In this paper, we consider human handedness and cerebral lateralization in a general biological context, and attempt to arrive at some conclusions common to the growth of human laterality and of other structural asymmetries. We suggest that many asymmetries appear to be under the influence of a left-right maturational gradient, which often seems to favor earlier or more rapid development on the left than on the right. If the leading side is damaged or restricted, this gradient may be reversed so that growth occurs with the opposite polarity. A mechanism of this sort appears to underlie the phenomenon of situs inversus viscerum et cordis, and the same principle may help explain the equipotentiality of the two sides of the human brain with respect to the representation of language in the early years of life. However we must also suppose that the leading side normally exerts an inhibitory influence on the lagging side, for otherwise one would expect language ultimately to develop in both halves of the brain. Examples of an inhibitory influence of this kind can also be found in other biological asymmetries; for instance, in the crab Alpheus heterochelis, one claw is normally greatly enlarged relative to the other, but if the larger claw is removed the smaller one is apparently released from its inhibitory influence and grows larger.

This last example is particularly interesting because it suggests a mechanism comparable to that proposed by Annett to account for the distribution of handedness in the human population. She argued, in effect, that there is a “right shift” factor among the majority of the population, but that among a minority who lack this factor handedness is determined at random. If it is supposed that cerebral lateralization is also determined at random among this recessive minority, the model can be extended to provide a reasonable fit to the data on the correlation between handedness and cerebral lateralization. However this genetic model (or any other) still fails to account for the near-binomial distribution of handedness among twins and among nontwin siblings. We suggest that right-handedness and leftcerebral dominance for language are manifestations of an underlying gradient which is probably coded in the cytoplasm rather than in the genes. We must leave open the question as to whether departures from this pattern are due to a recessive gene which effectively cancels the asymmetry to environmental influences, or to both genetic and cytoplasmic factors.

This paper focuses on the inheritance of human handedness and cerebral lateralization within the more general context of structural biological asymmetries. The morphogenesis of asymmetrical structures, such as the heart in vertebrates, depends upon a complex interaction between information coded in the cytoplasm and in the genes, but the polarity of asymmetry seems to depend on the cytoplasmic rather than the genetic code. Indeed it is extremely difficult to find clear-cut examples in which the direction of an asymmetry is under genetic control. As one possible case, there is some evidence that the direction, clockwise or counterclockwise, of rotation of the abdomen in certain mutant strains of Drosophila is controlled by a particular gene locus, although there appears to be some degree of confusion on this point. By contrast, it is much easier to find examples in which the degree but not the direction of asymmetry is under genetic control. For instance, there is a mutant strain of mice in which half of the animals display situs inversus of the viscera. The proportion has remained at one half despite many years of inbreeding, suggesting that the mutant allele effectively cancels the normal situs and allows the asymmetry to be specified in random fashion.

Although this account does not deny that the right hemisphere of humans may be the more specialized for certain functions, it does attribute a leading or dominant role to the left hemisphere (at least in most individuals). We suggest that so-called “right-hemisphere” functions are essentially acquired by default, due to the left hemisphere's prior involvement with speech and skilled motor acts; we note, for instance, that these right-hemisphere functions include rather elementary perceptual processes. But perhaps the more critical prediction from our account is that the phenomenon of equipotentiality should be unidirectional: the right (lagging) hemisphere should be more disposed to take over left-hemisphere functions following early lesions than is the left (leading) hemisphere to take over right-hemisphere functions. We note preliminary evidence that this may be so.