earticle

영어

This paper deals with velar softening occurring between base-final velar stops and non-low front vowels which initiate the process triggering suffixes in English. We adopt a mixed type of theory with derivation and constraints in order to explain examples which require both simple and multiple mapping between input and output. It is argued that velar softening is motivated by a markedness constraint, which bans a sequence of velar stops plus non-low front vowels across a morpheme boundary. The simple velar softening case, which results in [s] and [ʤ], is explained by ranking a specific marked constraint over a general markedness constraint. The procedural approach applies to opaque and diphthongization examples. The opaque case is explained by assuming that there is a word-final mid-front vowel /ɛ/, which triggers the process and is deleted. This type of non-surface apparent opacity needs a step-by-step account. The latter case is related to the verb-forming suffix initiating with /ɪ/, which is diphthongized in the output when it is stressed. This is explained by proposing a constraint calling for well-formed foot structure. Thus, the current approach can explain both simple and complex cases of English velar softening.