Gemination: Weight or Length? Evidence from Rural Jordanian Arabic

The aim of the paper is to evaluate how the implementation of geminates in Rural Jordanian Arabic (RJA) differs from that of clusters and by doing so, provided evidence for a prosodic weight. Geminates are represented and viewed controversially either as prosodic length (skeletal tier) or as prosodic weight (Moraic Theory) in the current theories of phonology. In this study, I argue that the geminate-singleton contrast word-medially and word-finally is better accounted for in terms of prosodic weight, within Moraic Theory, based on moraic weight, than to be represented in terms of prosodic length as timing units. This observation is based on two pieces of evidence from RJA. The first is that consonant clusters word-finally in RJA are not permissible and are usually broken up by an epenthetic vowel, while geminate consonants are permissible. Second, the attraction of stress onto the syllable that has a geminate consonant, whether word-medially or word-finally, is consistent with the weight representation; primary stress typically falls on the rightmost bimoraic syllable. Therefore, it will be more appropriate to account for this phenomenon under Moraic Theory and to view long segments as weight-bearing, and not as prosodic length/timing units.