F0 and plosive voicing in Afrikaans

Abstract

Afrikaans plosives are traditionally described as contrasting in voicing ([b d] vs. [p t]). Coetzee et al. (2014, J. Acoust. Soc. Am. 135, 2421), however, showed that the voicing contrast is collapsing word-initially, with voiced plosives merging with voiceless plosives. They also found that voicing loss does not result in loss of lexical contrast, which is preserved on the following vowel (high f0 after historically voiceless and low f0 after historically voiced plosives). That study investigated word-initial plosives, leaving unanswered whether word-medial plosives are also devoicing. The current study addresses this question. Acoustic analysis of data collected from nine Afrikaans speakers replicated the results of Coetzee et al. for word-initial plosives. For word-medial plosives, a robust voicing contrast was found along with a post-plosive f0 difference. The f0 difference in medial position was comparable to that in initial position in both magnitude and duration, and extended throughout the vowel. While lexical contrasts are hence cued by only f0 word initially, they are differentiated by both plosive voicing and the f0 of the following vowel word medially. The relevance of these results for theories of sound change and theories of phonological contrast will be discussed. [Work supported by NSF.]

Metrical conditioning of word-final devoicing in Polish

“…Word-final obstruent devoicing has been taken up in both phonological and phonetic/acoustic literature with respect to many languages, e.g. German (Brockhaus 1995), Russian (Dmitrieva, Jongman and Sereno 2010), Lithuanian (Campos-Astorkiza 2006), Dutch (Grijzenhout 2000), Afrikaans (Coetzee et al 2016), Turkish (Kopkalli 1993), Persian (Tofigh and Abolhasanizadeh 2015) and Setswana (Boyer and Zsiga 2013), among others. A detailed overview ofobstruent voicing options has been provided in Ashby and Maidment (2005).…”