An Agent-based Model of the Spread of Social Norms

From Santa Fe Institute Events Wiki

CSSS 2006 Santa Fe

Last modified: 6/16/06


In this project, I am hoping to illuminate how social norms emerge and spread in a population using an agent-based model(s) in Netlogo. Researchers in social psychology (Cialdini, 1990) have made a distinction between descriptive and injunctive norms. Descriptive norms are norms that people infer from patterns in other people's behavior, particularly in ambiguous and uncertain situations (this is related to the saying "When in Rome, do as the Romans do"). In my current thinking, you could implement this kind of norm by just having turtles use some simple update rule for setting their normative behavior (such as majority rule) that is dependent on what their immediate neighbors are doing.

Injunctive norms, on the other hand, are rules or beliefs that specify morally approved and disapproved conduct and are usually enforced by the threat of social sanctions ("When in Rome, you should do as the Romans do or else"). I think this idea can be realized in one or more agent-based models in two ways. First, there would be more than one turtle type:

Some turtles are just "followers" who imitate the predominant behavior of their neighbors (following the descriptive norm) or who do what they are told to do by top-down messages from authorities (following the injunctive norm; more on these "injunctive shocks" below). They might have weak preferences that make them somewhat reluctant to adapt their behavior to their surroundings and maybe push them to invest in moving to a more favorable patch environment.

Others are "zealots" or "rebels" who aren't easily swayed by the pressure of their neighbors and perhaps actively influence the behaviors of others either for or against the norm (spreading or undermining the injunctive norm on a local level). They might move around to spread the word about their beliefs to others or even enforce the norm at a cost to themselves and stubborn followers.

A second way to approach injunctive norms would be through applying periodic top-down "injunctive shocks" to the agent world, which would correspond to things like laws or messages from authority figures urging norm compliance. These shocks might convince a subset of followers to change their behaviors or even transform them into zealots. Others might change into zealots or rebels if they maintain their normative behavior in a favorable environment for a certain number of ticks.

Obviously, it's probably unrealistic to be able to implement all of these behaviors properly by the end of the month but I hope to develop one or two simple models that would yield interesting insights into the characteristics of agents, norms, and their relationships that allow for the rapid spread of a social norm through a population.

Group member:

Izzat Jarudi. Comments and suggestions are welcome. Email: izzat.jarudi (at) yale (dot) edu