Xygalatas 2013

From Projecting Power

Main Argument: The authors believe that there is not enough direct evidence to show the relation between ritual intensity and prosociality. In an effort to fix this problem, the author conducts a study to study this correlation.


This study took place in Mauritius, because it is multicultural and the citizens frequently negotiate between parochial ethnic-religious and inclusive national identities.

They examined two rituals: one, a low-ordeal ritual involving singing and collective prayer (Thaipusam), the other a high-ordeal ritual involving body piercing with multiple needles and skewers, carrying bamboo structures, dragging carts attached by hooks to the skin for over 4 hours before climbing a mountain barefoot (Kavadi).

The authors predicted that higher ritual intensity would lead to greater generosity for both performers and observers.

The authors also predicted an effect of ritual intensity on social identification


They observed 86 males

19 high-ordeal performers of Kavadi 32 high-ordeal observers of Kavadi 35 low-ordeal performers who took part in Thaipusam 3 days earlier

All participants took part in both rituals, although each was tested in only one, selected randomly to avoid self-selection biases. Both rituals had taken place at the same venue, and had similar durations and numbers of attendants.

Following the ritual, attendees entered a room near the temple and answered a questionnaire on a computer prior to engaging in a donation task.

Prosociality was measured on two axes: their behavior (charity), and attitudes (social identification).


The authors found that low-ordeal participants donated an average of 80.91 Mauritian rupees, high-ordeal observers donated 161.25 rupees, and high-orderal performers donated 132.63 rupees. This result showcases the increased generosity of high-orderal performers and observers relative to the low-ordeal performers. Furthermore, a bootstrapping analysis revealed that pain correlated with donations even after controlling for age, religiosity, and temple attendance. Upon splitting the high-ordeal group into four quartiles based on particular participants’ pain ratings, the authors found that the overall difference between quartiles was significant. IThe data shows that higher levels of perceived pain were directly associated with greater donations. There were no significant differences between experienced pain (performers) and perceived pain (obeservers).