Calling himself a scientific fundamentalist, evolutionary psychologist Satoshi Kanazawa does not shy away from controversy. Arguing with fundamentalists is usually futile, but when his latest book fell on my doormat I put aside my initial scepticism and started reading. The London School of Economics scholar presents his ideas forcefully, provides interesting examples and rightfully corrects some common misunderstandings about IQ. He introduces an evolutionary theory of general intelligence to explain a host of interesting phenomena such as infidelity, the tendency to stay up late, homosexuality, friendship, religion and political views. In each chapter, he presents data in support of his theory.
Unfortunately, Kanazawa's thesis failed to convince me. His evolutionary theory speaks to human nature in general, but the empirical support comprises predominantly weak associations in contemporary data. IQs are found to be associated with variables such as conservatism and religion, selected from large-scale surveys that contain well over 2,000 characteristics for each participant. For the most part, Kanazawa's results support his theory, but he would not be the first scientist (or fundamentalist) to ignore data that fit his ideas less well.
A good evolutionary theory enables the formulation of specific predictions, but the theoretical concepts in Kanazawa's theory are rather vague. For instance, he views general intelligence as a species-typical adaptation to deal with "evolutionarily novel" problems. General intelligence is something that varies between people (otherwise IQs would not predict anything), but Kanazawa fails to clarify how and why people differ in terms of this core adaptation. Similarly, his main idea is that humans have "difficulty comprehending and dealing with entities and situations that did not exist in the ancestral environment". This "ancestral environment" is the African savannah in which, according to Kanazawa, humans evolved before the Neolithic revolution (around 10,000 years ago). He defines comprehension as "true logical, and scientifically and empirically accurate understanding of how something works", but fails to specify what it means to "deal with" entities and situations. Dealing differs from comprehending. Although my three-year-old son is not (yet) as smart as his father, he effortlessly deals with an iPad or television. But I doubt he comprehends the inner workings of these evolutionarily novel gadgets (at least I don't). Because of such lack of specificity, empirical predictions from Kanazawa's theory can be easily fitted to a host of empirical phenomena.
Another core prediction from Kanazawa's theory is that intelligence as measured by IQ tests should not be related to "evolutionarily familiar" problems such as mating, parenting, bonding with others and staying alive. He provides some empirical support for his predictions, but "dealing with" and "evolutionarily familiar" are often used loosely. For instance, marriage could be viewed as evolutionarily novel (our ancestors on the savannah did not marry) but also as evolutionarily familiar (our ancestors on the savannah did pair up). So if divorce rates are found to be lower among those with higher IQs, Kanazawa could claim divorce to be evolutionarily novel. Dealing with an unsuccessful marriage may mean either divorce or staying together, but his theory is silent on which action is most evolutionarily adaptive or relevant. Similarly, when the number of close friends one has is positively associated with IQ, suddenly the quality of friendships (negatively associated with IQ) becomes Kanazawa's variable of interest. And when intelligent individuals do appear to be best at friendship or parenting, this is only so "in some ways in the current (evolutionarily novel) environment".
This lack of specificity was illustrated in a 2008 article in the journal Intelligence, in which Kanazawa pitted his evolutionary theory of general intelligence against a similar theory by his friend and fellow scholar J. Philippe Rushton. According to Rushton, the African savannah was harsh and unpredictable while in Kanazawa's theory it is the epiphany of stability. Whereas Rushton's theory states that longevity, marital stability and parental care are positively associated with general intelligence, in Kanazawa's, general intelligence is uncorrelated with these variables. Rushton and Kanazawa may be friends but their theories disagree. Interestingly, Kanazawa claimed to have found strong support for both theories in 2008, but since he no longer discusses the discrepancies, I continue to be confused.
Kanazawa is an original thinker and this book is an interesting read, but his scientific fundamentalism may stand in the way of rigorous testing of his theory. For many readers this poses no problem, but for those who want to understand human evolution, it does.
The Intelligence Paradox: Why the Intelligent Choice Isn’t Always the Smart One
By Satoshi Kanazawa
John Wiley & Sons, 2pp, £16.99
Published 10 April 2012