Dr Michael Ramsammy "Gender differences in sibilant voicing: the case of European Portuguese" | 30 March 2016
30 March 2016
Time: 9:05am
Venue: Francis Bancroft (FB) 3.40
"Gender differences in sibilant voicing: the case of European Portuguese"
In this talk, I present the results of ongoing experimental work conducted in collaboration with Patrycja Strycharczuk (CASL, Queen Margaret University) which examines the sibilant system of European Portuguese. Traditional descriptions state that the Central Standard dialect of European Portuguese contrasts four sibilants, namely /s, ʃ, z, ʒ/ (e.g. caça [kasɐ] ‘hunt’, caixa [kɐjʃɐ] ‘box’, casa [kazɐ] ‘house’, queijo [kɐjʒu] ‘cheese’). However, this contrast is restricted to syllable-initial environments: in syllable-final positions, the four-way contrast is suspended by neutralisation of place-of-articulation and voicing. The operation of neutralisation in this dialect is dependent on phonological contextual factors (i.e. prevocalic vs preconsonantal vs prepausal environments).
Whereas traditional descriptions assume that these generalisations should hold for all speakers of the Central Standard dialect of European Portuguese, our data reveal fine-grained patterns of variation that have not previously been reported. Analysis of acoustic experimental data shows that voicing is generally much more variable that has traditionally been assumed; and whilst all speakers maintain a four-way contrast between sibilants in prevocalic contexts, the phonetic realisation of this contrast differs significantly between men and women. Specifically, male speakers show a greater tendency to use full modal voicing in intervocalic [z] and [ʒ]-realisations, whereas female speakers display much less robust voicing. However, in other phonological environments –such as preconsonantal contexts (e.g. asno [aʒnu] ‘donkey’, castanho [kɐʃtaɲu] ‘chestnut’)– phonetic differences in the realisation of sibilants are not predictable on the basis of speaker gender.
Our current results therefore invite consideration of what causes gender differences in intervocalic sibilant voicing in European Portuguese. For example, crosslinguistic research has demonstrated very clearly that fine-grained differences in the realisation of sibilant consonants can carry important social meanings: see Pharao et al. (2014) for Danish; Crist (1997), Levon (2007), Munson (2007), Stuart-Smith (2007), Mack & Munson (2012), Levon & Holmes-Elliott (2013) for English; Rácz & Schépacz (2012) for Hungarian; Mack (2010, 2011) for Spanish. Thus, a relevant question for our data is to what extent gender differences in sibilant voicing may be linked to sociophonetic expression of gender. Alternatively, given that experimental phonetic work has also shown that voicing strategies differ in females and males –i.e. in terms of aerodynamic and articulatory control (Koeing 2000; Jessen 2009; Nadeu & Hualde 2014)– we also consider the possibility that the observed differences in sibilant voicing may originate in physiological differences between female and male speakers.