The non-linear analysis of Spanish phonology as proposed by Harris (1984) and others promises to augment the explanatory power of currently available phonological descriptions, and may offer significant insights into the description of phonological differences among dialects. However, the interaction between previously held notions of phonological processes and the new theoretical analysis has not yet been fully explored, and the claim that the latter apparatus must necessarily completely supplant the former is perhaps over ambitious. The present note will deal with competing analyses of two frequent processes in Spanish, velarization of underlying /n/ and aspiration of underlying /s/, both of which have been used in support of the latest theoretical proposals. I will attempt to demonstrate that the data used to formulate these proposals have been excessively idealized. The full range of complexity and variation which surrounds these and similar phenomena necessitates a more ramified analysis, and even an idealization which does not distort the fundamental nature of the two processes casts doubt on the simple analysis suggested as a replacement for orthodox generative and traditional structuralist analyses.
Labialization and palatalization in Judeo-Spanish phonology