Phonology requires shared actions in a community of speakers.

Imagine two agents, who are trying to accomodate their actions to one another, through imitation.

They produce actions whose constriction goal value is randomly chosen from some constriction continuum, for example the Constriction Degree (CD). Each time they happen to produce an action that matches the CD of their partner's action (within some tolerance), they increase the probability of producing that constriction.

Over time, they attune to one another. This is what happens:

For example, CD is patititoned into three modes: stop, fricative, approximant.