9 Types of Intelligence, Ranked by Science

You've seen the poster. Nine types of intelligence, each with its own little icon: word smart, number smart, picture smart, music smart, body smart, people smart, self smart, nature smart, and one for life's big questions. It hangs on classroom walls, it fills a hundred YouTube videos, and it tops most searches for "types of intelligence".

Almost every version presents the nine as settled fact. Very few mention that the man who drew up the list has said he never ran experiments designed to test it. So we did the obvious thing. We gave every type the same three tests from intelligence research, scored them all the same way, and put them on a tier list.

Only three of the nine pass all three tests. The other six aren't fake. Most of them are real talents, real skills or real parts of personality. They just aren't intelligences. And emotional intelligence, which usually gets added to the list as a tenth, walks in as one idea and leaves as two.

Where the nine types came from

The list comes from one person. Howard Gardner, a psychologist at Harvard, set out his theory of multiple intelligences in the 1983 book Frames of Mind. It began with seven intelligences. The naturalist came later, and in Intelligence Reframed (1999) he weighed up a ninth, existential intelligence, the one about life's big questions.

Teachers loved it, and it's easy to see why. It says every child is smart in some way, and it gives a teacher nine doors into a lesson instead of one.

What the poster leaves out is how the theory was built. Gardner assembled it from other people's findings: brain research, child development, studies of unusual talents and more. He calls the theory "strictly empirical" for that reason, and that's a fair defense. But he has also written this:

"Nor, indeed, have I carried out experiments designed to test the theory." Howard Gardner, 2016

That's an honest thing for a theorist to say. It also means the question of whether each type is really an intelligence is one somebody else has to test. So let's test it.

The three tests, and how a type scores

Every type faces the same three tests, in the same order.

1. Can you get it wrong? A real ability is measured with right and wrong answers. A quiz that asks you to rate yourself measures how you see yourself, which is a different thing.
2. Is it part of the family? Here's one of the most replicated results in psychology: people who do well on one kind of mental test tend to do well on the others. Vocabulary, puzzles, memory, speed. They rise together, like a family. The shared part, which researchers call g (general ability), accounts for roughly half of the differences between people across a broad set of tests. A type passes if it rises and falls with that family.
3. Does it predict anything? Does it forecast how people do at school, at work or in life?

A type that fails test two can still be completely real. It's just something else: a talent, a skill, or a piece of your personality.

The scoring is plain arithmetic. A pass is worth 1. A fail is worth 0. Where the evidence is split or missing, the type gets a question mark, worth half a pass. Three passes on strong, broad evidence is S. Three passes on weaker or narrower evidence is A. Two passes is B, one is C and none is D. Halves round down, so 2½ is a B and ½ is a D.

Number smart: the one that sets the bar

Logic and math. The friend who splits the bill in their head.

Can you get it wrong? Math puzzles have right answers. Pass. Is it part of the family? Reasoning sits right at the center of it. Pass. Does it predict anything? More than anything else on this list. A 2007 meta-analysis of long-term studies by Tarmo Strenze found that measured intelligence correlated .56 with how far people went in education, .43 with the level of job they reached, and .20 with income. A study of more than 70,000 English children found that general mental ability at age 11 tracked national exam results at 16 very closely (a latent correlation of .81). Pass.

At work the picture is more modest than it used to be. The classic estimate for cognitive ability and job performance was .51. A 2022 reanalysis by Paul Sackett and colleagues revised it down to .31, and the exact size is still debated. For comparison, the same reanalysis put conscientiousness at .19. Smaller than the old textbook number, but still a real prediction.

Score: 3 of 3, on strong evidence. S tier.

Word smart: S tier without trying

The one who always has the right word. Vocabulary tests have right answers. Vocabulary and general knowledge are among the strongest members of the family. And verbal ability tracks how far people go in school and the jobs they end up in. Pass, pass, pass.

Score: 3 of 3. S tier. It barely had to try.

Music smart and body smart: real talents, not intelligences

The most direct test of Gardner's types came from three Canadian psychologists. In 2006, Beth Visser, Michael Ashton and Philip Vernon gave tasks for each of Gardner's domains to the same 200 adults, then checked which tasks rose with general ability and which didn't. Two of the clearest results were about music and the body.

Music smart is the one who can play anything by ear. You can test pitch and rhythm, and you can get those wrong. Pass. But in Visser's data the music tasks barely moved with the family: their links to general ability were around .10 to .19, too small to be significant. Fail. Does musical ability predict how people do outside music? There's almost no research. Question mark.

Score: 1½. C tier. Music smart is a real talent. It just doesn't behave like a type of intelligence.

Body smart is the athlete, the dancer, the surgeon with steady hands. You can test balance and coordination, with right and wrong results. Pass. Then pull the family up, and body smart just stands there: in the same study, a balance task and a fine-motor task had links to general ability of about .03 and under .10. Fail. Does it predict anything? Coordination helps in some hands-on jobs, so: question mark.

Score: 1½. C tier. Body smart is a skill, and a great one. But it's a skill.

Nature smart: reasoning, wearing binoculars

The one who can name every bird in the garden. Before you read on, make a guess. Where does it land?

Higher than most people think. In Visser's study, the nature task was a sorting puzzle: a diagram test where you decide which category sits inside which (dogs go inside animals, and so on). It has right and wrong answers. Pass. And it was one of the two tasks most connected to the whole family, with a loading on general ability of about .75, alongside a math reasoning task. Pass.

That's not because naturalists are secret geniuses. It's because sorting things into groups is pure reasoning. Nature smart turned out to be reasoning, wearing binoculars. And it's one study, so treat the number with care.

Does nature smart predict anything on its own? There's almost no research on that. Question mark.

Score: 2½. B tier. Real intelligence. Just not a new type of it.

Picture smart: the most underrated mind on the board

The one who packs the car like a puzzle. Spatial tests ask you to turn a shape in your head and pick the match. Right or wrong. Pass. Spatial ability rises with the rest of the family. Pass.

Test three is the surprise. Jonathan Wai, David Lubinski and Camilla Benbow used Project TALENT, which tested about 400,000 American high-school students and followed them for 11 years. Spatial ability helped predict who went into science, technology, engineering and math, even after their math and verbal scores were already known. Pass.

But what it adds on its own is small (about four percent more of the variance explained) and narrow: one lane, science and building things. So by the rules, that's 3 of 3 on narrower evidence. A tier. And because most school and admissions tests don't measure it, which is exactly the point Wai and his colleagues made, it may be the most underrated mind on the board.

Street smart and creativity: the lightning round

Two internet favorites that aren't on Gardner's poster, but turn up in nearly every list. Verdict first.

Street smart: B tier. Tests of practical intelligence give people real-life problems with better and worse answers. Pass. Scores on them rise with the family. Pass. But what they predict turns out to be mostly the ordinary general ability inside them, plus job knowledge. In a long 2003 critique, Linda Gottfredson argued that the case for practical intelligence as a separate ability rested on "an illusion of evidence", built partly through "selective reporting". The research is contested, so: question mark. That's 2½, a B. Book smart and street smart are the same smarts, used in different places.

Creativity: C tier. The classic test asks how many uses you can find for a brick. There's no single wrong answer, but judges can score how original your answers are. Question mark. Creativity is linked to intelligence, but only a little: a 2021 meta-analysis put the correlation with divergent-thinking tests at about .25 (corrected). Question mark. And what those tests predict is still thin. Question mark. Three halves make 1½, a C.

The low tier says more about the tests, which are still rough, than about creativity. One popular claim does fall, though: the idea that creativity needs an IQ of about 120 and then stops depending on it. A 2020 reappraisal found no evidence for that threshold. Other work suggests intelligence may be necessary for high creative achievement without being enough on its own.

Self smart and people smart: the personal intelligences

Gardner called these two the personal intelligences. Self smart is knowing yourself. People smart is reading others. They walk in together, and leave a long way apart.

Self smart has already given itself an S. Which is the problem. How do you test how well someone knows themselves? Mostly by asking them, and asking has no right answer. Fail. When people rate their own intelligence, their guesses match their real scores only loosely: a correlation of .33 across 41 studies and more than 22,000 people (Freund and Kasten, 2012). No family link. Fail. Nothing it predicts. Fail.

Score: 0. D tier. Self-knowledge may well matter. There's just no way yet to score it right or wrong.

People smart does much better. There are right-answer tests of reading faces and feelings. Pass. And they rise with the family: in Visser's study, a social task loaded about .53 on general ability. Pass. Other research finds that social skills like perception and memory for people do relate to academic intelligence, but don't form one single "social intelligence". Whether those tests predict much in real life is still unclear. Question mark.

Score: 2½. B tier. And those face-reading tests have a famous big sibling.

Emotional intelligence splits in two

In 1995, Daniel Goleman's bestselling book Emotional Intelligence made the idea famous, and the notion of an EQ, an emotional score to sit next to your IQ, went everywhere. It's the one most people would bet on.

But EQ gets measured in two very different ways. One kind of test gives you problems with right answers. The other is a questionnaire about yourself. They share a name, so the tests split them in two.

Twin one: the ability test

What is this face feeling? What would calm this person down? These questions have right and wrong answers, scored against expert and consensus judgments. Pass.

Does it rise with the family? Yes. Scores on ability tests correlate about .34 with cognitive ability (Van Rooy and colleagues, 2005). When researchers model emotional ability alongside the other broad abilities, it loads strongly on general ability: .80 in a 2014 study of 688 people and .70 in a 2020 replication with 830. The leading model of intelligence now lists it as a "tentative" broad ability of its own. Pass.

Does it predict anything? Modestly, but yes. A 2020 meta-analysis found ability EI correlated .24 with school grades, and it predicts performance in some jobs. Pass.

Score: 3 of 3, on modest evidence. A tier.

Twin two: the EQ questionnaire

Question one: are you an intelligence? Strongly agree. A questionnaire like this has nothing to get wrong. Fail.

It barely rises with the family: self-report EQ correlates only about .13 with cognitive ability. Fail. And when researchers compare it with an ordinary personality test, the overlap is huge. A 2017 meta-analysis found EQ questionnaires correlated about .85 with the general factor of personality. It's mostly personality, wearing an intelligence name tag.

Does it predict anything? Here it's important to be fair. EQ questionnaires do predict how bosses rate people at work (a corrected correlation of .29). But a large 2015 meta-analysis showed what happens once you take out personality, cognitive ability, ability EI, self-efficacy and people's own ratings of how well they do: the questionnaire's contribution drops to essentially zero (β = −.02). For school grades the pattern repeats in miniature: .19 for mixed questionnaires and .12 for purely self-rated ones, against .24 for the ability test. Question mark, for the little that's left.

Score: ½. D tier.

Big questions: half an intelligence

And the last contestant: Big Questions. Why are we here? What happens when we die? It's meant to be next on stage, but it's outside, looking at the sky.

Gardner considered adding this one and never fully did. In his own words, from a 2000 interview: "So I say that I think there are eight and a half." Existential intelligence is the half. So, nine types. Well, eight and a half.

There's no test with right answers, no link to the family and nothing it predicts. Score: 0. D tier. Not because big questions lack depth. There's just nothing to measure.

The final board

Here is every contestant, its three marks, and where the arithmetic puts it.

Of Gardner's nine, three pass all three tests: number smart and word smart in S, picture smart in A. Counting everyone who walked in (Gardner's nine, plus emotional intelligence, street smart and creativity), four of twelve passed all three, because the emotional-intelligence ability test joins them in A.

And Gardner himself? In 2016 he wrote:

"I readily admit that the theory is no longer current. Several fields of knowledge have advanced significantly since the early 1980s." Howard Gardner, 2016

The twist: one mind, seen from three sides

Now look again at the three of Gardner's types that passed. Pull one up, and the other two rise with it. That's what passing test two means. They aren't three separate winners. They're one mind, seen from three sides.

That's why one number can sum them up. The leading model of intelligence, the Cattell-Horn-Carroll (CHC) model, describes 16 to 18 broad abilities sitting under general ability. A standard adult IQ test such as the WAIS-5 samples five of them (verbal comprehension, visual-spatial, fluid reasoning, working memory and processing speed) and adds them into a single full-scale score. Because the abilities it skips also rise with the family, that one score still captures much of what predicts school and work. It isn't "missing half your intelligence".

But a summary hides the shape. By the test makers' own numbers for the children's version, the WISC-V, a typical child's strongest and weakest areas sit about 25 points apart. A gap only counts as unusual at about 39 points, a size seen in roughly one child in ten. Two people with the same IQ score can have very different minds. The poster was right about that much. It was wrong about why.

What the rest really are

So what about the ones that failed? Look at who they are. Music and body: talents and skills. Self smart and the EQ questionnaire: mostly how people see themselves, and how they tend to behave.

Take the questionnaire twin. It didn't fail for being a questionnaire. It failed for calling itself an intelligence. How you tend to act has no right answers, so asking you, carefully, is the honest way to measure it. That's what personality is, and it's what a good personality questionnaire is built to do.

And personality matters in its own right. A 2007 review by Brent Roberts and colleagues found that personality traits predict how long people live, whether their marriages last and how their careers go, with effects on a par with IQ and socioeconomic status. (The trait model that research uses is the Big Five. If you've only met personality through four-letter types, we compared the two in Big Five vs MBTI.)

Twelve walked in. Four passed all three. The rest weren't wrong about you. They just had the wrong name tag.

Frequently Asked Questions

What are the 9 types of intelligence?

The nine types come from Howard Gardner's theory of multiple intelligences: linguistic (word smart), logical-mathematical (number smart), spatial (picture smart), musical (music smart), bodily-kinesthetic (body smart), interpersonal (people smart), intrapersonal (self smart), naturalist (nature smart) and existential (big questions). Gardner proposed the first seven in 1983 and added the others later. He counts existential as only half an intelligence.

Is Gardner's theory of multiple intelligences supported by science?

Only partly. Gardner has said he never carried out experiments designed to test the theory, and in 2016 he wrote that it is "no longer current". Scored on three tests from intelligence research (right-and-wrong measurement, a link with the other mental abilities, and prediction of real outcomes), three of the nine pass all three: logical-mathematical, linguistic and spatial. Most of the others are real talents, skills or parts of personality rather than separate intelligences.

Is emotional intelligence real?

It depends on how it is measured. Emotional intelligence measured with right-answer tests, such as judging what a face is feeling, behaves like a real ability: it rises with the other mental abilities and modestly predicts school grades. Emotional intelligence measured with self-report questionnaires overlaps heavily with personality, correlating about .85 with the general factor of personality.

What is the difference between emotional intelligence and IQ?

IQ tests measure reasoning, knowledge, spatial skill, memory and speed with right and wrong answers. Ability tests of emotional intelligence measure reading and managing emotions the same way, and their scores correlate about .34 with cognitive ability. "EQ" questionnaires correlate only about .13 with cognitive ability, because they mostly measure personality rather than ability.

Are multiple intelligences the same as learning styles?

No. Gardner himself has said that multiple intelligences are not learning styles. Learning styles, the idea that people learn best when teaching matches their preferred style, were reviewed in 2008 and found to have no adequate evidence base.

Does an IQ test measure every type of intelligence?

No. A standard IQ test such as the WAIS-5 samples about five broad abilities, out of the 16 to 18 in the leading model of intelligence, and adds them into one score. Because the abilities rise together, that score still captures much of what predicts outcomes. But it hides the shape: in the test makers' own norms, a typical child's strongest and weakest areas sit about 25 points apart.

What is the rarest type of intelligence?

There is no evidence-based answer, because the question assumes the types are separate. The abilities that pass the tests rise together, so few people are strong in one and blank in the rest. What is common is an uneven profile: a gap of about 25 points between a child's strongest and weakest areas is typical, and only gaps of about 39 points or more are unusual.

Sources & References

1. Gardner, H. (1983). Frames of Mind: The Theory of Multiple Intelligences. Basic Books.
2. Gardner, H. (1999). Intelligence Reframed: Multiple Intelligences for the 21st Century. Basic Books. The "eight and a half" remark is from Gardner's interview with Salon, January 5, 2000.
3. Gardner, H. (2016). Multiple intelligences: Prelude, theory, and aftermath. In R. J. Sternberg, S. T. Fiske & D. J. Foss (Eds.), Scientists Making a Difference (pp. 167-170). Cambridge University Press. Quoted via researchED.
4. Gardner, H. (2013, October 16). "Multiple intelligences" are not "learning styles". The Washington Post. Harvard Graduate School of Education
5. Visser, B. A., Ashton, M. C., & Vernon, P. A. (2006). Beyond g: Putting multiple intelligences theory to the test. Intelligence, 34(5), 487-502. Brock University repository
6. Deary, I. J., Johnson, W., & Houlihan, L. M. (2009). Genetic foundations of human intelligence. Human Genetics, 126(1), 215-232. University of Edinburgh
7. Strenze, T. (2007). Intelligence and socioeconomic success: A meta-analytic review of longitudinal research. Intelligence, 35(5), 401-426.
8. Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13-21.
9. Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection. Journal of Applied Psychology, 107(11), 2040-2068. doi:10.1037/apl0000994
10. Wai, J., Lubinski, D., & Benbow, C. P. (2009). Spatial ability for STEM domains: Aligning over 50 years of cumulative psychological knowledge solidifies its importance. Journal of Educational Psychology, 101(4), 817-835.
11. Gottfredson, L. S. (2003). Dissecting practical intelligence theory: Its claims and evidence. Intelligence, 31(4), 343-397. doi:10.1016/s0160-2896(02)00085-5
12. Gerwig, A., et al. (2021). The relationship between intelligence and divergent thinking: A meta-analytic update. Journal of Intelligence, 9(2), 23. PMC8167550
13. Weiss, S., et al. (2020). A reappraisal of the threshold hypothesis of creativity and intelligence. Journal of Intelligence, 8(4), 38. PMC7709632
14. Karwowski, M., et al. (2016). Is creativity without intelligence possible? A necessary condition analysis. Intelligence, 57, 105-117. Erasmus University
15. Freund, P. A., & Kasten, N. (2012). How smart do you think you are? A meta-analysis on the validity of self-estimates of cognitive ability. Psychological Bulletin, 138(2), 296-321.
16. Conzelmann, K., Weis, S., & Süß, H.-M. (2013). New findings about social intelligence: Development and application of the Magdeburg Test of Social Intelligence (MTSI). Journal of Individual Differences, 34(3), 119-137.
17. Goleman, D. (1995). Emotional Intelligence. Bantam Books.
18. MacCann, C., Joseph, D. L., Newman, D. A., & Roberts, R. D. (2014). Emotional intelligence is a second-stratum factor of intelligence: Evidence from hierarchical and bifactor models. Emotion, 14(2), 358-374. UCF STARS
19. Evans, T. R., Hughes, D. J., & Steptoe-Warren, G. (2020). A conceptual replication of emotional intelligence as a second-stratum factor of intelligence. Emotion, 20(3), 507-512.
20. Schneider, W. J., & McGrew, K. S. (2018). The Cattell-Horn-Carroll theory of cognitive abilities. In D. P. Flanagan & E. M. McDonough (Eds.), Contemporary Intellectual Assessment (4th ed.). Guilford Press.
21. Van Rooy, D. L., Viswesvaran, C., & Pluta, P. (2005). An evaluation of construct validity: What is this thing called emotional intelligence? Human Performance, 18(4), 445-462.
22. MacCann, C., et al. (2020). Emotional intelligence predicts academic performance: A meta-analysis. Psychological Bulletin, 146(2), 150-186. doi:10.1037/bul0000219
23. van der Linden, D., et al. (2017). Overlap between the general factor of personality and emotional intelligence: A meta-analysis. Psychological Bulletin, 143(1), 36-52. Tilburg University
24. Joseph, D. L., Jin, J., Newman, D. A., & O'Boyle, E. H. (2015). Why does self-reported emotional intelligence predict job performance? A meta-analytic investigation of mixed EI. Journal of Applied Psychology, 100(2), 298-342.
25. Pearson. (2024). WAIS-5 overview. Pearson Assessments (PDF)
26. Courville, T., Coalson, D. L., Kaufman, A. S., & Raiford, S. E. (2016). Does WISC-V scatter matter? In A. S. Kaufman, S. E. Raiford & D. L. Coalson, Intelligent Testing with the WISC-V (pp. 209-225). Wiley.
27. Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning styles: Concepts and evidence. Psychological Science in the Public Interest, 9(3), 105-119.
28. Roberts, B. W., Kuncel, N. R., Shiner, R., Caspi, A., & Goldberg, L. R. (2007). The power of personality: The comparative validity of personality traits, socioeconomic status, and cognitive ability for predicting important life outcomes. Perspectives on Psychological Science, 2(4), 313-345.