In Defense of Juncture Rules/Constraints

Abstract

Domain juncture rules which apply between two prosodic domains have been problematic in that they exhibit a gap in their distribution. There have been attempts to derive the apparent word-juncture effects of Sanskrit External Sandhi from the syllable-coda effects. The first part of the paper examines the arguments for the coda analysis, and concludes that, in view of cross-linguistic sandhi phenomena, the simplicity is bought at the expense of explanatory power. In addition, Italian Raddoppiamento Sintattico, another showcase example, is reanalyzed in terms of Optimality Theory, where there is no need for the junctural information necessary for consonantal gemination at the word-edge. The paper concludes that there are genuine cases of phonological juncture rules (the /n/-O alternation in Korean) which cannot be reanalyzed as domain span or domain limit rules like the Sanskrit and Italian phenomena.