Why is this of any import in the current clime? Of late, the most popular theory of human abilities has been the Multiple Intelligence (MI) view of Gardner (1983). The original presentation of the theory of multiple intelligences artfully combined information from a half dozen fields into an engaging set of essays on the initial set of seven intelligences. Each account was liberally illustrated with events from the lives of famous individuals or clinical cases. Indeed, insightful psychobiography seems Gardner’s particular forté.
MI theory has always been more popular with lay readers and practitioners than with either cognitive or differential psychologists. Cognitive psychologists were troubled by the theory’s extreme modularity, particularity the claim that a central working memory was unnecessary (Messick, 1992). Rather, each “intelligence” was seen as having its own working memory. This may sound like a relatively minor point, but it is not. Working memory is a pivotal construct in all models of cognition. Theorists debate how working memory functions, and what sort of modality-specific slave systems it might have. But most do not debate whether it is a useful or needed construct.
Differential psychologists, on the other hand, were troubled by Gardner’s dismissal of 80 years of research on the organization of human abilities (Carroll, 1993; Gustafsson & Undheim, 1996). The reason given is that, as Gardner sees it, evidence for G is provided almost entirely by short-answer multiple choice, paper-and-pencil tests of the sorts of linguistic and logical intelligence that are at best useful for predicting success in the narrow domain of conventionally structured schools (Gardner, 1993, p. 39). Yet even the most cursory examination of human abilities literature shows that every one of these claims at best overstates and at worst is simply false. But there is enough truth to them, and enough dissatisfaction with standardized tests of all sorts that even those who were troubled by the claims were willing to wait and see what sort of evidence would be produced by the new assessment procedures Gardner advocated. They are still waiting.
Lessons learned from performance assessments in other areas of education will probably hold here. We have found that so-called authentic tests have a beneficial effect on the curriculum and can measure aspects of knowledge and skills not tapped by surrogate measures. But we also have discovered that there is vastly more overlap with conventional tests than difference from them, that performance measures are less reliable, more time consuming to administer and score, and vastly more expensive. Most importantly, there is little reason to think that performance assessments show no overlap with each other, and thus support the notion of an independent set of intelligences.
Gardner’s theory appeals to teachers and parents because it reinforces the idea that at root all children are special, and that giftedness is a multidimensional—not undimensional—affair. Surely both of these are noble goals. However the theory also appeals the perverse human tendency to think categorically rather than probabilistically. To see two types or seven types of people in the world is surely an advance over seeing one type. But typologies mislead more than they lead when the underlying structure in the domain is not categorical. Indeed, the critical failure of Gardner’s theory is not just that it fails to acknowledge or explain why abilities are correlated, but why this correlational structure is implies a hierarchy. In other words, it does not explain why some “intelligences” are more intelligent than others.
