When conversations turn to religious diversity, it does not take too long before someone says something like “How can they believe such things?” where “they” refers to participants in some religion other than that of the speaker. Devoutly religious people regularly express astonishment that other devoutly religious people could possibly believe some outlandish or downright silly thing.

This exhibits what the philosopher Neil Van Leeuwen dubs the “puzzle of religious rationality” in his brilliant new book, Religion as Make-Believe: A Theory of Belief, Imagination, and Group Identity. That puzzle concerns how people who seem otherwise impressively rational on virtually all other fronts subscribe to religious beliefs that seem “anything but rational.” Crucially, each group is often completely blind to the outlandishness or downright silliness of some things that they themselves claim to believe.

A New Theory of Cognitive Attitudes

Van Leeuwen’s answer to this puzzle is that the English term “belief” is used in two very different ways to label what he argues are dissimilar cognitive attitudes. He distinguishes what he calls "groupish" beliefs (important examples include what he calls “religious credences”) from "mundane" beliefs (the most familiar examples of which are factual beliefs).

His principal rationale for this distinction is that groupish and mundane beliefs have different explanatory roles. A groupish belief constitutes a cognitive attitude that contributes to explaining why people belong to groups associated with that cognitive attitude’s contents. John’s groupish belief that Jesus is the Son of God explains why John belongs to a Christian church. A mundane belief constitutes a cognitive attitude that contributes to explaining why people are disposed to act in ways whose success turns on the contents of that cognitive attitude being true. John’s mundane belief that the pencil is on the table, when the pencil is, in fact, on the table, will enable John to find the pencil when he needs it.

Other Contrasting Features

Van Leeuwen contrasts the cognitive processing that these different cognitive attitudes involve. On the front end of that processing, groupish beliefs are voluntary and unresponsive to evidence, whereas mundane beliefs are the opposite. Regarding voluntariness, many religions insist that people must choose to accept God, but people do not choose to believe that the pencil is on the table. Regarding responsiveness to evidence, as so many religious people insist, nothing can shake their faith, but, by contrast, if they see that the pencil has rolled off the table, they will no longer hold the belief that it is on the table.

On the back end of that processing, both sorts of belief inspire behaviors, including speech, that communicate. With mundane beliefs, the communication is a means of signaling instrumentally useful information for achieving some goal (finding the pencil), whereas the aim of the behaviors associated with groupish beliefs is just the signaling in and of itself. Affirming groupish beliefs is about indicating a person’s identification with a group (publicly reciting the Apostle’s Creed signals that someone is a Christian). Groupish beliefs may sometimes seem instrumental (e.g., prayers for healing), but anthropologists, from Malinowski to Luhrmann, have stressed that such actions typically supplement straightforward instrumental actions (e.g., taking doctor-prescribed medications) and, thus, only appear to be instrumental.

Van Leeuwen also notes that mundane beliefs, but not groupish beliefs, exert “cognitive governance” over what we infer in our imaginings. People with celiac disease do not forget that the transubstantiated host is made of grain. People cannot abandon their factual beliefs about the mundane properties of their symbolic props. Mundane beliefs guide action across the board, while groupish beliefs are practical-setting-dependent. Treating them otherwise is what we call fanaticism.

One Consequence of Groupish Beliefs’ Distinctiveness

Neither groupish beliefs nor mundane beliefs need to satisfy the other’s characteristic feature to fulfill their respective explanatory roles. In fact, they usually don’t. The explanatory success of groupish beliefs turns not on their truth but, instead, on their distinctiveness (for establishing group identities). The less controversial they are, the less distinctive they are—hence, the less likely they are to be effective as group identifiers. Groupish beliefs tend to fulfill their explanatory role better precisely when they deviate (e.g., Moses heard God’s voice from a burning bush that was not consumed by the fire) from widespread conceptions of reality (e.g., bushes do not emit voices and are consumed when burnt). In short, groupish beliefs are, on balance, more likely to serve their explanatory function precisely when they are false.