Computing Religious Devotion: How Reinforcing Supernatural Beliefs Affects Normative Models in the Mind (CREDO)
Religions permeate the lives of billions and are hypothesised to play an essential role in normative behaviour. Yet, little is known about how religious devotion penetrates cognitive computations during decision-making. The CREDO project fills this lacuna by proposing a computational model of religious decision-making. In this model, religious belief, forged through religious practice, forms strong priors in the mind. When the mind simulates possible actions during decision-making, religious actions become readily available and likely selected due to their high value. To empirically develop the model, the CREDO project will determine how religious belief affects the strength of religious priors during normative decision-making in various laboratory and field studies. Moreover, in a large-scale cross-cultural study, the project will establish how beliefs and practices of different religious traditions affect cognitive computations during normative decisions. The goal of the CREDO project is to provide one of a kind, comprehensive model of how religions shape the mind.
CREDO is a five-year project (2024-2028) funded by the Czech Science Foundation.
The cross-cultural project:
Religious commitment is among the most reliable predictors of within-group cooperation, yet it is also implicated in some of the world’s most persistent forms of intergroup violence. These two patterns often characterize the same communities, the same institutions, and sometimes the same individuals. How can the same religious traditions that sustain trust, generosity, and sacrifice within groups also intensify exclusion, hostility, and violence across group boundaries?
We propose that this pattern reflects a fundamental mechanistic gap in current research. Existing accounts and interventions have focused primarily on explicit attitudes, narratives, and deliberative judgments, yielding important insights into what people profess but far less into the computations that generate differential treatment in real time. Yet the persistence of religious intergroup bias may depend on how social evidence is initialized, weighted, and updated during decision-making and learning. Explaining these computations is therefore essential for understanding why the same religious institutions can support both extraordinary cooperation and enduring hostility, and for opening a more precise pathway to intervention.
The same cognitive mechanisms that sustain trust, coordination, and sacrifice within groups can also amplify boundary maintenance with toward outsiders by labeling both cooperation and hostility as sacred duties. Computationally, this asymmetry can be understood as increased precision of culturally learned priors, which are resistant to change. We formalize these cognitive processes using a combination of drift-diffusion modeling (DDM; Ratcliff & McKoon, 2008) and reinforcement learning (RL; Sutton & Barto, 2018), the two dominant computational frameworks for modeling decision-making and its temporal updating.
In the planned study, we will estimate pre-learning decision bias, a learning parameter from learning trials, and post-learning decision bias in a structurally identical task with novel partners. Such data will allow us to quantify how much outgroup evidence shifts the starting state of the decision process versus how much it is discounted during learning, and to map those effects onto individual- and community-level variation in predictors of religious orthodoxy across countries and traditions.
The Team
WE ARE HIRING! See the call below
https://www.phil.muni.cz/en/careers/available-positions/80767
The core team members are two post-doctoral researchers, two PhD students, and a project manager. The team is carefully assembled to represent the various skills needed for the successful implementation of the objectives of the CREDO project: one post-doc is an experienced modeller focusing on the computational models of the human mind, while the other post-doc will be an expert on human social cognition and the various ways we can examine its workings. These post-docs are paired with two Ph.D. students. The PhD students are hired for the full course of their PhD with their dissertational research dedicated to the CREDO project.
Moreover, we have additional PhD students, and post-docs advising on the project, including external collaborators. For the cross-cultural component of the CREDO project, we will search for a team of 20 collaborators from different countries helping us to collect data from a wide sample of participants.
