Variation and preservation of Mapudungun dental fricatives

Despite being common amongst the languages of Europe, dental fricatives are relatively rare phonemes crosslinguistically. Voiced forms show up in only 5% of the phonemic inventories in the Phoible database (Moran & McCloy 2019), and voiceless ones in just 4% of them. Indeed, most languages of the world tend to have only one coronal place of articulation, most frequently alveolar. The implication is that dental/alveolar contrast is difficult to maintain, a notion that seems to be corroborated by the tendency for European dental fricatives to undergo mergers (cf. English TH-stopping and TH-fronting as well as Spanish seseo).

Enter Mapudungun, a (presumed) linguistic isolate spoken in Chile and Argentina, with a full series of dental/alveolar contrasts including /t̪~t/, /n̪∼n/, /l̪~l/ and /θ∼s/ (Sadowsky et al. 2013). Not only is this pattern unique on a global scale, but there seems to be little evidence for regional support for dental/alveolar contrasts more generally, and for dental fricatives in particular.

The core of this paper discusses the historical evidence for dental fricatives in Mapudungun, as gleaned from the Corpus of Historical Mapudungun (CHM — Molineaux in prep.) spanning the language’s 400-year documentary record. I will show that contrast between dentals and alveolars is not new, and can be traced to the earliest written records in the 17th century. I will also give evidence for the fact that, throughout the record, a definite imbalance existed between the two coronal consonant series, such that minimal pairs were rare, and the lexical incidence for the dental series among stops, nasals and laterals was far smaller than that for their alveolar counterparts. In the case of fricatives, however, I will show that the opposite pattern emerges: with alveolar /s/ (found in 67% of Phoible inventories) being rare (predominantly in Spanish and Quechua borrowings), while dental /θ/ is frequent and widespread throughout the lexicon.

Data will also be presented to show that, over the course of the 20th century, the peripheral dialects of Mapudungun have begun to loose the dental/alveolar contrast. In the southernmost dialect, Huilliche, dental fricatives have mostly merged with the alveolar, as have the other dental segments (Sadowsky et al. 2015). However, in the north and east (Picunche and Pe- huenche), the dental/alveolar contrast is maintained among fricatives, despite merger — on the alveolar — for stops, nasals and laterals (Salamanca & Quintrileo 2009). The most vital, central varieties retain the contrast throughout, even if /θ∼s/ alternation is increasingly frequent.

These patterns of variation and loss are related, no doubt, to contact with Spanish, as well as loss of vitality of the key dialects. However, I will claim that the preservation of dental fric- atives in the northern and eastern dialects is likely to be an artefact of a longstanding dialectal difference in Mapudungun: the voicing of fricatives. Indeed, despite the fact that Mapudungun does not contrast consonantal voicing — a broader areal feature for the Southern Cone —, the fricatives of central and southern dialects are generally voiceless, while in the north and east they are voiced. The main exception to this clear dialectal pattern is /s/ which surfaces on both sides of the isogloss. The voiceless realisation of /s/ in the fricative-voicing varieties is suppor- ted by Spanish, creating an additional dimension of contrast with /ð/, and hence keeping the two sounds distinct in those dialects.

Ultimately, then, the Mapudungun data points to the fact that dental/alveolar contrast can be maintained long-term in contexts of linguistic vitality, even with no significant areal support and substantial imbalance in frequency. In cases of loss of vitality, nonetheless, the contrast is lost fairly quickly, unless additional features can be relied upon for its maintenance.