Romancing with tone: Prosodic systems in language contact

Abstract

There is a widespread tendency to see tone as a crosslinguistically marked feature that either gets lost or is “reduced” in language contact (e.g. McWhorter 2001). This view is Eurocentric. It suggests that stress, which happens to prevail in Europe, is a norm towards which contact varieties (pidgins, creoles, transplanted languages, etc.) converge. It is unempirical because the evidence suggests otherwise. We propose an alternative view based on two hypotheses: (1) Contact varieties develop lexical tone or stress, in accordance with the dominant type in their respective ecologies. What needs to be explained are the exceptions, e.g. why do some contact languages feature stress in spite of predominantly tonal substrates/adstrates? (2) The tonal systems that emerge in contact varieties will reflect the possibilities and limitations of stress-to-tone mapping. Tonal contrasts might therefore appear to be more predictable than those of other languages in the ecology. The analysis of a corpus of primary field data in the African Romance varieties of Central African French (CAF) and Equatorial Guinean Spanish (EGS) suggests that they are lexical tone languages (e.g. Bordal Steien 2012). CAF and EGS differ from their “intonation-only” (i.e. “stress”; cf. Gussenhoven 2004) siblings European Standard Spanish (ESS) and French (ESF) in significant ways: Every syllable in CAF and EGS bears a low or high tone and we find tonal minimal pairs; a lexical tone may not be altered for pragmatic purposes; only utterance-final boundary tones fulfil the functions associated with intonational melodies in ESS and ESF. The opposition between stressed and unstressed is converted into a two-tone distinction and the largely phonotactic conditioning of stress assignment in individual words translates into equally “predictable” H tone placement in the contact variety. We propose a dynamic model for the development of prosodic systems during contact in accordance with the nature of the ecology: Tone systems emerge through similar mechanisms as in CAF and EGS in tonal ecologies. Intonation-only systems are, in contrast, found in ecologies where tonal languages played no, or a minor part (e.g. Cape Verdean Creole). In Afro-European and Euro-Asian creoles, the intensity and duration of contact with the non-tonal superstrate on the one hand, and with tonal substrates and adstrates on the other indicates whether a tonal prosodic system emerges or survives. Continuous contact with African adstrates has ensured the retention of such systems in African contact languages (e.g. the West African English-lexifier Creoles, CAF, EGS, Ghanaian and Nigerian Standard English, as well as all Niger-Congo- lexicon creoles, e.g. Criper-Friedman 1990; Yakpo 2009; Gussenhoven & Udofot 2010). Lack of contact with African tonal adstrates and simultaneous intense contact with a non-tonal superstrate has led to (the abandonment of tone systems and) emergence of intonation only systems (e.g. Jamaican, Haitian). Languages without contact with African tonal adstrates that have for a long time been isolated from non- tonal superstrates can however retain tonal systems under specific circumstances (e.g. Saramaccan and Ndyuka).