ABSTRACT
The lenition surveys (see Chapter 1 and Appendix) support a generalization concerning spirantization: unaffricated stops never lenite to strident fricatives, such as [s] or [f]. Rather, stops typically spirantize to weakly fricated or approximant continuants such as [β, ð, ɣ] or [ https://www.w3.org/1998/Math/MathML"> β T https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315023731/b9622830-b831-4769-958e-8135ab388aed/content/b1_B.tif" xmlns:xlink="https://www.w3.org/1999/xlink"/> , ð, ɰ]. Although assibilatory spirantization processes (e.g. t → s,ʃ) are attested, these are restricted to contexts in which the stop is inherently somewhat affricated. In this chapter, I document these generalizations, and demonstrate that they fall out from the effort-based approach, coupled with certain plausible assumptions concerning the effort cost of strident relative to nonstrident continuants. Moreover, this generalization is reconciled with the apparently conflicting observation, from studies of segment inventories (e.g. Maddieson 1984, Ladefoged & Maddieson 1996) that these nonstrident continuant consonants, which are favored as outputs of lenition, are highly disfavored (i.e. “marked”) relative to strident fricatives in the general case.
