In many European cities, adolescents growing up in linguistically and culturally diverse neighborhoods develop ways of speaking that differ considerably from the traditional vernaculars. According to Clyne (2000: 87), such ‘multiethnolects’ are spoken by “several minority groups” and sometimes also by “members of the (dominant) ethnic group”. In this study, we examine a phonological constraint of traditional Zurich German which only allows lenis (i.e., short) fricatives in word-initial position (Fleischer & Schmid 2006). This constraint does not seem to apply to multiethnolectal Zurich German (Schmid 2012). The current contribution aims at verifying if this sociophonetic feature persists across speaking styles.
In a previous study (Morand et al. 2019), we correlated acoustic measures extracted from read speech (in total, 931 tokens) with results of a sociophonetic perception experiment, in which speech samples of 48 adolescents were rated on how multiethnolectal they sound; the speakers were rated by 40 other adolescents on a 7-point Likert scale, ranging from ‘not at all’ to ‘completely’ multiethnolectal (rating score). There was a strong positive correlation between mean fricative duration of the speakers and their mean rating scores (r = 0.54, p < .001). A subsequent duration normalization (fricative duration - mean consonant duration) yielded very similar results (r = .51, p < .001). Adolescents who were perceived as speaking rather multiethnolectal Zurich German (high mean rating score) produced longer word-initial fricatives than adolescents who were perceived as speaking rather traditional Zurich German (low mean rating score).
To investigate whether fricative duration is a multiethnolectal feature that is used in different speaking styles, we now compared these results (fricative duration in read speech and rating score) with data obtained from spontaneous speech (in total, 483 tokens). The same 48 adolescents who produced the read speech also played a game of spot-the-difference in pairs using so-called Diapix (Baker & Hazan 2011).
Now, we find a strong positive correlation between fricative duration in read and spontaneous speech (r = 0.67, p < .001). More importantly, there is a significant positive correlation between mean fricative duration in spontaneous speech and the mean rating scores (r = 0.33, p = .02). However, these findings only hold partially for the analyses with normalized fricative durations. While we still observe a positive although weaker correlation between speaking styles (r = 0.3, p = .04), normalized fricative durations in spontaneous speech no longer correlate significantly with the mean rating scores (r = 0.25, p = .08).
Our findings show that speakers with shorter mean fricative durations in read speech also tend to have shorter mean fricative durations in spontaneous speech. Also in spontaneous speech, adolescents who were perceived as speaking rather multiethnolectal Zurich German produced longer word-initial fricatives than adolescents who were perceived as speaking rather traditional Zurich German, but this relation is not as strong as in read speech. All in all, the results suggest that fricative duration is a multiethnolectal feature that is used as a sociophonetic marker (Labov 1972) regardless of speaking style.