The Intelligence Paradox: Why the Intelligent Choice Isn't Always the Smart One - Satoshi Kanazawa 2012

Where Does Religion Come From?
Why Atheists Are More Intelligent than the Religious

While religion is a cultural universal—humans in all known societies practice some religion¹—recent evolutionary psychological theories suggest that religiosity—belief in higher powers—may not be an adaptation in itself.² Religiosity may instead be a by-product of other evolved psychological mechanisms, variously known as “animistic bias”³ or “the agency-detector mechanism.”⁴ Now, what in God's name does that mean?

Imagine you are our ancestor living on the African savanna 100,000 years ago, and you encounter some ambiguous situation. For example, you heard some rustling noises nearby at night. Or you were walking in the forest, and a large fruit falling from a tree branch hit you on the head and hurt you. Now what's going on?

Given that the situation is inherently ambiguous, you can either attribute the phenomenon to impersonal, inanimate, and unintentional forces (for example, wind blowing gently to make the rustling noises among the bushes and leaves, or a mature fruit falling by the force of gravity and hitting you on the head purely by coincidence) or attribute it to personal, animate, and intentional forces (for example, a predator hiding in the dark and getting ready to attack you, or an enemy hiding in the tree branches and throwing fruits at your head to hurt you). The question is, which is it?

Figure 6.1 Error management theory applied to religiosity

As you can see in the 2 × 2 table in Figure 6.1, there are four possible outcomes. In the two diagonal cases, you have made the correct inference. You inferred that the cause of the ambiguous situation was personal, animate, intentional, and it was; or you inferred that the cause of it was impersonal, inanimate, unintentional, and it was. There are no negative consequences if you made the correct inference.

Given the insufficient information you have, however, you cannot always make the correct inference. Sometimes you make mistakes in your judgment. In the two off-diagonal cases, you have made incorrect inferences. If you inferred that the cause was personal, animate, and intentional, whereas its true cause was impersonal, inanimate, and unintentional, you have made what the statisticians call the “Type I” error of false positive. You thought the danger was there, when it wasn’t. In contrast, if you inferred that the cause was impersonal, inanimate, and unintentional, whereas its true cause was personal, animate, and intentional, you have made what the statisticians call the “Type II” error of false negative. You didn't think there was danger, when there was.

All errors in inference have negative consequences, but these two types of errors—Type I error of false positive and Type II error of false negative—have very different negative consequences. The consequence of Type I error is that you become paranoid. You are always looking around and behind your back for predators and enemies that don't exist. The consequence of Type II error is that you are dead, being killed by a predator or an enemy when you least expect them. Obviously, it's better to be paranoid than dead, so evolution should have designed an inference system that overinfers personal, animate, and intentional forces even when none exist. Evolution should build an inference system that minimizes the chances of making Type II errors.

Here's the catch. An inference system cannot simultaneously decrease the chances of making Type I error and Type II error. Any inference system that decreases the probability of making Type I error must necessarily increase the probability of making Type II error. Conversely, any inference system that decreases the probability of making Type II error must necessarily increase the probability of making Type I error. So if the human mind has been selected to minimize the probability of making Type II errors, so that they could never be caught off guard and attacked by predators and enemies that they assumed didn't exist, then the human mind must necessarily make a large number of Type I errors.

You cannot simultaneously be paranoid and oblivious (or relaxed). The more paranoid you are, then, necessarily, the less oblivious (or relaxed) you are. The more oblivious (or relaxed) you are, then, necessarily, the less paranoid you are. In the face of a potentially dangerous yet ambiguous situation, the human mind is designed to be more paranoid and less oblivious.

Think of a smoke detector, which is designed, not by evolution by natural and sexual selection, but by engineers.⁵ Just like the human mind's inference system, smoke detectors can make errors of inference. It can sound the alarm, “thinking” there is fire, when there isn't (Type I error of false positive), or it can remain silent, “thinking” there is no fire, when there is (Type II error of false negative). The consequence of Type I error is that you are woken up in the middle of the night by the fire alarm, when there is no fire. The consequence of Type II error is that you sleep through the fire and are potentially burned to death.

Figure 6.2 The smoke detector principle

As annoying as it is to be repeatedly woken up by false alarms at three o’clock in the morning, the annoyance, even repeated annoyance, is nothing compared to what could happen if the smoke alarm makes one Type II error of not sounding the alarm when there is fire. So the smoke detector designers and engineers deliberately design smoke detectors to make lots of Type I errors of sounding alarm when there is no fire, in order to make sure that it would never ever make a single fatal Type II error of remaining silent when there is fire. Smoke detectors are therefore designed to be extremely sensitive for any potential smoke or fire. Just like the human mind, smoke detectors are designed to be “paranoid.” This is known as the “smoke detector principle.”⁶

Recent evolutionary psychological theories therefore suggest that the human inference system may have been designed by evolution to operate like a smoke detector. It may be designed to make as few Type II errors as possible, and, as a necessary and unavoidable consequence, to make many Type I errors. These theories suggest that the evolutionary origins of religious beliefs in supernatural forces may have come from such an innate cognitive bias to commit Type I errors rather than Type II errors, and thus to overinfer personal, animate, intentional forces behind otherwise perfectly natural phenomena. This tendency underlies what some researchers call the “animistic bias” or the “agency-detector mechanism.” These tendencies happen because evolution employs the same “smoke detector principle” that engineers use.

You see a bush on fire. It could have been caused by an impersonal, inanimate, and unintentional force (lightning striking the bush and setting it on fire). Or it could have been caused by a personal, animate, and intentional force (God trying to communicate with you). The “animistic bias” or “agency-detector mechanism” predisposes you to opt for the latter explanation rather than the former. It predisposes you to see the hands of God (an animistic and intentional agent) at work behind natural, physical phenomena whose exact causes are unknown.

In this view, religiosity—the human capacity for belief in supernatural beings—is not an evolved tendency per se; after all, religion itself is not adaptive. It is instead a by-product of the animistic bias or the agency-detector mechanism, the tendency to be paranoid, which is adaptive because it can save your life. Humans did not evolve to be religious; they evolved to be paranoid. And humans are religious because they are paranoid.

Some readers may recognize this argument as a variant of “Pascal's wager.” The 17th-century French philosopher Blaise Pascal (1623—1662) argued that given that one cannot know for sure if God exists, it is nonetheless rational to believe in God. If one does not believe in God when He indeed exists (Type II error of false negative), one must spend eternity in hell and damnation. In contrast, if one believes in God when he actually does not exist (Type I error of false positive), one only wastes a minimal amount of time and effort spent on religious services. The cost of committing Type II error is much greater than the cost of committing Type I error. Hence one should rationally believe in God.