This study investigates to what extent Arabic speakers, whose L1 has predictable lexical stress, are “stress deaf” and whether their L1 prosodic properties influence their L2 stress perception. Arabic L2 English speakers and L1 English speakers performed an identification task with English nonce… read more
The production of schwa in French is highly variable. Some of this variability is predictable based on sociolinguistic (e.g., dialects), phonological (e.g., segments, word position), and stylistic (e.g., reading) factors. We investigate the effect of prosodic structure on the production of… read more
From the perspective of word prosody, fixed stress languages such as Hungarian may seem rather uninteresting: stress, by definition, always falls on the same position in a word. This paper examines the acoustic properties of Hungarian stress based on a large, systematically collected, corpus and… read more
Linguistic rhythmic or timing categories, usually defined in terms of isochrony, remain controversial as a meaningful typology for classifying languages, despite decades of research. Romance languages offer an opportunity to address this question since closely related languages are proposed to be… read more
The acoustic properties of stress and focus prominence are examined in a large, systematically structured corpus of Arabic collected in Amman, Jordan. A modified version of the Functional Load Hypothesis correctly predicts that duration, a contrastive property of Arabic vowels, will not constitute… read more
The acoustic properties associated with prominence (e.g. duration, F0) may also serve for “phonemic” contrasts. The question is thus how speakers correctly interpret these properties. We address this question in terms of an extension of the Functional Load Hypothesis (FLH): given that vowel length… read more
This chapter examines a number of phonological properties of compounds. These include the limitation of many phonological phenomena to the individual members of compounds, the application of others at the juncture of the members of compounds, and the distribution of prosodic properties over an… read more
When speakers produce a prepared utterance, the amount of time required to initiate the utterance reflects the number of units in the utterance. In this paper, we investigate the nature of this unit on the basis of Italian experimental data, and compare our findings to previous studies of Dutch. We… read more
In Romanian, word final /-i/ is usually not pronounced as a vowel, but rather as a type of glide, or palatal property associated with the preceding segment. While this palatalization or “desyllabification” has been assumed to apply at the right edge of a word, we demonstrate that the word boundary… read more