Abstract

REGISTER HERE

ABSTRACT

In this talk I present early results from SPRINT (ERC-ADG-835263) that examine the accentual contrast between H✴︎ and L+H✴︎. This contrast has long been disputed in the literature on English intonation. Some argue that the two pitch accents are distinct, with H✴︎ (realized as high pitch) typically used to highlight new information, while L+H✴︎ (realized as a pitch rise) is used to highlight contrastive information. Others, however, argue that the noticeable rise of the L+H✴︎ is an epiphenomenon of added emphasis that results in increased pitch span, and thus that the distinction is not contrastive or related to information structure. Resolving this controversy is important for the description of English intonation and has repercussions both for understanding the relation between categorial and gradient aspects of F0 use, and the role of intonation in encoding information structure. I will present results from scripted and unscripted speech elicited from eight SSBE speakers and show that they support an analysis based on a phonological contrast between H✴︎ and L+H✴︎. In addition, I will show how using Functional Principal Component Analysis (fPCA) to understand variation in F0 while also investigating the role of additional parameters (such as duration and energy) can help us draw informed conclusions about phonological categories of intonation and their phonetic realization. More generally, these findings help shed light on the relation between the phonetics and phonology of intonation and suggest a path towards disentangling linguistic from paralinguistic aspects of F0 use.