It Is Not the God Engine That Sputters…

James E. Alcock

I am puzzled by the tone of Gregory Paul’s criticism of my article “The God Engine.” While respectful and constructive criticism is always welcome, his scornful and dismissive condemnation is both unwarranted and unhelpful. Enough said.

I am also puzzled by his misunderstanding of my article. He has set up a strawman based both on his misinterpretation of what I have written and his apparent unfamiliarity with the careful and extensive body of empirical research that underlies it. (That research is discussed in greater depth in the book chapter from which my article is derived, and extensive references to the research literature are provided.)

I am also disturbed by his obvious misunderstanding of the limitations of the data that he adduces and his mistaken notion that they can be used to “scientifically” test theories about the acquisition of supernatural beliefs in childhood.

And, finally, I am disappointed that he has ignored the opportunity to make common cause. His “attack mode” stance appears to have blinded him to the fact that there is no contradiction between the demographic data he presents and what I have written about how normal cognitive developmental processes render children prone to supernatural beliefs. 

The strawman: Paul misunderstands or misrepresents what I have written when he accuses me of reflecting the “common belief that religion is highly popular and enduring because it is a natural consequence of the way brains work from childhood on up.”  First, as I pointed out in the article, “religion” has to be taught; children do not invent deities or Santa Claus or tooth fairies on their own. If a child is not taught about Lord Krishna or Jesus, he or she will never develop that specific religious belief. However, because of the normal course of cognitive development, supernatural beliefs of one form or another come automatically to children, and they are very receptive to religious beliefs taught by trusted adults. Paul actually seems to agree with me on this point when he states that “children are highly gullible for assorted reasons including overactive pattern recognition that leads to unsubstantiated but often ardent belief in undetectable, magical items such as deities” and when he writes that “Another problem with the hypothesis that mass religion stems from children being automatically prone to magical and supernatural thinking is that they are taught such notions by adults, so it is more likely that kids are being contaminated by grownups than the reverse.” But he then goes on to accuse me of “the assumption that adults are highly prone to retaining juvenile gullibility, which is exaggerated to the point of an unsubstantiated mythology.” He stuffs more chaff into his strawman when he erroneously attributes to me the suggestion that our brains are “genetically preprogrammed” to be religious, that “almost all humans are inherently religious … ,” and that theism of some form or another is “truly universal to human brains … .”

I did not suggest such things. As do beliefs in general, supernatural beliefs, including theistic ones, thrive or wither depending on their exposure to logic, their seeming reinforcement through experience, and whether they are stimulated or discouraged within the individual’s important social groups. And when children are encouraged as they grow older to submit supernatural beliefs to logic, as they do with Santa Claus (even though some children struggle at first if informed by parents that there is no Santa), then they are likely to abandon specific theistic beliefs as well. On the other hand, when children are taught not to question and not to apply logic to their religious beliefs, challenging them becomes difficult, and it is often easier to “go with the flow” when part of a community that endorses such beliefs.

A more important concern about Paul’s comments is his apparent misunderstanding of the limits of survey data. When he suggests that one can understand how religious beliefs originate by looking at surveys of professed belief and attendance at institutionalized religious services in adulthood, he is sadly in error. Such data can no more inform us about the origins of theism or atheism than can surveys of adult literacy rates help us understand the developmental processes involved in learning to read. When Paul goes on to write that “Alcock presents no statistical data relevant to the popularity of theism versus atheism either longitudinally over time, or lateral over nations, to confirm or deny his psychology-based hypothesis,” this again demonstrates his failure to understand that such statistical data are unable in principle to confirm or deny such a hypothesis. (I should also add that it is not my “psychology-based hypothesis” that he is attacking but, as noted earlier, a widely shared conclusion based on a large body of careful psychological research carried out by a number of researchers.)

It is important to remember that self-reports about religious beliefs, records of attendance at religious gatherings, and so forth, are subject to a number of limitations in terms of the inferences they generate. 

  • First, such data are not direct measures of people’s beliefs but are only measures of behavior (verbal and otherwise), and there is abundant evidence that behavior is often inconsistent with beliefs. For example, were one to conduct a belief survey in North Korea with regard to how wonderful a leader Kim Jong-un is, or to observe attendance at rallies organized to honor him, it would be most unwise to interpret a near-100 percent endorsement as evidence that all those people truly believe what they appear to profess.

In this respect, Paul is inconsistent. On the one hand, he expresses complete confidence in survey data when they support his argument that atheism is increasing, but on the other, he challenges a similar kind of data concerning creationist beliefs, stating that “the oft cited Gallup result that a very large minority of Americans think humans date back no more than ten millennia does not mean that all said citizens think the planet is that young, because it is very probable that many who answer yes to people being recent are fine with the earth being many millions of years old.” To be clear, I’m not challenging the survey data, only their overinterpretation.

  • Second, one needs to consider with some care the questions each survey asks, for some will fail to gauge the full range of supernatural/theistic beliefs. For example, someone who does not believe in the god or gods that are worshipped in institutionalized religions is likely to respond in the affirmative when asked if he or she is an atheist but might simultaneously hold the belief that there is some sort of amorphous intelligence that has had a hand in setting up the universe. 
  • Yet, most important, Paul clearly misunderstands fundamental principles of “scientific testing” when he states that “Scientifically testing the problem is not at all hard to do. For theism of some form or another to be truly universal to human brains like language, materialism, and sex, it must be consistently the near-total opinion of all societies over all of time, with atheism being rare and hard to get to.” (Please remember that this is a strawman he is attacking, as I never suggested that theism is “truly universal to human brains.”)  His argument that such information represents some sort of scientific testing is without merit, and the data he presents cannot by their very nature tell us anything about the origins of religious belief or atheism. He again confuses theistic belief as expressed in adults with the predilection toward such supernatural belief in childhood. Consider this example: It is probably true that it is the “near-total opinion” among all contemporary peoples that Earth is a globe and that it goes around the Sun. This is of course based on what people have been taught rather than direct experience or logic. However, no one would mistake such data for “scientific testing” of a claim that the heliocentric model of the solar system is “universal to human brains like language … .”   

He compounds his misunderstanding about the limits of his data when he states that “The data are definitive. There is not a powerful ‘god engine’ inside human minds juvenile or adult that is driving the masses to be pious.” Not only are the data not “definitive,” they tell us nothing about how or why people come to believe or disbelieve as they do.

Paul informs us that he has written his piece out of frustration with the skeptical community’s ignorance of rates of atheism in various parts of the world. Fair enough. We must not throw out the baby with the bathwater, and it is essential to recognize the importance in its own right of the survey information he presents. I fully support his effort to bring this information forward, and while it is true that in my article, as Paul writes, “There is no mention by Alcock of the tremendous disparity in popular non/religiosity in developed democracies … ,” that omission is due to space considerations. I do indeed recognize the importance of such data and discuss them at length in the book chapter from which my article is condensed.

In closing, I encourage the reader to go beyond Paul’s diatribe against “The God Engine” (for in reality there is no incompatibility between the data that he cites and what I have written) and instead to pay close attention to the data he presents, which reflect significant shifts in proportions of religious faith and atheism. Yet remember that while these data most likely reflect changes in the larger society in terms of a number of factors, including socioeconomic conditions, general anxiety, traditionalism versus modernism, individualism versus collectivism, the importance given to logic, and so forth, they do not allow for conclusions about causality.

James E. Alcock

James E. Alcock, PhD, is professor of psychology at York University, Toronto, Canada. He is a fellow of the Canadian Psychological Association and a member of the Executive Council of the Committee for Skeptical Inquiry and the Editorial Board of the Skeptical Inquirer. Alcock has written extensively about parapsychology and anomalous experience and has for many decades taught a psychology course focusing on these topics. His most recent book is Belief: What It Means to Believe and Why Our Convictions Are So Compelling (Prometheus Books, 2018).


I am puzzled by the tone of Gregory Paul’s criticism of my article “The God Engine.” While respectful and constructive criticism is always welcome, his scornful and dismissive condemnation is both unwarranted and unhelpful. Enough said. I am also puzzled by his misunderstanding of my article. He has set up a strawman based both on …

This article is available to subscribers only.
Subscribe now or log in to read this article.