Final lengthening at prosodic boundaries in Dutch

In this paper we describe an experiment that was set up to measure segmental lengthening before five types of prosodic boundaries, ranging from the Prosodic Word boundary to the Utterance boundary. INTRODUCTION It has been shown by several researchers [1], [2] that segments are longer at syntactic boundaries, and that the amount of lengthening increases with the boundary’s place in the syntactic hierarchy. However, we assume that it is prosodic structure that regulates the rhythm of language, and that final lengthening therefore occurs at prosodic boundaries. In earlier experiments we have found this to be true for boundaries below the word level [3], In the experiment described below we investigated final lengthening at boundaries ranging from the Prosodic Word boundary to the Utterance boundary. METHOD Our experiment was set up to test the influence prosodic boundaries have on the durations of the segments that precede them. We based our definitions of the relevant prosodic boundaries in Dutch on [4], We devised five carrier sentences in which target words could be placed before one of five prosodic boundaries. The lowest boundary we tested was a Prosodic Word boundary within a compound. The next boundary was a Prosodic Word boundary at the end of a morphological word, for which the target word was an adjective within an NP. In prosodic theory there is no d ifference betw een these two b o u n d a rie s , a lth o u g h m orphosyntactically there is. The next higher boundary to be tested was the Phonological Phrase (PPh) boundary, which occurred at the end of an NP in our material. The highest boundary was the Utterance boundary. To rule out any possible effect of sentence length we made sure that all carrier sentences had the same number of words before and after the target word position. Since it is not clear whether the shortening effect of the number of words following a target word can pass the Utterance boundary, or alternatively, whether the Utterancefinal lengthening effect is distinct from the lengthening before the end of a discourse, we added a small sentence after the Utterance boundary, consisting of the same two words that followed the PPh-boundary. In order to be able to answer this question we also included the Utterance boundary without this following sentence in our materials.