I am beginning a series here attacking IQ, from an Alt Right perspective. To get the ball rolling, here is my first argument.
See: http://www.majorten.com/entertainment/27-celebs-fancy-degrees/16/, for an account of the IQ of Hollywood stars and celebrities.
There we are told that Stephen Hawking, the physicist has an IQ of 160, which is fair enough. But Sylvester Stallone, also has an IQ of 160! Some Hollywood cartoon characters have even higher IQs!
Now the IQs could be wrongly reported. But various other sources indicate, that people who most of us would regard as well, not exactly geniuses, end up coming out as geniuses at IQ tests. Surely this shows that it is the tests themselves, and what they are used for which is flawed? IQ tests thus yield paradoxical results, so by reductio ad absurdum, the tests must be incoherent.
Australian Aborigines, full-bloods, have average IQs of about 70, as do Africans, according to IQ guru Richard Lynn, who sees Jews and East Asians as intellectual superiors to all. The IQs of Africans and Aboriginals are at “retarded’ level according to Lynn. But there has to be something fundamentally wrong with a measure that condemns entire races of people as “retarded.” It makes no sense at all because from an evolutionary perspective, such people would have been eliminated, but the Aborigines and Africans survived and thrived in their native environments. Even given a cultural bias in the tests, this result should not occur, because it is contrary to evolutionary fact. Therefore, the IQ tests are flawed.
Consequently, it is the IQ measure itself, which is biased, embodying the values of the cognitive scientists at the time. Questions involving patterns and missing shapes are ambiguous, and often permit mathematically more than one answer if one thinks deeply about the problem, but of course there is only one “right’ answer, and there is no time for critical analysis because speed is of the essence in doing the test. This is not an exploration of intelligence, but of being a process worker, or a meat machine; a robot, not a creative individual.
And, there are deeper problems for Richard Lynn’s psychological theology. P. Bird, “The Impossibility of IQ,” Economic Letters, vol. 2, 1979, pp. 95-97, gives the following refutation of IQ, based on the mathematical result of Arrow’ impossibility theorem:
“We proceed by assuming that there exist n individuals, x, y, z, . . . . to be assessed by a psychologist who considers m specific abilities, i = 1,2,3, . . . . m, to be components of general intelligence. The psychologist is assumed capable, for each ability i, of ordering the individuals in some way, the order relation corresponding to the notion ‘is at least as intelligent as’.
An IQ function is then defined as a functional relation that for any set of orderings of the individuals by the various qualities determines a final general intelligence, or IQ, ordering. Formally, let X be the set of individuals.
The ordering by the psychologist of the individuals according to the ith ability is Ri, and let there be m such abilities. Let Q denote the desired IQ ordering. It is assumed that for each i there is an ordering Ri that is reflexive, transitive, and complete. Then an ZQ function is a functional relationship f such that for any set of m orderings of the individuals R 1, . . . . R, (one for each quality), one and only one IQ ordering, Q, is determined. The four Arrow conditions can now be imposed upon the IQ function. Condition U (Unrestricted Domain).
The domain of the function f must include all logically possible combinations of individual ability orderings. Condition P (Pareto). For any pair (x, y) in X, x Riy for all i implies x Q y. Condition I (Independence of Irrelevant Individuals). Let Q and Q’ be two IQ. orderings determined by the IQ function f corresponding respectively to two sets of orderings (RI, . . . . R,) and (R; , . . . . Rk). If for all pairs of individuals (x, y) in a subset S of X, x Ri y impliesx R: y for all i, then C(S, R) and C(S, R’) are the same, where C(S, R) is the choice set corresponding to the relation R. Condition D (No Dominant Ability). There is no ability i such that for every element in the domain off, for all (x, y) in X, x Ri y implies that x Qy.
Such a formulation parallels exactly Sen’s (1970) statement of the social choice problem. It thus follows directly that:
There is no IQ function that can satisfy simultaneously conditions U, P, I and D.
The relevance of this IQ Impossibility Theorem depends upon two conditions. The first, and the crucial, condition is that in any assessment of ability, the tester can apply to subjects only an ordinal and not a cardinal measure.
The second is that any meaningful ordinal measure of general intelligence must necessarily involve, explicitly or implicitly, constructing a general ordering from a number of particular orderings.
Both these propositions are valid.
To illustrate the first condition, consider three ways in which an apparently cardinal measure of a particular ability can be constructed. The first is that used by psychologists in reporting test scores. The rankings of a standardization sample are mapped directly to a normal distribution, with mean of 100 and standard deviation of 15. The numbers so obtained are far from objective measures of intelligence; rather they are statistical artefacts. A similar procedure is implicit in the grading in universities of students’ examination scripts.
A second road to cardinality is provided by the multiple-item test. The cardinal score is the aggregate of correct answers. But cardinalization is obtained at the expense of Condition U: the domain of the IQ function is restricted to that small subset of abilities by which individuals can be assessed on a possess/do not possess basis.
A third procedure is to cardinalise with some mechanical proxy, such as the time taken to solve a problem. i Again, the illusion of cardinality is false: the set of such times can be no more than a monotonic transformation of the ranking of individuals according to the ability.”
Hence an IQ function cannot exist.
I will have more to say in my next paper which addresses the question of alleged East Asian IQ superiority, in reply to a Chinese critic of one of my previous papers at this site, who obviously is happy with Lynn’s results, which have yet to be taken apart with the scrutiny they deserve. I have not forgotten you my critic, but in these times, battles need to be prioritised. In the meantime, the IQ cult can chew on Bird’s theorem.