Howard Gardner

Who they are

The original list was seven intelligences, with a naturalist intelligence added later and an existential one floated and never confirmed. His stated evidence base is neurological, evolutionary, developmental and anthropological rather than statistical, and the criterion he leans on hardest is selective impairment: a capacity that can be knocked out on its own by brain damage has a claim to being a faculty in its own right. The theory became one of the most widely adopted psychological ideas in education of the late twentieth century.

The critique is real and it is the strongest test anyone has run. Visser, Ashton and Vernon, in Intelligence in 2006, built a battery of tasks intended to tap each of his intelligences, gave the whole thing to the same people, and found most of the tasks loading on a general factor rather than standing apart. Waterhouse, in Educational Psychologist the same year, argued there was no adequate empirical validation of the intelligences as distinct entities at all.

His reply is not a dodge and it cannot be lifted out of the verdict. He has said the theory was not built as a psychometric claim in the first place, so a psychometric battery is not the test it has to pass. He questions the construct validity of the general factor directly, in the other direction. And he argues that the batteries used against him are contaminated by verbal and logical-mathematical demands, so they are guaranteed in advance to recover a general factor whatever the truth of the matter. That objection is principled and it is testable, and the battery that would test it has not been built.

Where this site lands. The psychometric evidence for independent intelligences is weak, and the theory is not supported as a structural model of ability. Its real and lasting contribution was the pressure it put on a single-number account of what a mind is good at, and that pressure was worth applying whether or not the seven hold up.

One thing is routinely laid at his door that is not his. The learning-styles movement is conflated with multiple intelligences constantly, and he has objected to the conflation himself: an intelligence in his sense is a computational capacity, not a preferred channel for receiving information. The weakness of the learning-styles evidence is not evidence against him.

Key papers

His statement of the theory is a book rather than a paper, and the primary has not been read in this vault. Nothing is attributed to it here beyond authorship and the standard account of its argument.

Gardner, H. (1983). Frames of Mind: The Theory of Multiple Intelligences. The original seven intelligences, the criteria for admitting one, and the case from neurology, development and anthropology rather than from factor analysis.

The papers that tested it, both from 2006, and both of which belong next to his reply rather than on their own.

Visser, B. A., Ashton, M. C., & Vernon, P. A. (2006). Intelligence. A battery built to tap each intelligence and administered to the same participants. Most tasks loaded on a general factor.

Waterhouse, L. (2006). Educational Psychologist. Argues there is no adequate empirical validation of the intelligences as distinct entities.

Key ideas

Multiple intelligences: human ability is several relatively independent capacities rather than a single general one. The original list was seven, with naturalist added later.

Selective impairment as the admission criterion. A capacity that brain damage can remove on its own has a claim to being a faculty.

His standing reply, which must travel with any statement of the critique: the theory was not proposed as a psychometric model, and a psychometric battery loaded with verbal and logical-mathematical demands will recover a general factor whatever the truth of the matter.

The contribution that survived the verdict is the pressure the theory put on a single-number account of ability, in education above all.

Learning styles are not his and he has said so. An intelligence is a computational capacity, not a preferred input channel.

---