The debate between general intelligence (g) and Howard Gardner's theory of multiple intelligences isn't just academic. It shapes how schools teach, how organisations hire, and how people think about their own capabilities and limitations. Both frameworks describe something real about human cognitive variation, but they describe it differently and with different implications. Understanding what each actually claims — as opposed to the popularised caricature of each — is essential for thinking clearly about intelligence.
What IQ Measures and What g Is
IQ tests measure performance on a range of cognitive tasks — verbal reasoning, numerical reasoning, spatial reasoning, working memory, processing speed. The observation that performance on all these tasks tends to correlate positively across people led Charles Spearman to propose a general intelligence factor, which he called g, in 1904. Someone who does well on verbal reasoning tasks tends also to do better than average on numerical reasoning, spatial tasks, and so forth. This positive manifold — the pattern of correlations — is one of the most replicated findings in differential psychology.
g is not the same as IQ score. IQ is a norm-referenced score derived from tests; g is a statistical construct extracted from the pattern of correlations between different cognitive tests. g predicts a striking range of life outcomes — academic achievement, job performance across different domains, health outcomes, and longevity — with effect sizes that are modest but consistent across many studies.
The evidence for g as a meaningful construct is strong. It has substantial heritability, correlates with brain imaging measures, and shows relative stability from childhood through adulthood. Psychologists who work with cognitive ability data broadly accept g as a real and important dimension, whatever arguments persist about its interpretation.
Gardner's Theory: What He Actually Claimed
Howard Gardner proposed his theory of multiple intelligences in Frames of Mind (1983). He argued that human cognition is not well-described by a single general factor but instead involves at least seven (later expanded to eight or nine) relatively independent intelligences: Linguistic, Logical-Mathematical, Spatial, Musical, Bodily-Kinaesthetic, Interpersonal, Intrapersonal, and Naturalist.
A critical point that often gets lost: Gardner was not making a psychometric claim. He was explicitly not claiming to have identified separate measurable factors that replace g in a statistical model. His theory is more philosophical and neuropsychological — he argued that the intelligences meet certain criteria including specific neural localisation, the existence of populations with isolated strengths or deficits (savants, people with specific brain injuries), a distinct developmental history, and cultural value across different societies.
Gardner himself has been critical of the "learning styles" applications of his theory, which typically caricature it as "some people are visual learners, others are auditory." That's not his claim. His claim is that different cognitive capabilities have different profiles of support and development, not that information is best processed in the modality that matches a learner's "type."
The Empirical Critique of Multiple Intelligences
Mainstream psychometrics has been largely sceptical of Gardner's theory for a specific reason: when the multiple intelligences are operationalised and measured, they tend to intercorrelate positively — which is the pattern that indicates g, not independence. If the intelligences were truly separate, knowing someone's linguistic intelligence would tell you nothing about their logical-mathematical intelligence. In practice, they predict each other substantially.
Gardner's theory also hasn't generated the body of empirical research that the g framework has. It's difficult to measure the intelligences in a psychometrically rigorous way, and the theory's criteria for what counts as an intelligence are sufficiently loose that almost any human capability could qualify. The theory has generated many educational programmes and much enthusiasm but relatively few testable predictions.
What Each Framework Is Actually Good For
These critiques don't mean the multiple intelligences framework is worthless. They suggest it does different things from g:
g is a better predictor. For predicting academic and occupational performance, g and IQ scores are among the best predictors we have. The predictive validity is real and robust across very different populations and domains.
Multiple intelligences is a better describer. For describing and acknowledging the range of human capability that formal academic tests don't measure — spatial creativity, kinaesthetic skill, musical ability, interpersonal skill — Gardner's framework provides useful vocabulary. It's more useful for describing what a person does well than for predicting their performance on any particular cognitive task.
The political and motivational dimensions are real. Decades of experience in education suggest that telling children they're "not intelligent" based on g-type measures has real negative effects on motivation and self-concept. Gardner's framework, whatever its psychometric limitations, supports a more expansive view of human capability that has genuine motivational value. This doesn't make it scientifically equivalent to g theory; it means the two frameworks serve different purposes.
To see how your own cognitive profile sits across multiple reasoning domains, our free multiple intelligences test gives a detailed profile across all eight areas.
Frequently Asked Questions
Is Gardner's theory of multiple intelligences scientifically valid?
As a framework for educational practice and describing human capability broadly, it has value. As a psychometric claim that there are eight independent intelligence factors replacing general intelligence, it's not supported by the pattern of cognitive test correlations, which show positive relationships consistent with g. Gardner himself distinguished his theory from psychometric intelligence research.
What is the difference between IQ and multiple intelligences?
IQ is a norm-referenced score from standardised cognitive tests that reflects largely general intelligence (g). Multiple intelligences is a theoretical framework proposing eight relatively independent cognitive capabilities. They make different claims and serve different purposes: IQ predicts performance outcomes; multiple intelligences describes the range of human cognitive variation more broadly.
Can someone be high in one intelligence and low in others?
Evidence from savants and people with specific brain injuries shows that some cognitive capabilities can be selectively preserved or impaired. But in typical populations, the correlations between different cognitive abilities are positive — someone with strong verbal ability tends to have above-average logical-mathematical ability. The "isolated intelligences" are more visible at the extremes than in typical range.
Does IQ capture all important mental abilities?
No. IQ captures g well and predicts a range of academically and occupationally relevant performance. It doesn't directly measure creativity (as typically defined), practical social intelligence, emotional processing, domain-specific expertise built through practice, or many of the capabilities Gardner's framework includes. The predictive validity of g doesn't mean it's a complete description of human cognition.
What does neuroscience say about multiple intelligences?
Neuroscience shows that different cognitive functions rely on different neural networks, which supports the idea that cognition is not unitary. However, the networks involved in different "intelligences" interact substantially — the brain doesn't operate in cleanly isolated modules corresponding to Gardner's eight categories. Neuroscience evidence is consistent with both a general factor and multiple relatively independent components; it doesn't definitively resolve the debate.