| Popis: |
Work on speech perception shows a bi-directional, but not necessarily reciprocal, relationship between linguistic features and speakers. On the one hand, social information has been shown to influence the interpretation of linguistics features (e.g., Rubin 1992, Babel & Russell 2015, D’Onofrio 2018). On the other hand, variants of linguistic features – phonetic, morpho-syntaxic or prosodic – can also be associated with different social information. For instance, the variable ING has been extensively examined in many varieties of English (e.g., Labov 2001, Campbell-Kibler 2007, 2009, Levon & Fox 2014) and the velar variant [iŋ] has been shown to be widely associated with ‘education’ or ‘intelligence’. At the same time, the alveolar variant [in] is frequently linked to ‘casualness’ in Mainstream American English. This being said, a variant can also lead to multiple social associations, depending on contexts and/or listeners (e.g., [in] can also index laziness, relaxedness, being layed-back, non-exhaustive list). The concept of ‘indexical fields’ (Eckert 2008) specifically refers to flexible potential social meanings that linguistic variables can convey. It allows to account for the variation encountered both in production and perception of linguistic variants: meanings associated with linguistic features and particular variants are not fixed in time or in interpretation and can be multiple. Of specific importance for this project is the recent work on linguistic coherence, co-variation, and speech perception, where several trends have been observed. One possible scenario is that the combinations of linguistic features can affect social meaning of one specific feature involved (Levon 2014, Pharao & Maegaard 2017). Sibilance of /s/ has been shown to cue gayness and femininity, but these associations can be blocked when fronted /s/ co-occurs with features that stereotypically index opposite categories such as masculinity. Another possible scenario is that the features of interest, despite belonging to different registers and having distinct social meaning in isolation, become harmonious when co-occurring (Grondelaers & van Hout 2016). One could also imagine the possibility that combination of features results in the synchronic interpretation of all the features. This experiment tests the effect of co-occurrence of Southern features on French listeners, and particularly the potential different interpretations of word-internal schwas depending on its co-occurrence with different productions of nasal vowel /ɑ̃/. (1) Can co-occurrence of internal schwa with another southern phonetic feature influence interpretation of internal schwa? Overall theoretical grounding behind this: testing “perceptual harmony” (Grondelaers & van Hout 2016) vs compositionality of social meaning (i.e., social meaning of x + social meaning y = overall social meaning x and y) vs combinativity (i.e., altogether social meaning of x and y create meaning z). Because of the type of stimuli chosen, it is impossible to track the full list of traditional indexicalities measured in this type of work (status and solidarity measurements)? Adapted predictions are provided below. (2) Can *absence* of feature x influence perception of feature y? |
