Studies of Human Twins 



75 



greater number are usually too infrequent to 

 be useful for a general study of the nature- 

 nurture problem. 



Multiple human births may occur also as 

 a consequence of sexual reproduction. In 

 this case the twins produced start with two 

 separate eggs, each fertilized by a separate 

 sperm. Such twins are genetically different, 

 being, in this respect, no more similar than 

 siblings conceived at different times, and are, 

 therefore, called nonidentical or dizygotic 

 {fraternal) twins. 



These two kinds of twins provide another 

 natural experiment for determining the rela- 

 tive influence of genotype and environment 

 upon the phenotype. For monozygotic twins 

 furnish the identical genotype in two indi- 

 viduals, and both kinds of twins share very 

 similar environments before birth and, when 

 raised together, after birth. 



Accordingly, the phenotypic differences be- 

 tween identical twins reared together are, 

 barring the rare event of mutation, purely the 

 consequence of environment (Figure 11-1). 

 One can compare the average of these differ- 

 ences between identical twins with the aver- 

 age of the differences between identical 

 twins who, for one reason or another, were 

 reared apart, usually in separate families. 

 This would yield information regarding the 

 influence upon the phenotype of greater, as 

 compared with lesser, environmental differ- 

 ences. Since nonidentical or identical twins 

 reared together are exposed to environments 

 which vary to the same extent on the average, 

 a comparison of the average difference be- 

 tween identical twins and the average differ- 

 ence between nonidentical twins will give an 

 index of the role of the genotype in causing 

 the differences observed. However, in order 

 to collect data from twin studies, you can see 

 how essential it is to be able to recognize in 

 each case whether the twins are monozygotic 

 or dizygotic in origin. 



The best way to identify twins as non- 

 identical is to compare the two individuals 



phenotypically. They should be compared 

 with regard to a large number of traits known 

 to have a basis in genes that are 100% pene- 

 trant and of fairly uniform expressivity. These 

 would include such traits as sex, eye color, 

 ABO, MN, Rh, and other blood group types. 

 Naturally, only those traits for which at least 

 one parent is heterozygous can be of use in 

 testing the dizygotic origin of twins. Ignor- 

 ing the rare event of mutation, any single 

 difference in such traits would prove the twins 

 nonidentical. Of course, two such differences 

 would make your decision infallible for all 

 intents and purposes, since two mutations 

 occurring in the limited number of traits 

 being followed in a pair of identical twins 

 would be so rare as to be beyond any reason- 

 able probability of occurrence. With these 

 criteria, twins of opposite sex may be classi- 

 fied immediately as nonidentical. 



Classification of twins as identical is based 

 on the same procedure except that the greater 

 the number of traits for which no difference 

 at all is shown, the greater is the probability 

 they are identical. For if the number of traits 

 serving to test the genotypes of twins is 

 sufliciently large, it becomes nearly certain 

 that had they been dizygotic in origin they 

 would have shown one or more differences, 

 their failure to show any difference being 

 attributable to their genotypes being identical 

 because they were both derived from a single 

 zygote. 



We are now in a position to study the rela- 

 tive roles of genotype and environment in 

 producing specific traits. What is done is to 

 score the percentage of pairs of twins, reared 

 together, in which one or both twins have the 

 trait under consideration. Let us outline the 

 procedure which we might actually follow. 

 Suppose we wished to study the blood group 

 AB m this respect. What we would do first 

 is to eliminate from consideration all pairs 

 of twins in which neither individual was 

 type AB. We would have remaining, then, 

 twins which had at least one member of AB 



