“I must have a prodigious quantity of mind; it
takes me as much as a week sometimes to make it up.” (Mark Twain)
This page is probably the
most difficult one on my blog. But there is a point in any scientific
understanding of a subject beyond which popular verbalism simply cannot
proceed. A somewhat more technical and quantitative form of expression becomes
essential if one is to get even a glimmer of understanding of what the scientists
who work in the field are really talking about. I will try to keep the
exposition of the genetics of intelligence as basic, simple, and nontechnical
as possible, consistent with conveying those ideas most essential for
understanding the main themes of this blog.
1- Alfred Binet’s Point of View
I am going to start with an
extensive quote taken from Gould’s The
Mismeasure of Man, second edition
(1996). As I have already explained in this blog, I am not a big fan of Gould
because he denies the very concept of intelligence (in Spearman's reification of g, pp. 295-299). This is particularly
true in the first edition of The
Mismeasure of Man (1981). However, in this excerpt taken from the second
edition (Critique of the Bell Curve,
pp. 386-390), Gould quotes Binet abundantly, so by quoting Gould here I will be
essentially quoting Binet:
“Moreover, Binet feared that if teachers read the IQ
number as an inflexible inborn quality, rather than (as he intended) a guide
for identifying students in need of help, they would use the scores as a
cynical excuse for expunging, rather than aiding, troublesome students. Binet
wrote of such teachers: "They seem to reason in the following way: 'Here
is an excellent opportunity for getting rid of all the children who trouble
us,' and without the true critical spirit, they designate all who are unruly,
or disinterested in the school." Binet also feared the powerful bias that
has since been labeled "self-fulfilling prophecy" or the Pygmalion
effect: if teachers are told that a student is inherently uneducable based on
misinterpretation of low IQ scores, they will treat the student as unable,
thereby encouraging poor performance by their inadequate nurture, rather than
the student's inherent nature. Invoking the case then racking France, Binet
wrote:
It is really too easy to
discover signs of backwardness in an individual when one is forewarned. This
would be to operate as the graphologists did who, when Dreyfus was believed to
be guilty, discovered in his handwriting signs of a traitor or a spy.
Binet felt that this test could best be used to
identify mild forms of retardation or learning disability. Yet even for such
specific and serious difficulties, Binet firmly rejected the idea that his test
could identify causes of educational problems, particularly their potential
basis in biological inheritance. He only wished to identify children with
special needs, so that help could be provided:
Our purpose is to be able to
measure the intellectual capacity of a child who is brought to us in order to
know whether he is normal or retarded. . . . We shall neglect his etiology, and
we shall make no attempt to distinguish between acquired and congenital
[retardation]. . . . We do not attempt to establish or prepare a prognosis, and
we leave unanswered the question of whether this retardation is curable, or
even improvable. We shall limit ourselves to ascertaining the truth in regard
to his present mental state.
Binet avoided any claim about inborn biological limits
because he knew that an innatist interpretation (which the test scores didn't
warrant in any case) would perversely destroy his aim of helping children with
educational problems. Binet upbraided teachers who used an assessment of
irremediable stupidity to avoid the special effort that difficult students
require: "They have neither sympathy nor respect for [these students], and
their intemperate language leads them to say such things in their presence as
'This is a child who will never amount to anything . . . he is not intelligent
at all.' How often have I heard these imprudent words." In an eloquent
passage, Binet then vented his anger against teachers who claim that a student
can "never" succeed as a result of inferior biology:
Never! What a momentous
word. Some recent thinkers seem to have given their moral support to these
deplorable verdicts by affirming that an individual's intelligence is a fixed
quantity, a quantity that cannot be increased. We must protest and react
against this brutal pessimism; we must try to demonstrate that it is founded
upon nothing.
Finally, Binet took pleasure in the successes of
teachers who did use his tests to identify students and provide needed help. He
defended remedial programs and insisted that gains so recorded must be read as
genuine increases in intelligence:
It is in this practical
sense, the only one accessible to us, that we say that the intelligence of
these children has been increased. We have increased what constitutes the intelligence
of a pupil: the capacity to learn and to assimilate instruction.
How tragic and how ironic! If IQ tests had been
consistently used as Binet intended, their results would have been entirely
beneficent (in this sense, as I stated, I do not oppose mental testing on
principle, but only certain versions and philosophies). But the very innatist
and antimeliorist spin that Binet had foreseen and decried did become the
dominant interpretation, and Binet's intentions were overturned and inverted.
And this reversal—the establishment of the hereditarian theory of IQ — occurred
in America, not in elitist Europe. The major importers of Binet's method
promoted the biodeterminist version that Binet had opposed—and the results
continue to ring falsely in our time…
… IQ is a helpful device for identifying children in
need of aid, not a dictate of inevitable biology. Such aid can be effective,
for the human mind is, above all, flexible. We are not all equal in endowment,
and we do not enter the world as blank slates, but most deficiencies can be
mediated to a considerable degree, and the palling effect of biological
determinism defines its greatest tragedy—for if we give up (because we accept
the doctrine of immutable inborn limits), but could have helped, then we have
committed the most grievous error of chaining the human spirit.
Why must we follow the fallacious and dichotomous
model of pitting a supposedly fixed and inborn biology against the flexibility
of training—or nature vs. nurture in the smooth pairing of words that so fixes
this false opposition in the public mind? Biology is not inevitable destiny;
education is not an assault upon biological limits. Rather, our extensive
capacity for educational improvement records a genetic uniqueness vouchsafed
only to humans among animals…
Biology is not the enemy of human flexibility, but the
source and potentiator (while genetic determinism represents a false theory of
biology). Darwinism is not a statement about fixed differences, but the central
theory for a discipline—evolutionary biology—that has discovered the sources of
human unity in minimal genetic distances among our races and in the geological
yesterday of our common origin.”
2- Heredity vs. Heritability
The terms “heredity, hereditary, and inherited”
simply mean that a given trait or characteristic of individuals depends on the
presence of certain genes or that the gene or genes affecting the trait are
transmitted from parents to offspring. Virtually all babies are born with one
head, two hands, and ten fingers, for example. As these human characteristics
are coded in the genes, they are hereditary characteristics.
According to the National
Human Genome Research Institute, we all share 99.9 percent of our three billion
DNA base pairs. All human beings are remarkably similar and they always give
birth to other human beings (not to whales, cats, or jelly fish). Additionally,
humans are far less genetically diverse than most other mammalian species.
The term “heritability” is only applicable (and,
in fact, only meaningful) if applied to hereditary characteristics for which
the population variance is not zero. When characteristics do not normally vary
among individuals, the concept of heritability is simply inapplicable to them. For
instance, ears have zero heritability in the human population because all
humans have ears. This concept of heritability may seem difficult to grasp but
it is not. Let me take an example:
I went to the supermarket
the other day with a magnifying glass. I picked randomly two similar bottles of
shampoo on a shelf and compared them in detail: I systematically looked for all
possible differences. This is what I found:
I first noticed that the
level of shampoo in the two bottles was not exactly the same. Also, in one
bottle I could see little bubbles but not in the other one. The two tops had
not been screwed exactly in the same position; one was headed in one direction
while the other was headed in a different direction. Then I took my magnifying
glass and found that there were many small scratches all over the two bottles.
I started studying the different patterns of scratches: very few of the dozens
of scratches on the two bottles were on the same locations. By the time I
finished I had established a long and impressive list of differences between
the two bottles.
Upon analyzing my data, I
discovered that 80% of all the differences between the two bottles were mostly due
to scratches uniquely located on one or the other bottle. I called these
differences environmental differences (i.e. differences that happened post
production and that were mostly due to handling). 20% of the differences I
called production differences or differences due to the manufacturing process
itself (such as level of liquid or printing variations).
If you think that
systematically comparing two similar bottles of shampoo in a supermarket is of
little interest (some kind of hair splitting or “fly fucking” as we say in
French) you are absolutely right. But this is precisely what heritability aims
at doing: finding differences among similar things, differences that MUST BE
small because the things compared are similar. For instance, you will never
find a heritability study comparing the intelligence of humans and cats because
they do not belong to the same species/populations.
Let’s go back to our shampoo
bottle example. Knowing that 80% of the variations between my bottles of
shampoo are environmental and 20% production based is of absolutely no
importance since the two bottles are virtually identical. I will not wash my
hair better with one bottle or the other…
You may find some old
studies that claim that intelligence is “80% heritable” (the current figures
are closer to 40%) and 20% environmental (respectively 60%). This sounds
impressive until you realize that the differences are between the tips of very similar
icebergs. To make matters worse, the percentages given refer to standard
deviation units, which are not real units of intelligence. Last but not least,
intelligence is increasing, so the situation is dynamic.
To clarify the situation
concerning the heritability of intelligence, I am going to take the example of
height. Height is an undisputable measure which, in some essential ways, is
very similar to intelligence. There are two crucial differences though:
1- Height is measured on a
ratio scale of measurement: a scale with a true zero and equal intervals. So
far, the main measures of intelligence are standard deviations, which have no
true zero and no equal intervals (we only have the rank orders).
2- The heritability of
height is very high and well-known (90%), which is not the case for
intelligence.
Interestingly, height was the
measure of choice selected by Galton to establish what
he first called reversion to the mean
and later the law of filial regression to
mediocrity (a.k.a. regression to the
population mean). An exceptionally tall father, for example, had sons who
were shorter than he; and an exceptionally short father had sons who were taller
than he.
I have obtained data from
the French Army to document what is now known as the secular trend in human growth, maturation, and development. The
French Army has sent me the average heights and standard deviations of French
conscripts from 1880 to 1990 (French conscription was terminated in the 1990s):
For some unexplained
reasons, I haven’t found any serious research that would explain the Flynn
effect as a particular case of this broader human growth trend.
I have drawn below the
normal curves corresponding to the 1880 and 1990 data (with inches as units for
readers unaccustomed to the French metric system).
The average height in 1880
was about 65”, which by today standards would be considered short or very short.
The average height in 1990 was about 69”, which by 19th century
standards would be considered tall or very tall – only 5% of the 1880 male
population had a height equal or higher to 69”. In 1880, very few French males
(less than 2%) were taller than 70”, whereas in 1990, 36% of them were. If you
consider the number of French males at +1 SD (standard deviation) today (71.5”
or roughly the top 16%), i.e. people that can be considered tall, then 99.5% of
the 1880 male population was shorter than them. To be at +1 SD in 1880 (considered
tall), you had to be 67”. In France, a 71.5” of 1880 was equivalent (in
frequency terms) to a 76.5” of 1990. In the US of the 21st century
(with an average male height of 69.3” and a SD of 2.94”) a 77.5” man would look
like a 71.5” Frenchman of the 19th century.
While heritability may
explain the relative height differences between people at a given time, it
cannot account for the substantial height increases that took place over more
than a century. The height increase was large and almost entirely environmental
(because the gene pool of the French population remained basically the same).
Similarly, intelligence
could double or triple due to environmental factors and be “90% heritable” (to
be clear, the true heritability of intelligence is closer to 40%).
Conclusion 1: If you want to know by how much you can actually
increase your IQ, heritability is of no interest at all. Undoubtedly, heritability
is an inadequate and misleading indicator that has been overused by hereditarians.
Heredity does not determine a fixed level of intelligence, particularly not in
the case of human beings who have highly trainable and flexible minds.
Last but not least,
intelligence (“general cognitive ability”) is considered by geneticists to be a
very polygenic trait, i.e. a characteristic which depends on a great number of
genes (it is distributed all over our chromosomes).
Conclusion 2 : You cannot easily “breed” smart human beings, there
are far too many genes to manipulate. As Galton had already noticed,
eventually, it all reverts to the mean (a mean that luckily seems to be
constantly increasing).
If you wish to improve your
own IQ, you may contact me (Phil – the author) by clicking on the profile link
(above right) or by typing in the following email address:
