Cultural Dissemination

The following CA model is based on Robert Axelrod's paper: Disseminating Culture.

Assume each patch represents a cultural region, French-speaking Switzerland, for example.

Assume each patch has an attribute named culture, which is a list of three numbers in the interval [0, 256) representing the region's religion, language, and economy, respectively.

For example [100 50 222] might mean [Hindu Spanish pastoral].

We can easily translate culture into color as follows:

set pcolor culture

Initially, cultures are random.

The model is a modification of Life. Here's the initial screen shot, a random patchwork of cultures:

3500 ticks later and we see great islands of similarity:

Note that we also see isolated white patches surrounded by blue patches. This happens because the blue and white patches are too culturally distant to borrow from each other. They would rather simply put up walls.

During the observation phase, each patch remembers the culture of a random nearby patch, where nearby means any patch within radius range, where range is controlled by a slider. In real life range is related to communication and transportation technology. Cultures can't borrow from each other if they don't know about each other.

During the update phase, each patch adopts a random trait from this other culture, if the cultural distance is below some user-controlled threshold. This threshold might correspond to the tolerance of cultures to each other.

If we think of cultures as points in 3D space, then cultural distance is simply the spatial distance given by the distance formula from analytic geometry.