# Estimated costs and benefits of participation in an extreme ritual in Mauritius

## TL;DR

The paper explores why people in Mauritius participate in painful rituals, finding that cultural beliefs about supernatural benefits help explain their persistence.

## Contribution

The study introduces a modified costly signaling theory that incorporates differential benefit estimation and supernatural beliefs.

## Key findings

- Ingroup participants estimated ritual costs as higher than outgroups but physical costs as lower.
- Ingroup participants reported larger supernatural benefits compared to outgroups.
- Differential benefit estimation, not cost estimation, is key to the persistence of extreme rituals.

## Abstract

Humans often participate in physically harmful and demanding rituals with no apparent material benefits. Although such behaviours have traditionally been explained using the lens of costly signalling theory, we question whether the canonical theory can be applied to the case of human cooperative signals and introduce a modification of this theory based on differential benefit estimation. We propose that along with cooperative benefits, committed members also believe in supernaturally induced benefits, which motivate participation in extreme rituals and stabilize their effects on cooperative assortment. Using Thaipusam Kavadi as a prototypical costly ritual, Tamil (ingroup) and Christian (outgroup) participants in Mauritius (N = 369) assessed the cost and benefits of Kavadi participation or hiking. We found that ingroup participants estimated material costs as larger than outgroups, physical costs as lower, and benefits as larger. These findings suggest that estimated costs may vary by modality and cultural expectations (e.g. Kavadi participants are not supposed to display pain), while supernaturally induced benefits were consistently reported as larger by ingroups compared to outgroups. We conclude that differential estimation of ritual benefits, not costs, are key to the persistence of extreme rituals and their function in the assortment of committed members, underscoring the role of differential estimation in the cognitive computation of signal utility.