal of data gathered on both the away and together homes, and on the identical/fraternal nature of the twinships.

There are always some arguments that even identical twins do not have perfectly identical DNA once the long process of cell division and migration is complete.  This is a well taken point but actually adds strength to the arguments.  That is, if even identicals are less than perfectly identical and yet much more like one another than fraternals or siblings, the case for heredity is all the stronger.

The second major argument about the findings is that homes for twins adopted apart do not show the whole range of environments that are possible  That is, they are probably more stable and supportive than the full range of possibilities in the world.  That is likely true, suggesting that these studies do not measure the effects of truly abusive and non-supportive homes.  About all there is to say about that is that this is true enough, it being fairly impossible to place children deliberately in the worst of possible circumstances.  Nevertheless there was a fair range of income, education and cultural artifacts–books, music etc.–in the samples.  This simply says that we don’t know from these studies what the effect on temperament might be in really destructive environments.

So–what did they find?  The statistical analysis looks at the differences in scores in frateral twins with half their DNA in common and in identical twins with all their DNA in common in both the reared together and reared apart circumstances.  Using these data the estimated effect of heredity can be separated from the estimated effect of shared environment (the home and family they had together), the estimated effect of non-shared environment, and statistical error.  This very careful study arrived at a figure over the five temperament measures in the Big Five, of 41% as the effect of heredity.   The great surprise was that only about 7% of the variability appeared to come from shared variance–that is from the home, parents etc. that they twins had in common.  Non-shared variance was seen as those experiences that were unique to one twin–a long illness the other did not have, a memorable teacher that was not shared, and so forth.

That was pretty shocking to educators, developmental psychologists and any parents who happened upon this news.  I think the public by 1990 was ready to live with the idea that genes might be half of the story and parental nurture somewhere around the other half.  But 7%?  Well, how to explain this?  There are some other ideas about this, but I would love to hear from any of you out there with a theory before next week’s blog.  Any thoughts?  (You are out there-that I know, for my stats counter tells me so)!

This entry was posted in Anything/Everything-Temperament, In History and tagged , , , , , . Bookmark the permalink.

3 Responses to From Galton to Bouchard: Nature/nurture, twins, correlations and controversy

  1. Lars says:

    7% percent sounds really shocking. (You should have made this the title of the article 😉 )

    But you have to accurately think this through: its not 7% of the whole Personality, but 7% of the Big Five traits that are altered through shared variance (parents, home…).
    A human personality doesn’t only exist of Big Five traits, but also of other things like learned behaviors and skills, habits, predilections et cetera (ok here one could argument that these are also influenced by the Big Five).

    So i think its save to say that the enviroment doesn’t change the core personality (Big Five Traits) much, but still has considerably influence on the overall personality with all its other facets.

    I think i already read it some time ago, but cane you give me the source of the Minnesota study?

    • INTJ says:

      The original reference for that 7% figure is:

      Bouchard, T. J. (1994). Genes, Environment, and Personality. Science 264, 1700-1701.

      That stirred up lots of controversy, of course. A good reference for the reactions of developmental psychologists is:

      Collins, W. et al. (2000). Contemporary Research on Parenting: The Case for Nature and Nurture. American Psychologist, 55(2), 218-232.

      My own take on this is that where parent and child 1 differ dramatically in temperament and parent and child 2 do not, the two children really do not have the same shared parental environment.

  2. INTJ says:

    Hey readers out there! I am longing to see a comment on what it means when there is little or no effect on chldren of the shared environment. New blog is going up soon so write now.


Leave a Reply

Your email address will not be published. Required fields are marked *