Mega Questions for Renowned (and controversial) Psychologist Dr.Arthur R. Jensen posed by Christopher Langan (highest I.Q. in the US), subjects include intelligence, education, and creativity

Mega Questions for Renowned Psychologist Dr._Arthur R. Jensen

– Interview by Christopher Michael Langan and Dr. Gina LoSasso and

members of the Mega Foundation, Mega Society East and Ultranet

_________________________

[4]Arthur R. Jensen is a prominent educational psychologist who

received his PhD from Columbia in 1956. He did his postdoctoral

research in London with [5]Hans J. Eysenck, author of the absorbing

HIQ must-read, [6]Genius: The Natural History of Creativity. Jensen

is best known for a very controversial essay on genetic heritage that

was first published in the February 1969 issue of the Harvard

Educational Review. His research work on individual differences in

intelligence led him to conclude that intelligence is 80% due to

heredity and 20% due to environmental influences. Even more

controversial were his findings regarding robust and replicable ethnic

differences in fluid intelligence. The publication of the extremely

well-conceived and executed research findings reported in [7]The g

Factor: The Science of Mental Ability (1998) on the heels of

Herrstein & Murray’s very controversial work, [8]The Bell Curve, moved

the heritability debate into an arena where it could finally be

satisfactorily explored and challenged.

We contacted Dr. Jensen in May and introduced him to the Mega

Foundation, our work and our communities. We asked him if we might

forward to him a few member questions on the topic of intelligence.

Although he is in the process of writing a new book, Dr. Jensen very

kindly took the time out of his busy schedule to answer all 31 of our

member questions, edited by Christopher Langan. Many thanks to Bob

Seitz, Andrea Lobel, Garth Zeitsman, and others who took the time to

submit questions and a special thanks to Mega Foundation’s Coordinator

of Volunteer Services, Kelly Self, for help with transcription. This

extensive and fascinating interview will be serialized in Noesis-E,

beginning with the current issue.

Question #1:

 

Christopher Langan for the Mega Foundation: It is reported that one of

this centurys greatest physicists, Nobelist Richard Feynman, had an IQ

of 125 or so. Yet, a careful reading of his work reveals amazing

powers of concentration and analysispowers of thought far in excess of

those suggested by a z score of well under two standard deviations

above the population mean. Could this be evidence that something might

be wrong with the way intelligence is tested? Could it mean that early

crystallization of intelligence, or specialization of intelligence in

a specific set of (sub-g) factors i.e., a narrow investment of g based

on a lopsided combination of opportunity and proclivity – might put it

beyond the reach of g-loaded tests weak in those specific factors,

leading to deceptive results?

Arthur Jensen: I dont take anecdotal report of the IQs of famous

persons at all seriously. They are often fictitious and are used to

make a point – typically a put-down of IQ test and the whole idea that

individual differences in intelligence can be ranked or measured.

James Watson once claimed an IQ of 115; the daughter of another very

famous Nobelist claimed that her father would absolutely flunk any IQ

test. Its all ridiculous. Furthermore, the outstanding feature of

any famous and accomplished person, especially a reputed genius, such

as Feynman, is never their level of g (or their IQ), but some special

talent and some other traits (e.g., zeal, persistence). Outstanding

achievements(s) depend on these other qualities besides high

intelligence. The special talents, such as mathematical musical,

artistic, literary, or any other of the various multiple intelligences

that have been mentioned by Howard Gardner and others are more salient

in the achievements of geniuses than is their typically high level of

g. Most very high-IQ people, of course, are not recognized as

geniuses, because they havent any very outstanding creative

achievements to their credit. However, there is a threshold property

of IQ, or g, below which few if any individuals are even able to

develop high-level complex talents or become known for socially

significant intellectual or artistic achievements. This bare minimum

threshold is probably somewhere between about +1.5 sigma and +2 sigma

from the population mean on highly g-loaded tests. Childhood IQs that

are at least above this threshold can also be misleading. There are

two famous scientific geniuses, both Nobelists in physics, whose

childhood IQs are very well authenticated to have been in the

mid-130s. They are on record and were tested by none other than Lewis

Terman himself, in his search for subjects in his well-known study of

gifted children with IQs of 140 or above on the Stanford-Binet

intelligence test. Although these two boys were brought to Termans

attention because they were mathematical prodigies, they failed by a

few IQ points to meet the one and only criterion (IQ>139) for

inclusion in Termans study. Although Terman was impressed by them, as

a good scientist he had to exclude them from his sample of high-IQ

kids. Yet none of the 1,500+ subjects in the study ever won a Nobel

Prize or has a biography in the Encyclopedia Britannica as these two

fellows did. Not only were they gifted mathematically, they had a

combination of other traits without which they probably would not have

become generally recognized as scientific and inventive geniuses.

So-called intelligence tests, or IQ, are not intended to assess these

special abilities unrelated to IQ or any other traits involved in

outstanding achievement. It would be undesirable for IQ tests to

attempt to do so, as it would be undesirable for a clinical

thermometer to measure not just temperature but some combination of

temperature, blood count, metabolic rate, etc. A good IQ test

attempts to estimate the g factor, which isnt a mixture, but a

distillate of the one factor (i.e., a unitary source of individual

differences variance) that is common to all cognitive tests, however

diverse.

I have had personal encounters with three Nobelists in

science, including Feynman, who attended a lecture I gave at Cal Tech

and later discussed it with me. He, like the other two Nobelists Ive

known (Francis Crick and William Shockley), not only came across as

extremely sharp, especially in mathematical reasoning, but they were

also rather obsessive about making sure they thoroughly understood the

topic under immediate discussion. They at times transformed my verbal

statements into graphical or mathematical forms and relationships.

Two of these men knew each other very well and often discussed

problems with each other. Each thought the other was very smart. I

got a chance to test one of these Nobelists with Termans Concept

Mastery Test, which was developed to test the Terman gifted group as

adults, and he obtained an exceptionally high score even compared to

the Terman group all with IQ>139 and a mean of 152.

I have written an essay relevant to this whole

question: Giftedness and genius: Crucial differences. In C. P. Benbow

& D. Lubinski (Eds.)[9] Intellectual Talent: Psychometric and Social

Issues, pp. 393-411. Baltimore: Johns Hopkins University Press.

 

Question #2:

 

Chris Langan: For practical purposes, psychologists define

intelligence as problem-solving ability. But there are many kinds of

problem, and some of them appear to involve factors not measured by

standard IQ tests. For example, the problem of how to execute a

complex series of dance steps or athletic maneuvers clearly involves a

cerebellar factor. Some experts would object that intelligence implies

a level of abstraction not required to solve kinesthetic problems. But

if problems must be abstract in order to qualify for inclusion in

intelligence tests, why the correlation of IQ with chronometric

indices involving sensorimotor components and virtually no

abstraction, e.g. simple reaction time?

Arthur Jensen: This is the trouble with defining intelligence. If IQ

tries to estimate g, its not going to estimate every particular

ability, because g is a factor common to all mental abilities. Mental

abilities is a more useful term and the various mental abilities

measured by all sorts of tests can classified hierarchically by means

of factor analysis in terms of their generality, that is, the amount

of variance they have in common with other tests and other factors.

The factor called g (for general) is at the top of the hierarchy only

because it is the one factor that all other mental abilities have in

common (this is explained in detail in Chapters 3 and 4 of my book The

g Factor).

The g loading of a given test or of some lower-order

factor in the factor hierarchy isnt a measure of importance of the

given ability but of its generality. Pitch discrimination is an

ability with a low g loading i.e., (it is correlated only about .30

with), but it is a crucially important ability for a musician and is

totally unimportant for a mathematician. The ability to discriminate

hues also has a g loading of about .30 and it is very important for an

artist, but not at all for a musician or a mathematician. Various

abilities differ markedly in g loading, but one of the interesting

things about g that cant be said about other ability factors, is that

to succeed in almost any kind of intellectual pursuit, some minimum

threshold level of g ability is necessary, though it may not be

sufficient. A high level of some special ability combined with very

low g describes an idiot savant, but not a mathematician, musician, or

artist in any socially important sense. For many types of subject

matter and intellectual skills, achieving a high level of facility or

mastery depends upon a fairly high g threshold. Abstract types of

problems are usually included in IQ tests because they tend to be more

highly g loaded than simpler or less abstract problems, and it is more

efficient in terms of test length to include high-g items in IQ tests

that are intended to estimate an individuals standing on g in some

reference population. However, it is possible to measure g without

using abstract test items or even anything that seems very cognitive.

The inspection time (IT) paradigm is a good example. IT is the

average the speed (visual or auditory exposure time) with which a

person can correctly make an exceedingly simple discrimination. This

measure correlates about + .50 with IQ as measured by complex and

abstract test items. A combination of several such sensory-speed

tests will rank-order people about the same as does the conventional

IQ. But these chronometric tests are less efficient for most

practical purposes, because they require individual testing with

special laboratory equipment and require a longer testing session.

One can get essentially the same result with a 15-minute

paper-and-pencil test that can be administered to a large number of

people at the same time. Psychometrics has two main aspects: (1)

theoretical and research-oriented, and (2) practical and applied.

They are related, of course, but often look very different and are

usually engaged in by different personnel.

The key question is why are reaction times and simple

sensory-motor types of performance correlated at all with IQ derived

from tests composed entirely of complex, abstract problems. The

simple answer is that such different types of tests are correlated

because they all reflect g to some extent. It is the next question to

which we still have no good answer: What is this g? There are

theories and hypotheses, but none that has proved entirely convincing,

empirically proved, or generally accepted by experts in the field. It

has to be some property (or properties) of the brain that enters into

every kind of behavior that involves a conscious discrimination,

choice, or decision. The main focus of present-day research on

intelligence is the discovery of the nature of this property of the

brain that accounts for the empirical fact of g. It is already known

that a number of different physically measured brain variables are

correlated with g; but how they work together to cause individual

differences in abilities and their intercorrelations is still

mysterious. Several chapters of The g Factor are devoted to this

subject. Another recent book devoted entirely to this question is

excellent, but quite technical: Deary, I. J (2000). [10]Looking down

on human intelligence: From psychometrics to the brain. Oxford

University Press.

 

Question #3:

 

Chris Langan: As already observed, intelligence is the ability to

solve problems. But while one psychologist talks about fluid g, a

general intelligence factor that affects the solution of any problem

at all, another talks about multiple intelligences applying to

different kinds of problem. To some extent, the distinction between

intelligence factors and multiple intelligences appears to be

semantic; as you have observed, it is easy to overlook with regard to

the kinds of problem found on IQ tests, e.g. verbal problems, spatial

problems and quantitative problems. So aside from the fact that the

multiple-intelligences school effectively expands the meaning of

intelligence by expanding the meaning of problem to include those

encountered by (e.g.) athletes and dancers, what (if any) is the

difference between the two approacheswhich, as you point out in The g

Factor (p. 128), rely equally on the threshold nature of g? In your

conversations or correspondences with Gardner, has he ever explicitly

repudiated the mathematics of factor analysis?

Arthur Jensen: It would be better to call multiple intelligences

multiple factors. Some of the multiple intelligences named by Howard

Gardner havent yet been included along with a variety of other tests

in any large-scale factor analyses, so we dont know if they would show

up on already establishes factors or would add new factors to the

overall map of the factor structure of human abilities. In any case,

several of Gardners multiple intelligences would at best qualify as

lower-order factors (most probably first-order factors) in the well

establishes 3-stratum hierarchy of human ability factors (Carroll, J.

B [1993] [11]Human cognitive abilities: A survey of factor analytic

studies. Cambridge University Press). They are not measured by IQ

tests (although they may have low correlations with IQ) because IQ

tests are intended to assess the g factor and therefore they include

mainly test items that best reflect g. Theres something to be said

for measuring g in as pure a form as possible and using other tests to

measure various other factors as purely as possible, although it turns

out that no tests known so far exclude some degree of correlation with

g. The g factor, however can be mathematically regressed out of a

measure of some other factor that one wishes to measure independently

of g. Because the basic musical aptitudes (e.g., discrimination of

pitch, duration of tones, timbres, and memory for rhythms) are all

correlated with g, one may be interested in measuring these

independently of an individuals level of g. This would be done, for

example, in a study of the heritability of musical aptitudes. Because

g is highly heritable, the investigator would want to know if the

musical aptitude variable are heritable independently of g and would

use the statistical techniques of regression or partial correlation to

answer this question. As far as I know, Gardner doesnt measure his

proposed multiple intelligences in any psychometric fashion, but I

would bet that the development of any of them to a degree that would

make for expert or professional levels of performance requires an

above-average threshold level of g. The children who attended Yehudi

Menuhins school for musically talented students and had been selected

solely on the basis of their demonstrated musical talent on some

musical instrument, for example, had an average IQ of 127. Does

anyone want to bet that you could find a concert violinist or pianist

with a low IQ? The talent without the g ingredient to go with it

results at best in an idiot savant kind of performance, not a

musically intelligent performance. The same goes for art, and most

probably dance, although that has not been tested, to my knowledge.

I have taken part in two symposia with Howard Gardner

and have also had correspondence with him regarding g. His position

at that time (and also probably today) is that although he believes in

the existence of psychometric g, he simply doesnt think it is very

interesting or important. I, and many others, on the other hand,

think that discovering the nature of g is one of the scientifically

most interesting and important subjects in the quest to understand

human nature. Others, such as Professor Linda Gottfredson are

especially interested in the sociology of intelligence, or the effects

of individual and group differences on educational, social, and

economic aspects of the human condition.

I should add that I do enjoy reading Gardners books.

I especially recommend [12]Creating Minds (1993) as of special

interest to members of the Mega Foundation. This book also reinforces

my view that eminence depends very much on other factors besides g.

Gardner admits, however, that just on the basis of IQ alone at least

90% of the general population would be excluded from the category of

the creative geniuses he writes about in his book. To then try to

minimize the importance of g and its critical threshold property is, I

think, a serious mistake. That is my chief complaint with Gardner,

along with his disregard for any form of quantitative treatment of the

variables he discusses but which is necessary if his claims are to be

objectively tested by himself or by other researchers.

 

Question #4:

Chris Langan: Given that intelligence is problem-solving ability,

scant attention is paid to perhaps the most important problem of all:

selecting a problem worthy of ones time. Historically, the term genius

has been associated with people who have solved this problem, and

having solved it, went on to solve the very urgent, very complex

problem(s) they had chosen. Indeed, many of our best minds consider

themselves too busy with important problems to bother with the

relatively trivial items in IQ tests. This suggests that a more

realistic measure of genius might be obtained by studying a brilliant

subject in his or her natural habitat, analyzing the importance and

computational complexity of the real-world problems that he or she has

solved or failed to solve (and with further research, perhaps even the

intelligence factors required). What do you think of this alternative?

Arthur Jensen: This is a very important point and it is most

important in the up bringing and development of intellectually gifted

children. I know of true prodigies – children with IQs in the 170-190

range – who were able to graduate from major universities, with majors

in math and science, when most children their age are in junior high

school, yet their early adult lives have been spent in trivial, but

often quite lucrative, activities. It is interesting to note that not

one of the four financially most successful adults who as children had

been selected for Termans study of gifted children (IQs>139) ever went

to college. The moral of this story seems to be that if you are

really very bright and your main aim in life is to make loads of

money, you should get started early and dont wasted your time going to

college. But I surely wouldnt say that J. D. Rockefeller, Henry Ford,

Bill Gates, and their lives are not of great value to society. They

are geniuses in their way, and they have made great contributions to

the society.

No one really know why some children never acquire or

develop the important kinds of values, ambitions, and goals that we

consider laudable and most beneficial to society, while others of

comparable or even lesser intelligences may do so. And those who do

so in the extreme (e.g., Beethoven, Darwin, Gandhi, Einstein, and

other stars of the last millennium) are an exceptionally rare minority

among any cohorts with a comparable level of sheer cognitive ability.

It is known that interests and values, as assessed by

questionnaires and inventories, have considerably high heritability,

as shown by the high correlations between parents and their biological

as compared with the lower correlations between adoptive parents and

their adopted children, and by comparing the correlations between full

siblings with the correlations between unrelated children reared

together. Most of us feel disappointed to see individuals with

conspicuously high innate abilities accompanied by a set of interest

and values that scarcely correspond to what we would deem the best

fulfillment of the individuals potential for achievement. The issue

boils down to the question of to what degree interests and values can

be inculcated in young people. It may well be that what we would

consider greatness is such a unique constellation of abilities and

traits that it would be virtually impossible to inculcate all the

necessary qualities of the particular constellation in any given

individual picked on the basis of just one of these qualities, such as

a high IQ, or special ability such as musical talent. This is example

of what behavioral geneticists now refer to as emergenesis: the

exceptional achievement results from a particular constellation of

traits (including interests and values), and does not emerge if any

one of them is lacking. Thus, for example, the difference between

Richard Wagner and his son Siegfried Wagner (also a composer and

conductor, though light-years from his fathers level of creativity)

could have been Siegfrieds lack of one or two traits in the rare

constellation that permitted Richard Wagner to become recognized as

one of the world’s great geniuses. It might well have been Richard

Wagners notably high level of the trait psychoticism, which was not

evident in his sons relatively normal, low-key, mild-mannered, and

modest character (see the reference in my answer to Question #1 and my

answer to Question #11). The kind of study you propose is, in

effect, the biographical analysis of persons of great accomplishment.

There are a number of such biographical studies in the literature.

The leading researcher on this topic is Professor Dean K. Simonton in

his three fascinating books [13]Scientific Genius, [14]Greatness, and

[15]Origins of Genius (also of interest: [16]Genius, Creativity, and

Leadership: Histriometric Inquiry – Editors). The subject is treated

in a much more biographical and anecdotal, though very insightful, way

in Howard Gardners [17]Creating Minds (1993).

Question #5:

Chris Langan: The study of neural networks suggests that as soon as

we can explore the microscopic structure of the human brain and its

sensory pathways, including neural connectivity and neurotransmitter

concentrations, in vivo e.g., through new medical scanning procedures

we can achieve what amounts to a purely biological measure of

intelligence. Do you think that such a measure will ever be wholly

sufficient, or do you think that refinement by performance-based tests

will always be necessary?

 

Arthur Jensen: Im not at all sure about intelligence, which is a

poorly defined term, but the g factor, I believe, will eventually be

explainable completely in terms of brain physiology along the lines

suggested in your question. Given the present technology and with a

concerted effort this could probably be accomplished within the next

two or three decades. And it will be possible to measure g physically

in terms of brain variables. The practical measurement of abilities,

however, may remain at the psychometric level, because of its

demonstrated practical validity and ease of obtaining measures, as

compared with MRI brain scans, PET scans, evoked potentials,

laboratory tests of brain chemistry, etc. Performance-based tests

will always be necessary for assessing learned skills and achievements

(for which the rate and depth of acquisition will inevitably be

related to g as well as to motivational and personality variables and

environmental circumstances). But much of what is now under the

purview of psychometric assessment will be taken over by chronometric

measurement, which will have more scientifically meaningful links to

brain physiology than do conventional psychometric tests (see my

answer to Question #31).

 

Question #6:

Chris Langan: Certain high-ceiling intelligence tests, generically

called power tests, are composed of extremely difficult items

requiring higher levels of problem-solving ability than the items on

ordinary IQ tests. Since these items usually have no known algorithms,

their solutions cannot be looked up in a textbook, and where subjects

do not know each other, one must rely on intrinsic problem solving

ability. However, by virtue of their difficulty, these problems take

longer to solve sometimes days or even weeks. Accordingly, power tests

are untimed and unsupervised. This opens the door to factors like

motivation and persistence, which are not among the factors primarily

measured by standard IQ tests. On the other hand, virtually every

significant intellectual achievement of mankind has involved these

factors in great measure. So why does the psychometric community still

pay no attention to power tests or the statistics derived from them?

 

Arthur Jensen: There are many power tests (i.e., non-speeded or

untimed tests) in psychometrics, although not of the kind described in

this question. Such tests would have little practical use, although

they could be of scientific interest in studying the nature of

high-level problem solving. But people even capable of taking such

tests could be identified with some conventional tests, such as a

combination of the Advanced Raven Matrices and Termans Concept Mastery

Test. People with high scores on such tests can demonstrate their

problem solving ability in their careers. What is the need for prior

selection? They can make it into college and graduate school if

theyve got high IQs, and it will be their virtually unique

constellation of traits (g + special abilities + motivation +

character, etc.) that will determine whether the will, first of all,

identify important problems, and secondly, be able to sole them or at

least materially contribute to their eventual solution. Solving

problems, or even thinking up problems, for which there are presently

no algorithms, takes us into the realm of the nature of creativity.

There are as yet no psychometric tests for creativity in a nontrivial

sense. We cant (yet) predict creativity or measure it as an

individual trait, but can only examine its products after the fact.

At present, there are much more tractable problems for research in the

realm of human abilities, the most important of which, I believe, is

discovering the physical basis of g.

References

4. http://www.indiana.edu/~intell/jensen.html

5. http://www.indiana.edu/~intell/eysenck.html

6. http://www.amazon.com/exec/obidos/ASIN/0275961036/megafoundation

7. http://www.amazon.com/exec/obidos/ASIN/0275961036/megafoundation

8. http://www.amazon.com/exec/obidos/ASIN/0684824299/megafoundation

9. http://www.amazon.com/exec/obidos/ASIN/0801853028/megafoundation

10. http://www.amazon.com/exec/obidos/ASIN/019852417X/megafoundation

11. http://www.amazon.com/exec/obidos/ASIN/0521382750/megafoundation

12. http://www.amazon.com/exec/obidos/ASIN/0465014542/megafoundation

13. http://www.amazon.com/exec/obidos/ASIN/0521352878/megafoundation

14. http://www.amazon.com/exec/obidos/ASIN/0898622018/megafoundation

15. http://www.amazon.com/exec/obidos/ASIN/0195128796/megafoundation

16. http://www.amazon.com/exec/obidos/ASIN/1583484388/megafoundation

17. http://www.amazon.com/exec/obidos/ASIN/0465014542/megafoundation

http://www.megasociety.net/NoesisNov/JensenPartII.htm

Question #7:

 

Christopher Langan for the Mega Foundation: In science, theories and

the definitions comprising them are required to have models, and these

models are required to fit into an overall model of reality. For

example, in physics, the predicate velocity must be semantically

connected to real physical objects in relative motion, which must in

turn be embedded in a model of space and time supporting a

mathematical definition of motion (e.g. the analytic geometry of

classical mechanics). But this becomes problematic with respect to

psychological predicates with subjective components for which we lack

objective models, e.g. consciousness, qualia and emotions.

Intelligence, which is studied strictly in terms of its effectual

correlates, is to some extent such a predicate. Can we achieve a true

understanding of intelligence without a model of reality transcending

the absolute separation of mind and body associated with Cartesian

dualism?

Arthur Jensen: This is a profound question and gets right at the

heart of many of the problems of psychology and making it truly a

natural science. Of the important variables in psychology,

intelligence is one of the few that may lend itself to being

researched strictly as a natural science. Much of present-day

psychology is, at best, a kind of applied technology, some of it

highly useful. But even more of psychology is a kind of shamanism,

which will always be here in one form or another, with a relationship

to science much like that of alchemy and astrology. Unfortunately

this pseudo-scientific kind of psychology, is the only side of

psychology known to the general public, and it is something of an

embarrassment to those who are striving to advance psychology as a

natural science.

A Serious part of the problem is the importance of measurement in the

sense of measuring the behavioral phenomena of interest by means of

true physical scales, i.e., a ratio scale that is standardized to be

invariant across earthly time and space, so that something measured

in, say, Bombay in the year 2001 can be directly compared with

something measured in New York in the year 2050, just as we can say

that the average height of 18-year old male U.S. Army recruits in 1916

was, say, 59 and in 2000 was 510. There are almost no psychological

variables that can be measured on such a true scale on which values

can be expressed as ratios or on which nominally equal differences

between pairs of values in different ranges of the scale can be

treated as truly equal intervals. The mathematical and statistical

treatment of data without these true scale properties is thereby

seriously handicapped.. The most natural scale of true measurement

for some psychological variables, e.g. mental abilities, is in units

of time. It is now well established that certain kinds of timed

performance, measured in seconds or milliseconds, are correlated with

scores on psychometric tests, which are best ordinal (i.e.,

rank-order) scales of performance. I believe further developments in

the use of time-measured psychological variables, such as various

reaction time and inspection time paradigms (see Chapter 8 in The g

Factor), can help to advance truly scientific research on individual

differences in mental abilities. (See my answer to Question #31.) Of

course, psychology as a natural science can have no use for mind-body

dualism. I think I was born opposed to that notion.

Question #8:

Chris Langan: As academic performance falls, there is a growing

tendency among educational theorists to claim that there is no such

thing as a bad student, only bad teachers (common sense, of course,

says that there are both). Learning theory, currently the vogue among

educators, distinguishes the different learning styles of students and

offers various prescriptions for helping students perform up to

capacity. I was recently told by several graduating teachers that (1)

IQ is rapidly becoming a forbidden topic in educational curricula, and

(2) the current vogue is a combination of brain-based learning

(inspired by the Multiple Intelligences model) and cooperative

learning, in which students with different learning styles (e.g.

graphic, visual, auditory or kinesthetic) contribute to each others

learning process. What is your take on these strains of learning

theory? Do they constitute a valid approach to the problem of

declining scholastic achievement?

Arthur Jensen: The purported decline in academic performance in

schools and colleges is a terribly complex phenomenon to get a handle

on for serious discussion. It undoubtedly has many causes, mainly

associated with the very concept of universal education and the

difficult transition from different kinds and levels of education for

different segments of society and an increasing uniformity of

education for the entire population. Individual differences in

abilities are largely ignored by the educational system and the

conspicuously continuing effects of their presence in the educational

process therefore has given rise to forms of denial that blames

teachers, curricula, and institutions. It has also given currency to

theories that deny or minimize the reality of individual differences

or attributes their causes to supposed faults of the schools and of

society in general. The now known scientific facts about individual

differences ( and I emphasize the word individual here) have to be

faced and dealt with in the design of education. (Group differences

basically are simply aggregated individual differences.) In general,

a much more highly diversified educational system is call for. It is

still too early to give up trying different approaches to discover

just how the required diversity can be accomplished. But each of the

proposed approaches must be clearly described and its results assessed

in the nature of a true experiment. Educational practices tend to be

a parade of fads and we see new ones come around every year to replace

last years. Few if nay of these trial balloons face the real problems

confronting public education. In the whole scene, I believe the

individual classroom teachers are the least deserving of blames.

Question #9:

 

Chris Langan: The founders of Mensa, regarded by many as the original

high IQ club, complained that the group had forsaken its original

purposethat instead of pooling its intellectual talent to solve the

most urgent problems of society, it had fallen into aimless

socializing and dilettantism. Since then, a small number of more

rarified groups, known collectively as the UltraHIQ Community, have

advocated a return to the original vision. What is your opinion

regarding the concept of a pool of intellectual talent based strictly

on high levels of g and dedicated to finding solutions for some of

societys more urgent problems?

Arthur Jensen: Its hard to imagine how a group of high-IQ people with

little else in common besides their IQ and probably differing in many

other ways perhaps even more than a random sample of the population

can do much to effect social change or carry out and large project

with a unified aim. On the other hand, a group of persons with a wide

range of IQs from average to very high who have come together as a

group because they all have a similar philosophy and some realistic

goal based on it could be a force for some concerted kind of

achievement. If there were a subgroup of UltraHIQ individuals all

with a similar vision, aim, and dedication to achieve their common

purpose, that would be something!

But I wouldnt apologize in the least for any High-IQ society that was

intended as a purely social organization that qualified people could

join simply because the find each others company more congenial than

that of most of the people they would be apt to meet in other social

groups. I suspect that the zone of tolerance for the intelligence

levels of ones friends and spouses is probably, at the outside, about

ones own IQ +/- 20. People in the upper-half of the IQ distribution

are more closely assortative in this respect than are those in the

lower half. In the general population, spouse similarity in IQ is

about the same as full-sibling similarity. Assortative mating for a

given trait has the effect of increasing the genetic variance in that

trait in the offspring generation. It is estimated that some 15 to 20

percent of the population variance in IQ is attributable to the effect

of assortative mating.

 

Question #10:

 

Chris Langan: Intelligence is about solving problems. Because

problems consist of constraints to be satisfied by their solutions,

those with high IQs are good at working within the bounds of more or

less complex constraints. Yet some problems, especially those

involving lateral thinking, require creativitythe ability to break

free of apparent constraints. So to some extent, attributes like

creativity, novelty and originality seem paradoxically related to

intelligence. Have we had any success in relating creativity to IQ,

and specifically to g?

Arthur Jensen: About all I can say on this is that the level of g

acts as threshold for the possibility of creativity and that this

threshold differs somewhat for different fields of creativity,

particularly to the extent that the filed calls for a special talent

that somewhat outweighs the relative importance of g. The main reason

that a fairly high level of g acts as a threshold is that to be

creative in most fields, one has to master the basic knowledge,

techniques, and skills needed just to be able to work in the field, to

say nothing of being creative in it. The cognitive demands on

achieving the essential level of mastery of the working tools are

typically considerable and are often highly g-loaded. Hence you dont

find truly creative scientists, writers, musicians, etc., with low or

even average IQs. A music composer, for example, must master such

abstract and complex subjects as harmony, counterpoint,

orchestrations, and so on — all g-loaded subjects. Plus an

incredible amount of assiduous practice, so that much of this

knowledge and skill repertoire becomes automatized, thereby freeing

the individual for creative expression. Read the biographies of any

of the importantly creative people in history and youll find that the

prerequisites and necessary personal conditions for creativity are

above-average g plus an unusual capacity for work and persistence in

the face of difficulty or adversity.

 

Question #11:

 

Chris Langan: Many people believe that genius and insanity are

closely related. Indeed, history provides numerous examples of

creativity and insanity or (near-insanity) in close conjunction.

Statistically, does intelligence correlate either positively or

negatively with any kind of insanity or mental instability?

 

Arthur Jensen: The supposed relationship between creativity and

mental disorder has been well researched and is proven to be a fact.

Depression and bipolar disorder have a high incidence among creative

writers and artists than in the general population; schizothymic

characteristics are somewhat more frequent among philosophers,

mathematicians, and scientists. The late Professor Hans J. Eysenck

hypothesized a trait he called psychoticism which he thought is an

essential ingredient in major-league creativity. Psychoticism is not

itself a psychiatric disorder or disabling condition (although it is

associated with a proneness for such), but a constellation of

intercorrelated personality traits, most of which I have found in

virtually every famous creative genius Ive read about. Eysencks

theory and the evidence for it is the most interesting I have come

across in this field. This is a complex subject and I couldnt

possibly do it justice by trying to explain it all here, but I will

recommend the following two books, which are the best Ive come across

on this topic:

[8]H. J. Eysenck, Genius: The Natural History of Creativity. 1995,

Cambridge University Press.

[9]M.A. Runco & R. Richards (Eds.), Eminent Creativity, Everyday

Creativity, and Health. 1998, Ablex.

References

8. http://www.amazon.com/exec/obidos/ASIN/0521485088/megafoundation

9. http://www.amazon.com/exec/obidos/ASIN/1567501753/megafoundation

http://www.megasociety.net/NoesisJan02/JensenPartIII.htm

 

Question #12

 

Chris Langan: Even IQ tests with moderate ceilings can be upwardly

extrapolated, and there exist experimental high-ceiling tests that

appear to have much higher ranges than standard IQ tests when

anchor-normed on those same standard tests. Indeed, whatever the

limitations on its measurement, there would seem to be no a priori

ceiling on intelligence itself. Yet, some claim that the very idea of

an IQ in excess of +4s is meaningless. In your opinion, can it be

fruitful to consider IQs in excess of +4s? What, if any, is the

absolute upper limit on the measurement of IQ?

 

Arthur Jensen: I believe we have no means at present of determining a

ceiling for intelligence or for extrapolating existing scales to a

theoretically derived ceiling. Im not even sure if the idea of a

ceiling for intelligence is a meaningful concept. An upper limit for

the measurement of g may be more meaningful and +4s (IQ of 160) may

well be the highest level in which we can have much confidence that it

is g that is being measured. It has long been known that various tests

become less g loaded the higher one goes in the IQ distribution. That

is, if we gave a large battery of diverse tests to people with IQs

above, say, 120 (i.e., the top 10% of the population) and to people

with IQs below IQ 80 (the bottom 10%), we will find that the

correlations among the tests are considerably smaller in the high IQ

group than in the low IQ groups, and consequently the tests have less

in common (i.e., their general factor g) and hence lower g loadings in

the high than in the low group. This appears to be quite a linear

effect as we move up the IQ scale. If the IQ scale were a true

interval scale (we only assume it to be such), we could extrapolate

the linear trend to the point at which g loadings = 0. That, then,

would be the ceiling of the g factor. High IQ persons abilities become

more highly differentiated and specialized, hence are less correlated

with one another and afford a weaker basis of prediction of any

particular ability from a knowledge of the individuals standing on

some other ability. Yet this diverse or differential development of

mental abilities itself seems dependent on the possession of a fairly

high level of g, in the sense of superior performance on the kinds of

tests that are the most g loaded.

The problem in researching the uppermost region of human abilities is

that the further we go above the mean IQ, the smaller is the

proportion of the population that we can obtain as research subjects,

and, since research in this field depends a lot on statistical

inference, we would find it exceedingly difficult, or even impossible,

to obtain large enough subject samples to permit statistically

significant conclusions. The more highly selected the subject sample,

the smaller is the variance of the test scores and their reliability.

There are more tractable and scientifically more important things to

be researched at present. Because there is little if any practical

value in measuring ability levels above the 99th percentile in the

general population, hardly anyone, least of all the producers of

mental tests, is interested in doing so. The only interest I have ever

seen has been among some members of the high IQ clubs that are

offshoots of Mensa. I once tested a group of some 20 to 30 volunteers

from Mensa. On a standard psychometric test they averaged about 20 IQ

points or so above the average of U. C., Berkeley, undergraduates. I

was interested in whether the Mensa subjects would also show faster

reaction time (RT) than Berkeley undergrads, who on our RT averaged

about +1 s above the general population mean on such tests. The Mensa

subjects averaged considerably faster RT than the Berkeley students.

The fact that RT is monotonically related to IQ throughout an

80-points IQ range, from about IQ 60 to at least IQ 140, suggests that

it might be a useful tool in studying the upper reaches of ability,

strange as that may seem. But of course there is a physiological limit

to RT, determined in part by the limits on time for sensory

transduction of the stimulus and afferent and efferent nerve

conductive velocity. But RT has the advantages of measurement on a

ratio scale and also of being based on the very same test at all

levels of IQ (beginning at a mental age of about 3 years, below which

subjects typically have difficulty in performing the RT tasks without

training).

 

Question #13

 

Chris Langan: Intelligence is the ability to reason, i.e. to solve

problems. Problems are solved according to procedural schemata called

algorithms. Algorithms can be learned. Ergo, intelligence can to some

extent be learned. Equivalently, a mathematician specializing in

neural networks might say that since the intelligence which becomes

crystallized in synaptic weighting patterns is algorithmic in both

form and content, neural nets can be trained for intelligence. The

brains of children undergo structural development, and even adult

brains retain a certain amount of neural plasticity. So even though

statistics indicate that IQ tends to be stable throughout the human

lifespan, does it remain possible that under the proper conditions, IQ

can to some extent be learnedthat a protean set of high-level

algorithms can be burned into cerebral synapses? Would such an IQ

boost necessarily be hollow with respect to g?

Arthur Jensen: Yes, certainly. Various thinking or problem-solving

algorithms can be trained and even automatized through extensive

practice. These phenomena are associated with neural plasticity and

the innate capacity for learning. It is individual differences in

these brain attributes, rather than the acquisition of specific

algorithms for thinking and problem-solving per se, that are the basis

of the g factor. Algorithmic training is remarkably specific to a

particular subject-matter and has surprisingly little transfer beyond

the material on which it has been trained. This is one of the problems

with most conventional IQ tests, verbal and nonverbal tests alike: two

things are being measured: g + learned algorithmic thinking and

problem-solving skills, and these are completely confounded in the

total score on the test. Chapter 10 in [9]The g Factor deals with just

this problem, which is described as the confounding of the vehicle

(e.g., the knowledge and skill demands of a particular test) for

measuring a given construct and the construct itself (e.g., the g

factor). This is a big problem, often insufficiently recognized by the

users mental ability tests. It is much less a problem in explicit

achievement tests. A test in algebra, for example, may be a poor way

of assessing g, but a good way to find out where a person stands in

knowledge and use of algebra. If everyone tested had taken equivalent

courses in algebra, the scores on the algebra test would also be quite

highly g-loaded (i.e., correlated with g). For persons who have

completed high school, tests of reading comprehension measure g about

as well as most IQ tests, except for true dyslexics. One of the

potential advantages of chronometric tests (e.g., reaction time and

inspection time) is that they have some g loading yet have virtually

no intellectual or algorithmic content. Their disadvantage is that

they also measure, besides g, a large component of purely

sensory-motor abilities that fall entirely outside the domain of

mental abilities (as shown by their lack of correlation with any other

kinds of cognitive tests).

The learning of problem-solving and other algorithms is crucial in

most realms of intellectual work and it can be inculcated to a

considerable degree through training. It may even improve certain test

scores to some extent. But this is not the same as improving whatever

it is that makes for g. In fact, the level of algorithmic complexity

that can be acquired is limited by an individuals level of g. Before

children are exposed to any kind of maths, for example, one can make

fairly good predictions on the basis if IQ of which ones will or will

not top out in various levels of higher abstract mathematics,

regardless of educational opportunity, effort, and the like. Only

persons in the top 15% of the IQ distribution are employed as

mathematicians; that seem to be the absolute minimum threshold for

this occupation. Many students entering college whose ambitions are to

be rocket scientists or engineers soon discover they cant make the

math requirements despite their most earnest efforts to do so.

Question #14

 

Chris Langan: In [10]The g Factor, you state (regarding the Flynn

Effect) that Whatever causes the rise in IQ, it has its greatest

effect on those at the lower end of the scale, with a corresponding

shrinkage of the standard deviation. However, since it is unclear how

adult IQ scores above 100 were normed on older IQ tests that relied on

mental age, it is unclear whether the distribution to which you refer

is that characterizing ratio IQ or deviation IQ, where ratio IQ is

thought by some theorists to be lognormally rather than normally

distributed (e.g. Vernon Sare, University of London, 1951). Can you

clarify this point?

Arthur Jensen: The Mental Age/Chronological Age, or 100(MA/CA) = IQ,

has been virtually defunct since the 1940s. All professionally

constructed and published IQ tests today are based on deviation IQ

[i.e., z = (Raw Score – Mean )/SD, and IQ = 15z + 100]. The ratio IQ

becomes increasingly suspect as children get older. It is based on the

presumed (or demonstrated) linear relationship of the tests raw scores

to CA. But this relationship begins to depart from linear at around 12

to 13 years of age, and after age 15 (it used to be 16) it is so

nonlinear that the MA/CA ratio becomes increasingly meaningless with

increasing age. Often the raw scores on a test are converted to

normalized z scores and then converted to IQs, ensuring that the IQs

are normally distributed; at least in the standardization sample. If

we assume that intelligence should be normally distributed, and if the

IQ distribution is made perfectly normal (i.e., Gaussian), then we can

claim that IQ is an interval scale. But the assumptions are the

critical joker in this line of reasoning. There is nothing that

actually compels these assumptions; they are merely plausible and

statistically convenient.

The best single study of the Flynn Effect (i.e., the secular rise in

IQ over the past several decades) was done in Denmark with military

conscripts. The lower portion of the IQ distribution showed larger

gains than the higher end, probably because in the more recent decades

more of the lower portion under the bell curve received more

educational attention and better education, and also probably better

pre-and post-natal health care and nutrition.

As raw scores on mental tests are based simply on number of correct

answers (a function of item difficulty, i.e., percent of population

passing an item), which constitutes only an ordinal (rank-order) scale

of ability on the given test, any transformation of the scale —

normal, lognormal, hypergeometric, or whatever — really has the same

status as an ordinal scale, i.e., the raw scores or any transformation

of them could just as well be treated as ranks. These can be converted

to percentile ranks, a given percentile simply indicating the percent

of persons in the standardization group that fall below a given raw

score (number right). These percentiles can also be transformed to

normalized or lognormalized scores (or any other transformation) if

one wants to make assumptions about the form of the distribution of

the latent trait (e.g. intelligence) in the population; but not an

iota more of real information in conveyed by these transformed scores

than is present in the ranked scores. Now if our measure were true

physical measures (i.e., a ratio scale) but were expressed as ranks,

their rank order would covey less information than the raw scores

themselves. A true ratio scale (e.g., height, weight, reaction time)

is a necessary and sufficient condition for describing the form of

their distribution in a given population or random sample of some

specified population. Thats why the Flynn Effect for the increase in

the average height in the population has not created any controversy

as it has in the case of IQ. By having a ratio scale, the phenomenon

and its magnitude are clearly established by the raw measurements,

whatever may be their cause. But no one argues, Is it really height

that has increased? That is the whole argument about the Flynn Effect

and IQ — is it really intelligence that has increased, or only test

scores? When we get true ratio scales of mental abilities, we will be

able to answer the kind of question you are asking. The scientific

study of developmental trends in mental growth is greatly handicapped

by our lack of true ratio scales, without which the shape of the

growth curve of mental test scores is almost meaningless beyond saying

it is positively monotonic between any two points on the scale of

chronological ages, up to about age 20.

 

Question #15

 

Chris Langan: Why are IQ’s measured on relative scales rather than in

absolute terms? Saying that someone is brighter than than 99% of the

population is no more meaningful than saying that someone is taller

than 99% of the population. While raw scores on tests containing items

of low to moderate complexity provide an absolute measure of sorts,

they seem only indirectly related to intellectual speed and power. The

solution times of various problems, or the most complex problems

solvable without time constraints, would be more direct measures of

speed and power and thus more acceptable as absolute metrics. Are

there other absolute measures of intelligence, and if so, how do they

relate to IQ?

Arthur Jensen: This is a continuation of the previous question. I

think it quite informative to know a person’s percentile score

(assuming it as accurate), as it tells you where that person stands

with reference to some “normative” group on the trait in question. A

pediatrician can rather precisely measure an infant’s head

circumference with a tape measure (a ratio scale), but to interpret

this measurement he needs to look it up in a table of norms giving the

percentile equivalent of that measurement (and its standard deviation)

for the average infant of the same age. The only absolute measures of

intelligence I know of that are behavioral are various forms of

reaction time (RT) and inspection time (IT) measures, which we know,

are related to IQ because of their significant correlations with IQ.

Interestingly, the longer the average RT for a task beyond about 1

second (for young adults), the less it correlates with IQ. In more

complex tasks that take much more than 1 second to perform, other,

noncognitive factors enter in and “dilute” the RT measure with sources

of variance that do not represent whatever we mean by general

intelligence. Physiological measurements, which are a true scale, such

as latency and amplitude of the evoked brain potentials and rate of

glucose uptake by the brain while solving a problem (measure by PET

scan), and (in one study) the brain’s pH level, are all correlated not

just with IQ, but with the g factor per se. A combination of such

chronometric and physical variables will one day yield ratio-scale

measures of mental ability that are scientifically more meaningful

than those obtained from conventional IQ tests. The details of this

topic form, in part, my answer to [11]Question #31.

Question #16

Chris Langan: On most IQ tests, ceiling effects begin to occur above

the two-sigma level. Thus, ceiling effects can occur before deviations

from a Gaussian distribution become significant, effectively obscuring

the deviations. But for (e.g.) blacks, the ceilings are high enough

(in standard deviations) that significant differences ought to be

apparent and measurable. E.g., if the SD for blacks were 12.75 (85/100

X 15), the 5 SD level would come at IQ 149 and the 4.75 SD level (one

in a million) would be IQ 145.56. So blacks should be ideal for

studying the differences between ratio IQs and adult deviation IQs,

which seem to approximate lognormal and normal distributions

respectively. However, this raises some questions: is the black IQ

distribution normal, lognormal or Pearson Type IV, i.e. “abnormal”?

How has the Flynn Effect acted upon the black IQ distribution (where

insulated from the heterotic effects of miscegenation)?

 

Arthur Jensen: This is a clever thought, although it has become

increasingly difficult to get IQ data on blacks, at least in

sufficient numbers to study the top-level percentile in the black

population. In light of what I said in my answers regarding scales and

distributions, I don’t think it would be fruitful to pursue this issue

with conventional tests. I have looked at a great many distributions

of both white and black IQs in whole school populations. The black

distributions generally resemble the Pearson Type IV Distribution; it

is considerably skewed to the right. Not as much, if any, theoretical

significance can be attached to this observation as would be possible

if the mental measurements were a ratio scale.

 

Question #17

 

Chris Langan: It has been argued that the deviations from a normal

curve that occur among child IQs are simply a function of varying

rates of mental maturation. Thus, while the distribution of childhood

ratio IQs looks closer to lognormal than normal, and while the

distribution of some adult indices like AGCT-derived IQ scores shows a

frequency pattern agreeing closely with childhood ratio-IQ

distributions, the distribution of adult IQs is Gaussian. Now, if

specific individuals tend to regress toward the mean as they mature

but the overall distribution remains the same as it is for children,

then there must be “late bloomers” who rise to take their places in

order to keep the upper ranges of the distribution populated. Has this

phenomenon been studied? Do very high adult ratio IQ’s appear with

greater-than-Gaussian frequency as they do with children, or are the

distributions different?

Arthur Jensen: Yes, the variation in IQ (or relative standing in some

normative group) as individuals grow up from about age 2 (when IQs are

first reliably measured) to maturity has been studied quite

thoroughly. (The subject is treated at length in my book [12]Bias in

Mental Testing, Chapter 7, 1980 Free Press). (Also see [13]The g

Factor, pp. 316-318.) Individuals’ IQs fluctuate rather randomly up

and down throughout their development, but become increasingly stable

with each successive year. This has been studied by looking at the

matrix of correlations betweeen IQs measured every year from age 2 to

age 18 or so. The correlations are increasingly higher as a function

of age. Many early bloomers and late bloomers exchange their positions

in the IQ distribution, and in about equal numbers. Hence the overall

distribution of IQs remains fairly constant throughout the entire

developmental period.

 

Question #18

Chris Langan: It seems that research on the profoundly gifted has not

only been very limited, but that virtually none of it addresses the

question of how society can bring out the best in its brightest

members. One of our members, Bob Seitz, asks: During my years with

NASA and Georgia Tech, I casually wondered why there didn’t seem to be

a national registry of the very brightest, with attention to their

needs and their encouragement. But when, two years ago, I finally

discovered the ragged state of affairs vis-a-vis our brightest, I was

shocked.

It seems that as IQ’s rise from 75 to 125, dramatic changes occur in

life outcomes and socioeconomic statuses. But once intelligence

exceeds the upper part of that range, there seems to be little

correlation between IQ and success in even the most demanding

intellectual pursuits. This raises the possibility that high-IQ types

are being neither allowed to fully utilize their potential nor

rewarded in proportion to their abilities. One might expect this to

detract from their enthusiasm and level of performance. But even

though the costs to society may be immeasurable, no one seems to be

addressing or investigating the situation. Do you have any opinions on

this matter?

Arthur Jensen: This all goes back again to the fact that achievement

in a multiplicative (not additive) function of a number of critical

traits, of which g is only one, though a very important one. Given a

range of IQs sufficient for everyone within that range to be able to

learn the “tools of the trade”, then other personal factors become

more critical determinants of achievement. The more unusual the

achievement, the greater the number of different factors that have

acted multiplicatively to produce it. People do not tend to undervalue

intelligence so much as they undervalue the other multiplicative

traits that enter into achievement. Our expectations for achievement

are weighted too much for then effect of IQ and not enough for other

valuable traits. Because of its threshold nature, however, a low IQ is

a handicap, and even more so in our modern technological society than

in the more agrarian past. Higher IQ is always an advantage in the

multiplicative combination of factors required for outstanding

achievement. One of the things most lacking in education, and often

also in parental upbringing today, is inculcation of the kind of

values, including self-discipline, that are among of the essential

ingredients in the multiplicative formula involved in outstanding

achievement.

 

Question #19

 

Chris Langan: Aside from social ineptitude, perhaps the trait most

often associated with IQ > +4s is being a multimillionaire (Bill Gates

is a frequently-cited example). It seems that when the hyper-gifted

turn their hands to making money, they succeed in spades. But with

respect to social utility, this is often a waste. We need cures for

cancer, better ways to relate to each other, cures for Alzheimer’s and

Parkinson’s Diseases, a marriage between general relativity and quantum

mechanics. In short, we need real works of genius. But even though

society has a vested interest in fully utilizing the talents of its

geniuses, it continues to let itself be vastly outbid for their

services. We encourage real geniuses to squander their potential on

what often turns out to be pointless, inflationary acceleration of the

financial treadmill while discouraging those without academic

credentials from participating in the social and intellectual

mainstream, relying on the survivors of academic bureaucracy to solve

our most urgent problems. Unfortunately, academic politics is not a

valid test of intelligence. Is there any effort to understand what’s

going wrong in this area?

Arthur Jensen: I believe that, generally, multi-billionaires do have

plenty of “social utility” –the Rockefeller, Ford, Carnegie,

Sloane-Kettering, and Mellon foundations, for example, not to mention

the industries, jobs, and their products that have benefited the whole

society are indeed a boon to the whole society. These foundations

built on the fortunes of these billionaires are responsible for many

of the grants made to researchers working on Alzheimer’s, Parkinson’s,

caners, AIDS, and a great many other medical and humanitarian

enterprises. The industrial and financial achievements on the scale of

Gates, Rockefeller, Ford, Etc., it seems to me, are highly worthy of

our admiration.

I do agree that in today’s world, especially in the United States, the

job market places too much emphasis on academic credentials, and not

enough on the assessment of actual abilities. If I had to choose

between knowing a job applicant’s IQ or level of education, I’d pick

the IQ, assuming the job doesn’t require some specialized skills that

can only be acquired in college or graduate school. In today’s world,

however, one has to wonder about a high IQ individual who has not

finished high school or gone to college; one would want to know about

other achievements as well as their personality traits. In personnel

selection it is most valuable to have objective test scores both on g

and on subjects most relevant to the job as well as formal educational

credentials. They are usually in fair agreement, but when not, they

bear further looking into.

 

Question #20

Chris Langan: In working with some of the profoundly gifted, I’ve

encountered a few hints about how their extraordinary potentialities

become derailed. There seem to be major problems with the extremely

gifted in a society that isn’t geared to them, like the plight of an

eight-footer in a house with six-foot ceilings. How much attention has

been given to the social and emotional problems of the highly gifted

population?

 

Arthur Jensen: I know other psychologists who are better able to

answer this than I can, for example Professors Julian Stanley (John

Hopkins), David Lubinski, and Camilla Benbow (Vanderbilt). It is true

that most super-gifted children, especially as they approach

adolescence, are not as challenged or as happy about going to school

with their age-mates as they would be if they were entered into a

regular 4-year college with classmates who are six or seven years

older. The channeling that takes place in college and thereafter in

the world of work is such that people generally find themselves in the

company of others who are not all that different from themselves in

abilities, interests, and the like. The super-ability types usually

come to realize that people differ greatly in abilities, and that they

have to learn to live with this fact gracefully. Those who don’t learn

this lesson pay a price. I haven’t yet seen a good case made for the

idea that people become maladjusted simply because of their having a

very high IQ. Although IQ and mental health have only a slight

positive correlation with each other, it’s not in the least surprising

to come across high IQ persons with emotional and inter-personal

problems. But I doubt that any disability can be blamed on a person’s

having a high IQ per se.

I do feel sorry for those children whose parents have been told that

their child is gifted and never let their child forget it for one

minute. (The singled-out child’s siblings suffer as well in this

case.) It’s interesting to read the later volumes of Terman’s Genetic

Studies of Genius (based on subjects selected as school-age children

with Stanford-Binet IQ>139). A large majority of these “Termanites”

became fairly ordinary adults and some were less successful in life

than are many persons of average IQ. I have heard some educators

express concern that something must have gone terribly wrong in the

upbringing or education of many of the Terman group to cause the

average level of their apparent achievements as adults to be so

considerably less impressive than their IQ. But this IQ-achievement

discrepancy is exactly what one should expect in terms of the

multiplicative theory of achievement I have described in my answers to

some of the previous questions.

References

8. http://www.megasociety.net/MegaPress/index.html

9. http://www.amazon.com/exec/obidos/ASIN/0275961036/megafoundation

10. http://www.amazon.com/exec/obidos/ASIN/0275961036/megafoundation

11. http://www.megasociety.net/NoesisJan02/JensenPartIII.htm#Q31

12. http://www.amazon.com/exec/obidos/ASIN/0029164303/megafoundation

13. http://www.amazon.com/exec/obidos/ASIN/0275961036/megafoundation

http://www.megasociety.net/NoesisJan02/JensenPartIV.htm

 

Question #21

 

Chris Langan: As students, doctors and lawyers take tests like the

LSAT, their average IQs are found to be around 127, while in contrast,

mathematicians average around 140. Has any research been done relating

test scores to minimally acceptable professional performance in (e.g.)

medicine and law, as gauged by (e.g.) deaths attributable to

diagnostic error, cases lost, or judgments overturned? Since certain

studies have found that IQ is a better predictor of job performance

than educational credentials, shouldnt we (and our licensing bureaus)

be paying more attention to it? Is our failure to do so attributable

to affirmative action or other minority preference programs?

 

Arthur Jensen: Excellent question. Probably the answer to it might be

too politically incorrect for anyone to be able to risk the research

that could answer it, or to even obtain a grant to do such research.

There are plenty of anecdotes that one hears of, but I haven’t come

across any bona fide research studies that investigated the

relationship between test scores and performance catastrophes at a

professional level such as you mention. But it is hard to imagine that

such a relationship does not exist, since such a relationship has been

amply demonstrated by research on personnel selection in hundreds of

jobs in which test validity has been determined in terms of actual job

performance. The U.S. Employment Service, using the General Aptitude

Test Battery (GATB), has published the results of literally hundreds

of such test validation studies for predicting success or failure in

various job categories, not including doctors or lawyers or other

high-level professionals. And it is the g factor of the GATB that

carries most of the predictive power of this battery composed of

eleven diverse tests. It would be a safe bet that doctors (or other

professionals) who are fired because their performance is at a sub

threshold level of competence have a lower average IQ than the

competent majority of their profession. I intend to circulate this

question among some colleagues who are more expert on this topic than

I and will let you know if there are any studies that can provide a

more definite answer to your question. But the issue is so

contaminated by the need for political correctness that it may be

virtually impossible to obtain a valid answer in the present climate.

 

Question #22

 

Chris Langan: The generality of g reflects the fact that g is found in

conjunction with every other intelligence factor that, as you posit in

[3]The g Factor, it represents a combination of all of the

distributive criteria that contribute to intellectual processing

everywhere in the brain. Some of these criteria clearly have a genetic

basis, e.g. neural and synaptic density, neural conduction velocity,

neurotransmitter abundances and control mechanisms, glial density,

degree of axon myelinization and so on. Just as genetics dictates that

a rat is more intelligent than an insect and a man is more intelligent

than a rat, human beings differ in genetic constitution and may

therefore differ in these criteria. So g is biologically plausible as

well as empirically confirmed. But with the advent of the politically

correct Multiple Intelligences theory, it has fallen into disrepute

among educators and been rendered prematurely obsolescent. What is

your opinion of those who, being more enamored of political

correctness than common sense, deny the existence of g despite its

scientific basis? Do you see a light at the end of the tunnel?

 

Arthur Jensen: My answer to this question must already be obvious. The

“light at the end of the tunnel” is simply objective empirical

science. Those who would belittle the role of g in human cognition

could prove their case simply by showing that their tests, or

measures, or assessments of “multiple intelligences” are more highly

correlated with any important “real-life” criteria independently of g

than those criteria are correlated with g alone. But most researchers

of “multiple intelligences” don’t actually measure anything at all.

Their claims are based on purely literary, armchair psychology. So

there is no means of putting their theories to an empirical test. It

is simply non-science and just a part of the passing parade of

untested notions that so frequently attract educators and dilettantes.

That some of these fads are also perceived as PC, of course, adds to

their popular attraction.

Question #23

 

Chris Langan: With each passing year, it seems that popular culture

places a lower value on high intelligence. Intelligent or studious

children are called geeks, while intellectual mediocrity is regarded

as cool. So shamelessly do the popular media encourage this perception

that it sometimes seems as though the human race is being

systematically lulled into a state of intellectual degeneracy. In your

opinion, will this trend ever be successfully counteracted? If not,

what do you foresee as the long-term effect on the distribution of

intelligence in the general population?

Arthur Jensen: The trend you describe will be (or is already being)

successfully counteracted in some other countries, and as a result,

unless we soon get our own house in order, we’ll be the

losers–scientifically, culturally, and economically. There is nothing

in the Book of Nature that says the USA is automatically immune to the

possibility of devolving towards the conditions of Third World

countries. The advancing front of future civilization may well

gravitate eastwardly. I can’t say I ever really understood Oswald

Spengler, but the title of his famous book (Decline of the West, Ed.)

seems prophetic. But I don’t worry about it as long as civilization

will be preserved and developed somewhere on earth.

 

Question #24

 

Chris Langan: There is a certain amount of evidence supporting the

hypothesis that intelligent people, being better able to fill their

lives without raising families, are having fewer children.

Unfortunately, for every socially responsible, intelligent person who

decides to postpone or forego childbearing, ten others, many with

lesser genetic endowments, stand ready to fill his or her place in the

gene pool with their own progeny. Insofar as the net result would

appear to be dysgenic, is it ethical to continue to let this happen?

Arthur Jensen: Yes, it is likely that there is a dysgenic trend in g

level, at least in the USA. A plausible case can be garnered from U.S.

Census data over the last 3 decades. I don’t know whether it is or

isn’t ethical to neglect seriously investigating the possibility of a

dysgenic trend or, if it indeed exists, to do nothing about it. But a

dysgenic trend that affects the overall level of g in the society

would have ill-fated consequences for this country’s future welfare,

to say the least. Three facts have to be much more generally

understood: (1) There is a g factor, (2) the distribution and overall

level of g in the population is causally related to the level of

civilization and the quality of life in a modern society, and (3) g is

highly heritable (i.e., influenced by genetic factors). Given these

facts, a conclusion regarding dysgenics would depend on examining

birth rates in different segments of the distribution of the g factor

in the nation’s population. Depending on the conclusions from this

examination, it will be up to informed public opinion and the public

will need to decide what, if anything, should be done, or could be

done, about it in our free society.

Question #25

 

Chris Langan: Modern civilization grows increasingly dependent on

complex technology, and thus on people with the intelligence to

design, implement and maintain it. This places a higher level of

social utility on high intelligence, and thus on highly intelligent

people. This brings to mind a rather depressing joke: The problem with

the gene pool is that there is no lifeguard! Would intellectual

eugenics necessarily be a bad thing for humanity? Is there a danger

that this would lead to a Brave New World scenario?

Arthur Jensen: Right on target! “Brave New World” is of course pure

science fiction, which is invariably based on the science of the past

and rarely imagines anything like the actual scientific and

technological developments of the future. But there are even worse

scenarios – dysgenic ones – than are portrayed in Huxley’s novel. The

lower one-fourth (perhaps even the lower one-third) of the IQ

distribution, as we know its mental capabilities today, will have a

hard time finding gainful employment of the kinds that are needed in a

largely technological, information-intensive society. The USA is

already having to import workers, mostly from Asia, to fill these

kinds of positions, which would otherwise have to go begging for

applicants.

A serious question that is hardly ever put up for discussion is

whether a society should design itself in terms of the level of

ability (largely g) and work demands that could accommodate the vast

majority of its existing population or work toward raising the overall

level of ability to accommodate the increasing ability demands of our

trend toward a more technological and information-intensive society. A

number of symposia could be organized about this theme.

 

Question #26

 

Chris Langan: For some time now, Robert Plomin has been locating genes

associated with high IQ. The evolution of the human genome project

raises the possibility that even more of these genes will soon be

located. Meanwhile, genetic testing and engineering technology

promises to let people select their mates for complementary genetic

characteristics, and even to upgrade the DNA of their offspring in

vitro. Do you see this as harmful or beneficial to society?

 

Arthur Jensen: American behavioral geneticist Robert Plomin (now a

professor in the Behavioral Genetics Research Unit at the Institute of

Psychiatry in London, England), working with a large team of

colleagues specializing in genetic research, has already identified

several different sections of DNA (on chromosome #6) which reliably

differ between large groups of person of average IQ and of very high

IQ. This research is progressing at an accelerating rate as the

technology for identifying differences in specific sections of DNA

(not necessarily genes per se) is advancing rapidly. Inevitably many

more “IQ genes” will be identified within the very near future. No one

in the field is really surprised by Plomin’s findings, because the

heritability of IQ and of psychometric g (which is the main basis of

IQ heritability) has long been well established by the methods of

quantitative genetics based on the correlations of various kinships

reared together and reared apart. The importance of Plomin’s research

is that it yields specific information that will be used to trace the

pathways of genetic expression, i.e., discovering just how the

identified genes chemically affect the development of the brain

variables that cause individual differences in g. It is a necessary

complement to the approaches based on direct studies of brain

physiology, affording clues that narrow the search for the key causal

variables. Knowing precisely what a gene does and how it does it is a

major step toward understanding the workings of brain-behavior

phenomena. The history of such advances in scientific knowledge

strongly indicates that they most usually prove beneficial to

humanity. Plomin’s effort, I believe, is one of the most worthwhile

pursuits in present-day behavioral science.

 

Question #27

 

Chris Langan: It was suggested some time ago that pharmacological

methods, e.g. neurotransmitter loading, could boost mental

performance. More recently, the initial phase of the Human Genome

Project has begun to give way to the secondary proteomic phase, i.e.

tracing the biochemical pathways of genetic expression. As some of the

involved proteins are implicated in mental performance, new

IQ-boosting drug therapies may be discovered. Is there any reason to

be interested in genetic intellectual endowment when it may soon be

possible for the under-endowed to swallow higher intelligence in the

form of a pill?

Arthur Jensen: One important advantage of the purely genetic effects

on the development of intellectual functions, in contrast to

chemically induced effects in individuals, is obviously that the

genetic effects can be transmitted naturally from generation to

generation, whereas the chemical effects must be continually

reinstated anew in every generation. In a period of large-scale

catastrophe many of those who were dependent on the chemical treatment

would be deprived. I think it essential that the genetic mechanisms

involved in mental abilities to be further researched, because even

the discovery of effective chemical interventions for improving a

person’s level of g will depend in large part on an understanding of

the chemical pathways through which the genes affect individual

differences in g or other ability factors that may also be under

genetic influence.

 

Question #28

 

Chris Langan: Because genetic testing and engineering costs money,

only the wealthy can easily afford it. This raises the possibility

that intelligence will become increasingly correlated with

socioeconomic status that the central thesis of the controversial

bestseller The Bell Curve will be artificially amplified by genetic

tampering. Do you see this as a potential threat to social stability?

 

Arthur Jensen: This question raises serious concerns about the extent

to which, in a democratic society, the government should be involved

in control over science, its applications, and the lives of its

citizens in general. The thesis of The Bell Curve was met with

paroxysms of denial and it is doubtful whether the problem posed in

this question will, in the present political atmosphere, receive the

kind of serious discussion it deserves. The gap between the “haves”

and “have nots” in this country, to say nothing of the world at large,

is, I fear, already great enough to be “a potential threat to social

stability.”

Question #29

 

Chris Langan: Just as the human brain excels at certain intellectual

tasks, computers excel at solving other kinds of problem. Hence, the

idea of creating a superior intelligence by wiring together brain and

machine. Do you regard as ethical this potentially dehumanizing cyborg

approach to intellectual augmentation, which some regard as

inevitable?

 

Arthur Jensen: This still looks to me like science fiction. Many of us

are already quite tied to computers (I am in that condition at this

very moment!), although not through any direct line into the brain’s

circuitry. That possibility sounds a bit awful to me, but as a matter

of principle I won’t stop it if it became a reality. In my personal

philosophy I tend to be “pro-choice” all the way, and I only hope we

can preserve and promote that freedom!

 

Question #30

 

Chris Langan: As far as the evidence is concerned, the existence of g

is scientifically indisputable. But lets face it: this poses a problem

for minorities possessing statistically less of it per capita. After

all, if it is simply accepted that the mean IQs for colored people and

pure blacks are respectively one and two standard deviations below the

mean white IQ, employers and educators may be tempted to apply these

statistics in vocational and academic contexts, effectively leading to

discriminatory outcomes in which the minorities in question are

underrepresented. Accordingly, certain remedial principles of social

engineering are assigned a higher priority than the psychometric

findings themselves, resulting in reverse discrimination against

qualified people of European and Asian ancestry. Given that this

country is run by those with backgrounds in the social sciences rather

than in psychometrics, do you foresee any changes?

 

 

Arthur Jensen: I sense a growing tendency in our society in favor of

treating all persons as individuals, and I believe that increasingly

individual rights will trump group rights. The government itself

should not discriminate on the basis of race, ethnicity, national

origin, religion, sex, or sexual orientation. I believe the same

policy should be inculcated in the personal belief system of all

citizens. But of course this is one person’s ethical philosophy (and I

hope also that of a vast majority of Americans), although it has

nothing to do with scientific evidence. I believe that any kind of

quotas or discrimination in education or employment opportunities

based on an individual’s group membership rather than on that

individuals own characteristics only promote social conflict and

instability. A just society can help people in need without resorting

to discrimination on the basis of irrelevant criteria involving

group-membership. It also promotes ill will and social unrest if

members of minority groups have the perceptions that the majority is

not making a very real effort to shun group discrimination and to

treat people strictly as so-called “America’s race problem.”

 

Question #31

 

Chris Langan: Youre working on a new book. Can you please tell us

briefly what the working title is and what it will cover?

 

Arthur Jensen: The working title of the book I am presently writing is

“Mental Chronometry and Individual Differences.” It is not conceived

as a “trade book” in the least, but will be a highly specialized and

technical treatise for advanced students and professional doing

research in this field, or wanting to learn more about it. Mental

chronometry bridges the interface of brain and behavior and can

benefit both of these subjects of inquiry. To get a better hold on

brain-behavior connections, we need better behavioral measures of

individual differences than are provided by our present psychometric

tests that have no true scale and can only rank-order individuals. As

mentioned several times in response to previous question regarding

measurement problems, I believe we must measure individual differences

in mental abilities by means of true ratio scales, and these can be

made possible with mental chronometry. Models of brain activity built

on the time taken by various mental functions are already a venerable

area of research in experimental psychology and can provide a basis

for exploring the nature and dimensions of individual differences. The

burgeoning research literature on this is already surprisingly vast,

and it is a big job just getting it under control, even though I have

been working in this area for some 20 years. This research requires

very special instrumentation (now greatly aided by computers), and

individual testing of subjects under highly controlled laboratory

conditions. The time measurements obtained make much more sense in

relation to physiological and electro-physiological brain measurements

than do the ordinal-scale scores on psychometric tests. We are dealing

here with measurements in milliseconds, mostly in the range below one

or two seconds. These chronometric methods are of interest not only in

experimental and differential psychology, but are being increasingly

used in medical diagnosis and treatment. Chronometric variables are

fare more sensitive to subtle drug effects than are any psychometric

tests. Chronometric methods also can detect insidious brain conditions

long before they can be recognized through subjective self-awareness,

gross behavioral observations, or conventional psychological testing.

However, as a useful tool for studying individual differences in both

their normal and abnormal aspects, mental chronometry is still in its

bare infancy. I believe it should become a major branch of behavioral

science, and I hope my projected book will help it along this path.

Here is a link to one of the more difficult I.Q. tests, Ronald Hoeflin’s ‘Titan Test’  

http://www.eskimo.com/~miyaguch/titan.html
_________________________