What can toys teach us about actuation?

This chapter proposes that aspects of the actuation problem can be fruitfully elu- cidated if the problem is studied in the limited context of simple toy models of variation and change. One such model, the "AB toy", is examined in detail. This examination suggests that the actuation of linguistic change is likely to depend on an interaction of deterministic and stochastic factors. It is pointed out that (sim- ple) mathematical models make available strong probabilistic predictions, and that the solution to the actuation problem, if one exists, is likely to lie in the general system-theoretic properties of speech communities, rather than in the establish- ment of sufficient conditions for individual instances of change to occur.