Deconstructing markedness in sound change typology

Abstract

Many sound changes have been attributed to misperception (Ohala 1981, 1993). When two sounds A and B are perceptually similar, A can be misperceived as B and vice versa. One sound change attributed solely to perceptual similarity is θ > f (Blevins 2004). Misperception of [θ] as [f] yields θ > f, while hearing [f] as [θ] should lead to f > θ changes. Context-free shifts of θ > f are attested, but regular f > θ changes are rare. Recent research questions the existence of f > θ changes and the perceptual basis of θ > f changes. Historical, typological, experimental, developmental, and language contact data reviewed here support the original perceptual account of θ > f and f > θ, suggesting that the observed asymmetry can be explained phonetically and structurally, without reference to markedness (cf. Andersen 2008).