Lexical Stress Strength vs Macro-Rhythm Strength: An Inverse Relationship Between Prominence Cues

In Jun’s [1] model of prosodic typology, phrase-medial tonal rhythm, or macro-rhythm (MacR), is one of the parameters used to classify languages. A language has strong MacR if its intonation regularly alternates between high and low tones over a word-sized prosodic unit in a phrase [1:522]. The model predicts that languages with an Accentual Phrase (AP) marked by a rising or falling tone will have stronger MacR than languages with various types of pitch accents. Jun observed that languages with weak MacR tend to have strong acoustic correlates of stress while languages with strong MacR tend to have weak acoustic correlates of stress. Therefore, she predicted an inverse correlation between the strength of lexical stress and the strength of MacR. The current study tests this prediction by comparing the lexical stress and MacR correlates of English, Uyghur, and Bengali. The results of a production experiment found that English had the strongest stress correlates, followed by Uyghur, followed by Bengali, and the results of a MacR perception experiment found some support for the opposite ranking (i.e., Bengali > Uyghur > English). These preliminary results suggest a potential trade-off between prominence cues and support for the predicted inverse relationship.