A review of:

The Origins and Development of High Ability

CIBA FOUNDATION SYMPOSIUM 178 (Chairman: R.C.Atkinson)
(Editors: G.R.Bock (Organizer) & Kate Ackrill)
Chichester, UK : Wiley DePublisher. pp.viii + 258. ISBN 0 471 93945 5. Price £45-45.

Convened at the suggestion of Michael Howe (Exeter University), CIBA's 1993 élite London conference on giftedness was a three-cornered contest. Tom Bouchard, Robert Plomin and researchers of mathematical talents argued for the importance of general intelligence (g) and of genetic factors. Anders Ericsson, John Sloboda and Howe himself urged the relevance of practice -- especially of intensive, systematic "deliberate practice". Avoiding both Galtonian essentialism and Watsonian behaviourism, a Thurstonian multi-dimensionalism was championed by Howard Gardner (7 separate intelligences -- including 'bodily-kinaesthetic') and Robert Sternberg (3 triarchic abilities -- analytical, creative, practical -- plus "general resource capacity" plus five conditions of being "labelled as gifted"). Fourteen short papers provided the rationale for discussions which were lively enough for three participants to offer their own houses as wagers on whether pre-1970 studies of learning in dogs and rats attested heritable g. However, wives seeking evidence for the divorce courts will have their work cut out due to poor indexing of the present volume of the proceedings.

The three corners have very different attitudes to and problems with evidence. The Thurstonians' problem is that no more than Thurstone himself do they have any quantitative evidence for the independent aptitudes that they feel moved to claim. Thus, despite his fifteen years of high-profile taxpayer-funded research on abilities, Gardner amazingly presents mere biographical sketches of seven turn-of-the-century geniuses (including Martha Graham) who happen to suit him; and Sternberg rejoices in correlations between abilities and attainments that are low only because of range restriction and the internal inconsistency of modern university gradings.

Unlike Howe, who engagingly admits to having had no luck with data, the Watsonian Ericsson has plenty of evidence on the trainability of at least the technical aspects of artistic performance. Unfortunately for readers, as presumably for conferencers, he does not present it here but leaves admirers to look up the Psychological Review for themselves. Still more unfortunately, his data on musicians are anyhow no more relevant to the issue of actual 'origins' than are Sloboda's: gifted youngsters will train harder than others just because they find it more rewarding.

For Galtonians, relevant evidence abounds; but there is a problem in extracting from it anything more than the high importance and heritability of g. Using MZ and DZ co-twins' regression to the mean, Plomin calculates an impressive 'group heritability' of .67 for being above IQ 120; and he properly rehearses the important discovery that g's heritability increases with age. Yet the psychogeneticists have still to exemplify convincingly just how genes affect or (occasionally) interact with the environment. Gifted people average about two standard deviations high in g (IQ 130), and Benbow and Lubinski find g is still a strong predictor of which of the gifted will do better than others; but more is needed to explain four-standard-deviation superiorities in specific attainments.

The latest development in g-cognizant theorizing has been the re-acceptance of the Spearman/Burt differentiation hypothesis. Correlations among mental tests are somewhat lower in higher-IQ than in average-IQ people (using similar IQ ranges). Spearman interpreted this "curiosity" as evidence for a restricting "law of diminishing returns" as to g's influence; and Mike Anderson's Intelligence and Development suggested high g simply allowed true levels of other abilities to be seen more clearly. Yet the phenomenon could equally attest a more 'developmental' hypothesis: as Doctor Johnson supposed, true genius might be "a mind of large general powers accidentally turned in some particular direction." In the most substantial and integrative chapter of the book, Douglas Detterman summarizes recent work on the differentiation hypothesis and reports further confirmation from the Kaufman-ABC standardization sample: plainly the very structure and developmental dynamics of intelligence will now yield only to researchers who use decent g ranges and g measures.

In line with the Johnsonian idea, and in contrast with the dismal record of Headstart programmes with low-IQ children, Julian Stanley and William Fowler both find that accelerative programmes really kick-start the high-IQ. Undoubtedly the next move must be to see whether giftedness involves greater differentiation (e.g. larger standard deviations) in personality variables (e.g. emotionality) that multiply with g to yield the strongly skewed distribution of high human gifts. Sadly, with the exception of Nicholas Colangelo's fascinating revelations of the Pollyanna-ish conscientiousness and good-naturedness of 34 Iowa inventors (including 3 millionaires and no divorcées), personality and motivation achieve all too little coverage here amongst psychologists still fighting like cats in a sack about what has actually been a sensational victory for the London School.

CHRIS BRAND
DEPARTMENT OF PSYCHOLOGY, UNIVERSITY OF EDINBURGH

A review by Chris Brand of Camilla Benbow & David Lubinski's Intellectual Talent -- showing the advantages of fast-track-learning for high-g children -- will appear in Personality & Individual Differences 1998/9.

For more on Howard Gardner, Robert Sternberg and other critics of London School views of IQ, see The g Factor and PERSONALITY, BIOLOGY AND SOCIETY (Sections 8-11).

Last modified: 3 xii 1997