Skip navigation to content

Are Atheists Really More Intelligent Than Others?

Denis Alexander

This article first appeared in the Church Times on 26 March 2010. It appears below or can be dowloaded in PDF format.

Dr Denis Alexander has spent the past 40 years in the biological research community and is Director of The Faraday Institute for Science and Religion.


The head-line 'Atheists are found more intelligent' in the 5th March edition of the Church Times was surely designed to catch the attention of the highly intelligent readership of this publication. But delving beneath the head-lines can be informative.

The article described a recent paper by Dr Satoshi Kanazawa of the London School Economics entitled 'Why Liberals and Atheists Are More Intelligent' [Social Psychology Quarterly. 73: 33-57, 2010]. Dr Kanazawa has attracted some notoriety through his previous publications touching on issues of race and IQ. But as he himself has written: 'Nothing else should matter in science except the objective, dispassionate pursuit of truth, and scientists must pursue it no matter the consequences' [THES, 15th Dec. 2006]. So, keeping that excellent principle in mind, let us look a little more closely at Dr Kanazawa's article on atheism.

Savanna-IQ Interaction Hypothesis

The data presented are cast within the framework of the 'Savanna-IQ Interaction Hypothesis'. The basic idea has been around for a very long time, and suggests that as our ancestors were evolving on the Savanna plains of Africa during the Pleistocene era (viz. between 2.6 million and 10,000 years ago), so their brains became well-adapted to a hunter-gatherer life-style. Our brains today, so the familiar argument goes, are therefore often much better adapted for life on the Savanna than for pursuing unexpected or intellectually more challenging goals, described as 'evolutionarily novel'.

Dr Kanazawa further suggests that universal human nature is characterized by a kit of innate cognitive abilities that cope with basic needs, such as survival, way-finding, reproduction, and interpersonal relationships, requiring little intellectual challenge. By contrast, 'general intelligence' has evolved to allow individuals to cope with unusual situations, and so today more intelligent individuals generate evolutionarily novel human beliefs, attitudes and activities. Implicit in this theory is the controversial claim that those with higher 'general intelligence' have a different genetic profile when compared with those of lower intelligence.

Scientific hypothesizing

Since Dr Kanazawa claims this as a scientific hypothesis, a few general points about scientific hypothesizing may be in order. A good hypothesis should be plausible: it should look more convincing than other more obvious alternatives. It should be as simple as possible, without introducing unnecessary convolutions. It should be falsifiable in principle: there must be data that could count against it. And it should explain the data better than its rivals. This of course depends on having reliable data available. Reliable data can lead to a solid hypothesis like a Lego kit construction in which each piece of data plays a critical role in the ensemble. Unreliable data, or data suffering from speculative interpretation, can lead to a house-of-cards.

Unfortunately, when Dr Kanazawa seeks to apply his Savanna hypothesis to the relationship between atheism and intelligence, it soon becomes apparent that it's of the house-of-cards variety. The hypothesis depends on the claim that because religion is now a cultural universal (which it is), therefore it must always have been so, in particular during the Pleistocene era when our cognitive faculties were evolving.

The problem is that we have no idea if that is the case, nor, if there was religion, what form it took. Firm data about religious beliefs only appear with human literature, from 5,000 years ago, and religious interpretations of cave art, burial sites and other artifacts from 35,000 years ago are rife with speculation. Yet anatomically modern humans were emerging in Africa from 200,000 years ago. So even if we ignore the rest of our earlier Pleistocene cousins and focus on our own species, we only have reliable data concerning humanity's religious beliefs for less than 5% of our existence.

Furthermore, the notion of atheism in Kanazawa's paper is problematic. The idea is that atheism is evolutionarily novel, not part of the background 'default' position of presumed innate human religiosity. But the notion of 'atheism' (in its western sense) is very modern and refers to a range of different positions. The Greek term atheistos meant 'one who denies the traditional religion of the Athenian establishment' as Socrates found to his cost. Later the early Christians were accused of being atheistos because they refused to join in the cult of the Roman Emperor.

The meaning of atheism

Atheism did not begin to acquire its modern meaning ('denial of belief in the God of the Abrahamic faiths') until the 17th century when Christian apologists began to critique it, although historians have a hard time identifying any atheists from this period. Even in the 18th century the Scottish philosopher David Hume, doyen of modern atheism, once commented that he had never actually met an atheist. Virtually all the natural philosophers who established 'evolutionarily novel' modern European science were Christians, and presumably Isaac Newton and Albert Einstein, both believers in their different ways, were quite intelligent.

The data that Kanazawa actually reports are sets of correlations between increased aptitude in intelligence tests (specifically verbal intelligence, suggested to reflect 'general intelligence') and low religiosity scores in US populations.Despite the fact that the data are correlative, Kanazawa's paper is peppered with the language of cause and effect, as if 'intelligence' is a causal factor in levels of religiosity. But correlations are not the same as causations, and the data are open to a myriad straightforward sociological interpretations, local to the present structure of US society (in which only 1.6% self-describe as 'atheist':, requiring no recourse at all to evolutionary psychology.

For example, since religious belief in the USA often correlates with political conservatism, and atheists are such a tiny minority, it is quite possible that the more verbally intelligent secondary school pupils reported in Kanazawa's paper, whose low religiosity scores were based entirely on self-reported assessment, may perceive themselves as rebels against the status quo.

Verbal intelligence is a plausibly useful asset in independence of thought, and so a similar correlation might be found, for example, amongst a religious minority exerting their independence in another country in which atheism is the status quo. I do not know whether this is the case. The point is simply that proximal cultural factors explaining correlations between intelligence and religiosity are intrinsically more plausible than convoluted explanations that depend on evolutionary speculations about what might have been the case hundreds of thousands of years ago, especially if human genetic variation is supposed to have anything to do with it.

Of course at this juncture it is always open to a supporter of Kanazawa's thesis to say: 'of course, that's what you expect, independence of thought is 'evolutionarily novel' as well!' But by that stage it's clear that the emperor has no clothes. A theory that can explain everything ends up by explaining nothing.

New atheists

Most puzzling of all for Kanazawa's evolutionary interpretation is why, despite differences in 'intelligence' between individuals that must go back millennia, 'atheism' suddenly appeared upon the world scene as a relatively popular human conviction in the 20th century. Were 20th century communists supposed to be more intelligent? Kanazawa does not tell us.

One of the many reasons why his Savanna-IQ hypothesis lacks plausibility in this context is simply because there are good politico-historical reasons why atheism appeared in the 20th century: the Leninist-Marxist revolution of 1917 ushered in an era of state sponsored atheism. By the early 1980s around one-third of the world's population lived under communist regimes. Yet the collapse of Soviet communism in the late 20th century has been associated with a revival of religious belief. North and South Korea have been separate countries since 1948, insufficient time to establish any significant genetic differences between the two populations, yet the North is resolutely atheistic, whilst at least 50% of those living in the South are religiously committed.

The contemporary shrill 'new atheists' of the western world and the 1.6% atheist US population are a tiny side-show compared with the 20th century waxing and waning of atheism on the world-stage. In the 'dispassionate pursuit of truth' Kanazawa's Savanna-IQ hypothesis does not do very well in explaining such phenomena: rival explanations based on culture and politics do so much better.