Journal article

Ultrametric distribution of culture vectors in an extended Axelrod model of cultural dissemination

Alex Stivala, Garry Robins, Yoshihisa Kashima, Michael Kirley



The Axelrod model of cultural diffusion is an apparently simple model that is capable of complex behaviour. A recent work used a real-world dataset of opinions as initial conditions, demonstrating the effects of the ultrametric distribution of empirical opinion vectors in promoting cultural diversity in the model. Here we quantify the degree of ultrametricity of the initial culture vectors and investigate the effect of varying degrees of ultrametricity on the absorbing state of both a simple and extended model. Unlike the simple model, ultrametricity alone is not sufficient to sustain long-term diversity in the extended Axelrod model; rather, the initial conditions must also have sufficientl..

View full abstract


Awarded by Victorian Life Sciences Computation Initiative

Awarded by ARC

Funding Acknowledgements

We are grateful to Dr Jens Pfau for making his source code from Pfau et al. [18] available. Dr Gary Au assisted with obtaining Eurobarometer survey data. Prof. Philippa Pattison provided invaluable advice on opinion vector distributions, and Dr H. Paul Keeler suggested the use of the Cholesky decomposition. We thank Mr Francesco Picciolo and Prof. Diego Garlaschelli for clarifying details of split ballot handling used in Valori et al. [19]. This research was supported by a Victorian Life Sciences Computation Initiative (VLSCI) grant number VR0261 on its Peak Computing Facility at the University of Melbourne, an initiative of the Victorian Government, Australia. We also made use of the University of Melbourne ITS Research Services high performance computing facility and support services. This work was supported by ARC grant DP130100845 and The University of Melbourne NHMRC/ARC Research Grant Support Scheme.