Evidence of biological involvement in exceptional skills

There is a large body of mainly correlational research on the relationship between various measures of brain structure, function and activity and behavioural data. Performance has been linked to (a) electrocortical measures such as evoked potentials (Hendrikson & Hendrikson, 1980; Benbow & Lubinski (1993) and their components (McCarthy & Donchin, 1981), (b) hemispheric laterality data (Gazzaniga, 1985), (c) brain images (see Eysenck & Barrett, 1993), and (d) saccadic eye movements (Charlton, Bakan, & Moretti, 1989).

A number of correlates of high ability have been identified, including left-handedness, immune disorders, myopia (see Benbow & Lubinski, 1993), blood flow measures (Horn, 1986), neurohistology (Scheibel & Paul, 1985), prenatal exposure to high levels of testosterone (Geschwind & Behan, 1982), allergy, uric-acid levels, and glucose metabolism rates (see Storfer, 1990), and laterality (Eysenck & Barrett, 1993; Fischer, Hunt & Randhawa, 1982).

Sex differences in spatial abilities (Vandenberg, 1966; Humphreys, Lubinski & Yao, 1993) appear to contribute to sex differences in mathematical performance and are probably based on biological differences (Lytton & Romney, 1991; Collaer and Hines, 1995). Information-processing parameters involved in a number of human abilities, such as response speed, are at least moderately heritable (Bouchard, Lykken, McGue, Segal and Tellegen, 1990). Hereditary factors underlie various other individual differences in competence, such as working memory (Dark & Benbow, 1991). Enhanced ability to manipulate information in short-term memory has been observed in young people who are unusually successful in mathematics (Dark & Benbow, 1990). Moreover, since there are modest positive correlations between measures of special skills and heritable basic abilities such as general intelligence (Ackerman, 1988; Howe, 1989b), it is likely that some of the innate influences that contribute to variability in intelligence test scores also contribute to individual differences in special skills.

In general, the correlational evidence linking performance to brain characteristics suggests that innately determined biological differences do contribute to the variability of expertise in specific areas of competence. However, there is a large gulf between identifying neural correlates of behavioural differences and finding a neural predictor of talent. The relations between neural and performance measures are too weak to warrant conclusions about talent, and correlations diminish as tasks become more complex (Sternberg, 1993).

To provide support for the talent account, neural correlates of exceptional skills would have to be accompanied by (1) clarity about the direction of causality and evidence that the neural measure is (2) innately determined (rather than the outcome of differences in experience), (3) specific to an ability, and (4) selectively facilitates expertise in a minority of individuals. We are unaware of any neural measures that even come close to meeting these criteria. Nor has firm alternative evidence of early physical precursors of specific abilities emerged from studies of either pre-natal capacities or post-natal cognition (Hepper, 1993; Lecanuet, 1995; Papousek, 1995; Trehub, 1990).

Ericsson (1990; Ericsson & Crutcher, 1988) has argued that apparent indicators of structural precursors of ability may need to be interpreted with caution. He points out that individual differences in the composition of certain muscles are reliable predictors of differences in athletic performance, and that this fact has been widely held to demonstrate genetic determinants of athletic excellence. Ericsson notes, however, that differences in the proportion of the slow-twitch muscle fibres that are essential for success in long-distance running are largely the result of extended practice in running, rather than the initial cause of diferential ability. Differences between athletes and others in their proportions of particular kinds of muscle fibres are specific to those muscles that are most fully exercised in athletes' training for their specific specialisation (Howald, 1982).

Some individual differences in brain structure and function are the outcome of differences in experiences rather than being a primary cause. Experience can lead to changes in various parts of the mammalian brain, including the somatosensory, visual, and auditory systems (Elbert, Pantev, Wienbruch, Rockstroh & Taub, 1995). In violinists and other string players the cortical representation of the digits of the left hand (which is involved in fingering the strings) is larger than in control subjects. The magnitude of the difference is correlated with the age at which string players began instruction. Differences in early musical learning experiences may also account for the atypical brain asymmetries observed in musicians by Schlaug et al. (1995).

Although the evidence of a genetic contribution to human intelligence is consistent with the talent account, the correlations between general intelligence and various specific abilities are small (Ceci, 1990; Ceci & Liker, 1986, Howe, 1989c; 1990b; Keating, 1984). General intelligence need not limit final levels of achievement (Ackerman, 1988) and general intelligence may have little or no direct influence on specific abilities (Brynnner & Romney, 1986; Horn, 1986; Howe, 1989c). Moreover, there is no evidence of specific gene systems affecting high-level performance at special skills in the predictive and selective manner required by the talent account. Psychological traits are more likely to be influenced indirectly by genes in a probabilistic way (Plomin & Thompson, 1993). Even in the case of general intelligence, most of the research addresses the aetiology of individual differences in the normal range of ability. Relatively little is known about the genetic origins of high-level ability.

Knowledge about the genetic basis of specific high-level abilities is particularly limited (Plomin, 1988, Thompson & Plomin, 1993). In the Minnesota Study of twins reared apart self-ratings of musical talent correlated .44 among monozygotic twins reared apart, considerably less than the correlation of .69 for monozygotic twins reared together (Lykken, in press), suggesting that family experience makes a substantial contribution to self-ratings of musical ability. Similarly, in a study of musical abilities in twins, Coon & Carey (1989) concluded that among young adults musical ablility was influenced more by shared family environment than by shared genes. On a number of measures the correlations between dizygotic twins, which ranged from .34 to .83, were not much lower than those between monozygotic twins (.44 to .90).

The importance of general processing constraints diminishes as levels of expertise increase (Ackerman, 1988; Krampe & Ericsson, 1996), and some differences in basic skills are predictive of unskilled performance but less so of skilled performance (Ericsson, Krampe & Tesch-Römer, 1993). In Coon & Carey's study all eight relevant estimates of the heritability of musical ability were lower for participants who had taken some music lessons than for those who took no lessons at all; the average was less than .20 in the former group. Genetic differences that are initially relevant to expertise may be less important when large amounts of training and practice have been provided.

Date: 2014-12-29; view: 811