In this project we use a NetLogo
2D-CA to study the dissemination of culture. The model is based on Robert
Axelrod's paper: Disseminating Culture, which can be found in [Axelrod].
Assume each patch represents an ethnic region. Assume the
state of a patch, called its culture, is a list consisting of N cultural
features. For example:
position 0 = religion
position 1 = technology
position 2 = political organization
position 3 = economic system
position 4 = language
etc.
Assume the value of a cultural feature, called a trait, is
an integer, t such that:
0 <= t < M
For example, the trait at position 0 might indicate the type
of religion:
0 = Animism
1 = Hinduism
2 = Buddhism
3 = Christianity
4 = Judaism
5 = Islam
etc.
How many cultures are there in our model?
Let's assume the absolute value of the difference between
two traits corresponds to their cultural distance. For example, if stone age technology is 0 then information age technology
might be 8 indicating that the difference is very large. (Of course not all
traits can be ordered in a linear way.) How can the color of a patch reflect
its state in such a way that similar cultures have similar colors and dissimilar
cultures have noticably different colors?
Initially the state of each patch is random.
To update the model:
1. For each patch, p1, pick
a random neighbor, p2.
2. Compute s = the percentage of
features that p1 and p2 have in common.
3. Pick a random number n < 100. If
n < s, then p1 borrows a trait from p2
Ideally, p1 borrows a trait from p2 other than one they
already have in common.
Hint: In the RGB color space there are 256^3 = 2^24 colors.
Since 24 = 6 * 4, this suggests we can take N = 4 and M = 6. Of course there
are only 140 colors in the NetLogo color space. This
might suggest choosing N = 5 and M = 3.