Costly signaling cannot explain why nonbelievers find religious displays costly

Evidence

A third difference is that in signaling, the behavior is a reliable reflection of group commitment and/or cooperation, whereas in CRED, it reliably reflects an underlying culturally transmitted belief. One issue with signaling accounts of religious displays is that it is unclear why it is more costly for nonbelievers to perform the costly acts than for believers, since beliefs are culturally transmitted, and are quite unlike possessing a genetically fixed physical attribute (such as physical stamina, height or large body size, or a colorful tail). Therefore, it seems that explaining costly religious displays requires cultural transmission of beliefs underlying these behaviors, which brings us back to a cultural evolutionary consideration of how beliefs get transmitted in the first place. — Norenzayan, Ara, Big gods - how religion transformed…, 343

Categorization

• XREF: The ‘why would a nonbeliever pay the cost’ problem echoes the fumi-e logic — the trampling is cheap for Japanese officials to perform because they lack the belief; it’s costly only for the believing priest, which is exactly why the ritual works as a test.