To Stay or Not to Stay? Artificial Sociality in GRASP World

Abstract

This paper describes an agent-based model that investigates group longevity in a population in a foundational way, using theory on social relations and culture. The model is the first application of the GRASP meta-model for social agents, containing elements of Groups, Rituals, Affiliation, Status and Power. It can be considered an exercise in artificial sociality: a culture-general, content-free baseline trust model from which to engage in more specific studies. Depending on cultural settings for individualism and power distance, as well as settings for xenophobia and for the increase of trust over group life, the GRASP world model generates a variety of patterns. Number of groups ranges from one to many, composition from random to segregated and pattern genesis from rapid to many hundreds of time steps. Parallels are discussed between patterns found in GRASP world and patterns found in societies that differ on individualism, power distance and heterogeneity.