Multiple intelligences (MI) is Howard Gardner's claim, set out in Frames of Mind (1983), that human intellect is not one general capacity but a set of relatively independent faculties, each with its own profile in any given person. Against the idea of a single IQ-style ability, Gardner proposed distinct intelligences that vary across individuals, so a person can be strong in one and ordinary in another. The theory became enormously popular in schools, where it was read as a charter for valuing more than verbal and mathematical talent.
Gardner originally named seven intelligences and later added an eighth: linguistic, logical-mathematical, spatial, musical, bodily-kinesthetic, interpersonal (reading other people), intrapersonal (reading oneself), and naturalistic (discerning patterns in the living world). What made the proposal more than a list was the set of eight criteria a candidate ability had to satisfy to count as an intelligence: potential isolation by brain damage; the existence of savants and prodigies; an identifiable core operation or set of operations; a distinct developmental trajectory toward expert end-states; an evolutionary history; support from experimental psychology; support from psychometric findings; and susceptibility to encoding in a symbol system. The criteria draw on neurology, biology, and anthropology rather than test scores alone, which is both the theory's reach and, critics argue, its escape hatch.
Gardner himself has drawn a firm line that the classroom often erases. Writing in The Washington Post (2013), he objected to being credited with "learning styles" and to the routine conflation of the two. MI is a claim about the content of mind, the kinds of information a person can process well; a learning style is a claim about a person's manner of processing any content. The two answer different questions, and Gardner regards the matching prescriptions attached to styles as no part of his theory.
The recurring objection is that MI has been adopted on its appeal rather than its support. Lynn Waterhouse (2006, Educational Psychologist 41(4), 207–225) reviewed the literature and found no published studies validating MI as a set of distinct intelligences. Strikingly, the first empirical test built specifically around Gardner's eight domains, Beth Visser, Michael Ashton, and Philip Vernon (2006, Intelligence 34, 487–502), did not appear until 23 years after the theory. Giving 200 adults two tests per domain, they found a strong general factor, the same g that MI was meant to displace, with substantial loadings for the purely cognitive domains (linguistic, logical-mathematical, spatial, naturalistic, and interpersonal) and weaker loadings for others, bodily-kinesthetic most of all. The domains that look most like conventional cognition behaved like conventional cognition, correlating through g rather than standing apart as independent intelligences.
This drives the conflation-with-ability charge. Where Gardner's intelligences are measured with cognitive tests, they correlate, which is what a single underlying ability predicts and what an array of truly separate faculties does not. Critics including Robert Sternberg add that several of the eight, musical and bodily-kinesthetic in particular, are better described as talents than as intelligences, so the theory may have relabeled the ordinary range of human abilities rather than discovered a new architecture of mind.
A deeper worry is falsifiability. By spreading the evidence across eight heterogeneous criteria and declining to tie each intelligence to a definite measurement, the theory makes itself hard to disprove, and a claim that cannot in principle fail a test sits uneasily as science. Gardner has answered that MI was never meant to be reduced to a battery of psychometric tests, that he is skeptical of the g tradition on its own terms, and that the theory is a reorganisation of evidence from many disciplines rather than a single testable hypothesis. That defence preserves the theory's breadth at the cost of the sharp empirical confirmation its classroom popularity assumes.
Treat MI as a humane corrective, not a validated mechanism. Its real service was widening what counts as worth cultivating, pushing back on a curriculum that prized only the verbal and the mathematical, and that broadening is defensible on educational grounds whatever the science says. What is not supported is diagnosing a learner's "dominant intelligence" and routing instruction through it, or teaching the same content eight ways on the theory that each taps a different intelligence, since the evidence for independent intelligences with distinct optimal channels is not there. Plan instead from the demands of the material and from real, measurable differences among learners, the province of differentiation and of language aptitude in the language classroom. Keep MI distinct from learning styles, which Gardner disowns, and from the general-ability g factor it set out to replace but which keeps reappearing in the data.