Translations of this page:


Ethnocentrism execution


This is an implementation of the ethnocentrism model proposed by Hammond and Axelrod. It shows how ethnocentric behavior can overcome other strategies under different scenarios. To accomplish this, agents are placed in a spatial structure and interact with each others, gaining or losing points in this interactions. Then, the agents with the most points reproduce.
Studies show that in-group bias can be triggered by arbitrary group definitions. Also, evidences report that this bias is controlled by the hormone oxytocin, suggesting biological roots. Others suggest that this behavior appears as a consequence of social and cultural concepts.
This model then tries to find minimal conditions for group discriminating conduct emerge.


Each agent in the model has two characteristics: color, as a summary of all traces a person has, and trait, as the strategy that that agent uses when dealing with other agents. The color field can have any value, and the strategy field takes four different values:
Altruist: cooperates with everyone
Ethnocentric: cooperates with same color
Xenocentric: cooperates with different color
Egoist: doesn't cooperate

In each time step, the following sequence takes place:
1. New agents immigrate and settle in unoccupied places
2. Agents interact with neighbors in one-move prisoner's dilemma to give and receive potential to reproduce. Agents play in this game according to their strategy and each neighbor's color
3. Each agent reproduce according to their potential to reproduce
4. Some agents die

the author of the article mapped the results of the one-step prisoner's dilemma in two variables, cost and benefit, making the implementation easier to understand. The game table goes like:

——–j2 D j2 N
j1 D (2,2) (0,3)
j1 N. (3,0) (0,0)


In order to run the model it is necessary to set the following parameters: (in parenthesis the default values)

FIRSTGEN: initial number of agents (1)
COST: cost of cooperating (0.01)
BENEFIT: benefit of receiving cooperation (0.03)
IMMIGRATIONRATE: quantity of agents immigrating per step (1)
MUTATIONRATE: probability of mutation when reproducing (0,005)
DEATHRATE: death rate per step (0.1)
BASEPTR: initial potential to reproduce (0.12)
SIZE: space dimension (50)
TIME: total time steps (2000)


It is important to notice that in all test cases the dominant strategy is the ethnocentric, but the color varies. If the cost of giving help is greater then the benefit, the egoist strategy dominates.
Other interesting thing is that


Try varying the parameters, both one at a time and multiple at the same time;
Vary the FIRSTGEN parameter and see that nothing changes when the system is in balance;
Vary the quantity of colors in the model, the more colors the bigger the ethnocentric strategy dominance;
Insert noise using the MUTATIONRATE variable. Ethnocentrism still dominates;


To make the model more detailed, there could be more traits other than just the color, as adding them would make the simulation more realistic. Other ideas are: vary the neighborhood structure (standard is Von Neumann's); implement different probabilities of trait in new agents entering the space; implement reproduction as two agents mix of traits.


A lot of different TerraME features were used. Some of the most important are: Agent, CellularSpace, Placement, Trajetory, Group, Timer, Observers. One workaround was to make a function that returns the amount of agents with some attribute. Also, the values of these variables had to be calculated in each time step.
c = Cell {
altruists = function() return ALTNUMBER end,
ethnocentrics = function() return ETHNUMBER end,
xenocentrics = function() return XENNUMBER end,
egoists = function() return EGONUMBER end

Repast: Ethnocentrism
TerraME: Schelling


This model is inspired by the article "The Evolution of Ethnocentrism" by Ross A. Hammond and Robert Axelrod, published in the Journal of Conflict Resolution in 2006. \\The complete reference is shown below.
Hammond, R. A., & Axelrod, R. (2006). The Evolution of Ethnocentrism. Journal of Conflict Resolution, 50, 926-936.

In Other Languages
Translations of this page: