Are There Multiple Intelligences in the Human Mind?

Russell T. Warne

Jun 5, 2025

…I argue that there is persuasive evidence for the existence of several relatively autonomous human intellectual competencies, abbreviated hereafter as “human intelligences.” ...the conviction that there exist at least some intelligences, that these are relatively independent of one another, and that they can be fashioned and combined in a multiplicity of adaptive ways by individuals and cultures, seems to me to be increasingly difficult to deny. (Gardner, 2011, pp. 8–9)

Howard Gardner’s Frames of Mind: The Theory of Multiple Intelligences, published originally in 1983, is one of those works, like Sigmund Freud’s The Interpretation of Dreams or B. F. Skinner’s Walden Two, that has seeped into the wider culture and pop psychology. Even people who have never read Frames of Mind know of the theory of multiple intelligences and may identify themselves as having, for example, high logical-mathematical intelligence, low intrapersonal intelligence, or high bodily-kinesthetic intelligence. In almost all introductory psychology textbooks Gardner’s theory of multiple intelligences is summarized (Warne et al., 2018), and Frames of Mind is one of the most commonly cited works in those textbooks (Griggs, Proctor, & Cook, 2004).

Despite the popularity of Gardner’s theory, it is not a viable theory of human cognitive abilities because of two major types of problems. The first problem is empirical, where Gardner’s theory does not find support in the data from psychological research on cognitive abilities. The second is that the theory has fundamental flaws in its logic and construction that prevent it from being a useful scientific theory. I will explore these issues in this chapter.

Description of the Theory of Multiple Intelligences

Before discussing criticisms of Gardner’s work, it is important to summarize the theory of multiple intelligences. Originally, Gardner posited that there were seven intelligences, which were:

1. Linguistic intelligence: the ability to show dexterity in oral and/or written language.

2. Musical intelligence: the capacity to learn, perform, create, and/or interpret music.

3. Logical–mathematical intelligence: the skill of dealing with logical systems.

4. Spatial intelligence: the ability to handle stimuli in two or three dimensions.

5. Bodily-kinesthetic intelligence: the capability of using one’s body in culturally useful ways (e.g., in athletics or artistic expression).

6. Interpersonal intelligence: the skill in dealing with others’ behavior, such as in a leadership situation.

7. Intrapersonal intelligence: the capacity to engage in self-reflection to better understand one’s self.

In 1999, Gardner combined the interpersonal and intrapersonal intelligences and added the naturalistic intelligence, which is the ability to understand and have empathy for objects in nature. From time to time, Gardner has suggested possible additional intelligences, such as an existential intelligence (Gardner, 1999). Some adherents of multiple intelligences theory only count the original seven, while others include the naturalistic and/or existential intelligences. Depending on which iteration of Gardner’s theory proponents prefer, they may say that there are as few as seven or as many as nine intelligences. Gardner also recognizes that there may be other, heretofore undiscovered intelligences (von Károlyi, Ramos-Ford, & Gardner, 2003).

The changing number of intelligences in the theory is not necessarily a bad thing; good scientists change their opinions and alter their theories as new information comes to light. Some people (e.g., Ritchie, 2015) argue that Gardner invents intelligences arbitrarily, but actually he has specific criteria for when an ability warrants being called an “intelligence.” These criteria for the existence of an intelligence are (1) brain damage that impacts the intelligence but no other abilities; (2) the existence of savants, prodigies, and eminent individuals in an area; (3) central operations that must be executed to work in that intelligence; (4) a developmental trajectory starting in childhood with a possible endpoint of expertise; (5) a plausible evolutionary theory of how the intelligence formed in humans; (6) support from experimental psychology; (7) evidence from psychological testing; and (8) a symbol system in which a person can demonstrate their accomplishment in the intelligence.

As described in Frames of Mind, the theory of multiple intelligences posits that there are a number of “relatively independent” mental abilities that have separate biological foundations and manifestations. Gardner (2011, pp. xxxix, 337) explicitly rejects the concept of g and argues that it is an illusion resulting from how intelligence tests are created. Instead, Gardner believes that these intelligences can – and often do – work at the same time (von Károlyi et al., 2003). For example, when writing a song, one may draw upon musical intelligence for composing the music and linguistic intelligence for writing the lyrics. But Gardner argues that this is merely different modules in the brain operating simultaneously, and not evidence of the existence of a global intelligence.

Empirical Problems with Gardner’s Theory

If the theory of multiple intelligences is correct, then it should be supported by empirical data. To an extent, it is. Gardner cites hundreds of sources in Frames of Mind from research in education, psychology, anthropology, and other areas as supporting evidence of the existence of his intelligences. The problem is that Gardner habitually cherry picks evidence in his favor and ignores evidence that contradicts his theory (Bouchard, 1984; Messick, 1992; Scarr, 1985; Snow, 1985). Indeed, much of the evidence that Gardner cites is ambiguous and can support a variety of theories. For example, Gardner (2011, Chapter 5) has an excellent description of how language develops in childhood and becomes more complex as individuals age. This developmental trajectory can fit into many theories and does not prove the theory of multiple intelligences true, nor does it disprove any other theory.

When Gardner does confront the evidence that undermines his theory, he dismisses it, a tendency best shown in the following passage:

Several critics have reminded me that there are generally positive correlations (the so-called positive manifold) among tests for different faculties (for example, space and language). More generally, within psychology, almost every test of abilities correlates at least a little bit with other tests of ability. This state of affairs gives comfort to those who would posit the existence of “general intelligence.” I cannot accept these correlations at face value. Nearly all current tests are so devised as to call principally on linguistic and logical faculties. Often the very wording of the question can tip off the test takers. (Gardner, 2011, p. xxxix, paragraph break eliminated; see also von Káolyi et al., 2003, p. 100)

Whether Gardner can “accept these correlations” or not does not change the fact that the correlations exist, regardless of the culture, tasks, and examinee population (see Chapters 1, 2, and 4).

Moreover, neither of Gardner’s statements about intelligence tests is true. How a test of block span (where an examinee must touch a sequence of blocks in the same order that the examiner touched them) or reaction time would tap into linguistic and logical faculties is not clear. Additionally, the wording of items does not “tip off the test takers” in regards to the correct response. Indeed, for most item formats (e.g., vocabulary, matrix reasoning, digit span), it is not clear how the wording of the items can give hints that advantage some examinees over others.

Despite Gardner’s beliefs, the reality is that even when researchers attempt to measure the multiple intelligences, the result is a series of correlated variables that produce a general factor (e.g., Castejon, Perez, & Gilar, 2010; Pyryt, 2000). A g factor emerges from these scores, even though that is exactly what should never occur, according to the theory of multiple intelligences. Yet it does anyway.

Frames of Mind also has the drawback of relying heavily on anecdotal data. Case studies of people with brain injuries, ethnographies of non-Western cultures, and quotes from biographies of famous scientists and artists are common in Frames of Mind. What Gardner does not have is large, representative samples of individuals performing tasks grounded in his intelligences (Lubinski & Benbow, 1995). While many of the stories in Frames of Mind are vivid, case studies are inherently limited in their ability to provide scientific data. Loftus and Guyer (2002) explained the limited nature of case studies well:

Case studies therefore illuminate, but can also obscure, the truth. In many cases, they are inherently limited by what their reporter sees, and what their reporter leaves out. This is especially true if the writer is untrained in the scientific method, and thus unaware of the confirmation bias, the importance of considering competing explanations before making a diagnosis and so forth. To the scientist, therefore, most case studies are useful largely to generate hypotheses to be tested, not as answers to questions. (p. 26)

Most of Gardner’s examples of individuals with a high level of a particular intelligence come from non-scientific sources, such as biographies. There is also no evidence that Gardner sought case studies or examples of individuals who excel in more than one intelligence – the sorts of individuals who would provide evidence undermining his theory. Gardner discusses T. S. Eliot (linguistic intelligence) and Igor Stravinsky (musical intelligence). But he does not discuss Hedy Lamarr, who is best known as an actress (interpersonal intelligence), but was also an inventor (spatial and/or logical–mathematical intelligences). As another example, there is also no mention of Winston Churchill, one of the greatest leaders of the twentieth century (requiring high interpersonal intelligence), but whose Nobel Prize was in literature (an undeniable display of high linguistic intelligence). If Gardner’s theory were true, then people who are eminent in multiple areas should be as rare as a Leonardo da Vinci because these intelligences would be uncorrelated, making multiple high abilities in the same person exponentially more unusual than high abilities in just one area. But people with high abilities in multiple areas are not as rare as Gardner claims; many eminent people in history had high achievement in multiple areas (White, 1931), and most science Nobelists are more accomplished in non-scientific fields than the general public (Root-Bernstein et al., 2008). With some knowledge of history and biography, one could create a lengthy list of eminent individuals who have achieved prominence in more than one intelligence (e.g., William Blake, Julius Caesar, Sir Francis Galton, Florence Nightingale, Lewis Carroll, Stephen Sondheim, Galileo Galilei).

Another problem with Gardner’s examples is that they are all people who have high general ability – a minimum IQ of 120 by his own estimate (Jensen, 1998, p. 128). Thus, to be highly eminent in one of Gardner’s intelligences requires high general intelligence. This is not surprising; for nearly a century, psychologists have known that scores on tests are correlated less strongly among high-IQ samples than among low-IQ samples (Spearman, 1927, pp. 217–221; see te Nijenhuis & Hartmann, 2006, and Tommasi et al., 2015, for examples of this phenomenon from the twenty-first century). If they do exist, the multiple intelligences seem to be the playground of high-g people.

Another problem with Gardner’s anecdotes is his emphasis on eminent individuals, which makes the intelligences appear distinct and uncorrelated with each other. But when studying the entire population, these abilities are clearly correlated. If one considers – as Gardner (2011) does – these individuals in isolation, then the vast differences between composing a symphony and writing a clever poem become clear, and it does appear that Igor Stravinsky and T. S. Eliot have very different abilities. But when one considers the entire population, then it is apparent that musical ability and verbal accomplishment are positively correlated and not as independent as Gardner believes. (Indeed, in Spearman’s 1904 study, the two abilities were correlated r = .51.)

It is well known that studying outliers and other eminent individuals exaggerates differences. This is apparent when imagining space aliens who abduct an Olympic gymnast, weight lifter, and basketball player. The aliens would see huge differences in these people’s training regimens, body types, physical capabilities, and diets. These differences might be so large that the aliens would find it difficult to classify all of them as athletes. But in a more general population of athletes – ranging from weekend warriors and kiddie league participants to elite professionals – the similarities in abilities would be more apparent (Lubinski, 2004). Thus, Gardner erred when considering only outliers because this masked the similarities among individuals.

Eminent individuals are not the only outliers Gardner builds his theory on. He also uses savants (i.e., individuals with low IQ scores but who have extreme abilities in one area, such as artistic production, piano playing, or mathematical computation) as evidence that an intelligence can be independent of other cognitive abilities. However, Gardner oversimplifies this evidence; while savants can sometimes have amazing abilities, they do not display all facets of any of Gardner’s intelligences. For example, a musical savant may be able to play a piano piece after hearing it a single time, but a musical savant cannot compose, improvise, transpose, arrange, orchestrate, or engage in other high-level displays of musical ability (Klein, 1997). Savants seem to have a prodigious memory, which is scientifically interesting, but not evidence for Gardner’s theory.

Theoretical Problems with Gardner’s Theory

The empirical problems with Gardner’s theory are enough to disprove it, but the theoretical problems associated with his theory of multiple intelligences are also a serious threat to the theory. These problems show that the theory lacks important characteristics of useful scientific theories, especially because of the theory’s (a) vagueness, (b) incoherence, and (c) inability to make new predictions. I will address each of these ideas in turn.

One of the essential characteristics of a scientific theory is that it has to be specific enough to test. Unfortunately, Gardner’s theory of multiple intelligences is too vague for any scientific purpose (Jensen, 1998). How uncorrelated do abilities have to be in order for them to be “relatively independent of one another”? Is it r = .40? .25? .10? 0? Gardner never says. Additionally, Gardner (2011) recognizes that intelligences can be used in tandem and that few human activities require just one intelligence. But he never specifies how to distinguish between multiple intelligences operating simultaneously and a general ability that functions in different domains. Gardner also never explains how the intelligences can work together if they are “relatively independent” and do not overlap neurologically (Klein, 1997).

What is most staggering is that Gardner acknowledges that his theory cannot be tested like a real scientific theory. For example, he stated, “I’ve never felt that MI theory was one that could be subjected to an ‘up and down’ kind of test, or even series of tests. Rather, it is and has always been fundamentally a work of synthesis” (Gardner, 2011, p. xix). And elsewhere: “I do not believe that educational programs created under the aegis of MI theory lend themselves to the kinds of randomized control studies that the US government is now calling for in education” (Gardner, 2011, p. xxi). How convenient.

A second – and related – problem is that the theory is not coherent. One way the theory is incoherent is in its circular reasoning of how to identify intelligences. As an example, Gardner (2011, p. xxxiv) stated:

There is no “pure” spatial intelligence: instead, there is spatial intelligence as expressed in a child’s puzzle solutions, route finding, block building, or basketball passing. By the same token, adults do not exhibit their spatial intelligence directly but are more or less proficient chess players or artists or geometricians. Thus, we are advised to assess intelligences by watching people who already are familiar with and have some skills in these pursuits.

In other words, it is only possible to identify an intelligence once it has been developed (von Károlyi et al., 2003). Consequentially, identifying an intelligence becomes a circular process that goes something like this:

“Why does Anne perform well in chess?”

“Because she has high spatial intelligence.”

“How do we know Anne has high spatial intelligence?”

“Look at how good she is at chess!”

This circular definition is an “explanation” for Anne’s competence in chess that explains nothing and is too incoherent to function as a meaningful scientific theory. It is also untestable, to boot.

Another incoherence in Gardner’s theory is why the abilities he emphasizes must be “intelligences” at all. He recognized this ambiguity and explained that he chose the term because he wanted “to replace the current, largely discredited notion of intelligence as a single inherited trait” (Gardner, 2011, p. 300). However, by including physical abilities (in bodily-kinesthetic intelligence), personality traits (in interpersonal intelligence) and other non-cognitive traits, Gardner has stretched the word “intelligence” so much that it ceases to have any real meaning (Hunt, 2011; Jensen, 1998; Scarr, 1985). Gardner sees g as being a narrow concept that encourages “a limited view of intelligence” (von Károlyi et al., 2003), but believing in the existence of g does not preclude the existence of other abilities. In fact, both the CHC and bifactor models explicitly recognize the existence of non-g cognitive abilities (as this book’s Introduction makes clear).

Finally, any useful scientific theory must be able to make predictions about phenomena or – in the case of the social sciences – individuals. Multiple intelligences theory is unable to do this (Hunt, 2011). One reason is that Gardner never creates a feasible plan for assessing the intelligences in his theory; indeed, he has stated that he does not endorse any psychological test to measure the multiple intelligences (von Károlyi et al., 2003). In Frames of Mind he spends only a few pages (2011, pp. 404–406) explaining how to measure the intelligences. These tips are extremely impractical to implement on a large scale. For example, Gardner suggests giving children a wide variety of materials from different intelligence domains (e.g., puzzles, a computer, musical instruments) and then examining which areas a child excels in. Gardner estimates that this endeavor will take 5–10 hours per child over the course of a month. Multiply that assessment time by the number of children in an elementary school and the result is a time commitment that is too impractical for real-world implementation. Without a practical method of assessing the intelligences accurately, there is no way to predict who will be successful in a drama program, a basketball team, the math Olympiad, or a pottery class.

In later works Gardner developed more sophisticated ideas – such as using portfolios of rating scales – about how to measure the intelligences (e.g., von Károlyi et al., 2003), but these methods have profound deficiencies. Lubinski and Benbow (1995) explained that these new procedures were susceptible to producing inconsistent data (i.e., “low reliability,” in the technical jargon; see Chapter 7) and that there was no evidence that these scores had any educational utility. Klein (1997) argued that these sorts of assessments have the same problems as traditional intelligence tests – except now the problems exist for a series of tests for the intelligences, instead of a single test that measures g. Again, without meaningful data, the ability to make predictions based on multiple intelligences theory is nonexistent.

Gardner dismisses these concerns:

In speaking of measurement, I touch on the issue about which psychologists interested in intelligence have spilled the most liquid or electronic ink. Having put forth the theory, they maintain, I should be required to test it and, on the basis of the results of those tests, either revise or scuttle the theory. In their view the fact that I’ve elected not to become a psychometrician is no defense! (Gardner, 2011, p. xix)

But it is no defense. Without a coherent method of measuring a trait, it is impossible to study the trait thoroughly, make predictions based on the trait, and test the theory (Hunt, 2001; Waterhouse, 2006). The fact that a theorist does not want to define how a trait is measured does not excuse them from that duty because other scientists will never completely know whether they are truly measuring the theorized trait. Moreover, because the intelligences cannot be measured until they are already developed, the theory is utterly useless at predicting which children will profit from music lessons, need additional instruction in math, or have difficulty reading. This is the sort of information that teachers and parents want that Gardner’s theory will never be able to provide.

Conclusion

Despite its popularity in the education establishment, Howard Gardner’s theory of multiple intelligences has fundamental theoretical problems that make it incoherent, untestable, and unable to generate predictions. Moreover, empirical evidence is overwhelming that different cognitive abilities are not independent of one another and that a general mental ability – g – exists. I agree with Hunt (2011, p. 119), who stated that “there is virtually no objective evidence for the theory” of multiple intelligences.

Perhaps because the theory of multiple intelligences and its creator are so identified with one another, it has become impossible to avoid criticizing Howard Gardner’s behavior regarding the theory. I disapprove of Gardner’s resistance to testing his theory and his blasé dismissal of unfavorable evidence, both of which are not how scientists should behave with regard to their theories. One final quote encapsulates his attitude well:

And even if at the end of the day, the bad guys [who advocate for a general intelligence] 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 the multiple intelligences way of thinking about things will continue to be useful, even if the scientific evidence isn’t supportive.

(Gardner, 2009, 0:45:11–0:45:32)

I suppose that if a person wants to make studying multiple intelligences theory their pastime, then there is no harm in that. (Most hobbies don’t have a scientific basis.) But the theory of multiple intelligences lacks empirical support and a coherent theoretical foundation. Therefore, in situations where it could impact people’s lives – like in education and in scientific research – it should be completely abandoned.

From Chapter 5 of "In the Know: Debunking 35 Myths About Human Intelligence" by Dr. Russell Warne (2020)

We hope you found this information useful. For further questions, please join our Discord server to ask a Riot IQ team member or email us at support@riotiq.com. If you are interested in IQ and Intelligence, we co-moderate a related subreddit forum and have started a Youtube channel. Please feel free to join us.