Resumen de Understanding change through stability: a computational study of sound change actuation
Many approaches to sound change attempt to derive common patterns of sound change from universal pressures, such as physiological and psychoacoustic constraints on speech. Accounts of this type face the following problem: it is not clear why universal pressures only lead to changes in some languages, but not in others. This issue is part of the so-called actuation problem. The question of sound change actuation is usually addressed by referring to social factors and individual differences that may inhibit or encourage the spread of a sound change in a community. While this paper acknowledges the importance of such explanations, it argues that some aspects of sound change actuation can also be approached by looking at structural factors that are typically associated with the initiation of sound change. I use computational simulations to investigate the evolution of sound systems under multiple pressures. The simulated sound systems evolve towards stable states in adaptive landscapes defined partly by universal pressures (e.g. phonetic biases and contrast maintenance) and partly by language-specific factors (e.g. the relative frequency of specific phonetic environments). The former create common pathways of change, while the latter lead to cross-linguistic variation. As it will be shown, this approach can account both for stability and change. The simulations also demonstrate how language-specific factors can be used to make predictions about the stable states towards which sound systems converge.
