Typologizing phonetic precursors to sound change

Why are some linguistic patterns more common than others? This question of the relative frequency of pattern attestation has long puzzled typologists of all stripes. In the case of phonological typology, the skewing in frequency of sound pattern attestation may be attributed to the relative robustness of phonetic precursors (i.e. the likelihood of phonetic precursors being phonologized into sound patterns). This article focuses on this concept of relative phonetic precursor robustness. What is it and how is it measured? I argue that one way the robustness of a phonetic precursor can be assessed is in terms of its likelihood to create enough confusion for misperception-based sound change. A model for quantifying relative precursor robustness is proposed using a rational model of speech perception. To illustrate the utility of the present approach, the strengths of the phonetic precursors of velar palatalization, HH- and HV-interactions are assessed, thus demonstrating the viability of constructing a typology of phonetic precursor robustness.