# Multi-agent modeling of pop/youth subcultures dynamic

Modelling the Dynamics of Youth Subcultures by Petter Holme and Andreas Grönlund (2005), Journal of Artificial Societies and Social Simulation vol. 8, no. 3:

What are the dynamics behind youth subcultures such as punk, hippie, or hip-hop cultures? How does the global dynamics of these subcultures relate to the individual's search for a personal identity? We propose a simple dynamical model to address these questions and find that only a few assumptions of the individual's behaviour are necessary to regenerate known features of youth culture. (...) Our model is based on five major assumptions:

- The dynamics of the underlying social network is negligible
- An adolescent belongs to one subculture at a time.
- If the fraction of friends that have adopted a certain subculture is big enough then an adolescent will adopt that subculture too
- The attractiveness of a subculture decreases with its age.
- There is a certain resistance to changings subcultures.

Yes this is subculture modelisation ;) The following graph depicts the propagation of subcultures:

Illustration of the model on a small example graph. The colour and number in the circles indicates the current subculture of the vertex. The bar next to the circle shows the vertex' score-function values for the three subcultures (0, 1, and 2) the current time t. The starting time of a subculture t(i) is, in this particular example, identical to i. (b) shows the time-step after (a) in with the threshold value T = 0.5. In this case no other subculture is introduced, no vertex will change subculture as time tends to infinity. (c) is the corresponding plot with T = 0.3. In this case the system will, as time progress, reach a state where every vertex adopts subculture 2.

**Why do I blog this?** Even though the underyling assumptions could be discussed, I find fancy this sort of stuff, people then have a model for how subcultures spread in a population of adolescents. It would be interesting to compare this dynamic model to propagation of mainstream cultures, to see if there are different phenomenon (I expect that the assumptions might be different of course). Besides, would it be possible to expand the model with other parameters (more related to the 'quality' of the subculture or social aspects like the possibility that at a certain period of time there could be a repulsion towards a specific subculture)?