Handbook of psychology volume 7 educational psychology

30

Contemporary Theories of Intelligence

Africa views people of higher social class and distinction as

speaking less (Irvine, 1978). This difference between the

Wolof and Western notions suggests the usefulness of look-

ing at African notions of intelligence as a possible contrast to

U.S. notions.

In fact, studies in Africa provide yet another window on

the substantial differences. Ruzgis and Grigorenko (1994)

have argued that, in Africa, conceptions of intelligence re-

volve largely around skills that help to facilitate and maintain

harmonious and stable intergroup relations; intragroup rela-

tions are probably equally important and at times more im-

portant. For example, Serpell (1974, 1982, 1993) found that

Chewa adults in Zambia emphasize social responsibilities,

cooperativeness, and obedience as important to intelligence;

intelligent children are expected to be respectful of adults.

Kenyan parents also emphasize responsible participation in

family and social life as important aspects of intelligence

(Super, 1983; Super & Harkness, 1982). In Zimbabwe, the

word for intelligence, ngware, actually means to be prudent

and cautious, particularly in social relationships. Among the

Baoule, service to the family and community and politeness

toward and respect for elders are seen as key to intelligence

(Dasen, 1984). 

Similar emphasis on social aspects of intelligence has

been found as well among two other African groups, the

Songhay of Mali and the Samia of Kenya (Putnam &

Kilbride, 1980). The Yoruba, another African tribe, empha-

size the importance of depth—of listening rather than just

talking—to intelligence, and of being able to see all aspects

of an issue and of being able to place the issue in its proper

overall context (Durojaiye, 1993). 

The emphasis on the social aspects of intelligence is not

limited to African cultures. Notions of intelligence in many

Asian cultures also emphasize the social aspect of intelli-

gence more than does the conventional Western or IQ-based

notion (Azuma & Kashiwagi, 1987; Lutz, 1985; Poole, 1985;

White, 1985).

It should be noted that neither African nor Asian cultures

emphasize exclusively social notions of intelligence. In one

village in Kenya (near Kisumu), many and probably most of

the children are at least moderately infected with a variety of

parasitic infections. As a result, they experience stom-

achaches quite frequently. Traditional medicine suggests the

usefulness of a large variety (actually, hundreds) of natural

herbal medicines that can be used to treat such infections. It

appears that at least some of these—although perhaps a small

percentage—actually work. More important for our pur-

poses, however, children who learn how to self-medicate via

these natural herbal medicines are viewed as being at an

adaptive advantage over those who do not have this kind of

informal knowledge. Clearly, the kind of adaptive advantage

that is relevant in this culture would be viewed as totally

irrelevant in the West, and vice versa. 

Grigorenko and her colleagues (2001) have studied con-

ceptions of intelligence in this village in some detail. There

appear to be four parts to the conception.

First, the concept of rieko can be translated as intelligence,

smartness, knowledge, ability, skill, competence, and power.

Along with the general concept of rieko, the Luo people

distinguish among various specialized representations of this

concept. Some representations are characterized by the

source of rieko: rieko mar sikul (knowledge acquired in

school), or rieko mzungu (the White man’s technical powers);

others by different domains of action: rieko mar ot (compe-

tence in household tasks, including planning skills and re-

source management), or rieko mar kite (being versed in

traditional customs and rules). Other representations are

characterized by specific outcomes, such as rieko mar lupo

(fishing skills, including knowledge of magic to provide rich

catches), rieko mar yath (knowledge of healing with herbal

medicines), and so forth. 

Luoro is the second main quality of children and people in

general. It encompasses a whole field of concepts roughly

corresponding to social qualities such as respect and care for

others, obedience, diligence, consideration, and readiness to

share. Luoro has an unequivocal positive meaning and was

always mentioned as a necessity in response to questions

such as “What is most important for a good child to have?”

and “What should people have to lead a happy life?” When

people were asked to compare the relative importance for an

individual’s life of rieko and luoro, respondents generally

gave preference to luoro. It is interesting that the only two re-

spondents ranking rieko higher than luoro were outsiders to

the local community who had a tertiary education and con-

siderable wealth by village standards. Rieko and luoro  are

complementary. Rieko is a positive attribute only if luoro is

also present. Ideally, the power of pure individual abilities

should be kept under control by social rules. 

Third, paro overlaps with both luoro and rieko and,

roughly translated, means thinking. Specifically, paro refers

to the thought processes required to identify a problem and its

solution and to the thought processes involved in caring for

other people. A child with good thinking ( paro maber) could

thus, for example, be a child who is able to react rationally in

case of another person’s accident or one who is able to collect

wood, burn charcoal, and sell it favorably in order to help his

old grandmother. The concept of paro stresses the procedural

nature of intelligence. In essence, paro occupies an interme-

diate position between the potentiality of rieko (its ability as-

pects) and the partially moral connotation of an outcome (the

deed) done with or without luoro. Paro also reflects the idea

of initiative and innovation, for example, in designing a new

Contemporary Theories of Intelligence

31

technical device. Paro encompasses the process of thinking,

the ability to think, and the specific kind of thinking that an

individual demonstrates. 

Fourth, winjo, like paro, is linked to both rieko and luoro.

Winjo means comprehending and understanding. It points to

the child’s abilities to comprehend, that is, to process what is

said or what is going on. But it also involves the ability to

grasp what is appropriate and inappropriate in a situation,

that is, to understand and do what you are told by adults or to

derive from the situation what is appropriate to do. It shares

with the other key terms the feature that its meaning is a func-

tion of context. For a teacher in school it means that a child

runs an errand as told. In contrast, a grandmother teaching a

child about healing might emphasize the aspect of procedural

learning combined with attention to another person. 

A “good child” as well as a “good community member”

needs a balanced mixture of all positive qualities, in which

the contradictory aspects counterbalance each other. Specifi-

cally, the ambiguous powers of individual rieko (which could

be either positive or negative) need to be controlled by social

values and rules (luoro).

These conceptions of intelligence emphasize social skills

much more than do conventional U.S. conceptions of intelli-

gence, but at the same time they recognize the importance of

cognitive aspects of intelligence. It is important to realize,

again, that there is no one overall U.S. conception of intelli-

gence. Indeed, Okagaki and Sternberg (1993) found that dif-

ferent ethnic groups in San Jose, California, had rather

different conceptions of what it means to be intelligent. For

example, Latino parents of schoolchildren tended to empha-

size the importance of social-competence skills in their con-

ceptions of intelligence, whereas Asian parents tended rather

heavily to emphasize the importance of cognitive skills.

Anglo parents also emphasized cognitive skills more. Teach-

ers, representing the dominant culture, emphasized cognitive

skills more than social-competence skills. The rank order of

children of various groups’ performances (including sub-

groups within the Latino and Asian groups) could be per-

fectly predicted by the extent to which parents shared the

teachers’ conceptions of intelligence. In other words, teach-

ers tended to reward those children who were socialized into

a view of intelligence that happened to correspond to the

teachers’ own. 

Explicit Theories

A Psychometric Theory

The psychometric approach to intelligence is among the old-

est of approaches, dating back to Galton’s (1883) psy-

chophysical theory of intelligence in terms of psychophysical

abilities (such as strength of hand grip or visual acuity) and

later to Binet and Simon’s (1905/1916) theory of intelligence

as judgment, involving adaptation to the environment, direc-

tion of one’s efforts, and self-criticism.

Carroll (1993) has proposed a hierarchical model of intel-

ligence, based on a factor analysis of more than 460 data sets

obtained between 1927 and 1987. His analysis encompasses

more than 130,000 people from diverse walks of life and

even countries of origin (although non-English-speaking

countries are poorly represented among his data sets). The

model Carroll proposed, based on his monumental undertak-

ing, is a hierarchy comprising three strata: Stratum I, which

includes many narrow, specific abilities (e.g., spelling ability,

speed of reasoning); Stratum II, which includes various

group-factor abilities (e.g., fluid intelligence, involved in

flexible thinking and seeing things in novel ways; and crys-

tallized intelligence, the accumulated knowledge base); and

Stratum III, which is just a single general intelligence, much

like Spearman’s (1904) general intelligence factor. 

Of these strata, the most interesting is perhaps the middle

stratum, which includes (in addition to fluid and crystallized

abilities) learning and memory processes, visual perception,

auditory perception, facile production of ideas (similar to

verbal fluency), and speed (which includes both sheer speed

of response and speed of accurate responding). Although

Carroll does not break much new ground, in that many of

the abilities in his model have been mentioned in other theo-

ries, he does masterfully integrate a large and diverse factor-

analytic literature, thereby giving great authority to his

model. At the same time, his meta-analysis assumes that con-

ventional psychometric tests cover the entire domain of intel-

ligence that needs to be covered by a theory of intelligence.

Some theorists, discussed next, question this assumption.

Cognitive Theories

Cronbach (1957) called for a merging of the two disciplines

of scientific psychology: the differential and experimental

approaches. The idea is that the study of individual differ-

ences (differential psychology) and of cross-individual com-

monalities (experimental psychology) need not be separate

disciplines. They can be merged.

Serious responses to Cronbach came in the 1970s, with

cognitive approaches to intelligence attempting this merger.

Two of the responses were the cognitive-correlates approach

to intelligence and the cognitive-correlates approach.

Hunt, Frost, and Lunneborg (1973; see also Hunt,

Lunneborg, & Lewis, 1975) introduced the cognitive-

correlates approach, whereby scores on laboratory cognitive

tests were correlated with scores on psychometric intelli-

gence tests. The theory underlying this work was that fairly

32

Contemporary Theories of Intelligence

simple components of information processing studied in the

laboratory—such as the time to retrieve lexical information

from long-term memory—could serve as a basis for under-

standing human intelligence. Intelligence tests, on this view,

present complex problems whose solution nevertheless relies

on fairly simple information processing. Thus, a participant

in a cognitive study might be asked whether two letters, A

and a, are identical in identity (answer: yes) or identical in

case (answer: no). The tasks were directly out of the literature

of experimental psychology, including the letter-comparison

task, which is based on work by Posner and Mitchell (1967). 

Sternberg (1977; see also Sternberg, 1983) introduced

the cognitive-components approach, whereby performance

on complex psychometric tasks was decomposed into ele-

mentary information-processing components. The underlying

theory was that intelligence comprises a series of component

information processes. In contrast to the cognitive-correlates

approach, however, the underlying components were seen as

complex rather than as simple. For example, solving an anal-

ogy of the form A : B :: C : ? involves components such as

encoding the terms, inferring the relation between A and B,

applying this relation from C to ?, and so forth (see review by

Lohman, 2000).

The cognitive approaches of Hunt and Sternberg are now

primarily of historical interest. Both authors have expanded

their conceptualizations of intelligence since this work. They

were forced to do so. Neither approach yielded consistently

high correlations between the tasks and task components

and psychometric tests of intelligence used as criteria.

Moreover, sometimes the components showing the highest

correlations were the ones least expected to show them.

Sternberg and Gardner (1983), for example, consistently

found the regression-constant component to have the highest

correlations with psychometric test scores, leading them to

wonder whether they had rediscovered through information-

processing analysis the general factor that had been discovered

through psychometric analysis.

In the 1990s cognitive and biological approaches (dis-

cussed next) began to merge (Vernon, Wickett, Bazana, &

Stelmack, 2000). A prototypical example is the inspection-

time task (Nettlebeck, 1982; see reviews by Deary, 2000;

Deary & Stough, 1996). In this task, two adjacent vertical

lines are presented tachistoscopically or by computer, fol-

lowed by a visual mask (to destroy the image in visual iconic

memory). The two lines differ in length, as do the lengths of

time for which the two lines are presented. The participant’s

task is to say which line is longer. But instead of using raw re-

sponse time as the dependent variable, investigators typically

use measures derived from a psychophysical function esti-

mated after many trials. For example, the measure might be

the duration of a single inspection trial at which 50% accu-

racy is achieved. Correlations between this task and measures

of IQ appear to be about .4, a bit higher than is typical in psy-

chometric tasks. Much of this correlation may be mediated

by the visual ability component of intelligence (Gv). There

are differing theories as to why such correlations are ob-

tained. All such theories generally attempt to relate the cog-

nitive function of visual inspection time to some kind of

biological function, such as speed of neuronal conduction.

Let us consider, then, some of the biological functions that

may underlie intelligence.

Biological Theories

An important approach to studying intelligence is to under-

stand it in terms of the functioning of the brain, in particular,

and of the nervous system, in general. Earlier theories relat-

ing the brain to intelligence tended to be global in nature, al-

though they were not necessarily backed by strong empirical

evidence. Because these earlier theories are still used in con-

temporary writings and, in the case of Halstead and Luria,

form the bases for test batteries still in contemporary use,

they are described here briefly.

Early Biological Theories.

Halstead (1951) suggested

that there are four biologically based abilities, which he

called (a) the integrative field factor, (b) the abstraction

factor, (c) the power factor, and (d) the directional factor.

Halstead attributed all four of these abilities primarily to the

functioning of the cortex of the frontal lobes. 

More influential than Halstead has been Hebb (1949), who

distinguished between two basic types of intelligence: Intelli-

gence A and Intelligence B. Hebb’s distinction is still used by

some theorists. According to Hebb, Intelligence A is innate

potential, and Intelligence B is the functioning of the brain as

a result of the actual development that has occurred. These

two basic types of intelligence should be distinguished from

Intelligence C, or intelligence as measured by conventional

psychometric tests of intelligence. Hebb also suggested

that learning, an important basis of intelligence, is built up

through cell assemblies, by which successively more and

more complex connections among neurons are constructed as

learning takes place.

A third biologically based theory is that of Luria (1973,

1980), which has had a major impact on tests of intelligence

(Kaufman & Kaufman, 1983; Naglieri & Das, 1997). Ac-

cording to Luria, the brain comprises three main units with

respect to intelligence: (a) a unit of arousal in the brain stem

and midbrain structures; (b) a sensory-input unit in the tem-

poral, parietal, and occipital lobes; and (c) an organization

Contemporary Theories of Intelligence

33

and planning unit in the frontal cortex. The more modern

form of this theory is PASS theory (Das, Kirby, & Jarman,

1979; Naglieri & Das, 1990, 2002), which distinguishes

among planning, attentional, successive processing, and si-

multaneous processing abilities. These latter two abilities are

subsets of the sensory-input abilities referred to by Luria. 

The early biological theories continue to have an influence

on theories of intelligence. Oddly, their influence on contem-

porary psychometric work is substantially greater than their

influence on contemporary biological work, which largely

(although not wholly) has left these theories behind.

Contemporary Biological Theories.

More recent theo-

ries have dealt with more specific aspects of brain or neural

functioning. One contemporary biological theory is based on

speed of neuronal conduction. For example, one theory has

suggested that individual differences in nerve-conduction ve-

locity are a basis for individual differences in intelligence

(e.g., Reed & Jensen, 1992; Vernon & Mori, 1992). Two pro-

cedures have been used to measure conduction velocity, ei-

ther centrally (in the brain) or peripherally (e.g., in the arm).

Reed and Jensen (1992) tested brain-nerve conduction ve-

locities via two medium-latency potentials, N70 and P100,

which were evoked by pattern-reversal stimulation. Subjects

saw a black-and-white checkerboard pattern in which the

black squares would change to white and the white squares to

black. Over many trials, responses to these changes were an-

alyzed via electrodes attached to the scalp in four places. Cor-

relations of derived latency measures with IQ were small

(generally in the .1 to .2 range of absolute value), but were

significant in some cases, suggesting at least a modest rela-

tion between the two kinds of measures.

Vernon and Mori (1992) reported on two studies investi-

gating the relation between nerve-conduction velocity in the

arm and IQ. In both studies nerve-conduction velocity was

measured in the median nerve of the arm by attaching elec-

trodes to the arm. In the second study, conduction velocity

from the wrist to the tip of the finger was also measured.

Vernon and Mori found significant correlations with IQ in the

.4 range, as well as somewhat smaller correlations (around .2)

with response-time measures. They interpreted their results

as supporting the hypothesis of a relation between speed of

information transmission in the peripheral nerves and intelli-

gence. However, these results must be interpreted cautiously,

as Wickett and Vernon (1994) later tried unsuccessfully to

replicate these earlier results.

Other work has emphasized P300 as a measure of intelli-

gence. Higher amplitudes of P300 are suggestive of higher

levels of extraction of information from stimuli (Johnson,

1986, 1988) and also more rapid adjustment to novelty in

stimuli (Donchin, Ritter, & McCallum, 1979). However, at-

tempts to relate P300 and other measures of amplitudes of

evoked potentials to scores on tests of intelligence have led to

inconclusive results (Vernon et al., 2000). Indeed, the field

has gotten a mixed reputation because so many successful

attempts have later been met with failures to replicate.

There could be a number of reasons for these failures. One

is almost certainly that there are just so many possible sites,

potentials to measure, and ways of quantifying the data that

the huge number of possible correlations creates a greater

likelihood of Type I errors than would be the case for more

typical cases of test-related measurements. Investigators

using such methods therefore have to take special care to

guard against Type II errors.

Another approach has been to study glucose metabolism.

The underlying theory is that when a person processes infor-

mation, there is more activity in a certain part of the brain.

The better the person is at the behavioral activity, the less is

the effort required by the brain. Some of the most interesting

recent studies of glucose metabolism have been done by

Richard Haier and his colleagues. For example, Haier et al.

(1988) showed that cortical glucose metabolic rates as re-

vealed by PET scan analysis of subjects solving Raven

Progressive Matrices problems were lower for more intelli-

gent than for less intelligent subjects. These results suggest

that the more intelligent participants needed to expend less

effort than the less intelligent ones in order to solve the rea-

soning problems. A later study (Haier, Siegel, Tang, Abel, &

Buchsbaum, 1992) showed a similar result for more versus

less practiced performers playing the computer game of

Tetris. In other words, smart people or intellectually expert

people do not have to work as hard as less smart or intellec-

tually expert people at a given problem. 

What remains to be shown, however, is the causal direction

of this finding. One could sensibly argue that the smart people

expend less glucose (as a proxy for effort) because they are

smart, rather than that people are smart because they expend

less glucose. Or both high IQ and low glucose metabolism

may be related to a third causal variable. In other words, we

cannot always assume that the biological event is a cause (in

the reductionist sense). It may be, instead, an effect.

Another approach considers brain size. The theory is sim-

ply that larger brains are able to hold more neurons and, more

important, more complex intersynaptic connections between

neurons. Willerman, Schultz, Rutledge, and Bigler (1991)

correlated brain size with Wechsler Adult Intelligence

Scale–Revised (WAIS-R) IQs, controlling for body size. They

found that IQ correlated .65 in men and .35 in women, with a

correlation of .51 for both sexes combined. A follow-up analy-

sis of the same 40 subjects suggested that, in men, a relatively