The Interaction of Spirantization and s-Aspiration in Andalusian Spanish Joaquín Romero Universitat Rovira i Virgili, Tarragona, Spain joaquin.romero@urv.es The Western Andalusian dialect of Spanish presents some interesting phenomena which have to do with the interaction of two common rules of the language: spirantization and s- aspiration. Spirantization is a general phenomenon which is present in all dialects of the language by which voiced stops deocclusivize in a variety of contexts; s-aspiration, on the other hand, is more dialectally restricted and consists, in its simplest form, in the loss of the apical gesture of syllable coda /s/. Traditional phonological/phonetic accounts refer to the outcome of spirantization as continuant segments (fricatives or approximants), but more recent work has suggested that the term 'deocclusivized stops' (Honorof 2003, Romero & Honorof 2004) might be more appropriate to describe the phonetic nature of these resulting segments. Along the same lines, Romero (1995) provided evidence that spirantized /b, d, g/ could be interpreted as regular stops with perhaps an altered constriction degree specification. One important aspect in this line of reasoning has to do with the temporal dimension. Spirantized stops do not show temporal characteristics that would justify classifying them as continuants. In this sense, Soler & Romero (1999) showed that the manipulation of the temporal dimension of spirantized stops influenced the outcome, with slower rates disfavoring deocclusivization. The aspiration of syllable coda /s/ has been described as 'debuccalization' or loss of the supralaryngeal articulation of the consonant, resulting in the percept of /h/. While this might be an appropriate characterization in some contexts, as in word-final position followed by a vowel, it isn't by any means universal. In Andalusian Spanish, in particular, a whole variety of outcomes are possible depending on the context. In this study we concentrate on contexts such as /sb/, /sd/ and /sg/ in which s-aspiration and spirantization interact. According to Romero (1995), this interaction results in voiced fricatives //-/v/, // and //. These voiced fricatives are distinguished from deocclusivized /b, d, g/ exclusively along the temporal dimension, with no significant difference in constriction degree. In order to further test the validity of these findings and provide additional evidence on the non-continuant nature of spirantized voiced stops in Spanish, this study investigates the nature of /sb/, /sd/ and /sg/ combinations in Western Andalusian Spanish using two types of data. In the first experiment aerodynamic data were obtained of /b/, /d/, /g/, the combinations /sb/, /sd/, /sg/ and individual /s/ from 2 speakers of Western Andalusian, who read a series of sentences containing these segments at two speech rates: normal and slow. It was hypothesized that there would be significant airflow differences for speech rate but not for single consonants vs. clusters. In the second experiment, a perceptual test was devised that looked at the effect of manipulating the temporal dimension of /sb/, /sd/ and /sg/, the hypothesis being that a sufficient reduction of the duration of these clusters would favor the percept of single deocclusivized stops. Preliminary results provide evidence that indeed the temporal dimension plays a crucial role in distinguishing single consonants from clusters, both at the level of the aerodynamics and in the perception. This is taken as evidence that spirantized stops are not continuants and that the term 'deocclusivized' describes their phonetic nature much more accurately. In addition, the results confirm the fricative nature of /sb/, /sd/, /sg/, thus providing further evidence that both spirantization and s-aspiration are highly complex phenomena, the outcome of which can only be explained accurately by looking at the articulatory effects of contextual variation.