Sewall Wright Research Paper

1.    Early Life

Wright  was a  precocious  child.  He  could  read  and kept  a diary  before  starting  school.  He also learned arithmetic  and  astonished  his first-grade  teacher  by extracting   cube  roots.   Later   he  and   his  brothers printed  the first poems of Carl Sandburg, who was a student    in   their   father’s   composition    class.   At Lombard he  took  several  courses  from  his  father, including calculus. But, despite his later reputation as a formidable  mathematician, Wright  never took  advanced courses; his mathematics was largely self-taught.  His interest in genetics was kindled by reading Punnett’s  article  on  Mendelism  in the Encyclopedia Britannica.

Between his junior  and  senior years of college, he used his mathematical and surveying skills to work on a  railroad  construction in  the  Standing  Rock  Reservation  in South Dakota. It was an exciting time for him.  In  his 90s he enjoyed  telling  about  the  rough frontier  life with cowboys, Indians,  and outlaws,  and he still recalled a large number  of words in the Sioux dialect. This work was cut short  by a lung infection. While confined to a caboose, he used the time to study Tait’s book on quaternions. As a result of the damaged lung,  he  later   had   trouble   getting   life  insurance, something  that  he found  increasingly  amusing  as he continued  to live far beyond the usual life-expectancy.

Wright’s father, Philip, was a man of many talents. He taught  courses in astronomy, mathematics, economics, surveying, English composition, and  physical education.   He  loved  poetry   and   music,  and   was disappointed that his son did not take to them. Later he moved  to the Brookings  Institute  where he published a number  of books  on economics. Wright  had two  brothers  who  also  attained  distinction.  Quincy became  a leader  in international law and  Theodore was an aeronautical engineer and acting president  of Cornell University.

A  full  length  biography   of  Wright   is  available (Provine 1986), with full details of his life and scientific accomplishments. His scientific papers span the period from  1912  to  1988,  an  astonishing   76  years.  His Bibliography: comprises  more  than  200 articles,  including   a  four   volume   set  published   during   his retirement  (Wright 1968–78).

2.    Animal Breeding

Wright’s  thesis  at  Harvard analyzed  interaction  of coat colors in guinea pigs, a subject which he pursued throughout his USDA and Chicago years. He enjoyed guinea pigs and spent many hours taking care of them. Only  the  non-availability of  an  animal  house  prevented his continuing this after his move to Wisconsin. While at Harvard, he also began his theoretical studies of population genetics. On moving to the USDA,  he inherited a colony of guinea pigs that had been inbred for  many  generations.  His  analysis  of the  effects of inbreeding and crossbreeding set a standard for careful analysis; the paper could be written today with hardly a  word  changed.  During  this  time  he  invented  the inbreeding  coefficient,  with  a simple algorithm  now standard  textbook  material   (Wright   1922a).  Also during this period, he formulated the basic principles for analysis of quantitative traits.

Wright,  along with R. A.  Fisher   (see  Fisher,  Ronald  A  (1890–1962)), converted animal breeding into a quantitative science. Wright  was  especially  proficient  in  manipulating correlation  and  regression  coefficients.  One  of  his earliest papers (Wright 1917) foreshadowed the analysis of covariance.  His  major  statistical  contribution was developing the method  of path analysis. This is a way of measuring the relative contribution of different causal paths  to a quantity  of interest,  such as size or performance.  This  was  presented   in  a  diagram   in which  causal  paths  were designated  by  arrows  and correlations by curved,  two-headed  arrows  (Wright 1921). A path coefficient, measuring the influence of a path,  is a standardized partial  regression  coefficient. Wright formulated a set of simple rules that made the analysis easy to apply. This method enjoyed great popularity among livestock breeders, although recently it has been largely replaced by more sophisticated, computer-driven techniques that permit analysis of large bodies of data,  and measurement of the precision of the estimates.

The technique  is most  useful for nonexperimental situations and in recent years has found its greatest use in the social sciences. Wright himself used the method in  the  late  teens  to  analyse  510 correlations of  42 variables  involving  corn  yields and  prices  and  hog production. He could not get this monumental paper published  because an animal  husbandryman was not supposed   to  know   about   economics.   It  was  not published  until several years later, and only after the intervention of  Henry  A.  Wallace,  son  of  the  then Secretary of Agriculture.

3.    Evolution

After   moving   to   Chicago,    Wright    changed    his emphasis from animal breeding to evolution, although he still retained his interest in the physiological genetics of guinea pigs. Three people, R. A. Fisher (1929) and J. B. S. Haldane  (1932) in Britain, and Sewall Wright (1931) in the USA are mainly responsible for the ‘new synthesis,’ the welding of Mendelian  inheritance  and Darwinian natural selection  into  a  coherent  mathematical theory. Wright’s particular contribution  was the ‘shifting balance theory.’ He thought that a major difficulty in evolution  was the frequent  inability  of a well-adapted organism to evolve into a better adapted one   without   passing   through   maladaptive   intermediates.  In Wright’s metaphor, a population at an adaptive peak could not move to a higher peak without passing through less well adapted intermediate  stages.

Wright thought that a way out of this dilemma was to have a population broken up into partially isolated subgroups. Within  a  small  subgroup  random  genefrequency  changes  can occur  and  occasionally  these may produce  a combination of genes that  is close to that   of  the  genotypes   at  the  higher  peak.   Then, selection could take over and carry the population to the peak.  This highly adapted subpopulation would then expand and send migrants  to upgrade  the whole population, and the process could start over.

Wright’s theory became very popular among biologists.  Its emphasis  on gene interactions and  its offering a way of putting  together  adaptive  combinations of genes that might be deleterious in other combinations had a wide appeal.  On the other  hand, the   theory   was   less  popular  with   mathematical theorists.  Wright’s model required  a delicate balance of  population size, selection,  and  migration,   which would  be rare.  It  also  required  that  the  population spend  much  of  its  time,  during  the  trial  and  error period, in a less than optimally adapted state. Finally, there is a question  as to whether  the theory  is really needed; perhaps mass selection usually suffices for adaptive  evolution.

Wright’s theory was steadily refined during the rest of his life, but the essence remained  unchanged. The theory  became almost  an obsession with him and he delivered many talks on the subject and wrote paper after paper. It is fair to say that, as to the importance of his theory, the jury is still out. Wright’s latest paper (1988)  was  spirited  defense  of  his  idea,  while  still noting  the importance of the contributions of Fisher and  Haldane, and  of  Kimura   (see Kimura,  Motoo (1924–94))   who   argued   for   the   importance   of mutation-driven, neutral  evolution of molecular differences. He concluded that ‘all are valid.’

4.    Retirement  Years

After  moving  to  Wisconsin,  Wright  spent  the  first few years finishing the analysis of his extensive guinea pig studies. These were masterful analyses, but by that time the molecular revolution  had changed the nature of developmental genetics, and Wright’s methods were out of vogue. When this was completed,  in his 70s he began   his  monumental  four   volume  set,  the  last volume  of  which  was  published   when  he  was  88 (Wright 1968–78). Although his shifting balance  theory is conspicuous  throughout, the book  is far more than this. It summarizes, in four densely packed volumes, not only his life’s work, but the whole field of population genetics.  For  the  last  decade  of  his life Wright’s eyesight became progressively worse, and he used a reader that projected a greatly enlarged picture onto a television screen. He continued  to read and to enjoy conversations, but in his last years he found  it increasingly  difficult  to  keep up  with  molecular  advances and turned more to reading history and biography.

As a person, Wright was quiet, shy, and introverted. He had no small talk, and was hard to converse with. Yet,  once  the  conversation turned  to  a  subject  on which he had an interest—and there were many—he could go on at great length. His lectures were always full of details  and  regularly  ran  overtime.  He had  a charming sense of humor and was excessively generous with his time. His students and colleagues held him in great affection.

5.    Social Sciences And Philosophy

Wright’s  greatest  influence  in  the  social  sciences is his method of path analysis. It was taken up by several social scientists, such as Dudley Duncan  in sociology and Arthur  Goldberger in economics, and has become a standard part of the statistical methodology. One of Wright’s  earliest  applications of  path  analysis  used adopted children to separate genetic from environmental  influences on IQ scores, and  there have been many  follow-up  studies  extending  his  methods   to more complex analyses of behavioral  traits with better data. Finally, with the increasing interest in behavioral evolution, especially among psychologists, mathematical theories  of evolution  are  becoming  increasingly common and Wright takes a place in this, along with Fisher and Haldane.

Wright  was unusual  among biologists in taking  an active interest in philosophy. He had a somewhat Leibnitzian  view. He disliked any idea of emergence. He found no sharp borders between different levels of complexity, such as between embryo and adult or between mind and no mind. Thus if the mind does not emerge by magic, it must trace its development  into a continuous process back to the embryo,  the egg and sperm, and ultimately  the DNA  molecules. He called his concept ‘dual-aspect panpsychism.’ Mind is everywhere; so is matter.

The biological community either disagreed with Wright’s view, or was indifferent. Most regarded mind as  a  natural outgrowth of  appropriately  organized matter.  Wright held on to his dualistic view, however, and found a few kindred  spirits among philosophers, particular his long time colleague at the University of Chicago,  Charles Hartshorne.

Bibliography:

1. Fisher R A  1929 The  Genetical Theory  of  Natural  Oxford University  Press, Oxford,  UK
2. Haldane J B S 1932 The Causes of Evolution. Longmans  Green, London
3. Provine W B  1986  Sewall  Wright  and  Evolutionary  University  of Chicago Press, Chicago
4. Wright S 1917 The average  correlation within subgroups  of a population. Journal of the Washington Academy of Sciences 7: 532–5
5. Wright S 1921 Correlation and causation. Journal of Agricultural Research 20: 557–85
6. Wright S  1922a  Coefficients  of  inbreeding  and  American Naturalist 56: 530–38
7. Wright S 1922b The effects of inbreeding  and crossbreeding  in guinea pigs Bulletin 1090: 1–63, 1091: 1–60
8. Wright S 1931 Evolution  in Mendelian  Genetics 16: 97–139
9. Wright S 1968–78 Evolution and the  Genetics of  4 Vols. University  of Chicago Press, Chicago
10. Wright S 1988 Surfaces  of selective value revisited.  American Naturalist 131: 115–23