The phonological representation of edge tones in Spanish alternative questions

University of Cambridge 1. Introduction A good deal of research on intonation following the tradition set by the Autosegmental-Metrical Theory (AM Theory) proposes T- as one of the structural elements in the phonological representation of utterances from Spanish (Nibert 2000, Hualde 2002). However, there are also numerous accounts in the literature which, even within the AM Theory, exclude T- from the representation (Sosa 1999, Beckman et al. 2002). In this context, we embark on the adventure of studying the adequacy of such tone in the specific case of alternative questions in two dialectal varieties of Spanish, namely, Madrid and Canarian Spanish, since their respective proposals show a conflict in the following terms: Estebas Vilaplana (2007) includes H- as the phonological unit which accounts for the peak at the end of the first disjunct of an alternative question in Madrid Spanish, whereas Cabrera Abreu and Vizcaino Ortega (2007) resort to H% to account for such a peak in Las Palmas de Gran Canaria (LPGC) Spanish. In order to solve this conflict, we turn to Cruttenden (1986), who claims that there is a correlation between the duration of nuclear and post-nuclear syllables on the one side, and the end of an intonation group on the other side, in the following terms: the syllables are lengthened before a prosodic boundary. Bearing this in mind, in our research we measure the duration of those particular syllables so that we can determine whether H- or H% stands in the phonological representation. 2. Setting the context In this section, first, we present briefly a set of proposals which initially seem to be assigning different pragmatic functions to H- and H%, and therefore, justify their presence in phonological representation. However, a closer study of their pragmatic functions reveals that, in fact, they are quite similar, and consequently, the initial motivation for their proposal weakens