People who do the kind of research Kanazawa does often cannot get published. The get fired from their jobs at conservative as well as liberal think tanks. They have a hard time getting tenure at universities. There is little freedom of inquiry left in the realms in which Kanazawa works. For that reason alone it is delightful to see him publish such a politically incorrect work as this.
In his preface Kanazawa credits the giants who preceded him. Robert Trivers, Arthur Jensen, Philippe Rushton, just to name three. What I hoped in reading this book was to find a worthy successor to these, the latter two of whom died just last year. My hopes remain somewhat unfulfilled. While Kanazawa has the courage to beard any lion in its den, he has some shortcomings when it comes to structure and research.
The introduction to the book is a delight. It is a defense of academic freedom, the right of an academic, indeed, the obligation of an academic, to pursue the truth wherever it takes him. Quite specifically, this pursuit of truth should not be constrained by concerns about what the implications might be of the truths that are found. He talks about the great fallacies, the naturalistic fallacy and the moralistic fallacy. These are worth repeating right here in this this review, because they are so pervasive and academics today.
"The naturalistic fallacy, which was coined by the English philosopher George Edward Moore in the early 20th century, though first identified much earlier by the Scottish philosopher David Hume, is the leap from is to ought--that is, the tendency to believe that what is natural is good; that what is, ought to be. For example, one might commit the error of the naturalistic fallacy and say, "Because different groups of people are genetically different and endowed with different innate abilities and talents, they ought to be treated differently."
"The moralistic fallacy, coined by the Harvard microbiologist Bernard Davis in the 1970s, is the opposite of the naturalistic fallacy. It refers to the leap from ought to is, the claim that the way things ought to be is the way they are. This is the tendency to believe that what is good is natural; that what ought to be, is. For example, one might commit the error of the moralistic fallacy and say, "Because everybody ought to be treated equally, there are no innate genetic differences between groups of people." The science writer extraordinaire Matt Ridley calls it the reverse naturalistic fallacy."
The scientist cannot afford to be blinded by any of this. However, the frightening thing is that the vast number of them are. Read my review of "Race Decoded" for an example. Stephen Jay Gould's "Mismeasure of Man" is a classic example of a political screed masquerading as science. Its specious arguments still sway liberal-arts majors by the thousands.
Kanazawa depends vary greatly on statistical techniques, especially factor analysis. For a work that depends so heavily on statistics, he does very little to explain his statistical methods. He sometimes assumes that the reader knows nothing, and at others a great deal. On the nothing side, talking about intelligence distribution, he says that 5% of the population is very bright (IQ>125) and 20% bright (110-125). No statistical terms whatsoever. Elsewhere, however, he uses the statistical terms of mean and standard deviation, without a background as to how the latter is an expression of how the Gaussian distribution (bell curve) indicates how rare or common a given IQ score is. Here is the title of one of his graphs: "Figure 4.1 Partial association between lifetime number of sex partners and number of children among the less intelligent." The plot shows an upward slanting line amid a sea of dots. The index for children runs from -4 to 8; for sex partners from -100 to 300 (Wilt Chamberlain wasn't one of the dots.) A reasonable person might wonder how it is possible to have a negative number of either children or sex partners. It would help to provide said reasonable person with a discussion of "partialing out" variables in multiple linear regression. Without this help, the average reader will simply say, "Huh?"
Elsewhere, he sometimes normalizes his variables to a mean mean of zero and a standard deviation of one. That's well and good for statisticians, but he is writing a book with a real world audience. The topics he chooses, why homosexuals may be smarter than straight people, why and chosen people are not having children, are aimed to the general audience. He is expecting a higher level of understanding of statistics then he has any right to assume.
Kanazawa's intro says that "Most of the empirical analyses that are summarized in this book use three different data sets: General Social Surveys (GSS) in the US, National Longitudinal Study of Adolescent Health (Add Health) in the US, and the National Child Development Study (NCDS) in the UK." These are long term surveys. The GSS surveys 1500 people (3000 every other year) on a large number of topics. Add Health is a longitudinal survey - it follows 20,000 school children who were selected in 1994-95 through life. The NCDS follows all children born in Great Britain during one week in 1958 (17,000) through life. Each of the surveys includes an IQ component, or a good enough proxy for IQ to satisfy Kanazawa's purposes. As is usually the case in social science, he can be thankful to others' extensive labor in compiling what he got, but has to be mindful of the limitations. Each survey is constrained by the way the questions are worded and the questions that were asked. In each survey participants drop out in non-random ways. The surveys are culture-bound (US and UK) and bound in the case of longitudinal studies to late Baby Boomers (Great Britain) and late Gen X (United States). Societies change over time, and he has chosen cohorts each of which represents only one point on a time continuum. Most significantly, he generalizes from the US and UK to Europe, which might work, and to Asia, which probably does not. He could be accused of being influenced by the "streetlight effect," about which Wikipedia says "The streetlight effect is a type of observational bias where people only look for whatever they are searching by looking where it is easiest.
Surveys are by their nature are imperfect instruments. They generally ask the respondent to answer on a scale of 1 to 5 or 1 to 7, from "strongly agree" to "strongly disagree" or something similar. How the question is worded affects the answer. The respondent may be culturally inclined to avoid - or to take- extreme positions. Survey designers usually include several related questions, so they can compose aggregate answers which are more telling than the individual ones. Still, it is an imperfect art.
Kanawawa's overarching theme is that intelligence correlates highly with a love of novelty, and that our love of novelty often takes us in evolutionarily unsuccessful directions. It is pitched towards a fairly general audience - liberal arts majors. Therefore, the level of statistical understanding that he expects exceeds what can be expected of the type of audience he wants to attract.
The novel things he says we do include among other things homosexual practices, getting a university education, embrace the kinds of universal altruism represented by liberal dogmas, and using drugs and alcohol. To me the point seems a little bit of a stretch. I think that he is at least in part conflating trends which are taken place within American society, or all of Western society over the last century, with with his thesis of novelty.
Classic liberalism was not what we moderns would call liberal: more like libertarianism. The eugenics movement of a century ago was certainly what one would call conservative movement that was led by Victorian intellectuals. Fascism and communism were broad extensions of government, but certainly not consistent with modern liberalism. So to say that liberals are the ones who seek novelty, and are thus intelligent, is a somewhat shaky thesis. Novelty seekers have not always been liberal. An alternative thesis (America-Lite: How Imperial Academia Dismantled Our Culture (and Ushered In the Obamacrats)) would note that since liberals claimed control of the American campuses worldwide campuses about the 1960s and haven't let it go. Therefore everybody experiencing a university education in the West is brainwashed with an inordinate degree of liberal thought. The question is not why many of them turn out to be liberal, but how some of them manage to survive a college education these days and remain conservative. In other words, it should not attributed merely to the novelty of liberalism.
One observation I will make which Kanazawa should is that there are far more people of exceptional intelligence among North Asian populations than there are among Caucasians. Just by brute numbers, the Chinese and the Koreans and Japanese together number about time and a half the worldwide Caucasian population. As Kanazawa notes, they have a higher average intelligence. Put the two facts together, and the higher up you go in intelligence, 125 would be his cutoff, the greater the proportion of North Asians. By my rough statistics, about three times as many North Asians should have an IQ over 125 as Caucasians. In Excel, =1.5*(1-NORMDIST(125,105,15,TRUE))/(1-NORMDIST(125,100,15,TRUE)). Orientals don't display the same kind of liberalism, and they have a markedly different intellectual history than we do in the West. His thesis certainly should be tested against what's happening in the lands of his ancestors.
I found the book's treatment of homosexuality to be quite interesting. Kanazawa found some sources that I had not seen before, and his percentage attribution of homosexuality to genetics, environment in utero, and the environment in which a child grows up I found to be reasonably credible. He also had a useful four-way division of the definition of homosexuality. How people define themselves, what they do, their reports of the kind of erotic response they feel when exposed same-sex images, and objectively measured response to sexual stimuli. In other words not everybody's going to agree on who is a homosexual who is not. It's a question both of definite feelings and actions, and latent characteristics that may or may not be expressed. All that said, the differences he finds in intelligence between gays and straights seems to be seems to be fairly high. Again, one wonders what he would find in different societies.
One strength of his discussions about homosexuality is his reference to ethnographic literature citing homosexuality in 1500 societies around the world. Basically, it was not much noticed by anthropologists, and he defends the anthropologist's ability to notice these things. I would agree with them that homosexuality is much more evident today than it was even during my childhood 60 years ago. This may be a function of simply revealing latent traits that were there all along, but I suspected it also involves a greater element of choice than the homosexual community would like to admit.
The statistics he presents on the marked difference in intelligence between liberals and conservatives is quite surprising. I could believe this among white populations. Although he does not say which populations he is dealing with, his text reads as if he's talking about white people. However, although this book does not dwell on it, Kanazawa recognizes in his introduction that that black and Hispanic populations are significantly below the whites in average intelligence, and their voting is certainly what a modern would call liberal. Therefore if he is talking about all populations in the United States or in Western Europe, I would be suspicious of his case that liberals are more intelligent. Being conservative is contrary to the interests of less intelligent people. Nobody disputes that less intelligent people benefit more from government largess in government handouts, and most people are at least smart enough to vote their economic self-interest.
I conclude that at the top end of the intelligence distribution he is probably right. The collegiate liberals are people who feel they can be magnanimous with other people's money, and even some of their own. Father down the intelligence and income distribution I think he's got it wrong.
Kanazawa's treatment of religion seems to be quite simplistic. He says that the religious are less intelligent than atheists. I would quibble on several points. First of all, a great many people who claim not to believe in any identified religion cling religiously and without any intellectual curiosity to elements of their liberal dogma. Environmentalism is one of these things, vegetarianism, belief in holistic medicines, bottled water and the evils of circumcision - the list goes on. And they're adamant, as adamant as a fundamentalist with his rattlesnakes.
I pose a philosopher's question, what is religion? Another question would be about religion within cultures. Islam, Hasidic Judaism, and Christian fundamentalism are all quite different. The religion is part and parcel of the culture. I would bet that Mormons and Hasidic Jews are smarter than the average white person. So to generalize about religious belief I think glosses over some pretty important aspects of culture which are associated with religion. Simply to say that atheism is novel is nothing new. Again, there would be a lot to be gained by a discussion of North Asians. The Chinese and Japanese do not have religions in the same sense as people in the West. There is certainly no belief in the divinity of a person such as Christ or the status of a person such as Mohammed. There, religious leaders are regarded as teachers and seldom much more. So one follows the teachers, but one doesn't believe in their divinity. I would call them atheists. Any discussion of religion that doesn't include what the North Asians believe, is kind of off the track. One might well also include Indians, who have a vast number of beliefs, but are not points of belief in divinity in the same way that Christians Jews and Muslims, people of the book, believe in their God.
Certainly more intelligent people led the drug revolution in the US in the `60s. I was there - Berkeley. However, I note that where I now live in Ukraine, the 20-somethings from the top universities with whom I associate have no experience with drugs. Here, as in Brazilian favelas, it is the dead-enders in the lower rungs of society who do drugs. They sniff glue and do needle drugs. For them it is not novelty - it is culture.
As far as sexual experimentation goes, that too changes quite rapidly. I witnessed it go from quite prudish in the `50s, when knocking a girl up was a disaster, to very loose during the sexual revolution of the 60s, to increasingly guarded, first when herpes became widespread, then AIDS. Culture also played a part, especially changing, unpredictable feminist notions of what they wanted men to be. Whether and how a guy got sex changed like lady's fashions - because it was in part a fashion concern. The novel thing is the swiftness with which it changes - for everybody.
In conclusion, I would like to say that what this field of evolutionary biology needs is more people like Kanazawa. Specifically, Kanazawa needs more people against whom he can bounce his ideas. A lot of these ideas are kind of off the wall. They would have benefited by being polished by a larger number of associates with similar views. Unfortunately, the academy the universities absolutely discourage people think and talk about these subjects, and therefore the number who are bold enough to work in this field is quite limited. Kanazawa must feel like he is the last of the Mohicans, the Lone Ranger. I hope his bravado inspires others to stand up against the tyranny of political correctness and join him in doing real science. He has put forth ideas which merit deeper study. Hope somebody rises to the challenge, and helps Kanazawa polish his theses.