ABSTRACT
The case studies of American English and Mexican Spanish test a range of predictions about consonant strength derived from phonological theories of lenition and from the data base of strength alternations. Using a carefully controlled word list to find effects of stress and position in word, I examine the realization of a full range of consonants in these two languages with different phonologized patterns of weakening. I examine a range of acoustic and articulatory parameters for each consonant token. The evidence collected is used to test theoretical views of lenition and outline a new theory of lenition in which a broad range of phonetic properties play a role.
