An interesting aspect of vowel systems is that they seem to balance between articulatory ease and auditory contrast. This tension is often proposed as the cause of the remarkable overlap between the organization of vowels in various languages. This thesis aims to integrate self-organizational, agent-based models of vowel dispersion with an existing Optimality Theoretic model of non-teleological phoneme dispersion. To this end, a computer simulation combining both approaches was developed. Simulation results show that dispersed vowel systems still emerge in the artificial language of the agents, although its predictions with respect to vowel quality are not entirely accurate. The ability of the model to account for the diachronic process of chain shifts caused by vowel splits or mergers is also explored. The results confirm that innate constraints are not needed to model vowel dispersion and that these types of simulations may be helpful in investigating synchronic and diachronic phonological phenomena. However, the model described in this thesis needs to be enriched with more levels of representation to increase its explanatory power.