The pioneer work in this field of behaviour study, as in so many others, is due to the genius of Sir Francis Galton. His studies on the inheritance of mental capacities, begun in 1869 and including the significant contribution of the twin-study method, initiated an interest in this field which has rarely lost momentum since. It is not the present intention to dwell on the efforts to assess the size and nature of the genetical influence on human behaviour, but rather to evaluate the contribution that experimental work using animal subjects has made. But in passing it should be noted that progress in the former, real though it has been, still leaves much to be desired. Thus Cattell, whose recent work using multiple variance analysis (1953, 1959a, 1959b; Cattell, Blewett, and Beloff, 1955) represents a methodological advance in this field on the earlier use of the twin study and familial correlation methods independently of each other, writes as follows: âNo argument is necessary to convince the scientific psychologist that personality theory, clinical psychology, and learning theory are bereft of knowledge of about half of the influences at work, in the absence of an understanding of heredity. All learning and adjustment is limited by inherent properties of the organism. Only the unintelligent or the uninformed clinician or educator practices individual guidance without estimating what degree of difficulty a given adjustment or performance offers a particular personality. The general laws of psychology, explaining the genesis of abilities, temperament, and dynamic traits, need also to be more concerned with maturational trends, innate dispositional rigidities, and the effects due to hereditary limitsâ (Cattell, Stice, and Christy, 1957, p. 143).
But in turning to the animal field, which is an essentially more tract able one for experimental psychogenetics, because of the availability of genetically pure lines, and the control possible over mating, enabling the formation of others, it must be confessed that the situation is only slightly better. Apart from a few significant exceptions in the last decade, it is still possible to agree with Stone when he wrote: âTo many of us who have worked for a long time in the field of comparative psychology it is a matter of shame and regret that only an amateurish beginning has yet been made by psychologists in the utilization of pure lines of animals in fundamental research in the nature-nurture areaâ (1947, p. 344).
Methods of comparative psychogenetics
Psychological authorities are agreed that there are two main approaches to the study of the inherited determinants of behaviour in animal subjects (Stone, 1947; Hall, 1951; Russell, 1953a). The first is by selection and the second by the analysis of the characteristics displayed by different, preferably pure, strains. In a sense these methods are complementary; the object of the first, selective breeding, is to establish lines which differ in the behavioural characteristics for which they have been selected. The fact that such selection is possible is strong presumptive evidence that the characteristic in question has some hereditary determination. The second method employs established strains, ones selected originally for differences other than those in which a psychologist is interested, such as coat-colour, tumour susceptibility, etc. The criterion for purity of strain in mammals is usually taken as twenty consecutive generations of mating brother with sister (though see Carpenter, Griineberg, and Russell, 1957; Biggers, McLaren, and Michie, 1958) and can be checked by tissue transplants (Billingham, and Silvers, 1959). Again, the fact of differences between different genetically pure strains is presumptive evidence of the influence of heredity upon behaviour.
The end point reached by the use of both methods is the sameâits purpose is, or should be, to arrive at a position from which a genetical analysis can commence using the standard procedures of, for example, cross-breeding with pure strains and the analysis of heritability from selectively bred ones. Both these methods have been employed in the present work.
Strain differences
The use of different strains, particularly of rats and mice, has long been a favourite method of psychogenetics. Bagg was a pioneer in the application of âthe methods of genetics to the study of conductâ. He found strain differences between strains of mice in various learning situations (1916, 1920). The Yerkes, (1913, 1916) and Utsurikawa (1917) compared pure and outbred strains of rats, the latter noting that individual differences exceed strain differences. Sadovnikova-Koltzova crossed albino with wild grey rats and studied the first and second generations of offspring of this cross by means of the Hampton Court maze (1926), publishing further papers later (1928, 1929, 1931, 1933).
The studies on mice of Vicari (1929) and Dawson (1932) show more genetical sophistication but are deficient psychologically. The maze performance measures as used by Vicari are unnecessarily complicated characteristics for study, and it is not surprising that her analyses of three crosses from among four inbred strains did little to clarify the genetics of maze learning, especially since one of her strains, âwaltzerâ mutant, is now known to be complex genetically (Griineberg, 1952). Dawson used as a measure of âtamenessâ the speed of unrewarded running in a straight alley, a test which bristles with difficulties to administer and interpret, but for which he claims a reliability of 0.92. He used the standard type of Mendelian analysis with mammals by crossing the parental strains, usually designated P1 and P2, to produce a first filial or hybrid generation (Fj), a second filial generation bred from the first (F2), and then making the backcrosses from F1 to the P1 and the P2-B1 and B2 respectively. He does not specify the degree of genetical homogeneity (homozygosity) of the four strainsâthree tame and one wildâwhich he used, thus rendering only tentative his conclusions that wildness is determined by a few genes which are dominant and which have no linkage for sex, or for any of the five genes which he knew to be carried by his strains. That this absence of linkage is probably true for one of the genes, âshort earâ, has recently been confirmed (Bundy, 1950).
The tendency of young rats to climb up an inclined plane (negative geotropism) has been the subject of a series of studies by Crozier and Pincus (see Munn, 1950, pp. 44 if., for a review of this work, and the controversy over the nature of the behaviour observed). The main papers of psychogenetic interest are Crozier and Pincus, 1931-2a, 1931-2b, 1931-2c, 1935, 1936. They found that different strains of rats varied in the angle taken up on the inclined plane in response to variations in the angle from the horizontal of the plane itself. Crosses between two of the strains were bred, and backcrosses made to each parental line in turn. The F1 showed a pattern of geotropic behaviour which was interpreted as hybrid vigour (heterosis) affecting the rear limbs only and giving disproportionate development. The absence of this in the geotropic behaviour of adults of the cross was taken as confirming the hypothesis. Separate consideration is given to each back-cross (Crozier and Pincus, 1931-2c, 1936), and despite the very small numbers used it is claimed that the dominance of the genes from one strain can be shown.
Herter (1936) found that strains of mice differed in their preferred floor temperature. By means of Fiâs and backcrosses to a wild strain he was able to identify the important determining factor as the density of the belly fur (Herter, 1938; Herter and Sgonina, 1938).
While observations relating to differences between strains of unspecified degrees of genetical purity are still encountered (Schaefer, 1959; Myers, 1959), the post-World War II work on the inheritance of behaviour has concentrated on the use of the inbred strains of rats, and, more especially, mice, which are now available (Staats, 1958). The volume of work on these two species allows a consideration of it topic by topic.
The hereditary determinants of audiogenic seizures have been investigated by several workers; their findings are reviewed by Finger (1947) and Bevan (1955). An hypothesis of multifactorial genetical control seems to fit the experimental data rather better than those suggesting one- or two-gene action.
Strain differences in respect of the hoarding of food by mice were found by Smith and Powell (1955), and confirmed for rats by Stamm (1954, 1955) who attempted a genetical analysis using three pure-bred lines of different coat-colours. He obtained an F1 from two of the strains and backcrossed to only one of them (1956). The results did not enable him to specify a genetical mechanism for the hereditary effects disclosed.
The influence of inheritance on emotional responses in five strains of rats, which included the three used by Stamm, was confirmed by Broadhurst (1958b), and these findings will be discussed in Chapter 4. Lindzey (1951) had previously shown the same effect with five inbred strains of mice.
Such strains of mice (Carter et al. mult., 1952) have been used with good effect in the study of activity. Thompsonâs paper (1953b), employing fifteen of them in three behavioural situations, set a new standard in this field. He went on to confirm the findings on exploratory activity in five strains (1956), two of which had not been used in the previous study. A suggestive parallel exists between Thompsonâs ordering of these strains in activity and that determined by the thyroid output measured in some of them (Amin, Chai, and Reineke, 1957). The relationship suggests that the higher the output the greater the amount of exploration. The genetical analysis proceeded with the breeding of Fiâs (Thompson, 1954), F2âs and backcrosses to the two showing the highest and the lowest activity (Thompson and Fuller, 1957, 1960). This is the most complete analysis of this trait yet reported, and the results are of extreme interest. The heritability, or ânature/nurture ratioâ, a measure of the extent to which the character observed is under genetical control, gives a value around 60%, the rest being non-genetical or environmental. An estimate of the number of genes involved is four or more, and there is some suggestion of dominance for those determining higher activity. The use of two somewhat different situations for measuring the activity allows the additional conclusion that the description of a genetical mechanism is in part dependent upon the way it is measured.
McClearn (1959) has also investigated the psychogenetics of exploratory activity in the mouse, on a somewhat broader basis, using four different tests of exploratory behaviour in novel situations, and six mouse strains. Thompsonâs earlier findings were supported, and an cross of the two extreme strains, though not, unfortunately, the same strains as in Thompson and Fullerâs study, gave results which suggested that the trait as measured is determined multifactorially. The question of exploratory activity in the rat will receive further attention later; that it...
