Comfortable Fictions: The Myth of Multiple Intelligences

Why moral certainty and pseudoscience keeps bad ideas alive in our schools

In the early 1980s, Howard Gardner’s Frames of Mind (1983) was explicitly written as a rebuttal to the psychometric theory of intelligence. He wanted to challenge the idea that a single numerical score could capture human intellect. For Gardner, intelligence was plural and context-dependent; musical, spatial, interpersonal, and so on. On the surface, this appears to be a far kinder approach to childrens’ abilities than the cold, harsh instrumentalism of the psychometrician. But the problem is not that Gardner’s idea is unkind; it’s that it isn’t true.

Curiously, the most prominent critic of the theory is Gardner himself. Yesterday, I saw a clip of him make a startling admission about his own theory. He claimed that it may not be scientifically true, but that he believes it should remain influential nonetheless.

I couldn’t quite believe what I was seeing, especially his abdication of the scientific method and phrasing of his opponents as “bad guys,” so I wanted to find the context of the quote to see if he was taken out of context. The full video makes clear that Gardner was indeed serious, and in fact, the wider context only makes the claim worse. The actual section where he makes the claim starts here.

He was describing a theme park in Denmark called Danfoss Universe that features an “explorama” with 50 different games, each purportedly based on a different intelligence. After encouraging his audience to visit this attraction, he stated:

“I defy anybody to spend a morning or an afternoon at Danfoss Universe and not develop an intuition of multiple intelligences and even if at the end of the day the bad guys turn out to be more correct scientifically than I am, life is short and we have to make choices about how we spend our time and that’s where I think that the multiple intelligences way of thinking about things will continue to be useful even if the scientific evidence doesn’t support it.”

Now I believe we should apply the principle of charity to any views we object to, and so to steelman Gardner’s statement here I can only assume that by “bad guys,” Gardner appears to use humour to refer to researchers who emphasise IQ and general intelligence, those he characterised earlier in his lecture as having a more traditional, psychometric view of intelligence. But given his previous stance on IQ and his tone and the substance of what follows, even this charitable reading cannot salvage the statement.

This is not the language of science. It is the language of ideological conviction and exemplifies so much of what is wrong with university education faculties which unfortunately has a disproportionate number of ideologues defending narratives they find morally preferable to empirical reality.

That this was said at Rockefeller University, one of the world’s premier biomedical research institutions, before an audience including leading neuroscientists, makes the abdication of scientific standards all the more lamentable. It’s a remarkable inversion of intellectual responsibility: where truth is subordinated to utility, evidence sacrificed for comfort.

Science means accepting the findings of well designed experiments whether you like them or not.

Comfortable fictions and Pseudoscientific Kindness

In his seminal 1992 essay, Vine Deloria Jr. used the term “comfortable fictions” to describe the reassuring stories societies tell to preserve moral comfort in the face of contradiction. In he argued that these narratives about history, identity, and progress allow dominant groups to evade the moral costs of their own actions.

Indeed, in his 1983 New York Review of Books essay “State of the Child,” Jerome Bruner reviewed Howard Gardner’s Frames of Mind and offered a characteristically shrewd appraisal. Bruner praised Gardner’s ambition and the book’s creative scope but agreed with Gardner’s own admission that his proposed “intelligences” were “at most useful fictions… sets of knowhow.”

That phrase comfortable fictions neatly captures I think, the enduring problem with multiple intelligences theory (and edu-myths in general) more precisely than any later sound bite. Gardner’s idea was not merely unverified; it was an imaginative framework presented as if it were an empirical one.

What he saw as “useful,” however, has since become uncomfortably durable. The theory’s fictions have outlived their usefulness, hardening into orthodoxy within education that often mistake humane intentions for scientific legitimacy. Gardner’s later insistence that MI should remain “useful even if the scientific evidence doesn’t support it” is simply the logical endpoint of that same mindset: the belief that moral appeal can stand in for empirical truth.

It’s also how other comfortable fictions like learning styles and a host of other neuromyths took hold. Each began with an intuitive appeal and a moral promise, that every child is unique, that everyone can be intelligent in their own way, and each ended up diverting attention and resources from what actually helps students learn.

This gets to the heart of the matter: there is a particular kind of delusion that afflicts well-meaning people in education: the belief that because their intentions are good, their methods must be sound. Gardner’s theory is morally appealing, it tells us that all children are gifted, just in different ways. It offers educators a flattering self-image: they are not sorting children by ability, but discovering and nurturing each child’s unique intelligence. Who could object to such a humane vision?

But moral appeal is not empirical validity. And the most dangerous ideas are often those that make us feel virtuous for believing them. When a theory aligns perfectly with our moral preferences, we should be more sceptical, not less. We should ask harder questions, demand better evidence, and resist the seductive pull of narratives that confirm what we already wish to be true.

The tragedy is that comfortable fictions like MI theory often do the most harm to the very children they claim to help. As Douglas Carnine noted in his essay on evidence-based education, “The greatest victims of these fads are the very students who are most at risk.” When schools spend resources on unproven MI teaching strategies instead of evidence-based interventions, when teacher training programmes devote hours to identifying students’ “intelligences” instead of teaching effective instructional methods, disadvantaged students pay the price.

Consider Mississippi’s literacy transformation between 2013 and 2019 (

Karen Vaites

’ analysis is a must-read on this). The state moved from near the bottom of national rankings to dramatic gains in reading and maths scores, outpacing nearly every other US state despite high poverty rates. These improvements were not the result of identifying children’s multiple intelligences or matching teaching to learning styles. They came from a deliberate, statewide adoption of evidence-based reading instruction grounded in cognitive science: systematic phonics, phonemic awareness, and explicit teacher training. The students who gained most were those from disadvantaged backgrounds, precisely the children who would have been most poorly served by well-intentioned but ineffective approaches.

This reveals the cruelty hidden within comfortable fictions. To tell struggling students that they are “spatially intelligent” or have “high bodily-kinesthetic intelligence” when what they actually need is clear instruction in reading, writing, and numeracy is not kindness. It is abandonment dressed up as affirmation. It allows educators to feel good about themselves whilst failing to give children the tools they need to succeed.

What is the theory and why is it wrong?

Gardner proposed that humans possess eight independent brain-based intelligences: linguistic, logical-mathematical, visual-spatial, bodily-kinesthetic, musical, interpersonal, intrapersonal, and naturalistic. Each intelligence, he argued, operates according to its own rules in relative autonomy from the others, supported by dedicated neural networks. As Gardner himself stated, “MI theory demands that linguistic processing, for example, occur via a different set of neural mechanisms than does spatial or interpersonal processing.”

The problem is straightforward: no such dedicated neural networks have been found. Despite Gardner’s claims that “each of the intelligences has its characteristic neural processes,” and his prediction that researchers should develop “an atlas of the neural correlates of each of the intelligences,” no one has actually demonstrated that these separate intelligences exist in the brain. This is not a minor oversight. It is the central empirical claim of the theory, and it remains unproven four decades after Gardner first proposed it.

Lynn Waterhouse identifies three fundamental problems with MI theory. First, the intelligences are not independent. Factor analyses, the statistical method used to determine whether proposed abilities actually function separately, have repeatedly shown that the intelligences are intercorrelated. When Visser and colleagues tested 200 individuals on measures of the eight intelligences, they found that scores were correlated “despite representing different domains of Gardner’s framework.” This directly contradicts Gardner’s assertion that there are independent intelligence domains. As Gardner himself acknowledged, if there is “a significant correlation among these faculties, the supposed independence of the faculties would be invalidated.”

Second, studies of MI teaching strategies suffer from profound methodological problems. The research claiming to demonstrate the effectiveness of MI-based instruction has not controlled for obvious alternative explanations: increased repetition of material, the novelty effect of new teaching methods, greater individual attention from enthusiastic teachers, or the simple excitement generated by novel activities. We know that repetition enhances learning, that novelty activates specific attention systems in the brain, and that emotional arousal at the time of learning strengthens memory consolidation. Any of these factors could explain the positive results attributed to MI teaching strategies.

Crucially, Ferrero and colleagues’ analysis found that most MI teaching studies lacked adequate sample sizes, appropriate control groups, or reliable outcome measures. The researchers concluded that sound studies were needed before “its use in the classroom can be recommended or promoted.” Yet MI strategies remain widespread, taught in teacher training programmes and implemented in thousands of schools worldwide.

Third, and most devastatingly, there are no empirical studies of the brain basis of MI theory. None. Despite Gardner’s repeated claims about dedicated neural networks for each intelligence, no researcher has actually looked for them using modern neuroimaging methods. The alignment studies that have been conducted, which simply match proposed intelligence functions to known brain regions, are not validity studies. They do not measure brain activity during tasks that supposedly characterise a particular intelligence.

The Brain Does Not Work That Way

When Gardner developed his theory in 1983, many neuroscientists believed the brain was organised in discrete modules, each dedicated to specific cognitive functions. In The Modularity of Mind (1983), the philosopher Jerry Fodor described cognitive modules as “domain-specific, innately specified, hardwired, autonomous,” and “informationally encapsulated” within a “fixed neural architecture.” Gardner borrowed the language of modularity but dispensed with Fodor’s empirical discipline, replacing a tightly constrained cognitive model with a free-floating metaphor for human potential.

This is no longer tenable. Research over the past four decades has demonstrated that the brain is organised in complex multifunction networks. The same neural regions that process language also process mathematics, music, logic, and social cognition. Anderson and colleagues used a meta-analytic approach to assess data from numerous functional imaging studies and found functional diversity at nearly every brain location. They argue that regions do not correspond to single cognitive domains like language or mathematics; rather, every region supports a variety of tasks, and complex functions emerge from networks of overlapping, reused neural circuits

None of this stopped Multiple Intelligences being wildly successful. By the late 1990s, Gardner’s Project Zero at Harvard had received tens of millions of dollars in philanthropic and public funding. Major funders included the Atlantic Philanthropies, the Spencer Foundation, the Pew Charitable Trusts, and the Rockefeller Foundation, many of which supported MI-inspired projects such as Arts PROPEL, Spectrum, and Key School in Indianapolis.

The Key School (Indianapolis), founded in 1988, was explicitly built around MI principles. It became a model for over 200 replication schools in the US and abroad. Each of those schools received local and state funding, plus private donations, to implement MI-themed curricula and staff training.

In a 2013 Education Week interview, Gardner said he was “astonished” by how far MI spread beyond his control, noting that “millions of dollars have been spent by schools, districts, and ministries of education” on training and materials based on his ideas, often in ways he hadn’t endorsed.

But the success of MI theory was moral before it was scientific. It promised to democratise intelligence, to dethrone the tyranny of IQ tests and recognise the many ways in which human beings can be capable, creative, and worthy.

Teachers embraced it not because it worked but because it felt right. It told them that the child who struggled with algebra might simply possess “musical” or “bodily-kinesthetic” intelligence instead. Everyone, in this worldview, is brilliant at something.

This, as it turns out, is not how the brain works.

The Cost of Comfortable Fictions

Gardner has spent much of his career oscillating between two incompatible positions: claiming that his intelligences have “dedicated neural networks,” and later insisting that MI theory was never intended to be a neurological claim. The contradiction matters, because his early work explicitly grounded the intelligences in the brain. He cited cases of brain damage as evidence of separable capacities and even proposed “an atlas of the neural correlates of each of the intelligences.” None of this materialised .

Life may be short, but students’ educational opportunities are shorter still. We cannot afford to waste them on theories that prioritise their creator’s preferences over their validity. The science matters. And until multiple intelligences theory meets the standards that every mature profession demands of its foundational claims, it deserves no place in our schools.

The danger of comfortable fictions extends far beyond education. Sadly, we see it in public life too, where pseudoscientific ideas now masquerade as moral crusades. The same impulse that claims to be “saving” children, now fuels movements that reject vaccines, climate science, and evolutionary theory. Far more damaging is RFK Jr.’s pseudoscientific claims about vaccines and the idea that Tylenol causes autism. From fad diets to the superstitions of celebrity healers, the pattern is the same: a deep distrust of expertise disguised as enlightenment.

The real danger isn’t that teachers believe in theories like Multiple Intelligences, it’s that they stop searching for what’s actually true. And if you think this is some fringe belief that most teachers don’t believe, then prepare yourself for the cold splash of water to the face that some 9 out of 10 teachers believe in the pseudoscience of learning styles. (A finding that by the way, has been replicated many times.) Imagine if 9 out of 10 doctors believed in healing crystals or bloodletting? When pseudoscience fills the space where evidence should be, genuine progress becomes impossible. The comfort of myth crowds out the discipline of truth.

This is not science. It is not even good pedagogy. It is moral theatre, performed for an audience of educators who want to believe they are doing the right thing without having to prove it. And the applause comes not from improved student outcomes, but from the warm glow of shared moral conviction.

The great danger of comfortable fictions is not that they are wrong, but that they make us feel right. They provide the satisfaction of moral certainty without the inconvenience of empirical accountability. They allow us to believe we are helping when we may be harming, to think we are progressive when we may be perpetuating the very inequalities we claim to oppose.

---

For any work related enquiries, please contact me here.