Constraint rank border crossing

Roland Noske

It has often been observed that important systematic prosodic differences exist between Netherlandic Dutch and Belgian Dutch. However, there are no detailed descriptions of these differences, which is in stark contrast with the many descriptions of segmental differences between the two variants of the language. This may be due to the fact that the prosodic differences are generally felt as being difficult to describe (Willemyns 2003:330). Sometimes, however, prosodic differences have segmental reflexes, which then allows for their description. In addition, up till now, relatively little attention has been paid to differences in phonological processes between the North and the South.

This paper is based on two observations by the author, viz. (i) that in Belgian Dutch, there is no glottal stop is not insertion if a consonant final morpheme is combined with a vowel initial one (while this is the case in Northern Dutch and syllable initial glottal stop insertion before vowels does exist in Belgian Dutch) and (ii) that in Belgian Dutch, vowel deletion in pronouns takes place much more easily than in Netherlandic Dutch.

The first phenomenon will be analysed as a result of a difference in syllabification, and hence of a difference in prosodic organisation. This difference in its turn can be analysed as a difference in ranking between two specific constrains. Then, it will be shown that exactly this same difference in constraint ranking is responsible for the observed contrast in vowel deletion in pronouns between the North and the South.

Once this has been established, the question is raised about the origins of the difference in constraint ranking. A number of arguments will be advanced that show that the specific constraint ranking in Belgian Dutch has been adopted from Romance. This will confirm that borrowing from one language to another is not limited to words and sounds, but that it can also extend to deeply rooted settings in phonological organisation.