In the Robert Louis Stevenson novel, Strange Case of Dr. Jekyll and Mr. Hyde, written over one hundred years ago in 1886, a basically well-meaning, well-mannered doctor, Dr. Jekyll, makes a discovery that enables him temporarily to transform himself into a hideous monster of a man, Mr. Hyde. At first, the discovery is merely a curiosity. But gradually the alter ego, Hyde, begins to dominate Dr. Jekyll, until, of course, tragedy strikes. Eventually, Mr. Hyde takes over. Early death puts an end simultaneously to both Jekyll and Hyde, as it must, because the two individuals share but a single body.
The field of intelligence has many of the characteristics of a Jekyll and Hyde relationship. It has been and continues to be, in many respects, well-meaning and well-mannered, offering the possibility of doing good for science and for the public. But the field also has an ugly side, which continually seems to be trying to dominate its good side. The question remains as to whether one side will ultimately dominate the other, or whether, as is more likely, the two sides will continue to live together in an uneasy truce as time passes by.
The field of intelligence has had its Jekyll and Hyde sides for me personally, which is why I entered the field in the first place. I became interested in intelligence when, as an elementary-school student, I did poorly on IQ tests. In fact, I did so poorly that in sixth grade I was sent back to a fifth-grade classroom to retake the fifth-grade intelligence test. In a sense, my professional career has been an attempt to understand and come to terms with my own early failures on these tests!
It is important to realize that this Jekyll-and-Hyde dualism is not limited to the field of intelligence. Consider physics: Is nuclear power good or bad? Potentially, it is either or both. Or try biology: Is gene splicing good or bad? It can be either, or both. In field after field, there is the potential for good or bad. What decides which way things go is not the knowledge base, but the people who use it for wise or unwise ends (Sternberg, 1998). How knowledge is used is always a choice. Thus, students need not only acquire knowledge, but the wisdom to use the knowledge for good ends.
A Brief History Lesson
Thinking in the field of intelligence dates back at least to ancient philosophers such as Plato and Aristotle. For example, in Book 5 of the Republic, Socrates asks Glaucon whether he is not in agreement that a gifted individual is one who easily acquires knowledge and tends to remember it, in contrast to the less gifted individual who acquires knowledge only with great difficulty, and then tends to forget it. Glaucon, of course, agrees. Who can disagree with Socrates, other than those of his contemporaries who put him to death? Aristotle, in Posterior Analytics Book I, conceives of intelligence in terms of "quick wit," which is the notion of hitting on ideas instantly. Such ancient thinking was rather harmless, but some more modern thinking has not been.
What is Intelligence?
The issue of the nature of intelligence was important in ancient Greece and continues to form today a fundamental issue in the psychology of intelligence. What makes the issue so important?
Importance of Issue
First, we use many different kinds of tests to measure intelligence. But before we measure intelligence, certainly we ought to seriously consider just what it is we are measuring. Otherwise, we run the risk of measuring the wrong thing. If we are going to use these tests to make high-stakes decisions, such as about tracking in schools or admission to colleges, universities, and professional schools, we especially should make sure we are measuring the right thing. This problem is not limited to intelligence. If we want, say, to sort children or adults on creativity, "good personality," or "good values," we also need a sense of just what we mean by these things before we blindly sort by them.
Second, when we start measuring something before we know what it is, we run the proverbial risk of putting the cart before the horse. Instead of ideas driving measurement, measurement ends up driving ideas. We may end up manipulating our theories to fit our measurements, rather than the other way around. Most likely, we will measure what is easiest to measure, and then end up creating theories that justify these measurements after the fact.
Modern thinking about intelligence is usually dated back to Sir Francis Galton (1883), who proposed that intelligence could be understood in terms of the qualities of energy and sensitivity. Galton had much that was positive to say about those who were well endowed with intelligence, but much that was negative to say about individuals and groups of individuals in what he believed to be the lower ranges of intelligence. Galton thus exhibited early the Mr. Hyde lurking in the background of work in intelligence. For example, Galton believed that
The discriminative facility of idiots is curiously low; they hardly distinguish between heat and cold, and their sense of pain is so obtuse that some of the more idiotic seem hardly to know what it is. In their dull lives, such pain as can be excited in them may literally be accepted with a welcome surprise. (p. 28)
Given that Galton believed that idiots were ones who did poorly on psychophysical tests of visual, auditory, haptic, and other forms of acuity and discrimination, one shudders to think of the implications of Galton's views for those who were on the lower end of what Gardner (1983) has more recently dubbed "bodily-kinesthetic intelligence." You want to do them a favor? Give them some pain as a gift that will excite in them a welcome surprise. In this and other writings, Galton made no effort to hide the Hyde side. Fortunately, at least for a time, his views more or less died.
Binet and Simon (1916) devised a test to measure intelligence that was based not on psychophysical acuity and discrimination, but on judgment. Their test had a distinctly Jekyll-like character to it. Their purpose was to determine which students would learn well in school. They also were trying to save children who were behavior problems in school from being relegated to special classes or schools for the mentally retarded. At the time, teachers did not clearly distinguish between behavioral and mental problems, and often the teachers would assume that bad behavior was tantamount to mental retardation. Such an assumption had the Hyde-like advantage of giving teachers a quick way to get rid of students they didn't want in their classes. Thus, the intelligence tests of Binet and Simon were designed to protect students, in Jekyll-like fashion, from Hyde-like decisions that could ruin their careers and their lives.
The problem is that the two faces of selection and retention are so easily reversed. Just as tests can be used to preserve talent, so can they be used to rid society or other groups of individuals who, for one reason or another, come to be viewed as undesirables. Indeed, some people believe that intelligence tests and related tests have come to be used in just such a way: They can provide a pseudo-scientific smoke screen for rejecting people from the mainstream of society whom the society wants to reject for nonscientific reasons (Gould, 1981). Instead of rejecting the people on the basis of their ethnic or national or other group, one creates a test that appears to give an objective reason for their exclusion.
Gould's (1981) recounting of events is sometimes in devoted service of his message, and so one needs to exert caution in accepting some of his conclusions. But there can be little doubt that, de facto, tests have tended to favor certain groups over others, for whatever reason (Sternberg, 1996). This fact illustrates an additional complexity in the study of intelligence as applied to society. When certain groups do more poorly on intelligence tests than others, there are those who believe that the difference reflects the Dr. Jekylls of psychology who create the kind of fair meritocracy that our country should be, and there are others who believe that the difference reflects the Dr. Hydes of psychology who legitimize institutional discrimination against the unfortunate. In other words, unlike in the novel, it is not always totally clear who is Jekyll and who is Hyde.
Although a major conflict regarding testing was between Galton and Binet, a main conflict regarding theory was between Charles Spearman (1904) and Louis Thurstone (1938). Spearman believed that performance of cognitive tasks requires two kinds of abilities, a general ability (g) common to all of the tasks, and specific abilities each relevant to just a single task. Thurstone, in contrast, believed that no single ability existed, but rather, that there are multiple primary mental abilities, such as verbal-comprehension ability, numerical ability, and spatial ability. The conflict between unitary and multiple conceptions of intelligence has carried down even to the present day.
Contemporary Theories of Intelligence
Today, the question of what intelligence is remains as hot as it was in the days of the differences between the views of Galton and Binet (Sternberg, 2000). Some investigators are actively attempting to expand our notions of intelligence. Two examples are the work of Howard Gardner and of myself.
Gardner (1983, 1993, 1999) has proposed a theory of multiple intelligences according to which intelligence comprises not just a single entity, but multiple ones, including linguistic, logical-mathematical, spatial, bodily-kinesthetic, musical, interpersonal, intrapersonal, and naturalist intelligences. For example, when you write a paper, you are using primarily linguistic intelligence. When you solve calculus or other mathematical problems, you are using primarily logical-mathematical intelligence. When you try to figure out why you procrastinate in your work, you are using intrapersonal intelligence. Gardner has also speculated as to the existence of existential and spiritual intelligences. According to Gardner, each of these multiple intelligences is more or less independent of the others. Conventional tests of intelligence measure primarily linguistic and logical-mathematical intelligences, and to some extent spatial intelligence, but ignore the other intelligences. Moreover, even the intelligences that are measured are assessed in ways that are very limited, such as through fairly trivial multiple-choice kinds of questions. Thus, these tests can at best give only a limited picture of what children and even adults can do.
I, too, believe that conventional tests give only a limited picture of what people can do, although my beliefs originate in a somewhat different way. In my own work (Sternberg, 1985a, 1988, 1996, 1999b), I have suggested a triarchic theory of intelligence, according to which there are three major aspects of intelligence: analytical, creative, and practical. Conventional tests of intelligence measure primarily analytical abilities, and all but ignore creative and practical abilities. The result is that people who might have very important contributions to make to society may be derailed early in their lives because they do not do well on conventional tests.
I sometimes refer to my own theory as the theory of }successful intelligence," because I place emphasis not just on predicting success in schoolwork, but also, predicting success in life. Intelligence tests were originally created to predict success in school, but there is much more to life than school, of course. The question then arises as to what, exactly, success is. According to my theory, success is what you believe it is, so long as you are defining it in a prosocial way. So it success could be life as a teacher for one person, as a carpenter for another. But being an axe-murderer does not count because it is antisocial rather than prosocial.
Gardner's and my own theories are sometimes presented as though they are "alternative" theories of intelligence. Indeed, in almost all of the introductory-psychology texts, they are presented in this way. But in fact, they deal with somewhat different aspects of intelligence. Gardner's theory deals with domains of intelligence, my own with processes within (or between) domains. Thus, one can think analytically, creatively, or practically, for example, in the linguistic (or any other) domain, as when one analyzes a work of literature (analytic), writes a poem (creative), or discusses the relevance of the travails of a literary character for one's own life (practical). It is thus important in psychology to realize that views that are presented as being in conflict with each other often only have the appearance of being in conflict. We must examine them rather carefully to determine whether the conflict is real or apparent.
Relation to General Issues in Cognitive Psychology
The issue of what intelligence is may seem remote from the concerns of cognitive psychology, in general, but nothing could be further from the case. There are several reasons why this is so.
First, many theorists view intelligence as the central set of abilities that organizes all cognitive functions (see, e.g., Anderson, 1983; Newell 1990; Schank, 1980; Sternberg, 1977, 1996, 1997). Thus, an understanding of what intelligence is would help us understand how all of cognition is organized and effectively brought to bear upon the tasks we face in our daily lives.
Second, the tension between Galton, on the one hand, and Binet, on the other, is a tension that has existed in cognitive psychology since its inception, and that shows no sign of abating. This tension is between those who emphasize the sensory and perceptual functioning characteristic of bottom-up processing (as did Galton), and those who emphasize the judgmental and comprehension characteristic of top-down processing, (as did Binet). In bottom-up processing, one starts information processing at the level of elementary sensory processes, and then builds up to more complex cognitive processes, such as reasoning. In top-down processing, one starts information processing at a complex level, and then works down to more basic processes. The tension between the two kinds of processes can be seen in many areas of cognition. In the field of visual perception, for example, some theorists emphasize bottom-up perceptual processes (e.g., Gibson, 1979) whereas others emphasize top-down perceptual processes (e.g., Rock, 1983). A similar issue arises in speech perception, where some theories are more bottom-up (Massaro, 1987) and others are more top-down (Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967; Liberman & Mattingly, 1985). Theories of reading can also emphasize bottom-up or top-down processing, although modern theories tend to seek a balance between the two emphases (e.g., McClelland & Rumelhart, 1981).
Third, the nature of intelligence raises questions about the extent to which the mind is modular, a key question in contemporary cognitive psychology. Modularity refers to the minds being divided into multiple systems of processing information that are largely independent of each other. For example, is the intelligence you use to do your schoolwork in a different module, or functional part of the brain, than is the intelligence you use to decide whether you like someone? In a much-cited book, Fodor (1983) argued that the mind is largely modular, except for higher intellectual processes. Gardner (1983), however, argued that even the higher intellectual processes are modular. The problem for a theory such as Gardner's, however, is adequately accounting for what Spearman (1904) called the positive manifold--the tendency for tests of higher level intellectual abilities all to intercorrelate positively with each other. At the very least, a hierarchical structure that somehow integrates the lower order modules seems to be necessary to account for these intercorrelations (Carroll, 1993). Gardners theory, proposed more than 15 years ago, still has not generated even a single empirical test. Until it does, the theory must be viewed as intriguing but, from a scientific viewpoint, as highly speculative. In particular, research is needed to know whether the intelligences Gardner has proposed actually exist, are truly independent, and are measurable in reliable and valid ways.
Finally, it is not clear that there even is any one answer to what intelligence is. A cognitive psychologist, Ulrich Neisser (1974), pointed out that intelligence may be best conceived of as a prototype (Rosch, 1978), or as a fuzzy concept. According to this view, we have conceptions of what characteristic features of intelligence might be but there are no necessary features that categorically distinguish intelligence from everything else. We have invented the concept to make sense of differences we see among people in everyday life. Different cultures may have different prototypes for intelligence (Sternberg & Kaufman, 1998).
Relation of the Nature of Intelligence to Everyday Life
The question of whether intelligence comprises a more general ability (Jensen, 1998), perhaps with subsidiary abilities embedded hierarchically beneath it (Carroll, 1993; Cattell, 1971) or instead is modular with no general factor (or at least none of importance; Gardner, 1983; Thurstone, 1938) is not one just of theoretical importance. It has implications for our everyday lives. If intelligence comprises primarily just one ability, then a single IQ (intelligence quotient) score can tell us pretty much what we need to know about a person's intellectual abilities. If intelligence comprises multiple abilities, however, any single number (or two numbers, such as verbal and mathematics scores) will leave us with an inadequate account of a person's profile of intellectual abilities. The fight among theorists, therefore, has implications for the practical realm. People who emphasize multiple abilities may well see people who emphasize just a single ability as reactionary Mr. Hydes who allow their right-wing sociopolitical ideology to corrupt their thinking because of their desire to suppress individuals--often members of minority groups--whose talents are not well captured by a single number. People who emphasize a single ability, however, may see the multiple-ability theorists as radical Mr. Hydes who allow their left-wing sociopolitical ideology to corrupt the integrity of their scientific thinking.
It is important to realize that the same general principle potentially applies to all thinking. Arguments about the ethical justifiability of abortion and capital punishment go on and on, for example, and do not get resolved on the basis of intellectual discourse. Ideology creeps into these arguments, as surely as it creeps into arguments about psychological issues. We all always need to be aware of how our ideology unconsciously shapes our thinking.
Of course, questions about what intelligence is are not the only ones at the forefront of modern psychological investigations. Another key question is how intelligence, whatever it is, should be investigated.
How Should Intelligence be Investigated?
It would seem as though how one would go about investigating something--say, intelligence--would depend on just what the thing is that one was investigating. Thus, the answer to the question of how intelligence should be investigated would seem to depend on what intelligence is, much as the question of how, say, a crime should be investigated would depend on what the crime is. One would not investigate a murder in quite the same way as one would investigate a land swindle. So how should intelligence be investigated?
Importance of Issue
Historically, the question of methods of investigation has, if anything, preceded rather than followed the question of what is being investigated in the field of intelligence. Methods have become available, and they have driven, to a large extent, how intelligence is conceived (see Sternberg, 1990). Although methods can certainly be helpful in driving substantive research (Gigerenzer, 1991), they can also lead to the situation where the methods, rather than the construct, drive research. In other words, the cart is placed before the horse.
To a large extent, how one investigates something can determine what one can find out about that something. For example, if one is investigating a murder, reliance on eyewitness testimony may suggest a suspect different from the suspect suggested by DNA analysis. Neither kind of testimony is infallible: Recall of eyewitnesses is often quite poor (Loftus, 1975; Loftus & Ketcham, 1991), and evidence with DNA in it may be planted at the scene of a crime to point the figure toward a targeted suspect. Similarly no one method for studying intelligence or anything else is infallible: Ideally, we want to use converging operations that help us understand a construct as a result of multiple kinds of analysis (Garner, Hake, & Eriksen, 1956). The importance of method, then, is that it may in large part determine the outcome; but the users of the method may not recognize the contribution of method to the determination of that outcome. This issue is not just hypothetical. Multiple-choice tests tend to correlate with each other, but they do not correlate as well with essay and other performance-based tests. Which kind of test is used has a major effect on how a person's intelligence is evaluated (Sternberg, 1996; Sternberg, Ferrari, Clinkenbeard, & Grigorenko, 1996). Thus, who is identified as intelligent may depend in large part upon the method used to identify different levels of intelligence.
Contemporary Methods for Studying Intelligence
Many different methods have been used to investigate intelligence (Sternberg, 1982, 1990, 2000), but here I will concentrate on just two methodologies--factor analysis, which has its origins in differential psychology (Spearman, 1904), and cognitive analysis, which has its origins in experimental psychology (e.g., Donders, 1868).
The differential approach. Factor analysis is a technique that takes correlations--which represent patterns of individual differences shared across tests or other measurements--and attempts to find the latent mental (or other) structures that underlie, or give rise to these correlations. So, for example, if one were to give tests of vocabulary, verbal analogies, understanding of mathematical concepts, and mathematical problem solving, a factor analysis might plausibly reveal two factors, or latent sources of individual differences--verbal and mathematical abilities. The idea would be that underlying all the tests are just two fundamental abilitiestverbal and mathematical. You could be strong verbally but not mathematically, or vice versa. Or you could be strong or weak in both. Because the factors are separate, the abilities are largely independent.
Factor analysis has been widely used as a basis for understanding intelligence (see Carroll, 1982, 1993; Jensen, 1998), and continues to be a major source of information about abilities. Some psychologists, however, have been less than satisfied with this methodology as a sole or even primary basis for understanding mental abilities (e.g., Sternberg, 1977). Why?
First, the identification of abilities through factor analysis is dependent on there being individual differences in those abilities. If there are no individual differences, there can be no meaningful correlations between pairs of tests, and thus no factors can be revealed. But not all abilities yield salient individual differences. For example, the ability to use language is certainly a part of intelligence. One characteristic that separates human intelligence from that of most or even all other animals is the ability to use language. But virtually all humans have this ability, so the existence of linguistic ability is not susceptible to identification by factor analysis (although variations in levels of it usually will be).
Second, factor analysis may be useful as a structural model, but it typically tells us little or perhaps even nothing about the mental processes underlying intelligence. One can imagine factor analysis as revealing a map of the mind (Sternberg, 1990)--a representation of the terrain of how abilities are distributed. But factor analysis would not tell us how people navigate that terrain, or make effective use of it.
Third, conventional (so-called exploratory) factor analysis does not yield unique solutions. Imagine a map of the world. The locations of the sites on the map (cities, mountains, oceans, or whatever) are fixed, but the axes used to assign meaning to these locations are not. Typically, we use lines of longitude and latitude, so that we can make meaningful judgments about how far north or south, or east or west, a given site is. But there is no particular reason to use lines of longitude and latitude. We might use polar coordinates or some other system of coordinates. Similarly, in factor analysis, interpretation of results depends heavily on where factorial axes are drawn; but as with the map, there are an infinite number of axes we might draw that could be used uniquely to locate sites, in this case, in the factor space.
In sum, factor analysis is far from a perfect method for studying intelligence. Does cognitive analysis provide a more nearly perfect method of analysis, or at least, a better one?
The cognitive approach. The cognitive analysis of intelligence starts in a place quite different from that of the factor (differential) analysis of intelligence. The goal is to unpack variation in difficulties among tasks rather than variation in performance among people. In other words, the parameters that are isolated from the method of analysis are based upon stimulus rather than upon person properties.
For example, in some of my early work, I was interested in the role of inductive reasoning (a major part of what is called "fluid abilities" in the literature on differential psychology) in intelligence. I therefore had experimental participants solve fairly simple analogy, classification, and series-completion problems while they were timed by a machine (Sternberg, 1977; Sternberg, 1983; Sternberg & Gardner, 1983). Items were carefully selected for their stimulus properties so that they would vary in difficulty. For example, an analogy (such as DOCTOR : LAWYER :: PATIENT : a. CLIENT, b. JUDGE) might vary in the complexity of the relation between the first two terms of the analogy, or the first and third, and so forth. Ultimately, processes of inductive reasoning were identified such as inference--required to understand the relation between the first two terms of the analogy--andapplication--required to carry over the rule inferred in the first half of the analogy to solve the problem in the second half. These processes were identified on the basis of differential patterns of response times across items rather than across persons (as would be typical in factor analysis).
In the 1980s, the prospects for what at the time seemed to be a revolutionary approach to intelligence were rosy indeed. Those of us engaged in the cognitive analysis of intelligence (such as Carroll, 1976; Hunt, 1980; Pellegrino & Glaser, 1980; and Snow, 1980) thought that we had found the cure for the ills of the field of intelligence. Perhaps we were insensitive, though, to the dialectical nature of science, in general, and psychology, in particular (Sternberg, 1995, 1999a).
Relation to General Issues in Cognitive Psychology
In the dialectic, a given approach to a problem is offered--say, the psychometric one. In the terms of the dialectic, it is called a thesis. Proponents of it view themselves as the Jekylls of the field: They have an answer--one that will help enlighten and benefit all. But not everyone sees the thesis so positively. Skeptics come along, and view those promoting the thesis as the Hydes of the field. At best, they are unenlightened; at worst, they are reactionary enemies of progress. These skeptics pose an antithesis that is critical of the thesis on one and usually more than one dimension. Thus, the cognitivists viewed themselves as offering an antithesis to the differentialists--perhaps even the panacea intelligence research had been looking for.
As always happens, the supposed panacea proves to be nothing of the sort. Critics internal and external to the antithetical movement begin to see flaws. Thus, in my own work, I began to see cognitive psychologists as using pretty much the same kinds of test questions as had differentialists, with the main difference between the two groups in the way they were dividing up the same variance (Sternberg, 1985a). Moreover, information-processing analyses of intelligence were not working out the way they were supposed to. Often, the variation that best correlated with conventional psychometric tests was that in the regression constant--what was common to all of the test items being studied. It seemed almost as though general ability (g) had been rediscovered. Finally, just as some of the factor analysts had seemed to be mindlessly applying factor analysis to every data set in sight, so were cognitivists shown to be capable of applying cognitive analysis to every task in sight without asking whether the task was the right one to use. Cognitive analysis became just as task-based as had factor analysis (Sternberg, 1997).
Various attempts have been made to achieve some kind of synthesis between the differential and cognitive approaches, a call made over 50 years ago by Cronbach (1957). For example, my own triarchic approach represents one attempt at such a synthesis (Sternberg, 1985a). But there is no one right synthesis, and in any case, the whole idea of the dialectic is that the synthesis will eventually become a new thesis, itself to be criticized by those who will propose an antithesis, and so on.
This kind of dialectic is by no means limited to intelligence, cognitive psychology, or even psychology. It seems to be an aspect of reality in many and perhaps all fields of endeavor. It means, though, that we have to be careful in viewing our current paradigms or facts accumulated within these paradigms as in any sense final. Look at the textbooks of today and of ten, twenty, or thirty years ago, and you will see that general ways of thinking stay the same, but not paradigms and the facts accumulated within them. Most of those facts will never be shown to be wrong or other disproved. Rather, they will just come to seem uninteresting. The problems of interest to psychologists will change, and with them, the answers.
Truly, much of life has a dialectical property. At times, we tend to focus on one thing; then we realize we were focusing too much on that thing, so we try to focus on something different. Then we realize we need to integrate the two. For example, we may focus too much on our work at one point, and then rebel and focus too much on our personal life. Ultimately, we need to find an integration that incorporates both of these foci.
Relation to the Nature of Intelligence in Everyday Life
Methods used to study intelligence may seem to involve an issue quite distant from the concerns of everyday life. In fact, though, they are not so distant at all. If the assumptions underlying the psychometric analysis of intelligence are wrong, then the millions of ability tests given each year under different acronyms--for admission to private primary and secondary schools, or to colleges, graduate schools, or professional schools, not to mention for placement and diagnosis--are faulty and are possibly leading to questionable decisions. Where intelligence is most important is in its interface with everyday life (Sternberg et al., 2000).
Indeed, there is reason to believe that some of the assumptions underlying our use of tests are wrong. The factor structure of a test may be the same across cultures or subcultures, but it may be that the whole conception of intelligence differs from one culture or subculture to another, so that the test is not really measuring the same thing across cultures. For example, Okagaki and Sternberg (1993) found that different ethnic groups have different conceptions of intelligence, and that they socialize their children according to these conceptions. We found, for example, that Latinos tended to emphasize the social-competence aspect of intelligence more, whereas Anglos tended to emphasize the cognitive-competence aspect of intelligence more. These conceptions fit the school conception of intelligence to a greater or lesser degree. To the extent the fit is lesser, bright children may appear to be dull to their teachers. Even within the mainstream, different occupations have different conceptions of what it means to be intelligent in their field (Sternberg, 1985b). And in different countries, different views of intelligence may predominate (Yang & Sternberg, 1997). Indeed, many languages, such as Chinese or Hebrew, have no word in the languages that corresponds well to the word intelligence.
In sum, what might seem like arcane issues of methodology have a real impact on what happens from day to day in children's lives, and what opportunities are given to or taken away from them--on whether, as psychologists, we serve as Jekylls or Hydes. Psychologists and others consequently bear a major burden in ensuring that the tests of intelligence they are providing adequately represent the construct they are supposed to be assessing.
The battles over what intelligence is and what methods should be used to study it represent only two of the many issues that intelligence researchers will have to address during the years of the 21st century. Several other major issues loom on the horizon, and their answers are potentially high-stakes one. These are the issues that are being hotly researched today and are likely to be hotly investigated at least in the foreseeable future.
One of these major issues is the relation between intelligence, on the one hand, and heredity and environment, on the other. Although the issue is an old one, new research on it is proceeding at a rapid clip, and shows no sign of abating (see Sternberg & Grigorenko, 1997). Most researchers have passed the point where they merely wish to attach percentages to hereditary and environmental contributions to intelligence. They realize the important roles of covariance (forces that lead heredity and environment to have the same effects) and interaction. Various kinds of behavior-genetic designs have been proposed to tease out the various kinds of effects heredity and environment can have on intelligence, but all of them have their own idiosyncratic flaws. Almost certainly, converging operations are more informative than are single ones. But we need to remember that, regardless of the methodology used, the conclusion drawn can be as good only as the tests from which one draws those conclusions.
Recent findings are quite intriguing (see Plomin, 1997; Sternberg & Grigorenko, 1997). For example, we now know that heritability of intelligence, at least as measured by conventional tests of intelligence, increaseswith age. Most investigators previously had thought that it would decrease as environment had more and more of an effect. Instead, it appears that as time goes on, environment matters less and effects of genes matter more. Another interesting effect is that within-family rather than between-family differences appear to matter most for the development of intelligence. In other words, to the extent that differences in environment have an effect on the development of intelligence, it appears that the differences that matter most are those in the way different children within a given family are treated rather than differences in the way children are treated across families.
Some investigators, such as Plomin (1997), are attempting to go beyond quantitative-genetic studies to actual identification of the genes that are responsible for various aspects of intelligent behavior. Plomin believes that such research may eventually supplant the more conventional statistical studies. Perhaps, but at the moment, such studies yield little more than educated speculations. Links between genes and intelligence are so weak that it is unclear how much time it will take before we will be able to establish any strong and meaningful links. Certainly, such links would be important to establish, but they have yet to be discovered. Studies of the genotype of intelligence have the potential to bring us either into the land of Jekyll or that of Hyde. Will we use such knowledge to benefit humankind, or to justify the already questionable treatment of certain groups? Only time will tell.
A second major issue is group differences in intelligence. Although such differences are widely accepted as a fact of life (Sternberg, 1997), they depend on one's accepting the conventional notion of what intelligence is. And as the work of Herrnstein and Murray (1994) made clear, perhaps more important than the existence of such differences is their interpretation. The evidence in favor of a genetic interpretation for them is quite weak (Nisbett, 1995), but again, it is difficult to predict what the future will tell. At present, our best guess is that most differences between groups are better attributable to socialization than to genetic effects, or perhaps to covariance and interaction between socialization and genetics. But more research is needed to understand how socialization has its effects.
A third issue that I will mention is the modifiability of intelligence. Some investigators are utterly convinced that intelligence is modifiable (e.g., Feuerstein, 1980; Grotzer & Perkins, 2000; Nickerson, 1994; Ramey, 1994; Sternberg, 1996), at least in some degree, whereas other investigators are equally convinced that intelligence is minimally or not at all modifiable (e.g., Herrnstein & Murray, 1994; Jensen, 1969, 1998). It may be that we have not gone far in understanding the modifiability of intelligence because the kinds of tests we use do not give students much of an opportunity to learntthat is, to modify themselvestwhile they are being tested. Alternative testing techniques, called dynamic testing techniques, allow students to learn at the time they are tested (Grigorenko & Sternberg, 1998; Vygotsky, 1978). In such tests, individuals learn material while they are being tested, and then are tested on how much they have learned at the time of test. These techniques may give us a better understanding of peoples ability to modify their intelligence.
Whatever the answer, it is important to remember that heritability and modifiability are two completely independent issues. Something can be heritable and either modifiable or not. For example, height is highly heritable, and has also shown substantial modification in recent decades.
The distinction between modifiability and heritability, and yet its confusion not only in the mind of the public but even of specialists in the field, is noteworthy. To many people, it simply seems like a matter of logic that if something is inherited, then it is nonmodifiable. These people view genes as opposed to environment. Yet, this view is wrong. Even IQ, which appears to have a heritability of perhaps .5, has been shown to be modifiable if only because research shows that scores on tests that give rise to IQs have risen substantially over the past several generations (Flynn, 1987, 1994; Neisser, 1998). There is a lesson to be learned, and it is that we often make assumptions not because they are right, but because they are easy to make and sound so darn plausible.
The last issue I mention is the one that excites me the most: how to improve achievement in school and society by taking into account differences not only in amounts of intelligence, but in profilesof intelligence. Our research shows that if one teaches in a way that enables students to capitalize on their patterns of analytical, creative, and practical abilities, the achievement of the students increases (Sternberg, Ferrari, Clinkenbeard, & Grigorenko, 1996; Sternberg, Grigorenko, Ferrari, & Clinkenbeard, 1999). Moreover, teaching all students in a way that enables them to use their analytical, creative, and practical thinking and learning skills appears to result in higher school achievement for all students (Sternberg, Torff, & Grigorenko, 1998). It may be possible in the future to help students improve their achievement by teaching in ways that expand upon the teaching repertoires that most teachers currently use.
The field of intelligence is one of the most exciting to work in, because the stakes are so high, both theoretically and practically. But it is not a field for just anyone. Precisely because the stakes are so high, people who decide to join the fray need a thicker than average skin. There are so many different points of view that it is not a field in which there is any kind of work that is likely to please everyone; unless an investigator is prepared to take a certain amount of flack, he or she would do better finding another pursuit.
Working meaningfully in the field of intelligence also requires a broader background than might be the case in another field. Work in this field cross-cuts cognitive psychology, biological psychology, developmental psychology, differential psychology, educational psychology, personality psychology, cultural psychology, industrial-organizational psychology, and perhaps other areas of psychology as well. To keep up with the field and advance it, one must be able to understand and to integrate the contributions from these various aspects of the field.
Finally, I believe that the best work in this field, and perhaps any field, is that done by people who are willing to defy the crowd--to generate their own set of grounded beliefs and to fight for them (Sternberg & Lubart, 1995). Students like to believe that science is for the courageous, the willful, the strong of mind, and the towering intellect that will fight for truth. None of these stereotypes works awfully well, as Kuhn (1970) pointed out some years ago. For the most part, scientists follow the paradigms set by others, accept what they are told, and fill in the small gaps left to be filled by others. The result is that much and arguably most of the work that is done has little impact on a given field; the field would have changed little if at all had the work never been done.
We have no final truth in this or any other area of psychology. At best, we have good theories that will lead to new and hopefully better theories. It is the responsibility of the next generation of researchers to take up the dialectic where the previous generation left off--to build on these past theories, even if building on them means attacking them in the process. Perhaps curiously, even work that forms an antithesis to an existing thesis builds on the past work, because without the thesis, the antithesis could not exist--there would be nothing to attack. We need to encourage students to be scientific but bold, innovative but responsible--in other words, to show in their work the kinds of broader attributes of the intelligent people that the researchers themselves study.
Most of all, researchers in this field need to remember the Jekyll and Hyde character of work in their field. Work in the field of intelligence can do enormous good in providing opportunities for those who would not otherwise have them, or enormous harm in stealing opportunities from those who truly deserve them. Researchers need to think not only about the scientific contribution they have to make, but the contribution to science and society that they leave behind them.
Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press.
Binet, A., & Simon, T. (1916). The intelligence of the feeble-minded (E. S. Kite, Trans.). Baltimore: Williams & Wilkins.
Carroll, J. B. (1976). Psychometric tests as cognitive tasks: A new }structure" In L. B. Resnick (Ed.), The nature of intelligence. Hillsdale, NJ: Erlbaum.
Carroll, J. B. (1982). The measurement of intelligence. In R. J. Sternberg (Ed.), Handbook of human intelligence (pp. 29-120). New York: Cambridge University Press.
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. New York: Cambridge University Press.
Cattell, R. B. (1971). Abilities: Their structure, growth and action. Boston: Houghton Mifflin.
Cronbach, L. J. (1957). The two disciplines of scientific psychology. American Psychologist, 12, 671-684.
Donders, F.C. (1868). Over de snelheid van psychische processen. Onderzoekingen gedaan in het Physiologisch Laboratorium der Utrechtsche Hoogeschool, 1868-1869, Tweede reeks, II, 92-120.
Feurfstein, R. (1980). Instrumental enrichment. Baltimore, MD: University Park Press.
Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101, 171-191
Flynn, J. R. (1994). IQ gains over time. In R. J. Sternberg (Ed.), Encyclopedia of human intelligence (Vol. 1, pp. 617-623). New York: Macmillan.
Fodor, J. A. (1983). The modularity of mind. Cambridge, MA: MIT Press.
Galton, F. (1883). Inquiry into human faculty and its development. London: Macmillan.
Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic.
Gardner, H. (1993). Multiple intelligences: The theory in practice. New York: Basic Books.
Gardner, H. (1999). Intelligence reframed. New York: Basic Books.
Garner, W. R., Hake, H. W., & Erikson, C. W. (1956). Operationism and the concept of perception. Psychological Review, 63, 149-159.
Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton Mifflin.
Gigerenzer, G. (1991). From tools to theories: A heuristic of discovery in cognitive psychology. Psychological Review, 98, 254-267.
Gould, S. J. (1981). The mismeasure of man. New York: W. W. Norton.
Grigorenko, E. L., & Sternberg, R. J. (1998). Dynamic testing. Psychological Bulletin, 124, 75-111.
Grotzer, T. A., & Perkins, D. N. (2000). Teaching intelligence: A performance conception. In R. J. Sternberg (Ed.), Handbook of intelligence (pp. 492-515). New York: Cambridge University Press.
Herrnstein, R., & Murray, C. (1994). The bell curve. New York: Free Press.
Hunt, E. B. (1980). Intelligence as an information-processing concept. British Journal of Psychology, 71, 449-474.
Jensen, A.R. (1969). How much can we boost IQ and scholastic achievement? Harvard Educational Review, 39, 1-123.
Jensen, A. R. (1998). The g factor. Westport, CT: Praeger-Greenwood.
Kuhn, T. S. (1970). The structure of scientific revolutions (2nd ed.). Chicago: University of Chicago Press.
Liberman, A. M., Cooper, F. S., Shankweiler, D. P., & Studdert-Kennedy, M. (1967). Perceptions of the speech code. Psychological Review, 74, 431-461.
Liberman, A. M., & Mattingly, I. G. (1985). The motor theory of speech perception revised. Cognition, 21, 1-36.
Loftus, E. F. (1975). Leading questions and the eyewitness report. Cognitive Psychology, 7, 560-572.
Loftus, E. F., & Ketcham, K. (1991). Witness for the defense: The accused, the eyewitness, and the expert who puts memory on trial.New York: St. Martins Press.
Massaro, D. W. (1987). Speech perception by ear and eye: A paradigm for psychological inquiry. Hillsdale, NJ: Earlbaum.
McClelland, J. L., & Rumelhart, D. E. (1981). An interactive activation model of context effects in letter perception: Part I. An account of basic findings. Psychological Review, 88, 483-524.
Neisser, U. (1979). The concept of intelligence. Intelligence,3, 217-227.
Neisser, U. (Ed.) (1998). The rising curve. Washington, DC: APA Books.
Newell, A. (1990). Unified theories of cognition. Cambridge, MA: Harvard University Press.
Nisbett, R. (1995). Race, IQ, and scientism. In S. Fraser (Ed.), The bell curve wars (pp. 36-57). New York: Basic Books.
Okagaki, L., & Sternberg, R. J. (1993). Parental beliefs and children's school performance. Child Development,64(1), 36-56.
Pellegrino, J. W., & Glaser, R. (1980). Components of inductive reasoning. In R. E. Snow, P-A Federico, & W. E. Montague (Eds), Aptitude, learning, and instruction (Vol. 1): Cognitive process analyses of aptitude. Hillsdale, NJ: Lawrence Erlbaum Associates.
Plomin, R. (1997). Identifying genes for cognitive abilities and disabilities. In R.J. Sternberg & E.L. Grigorenko (Eds.), Intelligence, heredity, and environment (pp. 89-104). New York: Cambridge.
Ramey, C.T. (1994). Abecedarian project. In R.J. Sternberg (Ed.), Encyclopedia of human intelligence (Vol. 1, pp. 1-3).Rock, I. (1983). The logic of perception. Cambridge, MA: MIT Press.
Rosch, E. (1978). Principles of categorization. In E. Rosch & B.B. Lloyd (Eds.), Cognition and categorization. Hillsdale, NJ: Erlbaum.
Schank, R. C. (1980). How much intelligence is there in artificial intelligence? Intelligence, 4, 1-4.Snow, R. E. (1980). Aptitude processes. In R. E. Snow, P.-A. Federico, & W. E. Montague (Eds.), Aptitude, learning, and instruction: Cognitive process analyses of aptitude (Vol. 1, pp. 27-63). Hillsdale, NJ: Erlbaum.
Spearman, C. E. (1904). 'General intelligence ' objectively determined and measured. American Journal of Psychology, 15, 201-293.
Sternberg, R. J. (1977). Intelligence, information processing, and analogical reasoning: The componential analysis of human abilities. Hillsdale, NJ: Erlbaum.
Sternberg, R. J. (Ed.) (1982.) Handbook of human intelligence. New York: Cambridge University Press.
Sternberg, R. J. (1982).A componential approach to intellectual development. In R. J. Sternberg (Ed.), Advances in the psychology of human intelligence (Vol. 1, pp. 413j463).Hillsdale, NJ:Lawrence Erlbaum.
Sternberg, R. J. (1983).Components of human intelligence.Cognition, 15, 1-48.
Sternberg, R. J. (1985a). Beyond IQ: A triarchic theory of human intelligence. New York: Cambridge University Press.
Sternberg, R. J. (1985b). Implicit theories of intelligence, creativity, and wisdom. Journal of Personality and Social Psychology, 49, 607-627.
Sternberg, R. J. (1988). The triarchic mind: A new theory of human intelligence. New York: Viking.
Sternberg, R. J. (1990). Metaphors of mind: Conceptions of the nature of intelligence. New York: Cambridge.
Sternberg, R. J. (1995). In search of the human mind. Orlando: Harcourt Brace College Publishers.
Sternberg, R. J. (1996). Successful intelligence. New York: Simon & Schuster.
Sternberg, R. J. (1997). The concept of intelligence and its role in lifelong learning and success. American Psychologist, 52, 1030-1037.
Sternberg, R. J. (1998). A balance theory of wisdom. Review of General Psychology, 2, 347-365.
Sternberg, R. J. (1999a). A dialectical basis for understanding the study of cognition. In R. J. Sternberg (Ed.), The nature of cognition(pp. 51-78).Cambridge, MA: The MIT Press.
Sternberg, R. J. (1999b). The theory of successful intelligence. Review of General Psychology, 3, 292-316.
Sternberg, R. J. (Ed.) (2000). Handbook of intelligence.New York: Cambridge University Press.
Sternberg, R. J., Ferrari, M., Clinkenbeard, P. R., & Grigorenko, E. L. (1996). Identification, instruction, and assessment of gifted children: A construct validation of a triarchic model.Gifted Child Quarterly, 40(3), 129-137.
Sternberg, R. J., Forsythe, G. B., Hedlund, J., Horvath, J., Snook, S., Williams, W. M., Wagner, R. K., & Grigorenko, El. L. (2000). Practical intelligence in everyday life. New York: Cambridge University Press.
Sternberg, R. J., & Gardner, M. K.(1983). Unities in inductive reasoning. Journal of Experimental Psychology: General,112, 80-116.
Sternberg, R. J., Grigorenko, E. L., Ferrari, M., & Clinkenbeard, P. (1999). Triarchic analysis of an aptitude-treatment interaction. European Journal of Psychological Assessment, 15,1-11.
Sternberg, R. J., & Kaufman, J. (1998). Human abilities. Annual Review of Psychology, 49, 479-502.
Sternberg, R.J., & Lubart, T.I. (1995).Defying the crowd: cultivating creativity in a culture of conformity.New York: Free Press.
Sternberg, R.J., & Grigorenko, E.L. (Eds.) (1997). Intelligence, heredity, and environment. New York: Cambridge University Press.
Sternberg, R. J., Torff, B., & Grigorenko, E. L. (1998). Teaching triarchically improves school achievement.Journal of Educational Psychology, 90, 374-384.
Thurstone, L. L. (1938). Primary mental abilities. Chicago, IL: University of Chicago Press.
Yang, S.-Y., & Sternberg, R.J. (1997). Taiwanese Chinese peoples conceptions of intelligence. Intelligence, 25, 21-36.
Preparation of this article was supported under the Javits Act Program (Grant No. R206R950001) as administered by the Office of Educational Research and Improvement, U.S. Department of Education. Grantees undertaking such projects are encouraged to express freely their professional judgment. This article, therefore, does not necessarily represent the position or policies of the Office of Educational Research and Improvement or the U.S. Department of Education, and no official endorsement should be inferred.
Requests for reprints should be sent to Robert J. Sternberg, Department of Psychology, Yale University, P.O. Box 208205, New Haven, CT 06520-8205
Robert J. Sternberg is IBM Professor of Psychology and Education in the Department of Psychology at Yale University. He received the Ph.D. from Stanford University in 1975 and the B.A. summa cum laude, Phi Beta Kappa, from Yale University in 1972. He also holds an honorary doctorate from the Complutense University of Madrid and this school year will be awarded honorary doctorates by the University of Leuven, Belgium; the University of Cyprus; and the University of Paris V, France. Sternberg is the author of over 800 journal articles, book chapters, and books, and has received about $15 million in government grants and contracts for his research. The central focus of his research is on intelligence and cognitive development. Sternberg is a Fellow of the American Academy of Arts and Sciences, the American Association for the Advancement of Science, the American Psychological Association (in 12 divisions), and the American Psychological Society. He is listed in Who's Who in America and numerous other who's whos. He has won many awards from APA, AERA, APS, and other organizations. These awards include the Early Career Award and Boyd R. McCandless Award from APA; the Palmer O. Johnson, Research Review, Outstanding Book, and Sylvia Scribner Awards from AERA; the James McKeen Cattell Award from APS; the Distinguished Lifetime Contribution to Psychology Award from the Connecticut Psychological Association; the International Award of the Association of Portuguese Psychologists; the Cattell Award of the Society for Multivariate Experimental Psychology; the Award for Excellence of the Mensa Education and Research Foundation; the Distinction of Honor SEK, from the Institucion SEK (Madrid); the Sidney Siegel Memorial Award of Stanford University; and the Wohlenberg Prize of Yale University. He has held a Guggenheim Fellowship as well as Yale University Senior and Junior Faculty Fellowships. He also has held the Honored Visitor Fellowship of the Taiwan National Science Council and the Sir Edward Youde Memorial Visiting Professorship of the City University of Hong Kong. He has been listed in the Esquire Register of outstanding men and women under 40 and was listed as one of 100 top young scientists by Science Digest. He has been president of the Divisions of General Psychology, Educational Psychology, Psychology and the Arts, and Theoretical and Philosophical Psychology of the APA and has served as Editor of the Psychological Bulletin and is Editor of Contemporary Psychology. Sternberg is currently PI on a project with the College Board. Sternberg has consulted for hundreds of educational organizations on projects related to teaching and testing for intellectual skills. He is most well-known for his theory of successful intelligence, investment theory of creativity (developed with Todd Lubart), theory of mental self-government, balance theory of wisdom, and for his triangular theory of love and his theory of love as a story. Sternberg is a member of the Trustees Research Committee of the College Board and has served on a Research Advisory Committee for ETS.
The author may be reached at Robert_Sternberg@yale.edu