Who Owns Intelligence?

September 24, 2001 by Howard Gardner

Before intelligence can be enhanced or artificially created, it has to be defined; this excerpt from Howard Gardner’s Intelligence Reframed ponders the different ways in which intelligence is quantified and conceived.

Originally published 1999. Excerpts from Intelligence Reframed. Published on KurzweilAI.net September 24, 2001.

The theory of multiple intelligences has helped break the psychometricians’ century-long stranglehold on the subject of intelligence. ‘While we may continue to use the words smart and stupid, and while IQ tests may persist for certain purposes, the monopoly of those who believe in a single general intelligence has come to an end. Brain scientists and geneticists are documenting the incredible differentiation of human capacities, computer programmers are creating systems that are intelligent in different ways, and educators are freshly acknowledging that their students have distinctive strengths and weaknesses.

In this book, I have laid out a position that challenges the psychometric consensus. I have proposed a set of several intelligences, each resting on its own neurological substrate, each of which can be nurtured and channeled in specific ways, depending on a particular society’s values. I have listed the criteria for an intelligence and shown how these can be evoked in evaluating potential new intelligences. But where does one draw the line? Are my criteria the right ones? In the future, the new dimensions and boundaries of intelligence will likely be thrashed out on a pivotal battlefield. Now that the Scylla of the psychometricians has been overcome, we risk succumbing to the Charybdis of “anything goes”–emotions, morality, and creativity all being absorbed into the “new intelligence.” The challenge is to chart a concept of intelligence that reflects new discoveries and understandings and yet can withstand scrutiny.

The Stretch and Limits of Multiple Intelligences

One can think of intelligence as an elastic band. For many years no one effectively challenged its definition, and the band seemed to have lost its elasticity. Some of the new definitions of intelligence have expanded the band and renewed its resilience, even while incorporating the earlier work on intelligence that is still germane. Other definitions have expanded the band to the snapping point, rendering unusable the earlier foundational study of intelligence.

Until now, the term intelligence has been largely limited to linguistic and logical capacities, although (as I’ve argued) humans can process other elements as diverse as the contents of space, music, or their own and others’ psyches. Like the elastic band, conceptions of intelligence need to encompass these diverse contents–and stretch even more. We must move beyond solving existing problems and look more at the capacities of human beings to fashion products (like works of art, scientific experiments, classroom lessons, organization plans) that draw on one or more intelligences.

As long as intelligences are restricted to the processing of “contents in the world,” we avoid epistemological problems. So it should be. The concept of “intelligence” should not be expanded to include personality, motivation, will, attention, character, creativity, and other valued human capacities. If we conflate intelligence with creativity, as we have seen in chapter 8, we can no longer distinguish between the expert (the person highly skilled in a domain) and the creator (one who expands a domain in new and unexpected ways). We would also fail to recognize that creative individuals stand out particularly in terms of their restless temperament and personality, whereas experts efficiently process informational content and accept the status quo.

Consider also what would happen if we stretched intelligence to include good or evil attitudes and behaviors. By making that incursion into morality, we would confront human values within a culture. Granted, a few values probably can be expressed generically enough so that they command universal respect: One promising candidate is the Golden Rule (in its biblical version, in other religions’ versions, or in the contemporary version introduced by the sociologist Amitai Etzioni: Respect the mores of your society). However, most other values–even such seemingly unproblematic ones as the rejection of theft, killing, or lying–turn out to be specific to cultures or subcultures.

If we conflated morality and intelligence, we would need to deal with widely discrepant views of what is good and bad, and why–vexing questions about abortion, capital punishment, holy wars, marriage between relatives, patriotism, treatment of strangers, and more. Consider too that people who score high on tests of moral reasoning often act immorally outside the test situations, even as courageous and self-sacrificing people turn out to be unremarkable on tests of moral reasoning. Many of those who hid Jews or other persecuted people during World War II lacked education or sophistication. In contrast, eight of the fourteen men who laid plans to implement the Final Solution held doctoral degrees from major European universities.

Furthermore, Adolf Hitler and Joseph Stalin probably knew full well which situations were considered moral in their culture, but they either did not care (Stalin commented, “How many divisions does the Pope have?” and Hitler extolled the Big Lie) or embraced their own peculiar codes (“Wiping out a generation is a necessary, indeed inevitable, move if you are committed to the establishment of a Communist state.” Or “Eliminating Jews is the moral imperative in quest of an Aryan society.”)

The notion of an “emotional intelligence” proves problematic in certain respects (as discussed in chapters 1 and 5). Unlike language or space, emotions are not “contents” to be processed. Rather, cognition has evolved so that we can make sense of human beings (self and others) who have and experience emotions. Emotions do accompany cognition, and they may well prove more salient under certain circumstances; they accompany our interactions with others, our listening to music, and our efforts to solve mathematical puzzles. Calling some intelligences emotional implies that other intelligences are not, and that implication flies in the face of experience and empirical data. Further problems arise when we conflate emotional intelligence with a certain recommended pattern of behavior–a temptation to which Daniel Goleman sometimes succumbs in his otherwise admirable Emotional Intelligence. Goleman singles out as “emotionally intelligent” people who use their understanding of emotions to make others feel better, solve conflicts, or cooperate in home or work situations. I certainly cherish such people, but we cannot assume that being emotionally intelligent means those skills will be used for socially desirable ends.

For these reasons, I prefer the term emotional sensitivity, which applies to those who are sensitive to emotions in themselves and others–that is, individuals who exhibit the personal intelligences (in my own terminology). Presumably, clinicians and salespeople excel in sensitivity to others; poets and mystics, in sensitivity to the melodies of their own psyches. And there are others–autistic or psychopathic persons, for example–who seem completely deaf to the emotional realm. I insist, however, on a strict distinction between being emotionally sensitive and being a “good” or “moral” person, since someone who is sensitive to others’ emotions may still manipulate, deceive, or create hatred. I call, then, for a delineation of intelligence that includes the full range of contents to which human beings are sensitive, but excludes such valued but separate human traits as creativity, morality, and emotional appropriateness. This delineation makes scientific and epistemological sense; it reinvigorates but does not break the elastic band; and it helps resolve two remaining struggles: how to assess intelligences and how to connect intelligences to other virtues.

The Assessment of Intelligences

All societies want to place the most appropriate people in positions of importance, but the most desirable niches often have far more candidates than can be accommodated. Hence, some forms of assessment are almost inevitable. Once we restrict the definition of intelligence to human information-processing and product-making capacities, we can use and supplement the established technology of assessment. We can continue to use paper-and-pencil or computer-adapted techniques, while simultaneously looking at a broader range of capacities, such as sensitivity to musical patterns or the understanding of people’s motivations. And we can avoid ticklish, and possibly unanswerable, questions about the assessment of values and morality. But even with a limited definition of intelligence, important questions remain about which assessment path to follow. Here I hold strong views. I consider it a fool’s errand to embrace the search for a “pure” intelligence–whether general intelligence, musical intelligence, or interpersonal intelligence. I do not believe that such alchemical cognitive essences actually exist; they are an outcome of our penchant for creating (and then attributing reality to) terminology rather than searching for determinable, measurable entities. Moreover, the correlations that have been found between allegedly “pure measures” (like certain brainwave patterns that purport to measure intelligence directly) and the skills we actually value in the world (like mathematical problem solving and good writing) are too modest to be useful.

What matters is the use of intelligences, individually and in concert, to carry out tasks valued by a society. Accordingly, we should be assessing people’s success in carrying out valued tasks that presumably involve certain intelligences. For example, instead of testing musical intelligence by looking at evoked cortical responses when someone is listening to pure tones, we should teach people to sing songs, play instruments, or compose or transform melodies, and then determine how well they have mastered these tasks. By the same token, we should not search for immaculate emotional sensitivity–for example, with tests of galvanic skin response to a word or photograph. Rather, we should observe people in real-life situations where they have to be sensitive to the aspirations and motives of others. For example, we can see how someone handles a fight between two teenagers or convinces a supervisor to change an undesirable policy. These are realistic contexts for assessing mastery of the emotional realm.

Increasingly, we have another assessment option: simulations. We are now in a position to use technologies that not only can present realistic situations or problems but also can also measure performance through virtual realities, and even “intelligently” select subsequent steps in light of responses on earlier phases of the simulations. Thus, presenting a student with an unfamiliar tune on a computer and having him learn the tune, transpose it, and orchestrate it, can reveal much about his intelligence in musical matters. Similarly, we can learn about interpersonal or emotional sensitivity by simulating human interactions and asking people to judge the shifting motivations of each actor. For example, subjects can give their running reactions to members of a jury who are attempting to reach a verdict on a sensitive case. Or one can create an interactive hypermedia program–for example, a program that features members of an organization as they are grappling with a major change in corporate strategy–and ask respondents to react to the virtual (or “real”) people’s moves even as those moves are being altered by the program.

An increase in the breadth, or elasticity, of our concept of intelligence, then, should open the possibility for innovative forms of assessment that are far more authentic than the classical short-answer examinations. Why settle for an IQ test or an SAT, on which the items are at best remote “proxies” for the ability to design experiments, write essays, critique musical performances, or resolve a dispute? Why not instead ask people to do the things–either in person or online? As long as we do not open the Pandora’s box of values and subjectivity, we can continue to use established insights and technologies judiciously. Of course, if we used the psychometricians’ traditional armamentaria, we could create an instrument to test any conceivable virtue (or vice), including morality, creativity, and emotional intelligence. Indeed, since Goleman’s landmark book, there have been dozens of efforts to create tests for emotional intelligence. But such instruments are far more likely to satisfy the test makers’ desires for reliability (that is, each testee would get roughly the same score on two separate administrations of an instrument) than the need for validity (that is, the test measures the trait it purports to measure, such as emotional sensitivity within one’s family or at the workplace).

These kinds of instruments are questionable for two reasons: First, it is too difficult to agree on what it means to be emotionally intelligent–consider the different interpretations that might be given by Jesse Helms and Jesse Jackson, or Margaret Thatcher and Margaret Mead. Second, “scores” on such tests are more likely to reveal test-taking savvy (people’s skills in language and logical inference) than fundamental emotional acuity.

We are at a turning point. A tight view of assessment is likely to produce reliable instruments that correlate well with one another but do not broaden the sample of talents to be surveyed nor the range of individuals who will stand out. A subtler view opens up many new and exciting possibilities. We will be able to look directly at the skills and capacities we value, and we will give people a variety of ways to demonstrate what they know and what they can do. Rather than just selecting one kind of a person, we may help to place many kinds of people in positions well matched to their skills and aspirations. If assessment is to be reinvented, such innovations point the way.

Connecting Intelligences to Other Virtues

While broadening the definition of intelligence, I have steadfastly argued that the expansion of the band must be regulated. We cannot hijack the word intelligence so that it becomes all things to all people–the psychometric equivalent of the Holy Grail. Yet the problem remains: How, in a post-Aristotelian, post-Confucian era, one in which psychometrics still looms large, should we think about the virtuous human being–the human being who is justly admired because of his or her personal qualities? One promising approach is to recognize that intelligences, creativity, and morality–to mention just three commonly recognized virtues–are separate. Each may require its own measurement or assessment, and certain species or subspecies of these virtues will prove far easier to assess objectively than others. Indeed, with respect to creativity and morality, we are more likely to rely on overall judgments by disinterested experts than on any putative test battery. At the same time, we might well look for people who combine attributes: people who have musical and interpersonal intelligence, who are psychometrically Intelligent and creative in the arts, who combine emotional sensitivity and a high standard of moral conduct.

Consider that selective colleges pay much attention to scholastic performance, as measured by the College Entrance Examination Board and secondary school grades; but they also weigh other features, and sometimes a student with lower test scores but high value in citizenship or motivation is chosen over one who has “aced” the tests. Admissions officers do not confound these virtues (indeed, they may use different “scales” and issue different “grades”), but they recognize the attractiveness of candidates who exemplify two or more of these desirable traits.

We probably will never re-create an Eden where intellectual and ethical values commingle, and we should recognize that these virtues can be separate. Indeed, despite the appeal of the Confucian or Grecian hero, these virtues are often all too remote from one another. Thus, when we attempt to aggregate the virtues, through phrases like “emotional intelligence,” “creative intelligence,” or “moral intelligence,” we should realize that we are expressing a wish rather than describing a probable reality. Despite this caution, it is important to recognize that there are powerful models–people who successfully exemplify two or more cardinal human virtues. In the recent past, one can without reservation name the scientist Niels Bohr, the writer Rachel Carson, the athlete Arthur Ashe, the statesman George Marshall, and the musicians Louis Armstrong, Pablo Casals, and Ella Fitzgerald. In our own time, few would challenge the singling out of Nelson Mandela.

Examining lives like these reveals human possibilities. Young people learn primarily from the examples of powerful adults around them, ones who are admirable as well as ones who are simply glamorous. Sustained attention to admirable examples may well increase the incidence of people who will eventually link capacities now considered scientifically and epistemologically separate.

First-hand acquaintance with exemplary models probably constitutes the first step in becoming a person of multiple virtues, but exposure is not enough. Capacities must be trained. Threats to morality and decency must be identified and confronted. We need practice, with feedback, in handling morally charged situations, dilemmas that pull us in competing directions. We must learn from others but also recognize that we must sometimes go our own way. And ultimately, we must be ready to serve as role models for younger people.

The British novelist E. M. Forster counseled: “Only connect.” Some expansionists in the territory of intelligence have prematurely claimed connections that do not exist. But it is within our power to help forge connections that are important for our physical and psychic survival. Just how the precise borders of intelligence are drawn is a question we can leave to scholars. But the imperative to broaden the definition of intelligence responsibly goes well beyond the academy. Who “owns” intelligence promises to be an even more critical issue in the twenty-first century than it has been in this era of the IQ test.

Remaining Puzzles: The Research Agenda

The first half of the twentieth century was the period of physics; the second, the period of molecular biology and genetics. Few doubt that the next century will highlight the study of the brain and mind. And, of course, exploring the nature of intelligence will be an important part of the research agenda. The research of the coming years will explore three major areas–each with two main thrusts.

The Basic Sciences of Intelligence

While intelligence has belonged largely to psychology, I see it increasingly being explored by other disciplines. On the one hand, those who work at the cellular and genetic levels are asking which genes control which aspects of intellectual functioning and how the genes work together to produce intelligent behavior. On the other hand, there is a growing interest in, and knowledge of, the ways human intelligence is applied in different social and cultural contexts.

We already know of genes or gene clusters that code for specific cognitive abilities like reading and spatial capacities, and there may be others that are critical in the attainment of high IQ-test scores. The interest in the structure of specific human abilities is magnified by new imaging techniques; we now can examine the neural structures involved in particular aspects of language (like reading, naming, and learning foreign languages), music (like rhythm and tonal perception), and even the understanding of people’s minds, which proves crucial in interpersonal intelligence. At the other end of the scientific spectrum, within our own society, ethnographic investigators are studying different work settings and trying to determine which intelligences people use alone and in concert to accomplish important tasks. Studies of the building of a computer and of the navigation of a huge carrier reveal, for example, that no single individual understands the entire process; rather, this type of intelligent behavior depends upon capacities distributed across numerous individuals.

Cross-cultural studies continue to challenge our notions of human intellect. We look at contemporary societies and note the different emphases placed, for example, on sensitivity to others, on the capacity to cooperate with strangers, or on various putative psychic capacities like meditation or healing. And retrospectively we can also study intelligences of an earlier era. The archaeologist Stephen Mithen, for example, has described the naturalist and technological intelligences that may have been important for the forerunners of Homo sapiens 250,000 to 500,000 years ago.

These lines of study help us appreciate the limitations of singular views of intelligence, formulated largely in terms of the capacities needed to exist in a certain kind of European or American school one century ago.

The Operations of Intelligences

Information-processing techniques and computer simulations offer powerful ways of learning how people perform specific tasks, from understanding a foreign language, to creating a piece of music. Such studies not only will facilitate the development of software that excels at these tasks but also will suggest the kind of training that can improve the performance of ordinary (and extraordinarily talented or impaired) people. It is equally important to understand those capacities that extend beyond the operations of single intelligences or subintelligences. Cognitively oriented researchers will probe in two directions. They will investigate the ways in which particular intelligences work together, in general or on specific tasks. And they will explore those capacities that seem to cut across different intelligences–the making of metaphors and analogies, the capacity to synthesize information, the emergence of wisdom. Unfortunately, MI theory has not made much progress in explicating the nature of these transintellectual capacities. And these capacities have also eluded investigators from other psychological camps. It is probably true that some capacities, like the making of metaphors, will turn out to be part of the basic cognitive equipment of all human beings, while others, like the capacity to synthesize different bodies of information, require a culture that has cultivated these encyclopedic skills over a long period.

Shifting Demands

Two 1997 events symbolize tidal shifts in world culture: The defeat of chess champion Gary Kasparov by the IBM “Deep Blue” computer program proved once and for all that a machine could be “smarter” than the cleverest human performer in a domain long cherished by the intelligentsia. And the cloning of the sheep Dolly by the Scottish scientist Ian Wilmut and his colleagues demonstrated the potential to engage in the most profound experiments of genetic engineering. Some people would like to turn their backs on these events, because they fear a world in which machines dominate human beings or because they don’t want to see people play God by controlling the genetic options of future generations. I share these reservations, though I doubt whether it is possible to prevent human beings from exploring such possibilities. However impressive, Dolly and Deep Blue are products of technology; they are neither good nor evil in themselves. We humans, operating individually and as a corporate body, must judge how the technology that spawned these “creatures” should be used or not used.

Many of us would welcome a society in which drudgery is eliminated because robots carry out mindless work. But a society freed of much human labor can turn in two opposite directions. We might be freed to exercise higher powers of mind in the arts or in other creative spheres, or we might surrender to the pursuit of mind-numbing entertainment, either benign (like television comedies or soap operas) or malignant (the bread and circuses held in amphitheaters during the decline of Rome). As a species, we remain free to become smart or stupid, moral or immoral, in various ways.

The options offered by genetic engineering are even sharper. Even those who are repelled by the idea of cloning or more aggressive forms of eugenics understand the appeal of testing for the lethal gene of, say, Huntington’s chorea and, if possible, turning it off. But the decisions first exercised in the realm of bodily disease will sooner or later reverberate in the corridors of personality and intellect. We have to ask whether we want to eliminate the genes that give rise to dyslexia, and we may have to ask whether we will tolerate genetic engineering aimed at producing individuals who excel at mathematics, chess, music making, or the less appealing capacity to manipulate others. The identification of new intelligences and a superior understanding of how they operate will stimulate geneticists to probe the biological underpinning of these capacities and, by the same token, geneticists’ discoveries will alert us to the possibility of new or different configurations among human intellectual capacities.

No single authority has the right to make decisions in crucial realms like artificial intelligence or genetic engineering. But this does not mean that the opposite is true: that these decisions do not belong to anybody or that marketplace forces should be allowed to determine what is done. We cannot have a society in which people abdicate personal responsibility, dismiss the need to debate issues, and reject outright approaches that might be reasonable. Human societies can and must participate actively in decisions that affect the health and well being of the planet. The primary responsibility rests upon those who actually work in fields like engineering or genetics. They know the field best and therefore have the potential to discern misapplications before those become clear to outsiders. But because intimacy is rarely correlated with disinterestedness, I place an equal burden of responsibility on those who work in neighboring fields, who have enough familiarity to make informed judgments, and also can take a neutral or broader stance. We must not remove responsibility from those who are closest to the action. But ordinary citizens sometimes have better instincts than those whose lives are immersed in a discipline. The problem is that ordinary citizens are typically poorly informed and so are easy prey to misrepresentations and deceptions.

I place hope in four groups: better-trained journalists who can clarify options for the public, political leaders who have studied the issues and are able to explain them, ordinary citizens who are willing to inform themselves about the issues enough to share the decision-making burden with experts, and leaders or “trustees” of domains who will put aside their own ambitions to promote a wider good. Unfortunately, rewards in our society today favor none of these options. Yet the stakes could not be higher: If we do not make the most informed decisions–I might even say the most intelligent decisions–about our genetic and cultural destinies, we may find it has become too late.

For the next millennium, I nominate a new virtue: species humility. In the past, we honored those (like St. Francis) who were humble, even as we scorned leaders and groups guilty of hubris. But we are now one, inextricably bound world, and the unimaginable, in many forms, has become possible. As a species, we must somehow arrive at decisions about what we will do and what we will not, about which Pandora’s boxes to open and which to keep shut. We have eliminated smallpox and polio, and we stand on the verge of eliminating biological warfare and land mines. Perhaps we can also agree not to manipulate the intellectual capacities of future generations.

Greater Individuation: A Challenge for the Future

A wondrous feature of life is that we humans differ from one another, and, despite the homogenization of the world, our differences show no sign of declining. Indeed, the opposite is the case. Humans evolved to live in small groups, with similar experiences from one day to the next and from one generation to the next. In such milieus, the number of “models for living” were small. We now live in a global village, with rapid change and constant contact with thousands of others. The more experiences we have, the more media we are exposed to, the more people we interact with, the greater the differences that are likely to emerge. Diversity is the order of the millennium.

If the past millennium ushered in greater democracy, this one should usher in greater individuation–individuation, not in the sense of selfishness or self-seeking, but in the sense of knowledge about and respect for each individual. We already can know a great deal about individuals, and we are learning even more from genetics, psychology, and other behavioral and biological sciences. Information widely available in databases will allow us to determine how we resemble and differ from one another and will empower us to make more judicious life decisions.

We cannot avoid moral issues. Individual information can be used to manipulate us; intelligent programmed agents can serve us what they think we want or what they want us to want. And there is no guarantee that we will make sensible use of information about ourselves. In certain spheres we will not want to dote on our individuality; we hope, for example, that everyone will honor the laws of driving and will show civility to each other on the road. But when it comes to learning, using our minds well, and informing others and being informed by others, there need be no limitations. Knowledge need not be competitive; we can all increase our own knowledge and the knowledge of others without end, without the peril of zero-sum situations. Indeed, information about our own minds and the minds of others can be mobilized to broaden our understandings in myriad ways and to open up new vistas.

Everyone acknowledges the importance of science and technology, but it is also important to remember the necessity of the arts and the humanities. The sciences deal with general principles, universal laws, and broad predictions; the arts and humanities deal with individuality. We learn about seminal historical figures in their individuality; we explore the psyches of diverse (and often perverse) characters in literature; we gain from artists’ and musicians’ reflections of their own emotional lives through their works. Every time we are exposed to a new individual–in person or in spirit–our own horizons broaden. And the possibilities of experiencing different consciousnesses never diminish. The humanist of classical times said, “Nothing human is alien to me”; and the saga of individual consciousnesses cannot be reduced to formulas or generalizations.

Here, we connect to multiple human intelligences. Granted only our species membership, we are fundamentally alike. Factoring in each person’s unique genetic blueprint, we become capable of achieving different potentials, and our different family and cultural milieus ensure that we will eventually become distinct human beings. Because our genes and our experiences are unique and because our brains must figure out meanings, no two selves, no two consciousnesses, no two minds are exactly alike. Each of us is therefore situated to make a unique contribution to the world. In the recognition of our individuality, we may discover our deepest common tie–that we are all joint products of natural and cultural evolution. And we may discover why we must join forces, in a complementary but synergistic way, to make sure that Nature and Culture survive for future generations.


The Golden Rule is discussed in A. Etzioni, The New Golden Rule (New York: Basic Books, 1996).

Moral reasoning tests are compared with moral behaviors in A. Colby and W. Damon, Some Do Care (New York: Free Press, 1992).

On the men who laid the plans to implement the Final Solution, see D. Patterson, When Learned Men Murder (Bloomington, Ind.: Phi Delta Kappan Educational Foundation,1996).

See D. Goleman, Emotional Intelligence (New York: Bantam, 1995).

Identifying the genes for general intelligence: M.J. Chorney, K. Chorney et al., “A quantitative trait locus associated with cognitive ability in children,” Psychological Science, 9, no. 3 (May 1998): 159-66.

Intelligences used in concert: E. Hutchins, “The Social Organization of Distributed Intelligence,” in Perspectives in Socially Shaped Cognition, ed. L. B. Resnick, J. M. Levine, and D. Teasley (Washington, D.C.: American Psychological Association, 1995), pp. 283-307; T. Kidder, The Soul of a New Machine (Boston: Little, Brown, 1981).

The forerunners of Homo sapiens are discussed in S. Mithen, The Prehistory of the Mind (London: Thames & Hudson, 1996).

On the cloning of the sheep Dolly, see G. Kolata, Clone: The Road to Dolly and the Path Ahead (New York: Morrow, 1998.)