Projects
All languages change through time. But to what extent, and along what parameters, do the types of changes that languages undergo vary? My research explore this question, by testing and refining traditional models in historical linguistics with data from the understudied languages of west New Guinea. These languages contribute a novel perspective to theoretical questions about language change: the local linguistic ecologies in which changes are initiated and spread are highly diverse; and the speaker populations are often unique when compared with most other areas of the world in terms of mobility, levels of multilingualism, and language attitudes and ideologies.
This was a project I carried out during my British Academy Postdoctoral Fellowship, in which I looked at several issues in the Raja Ampat-South Halmahera (RASH) branch of South Halmahera-West New Guinea, itself an understudied—and mostly endangered—branch of Austronesian. For this project, I carried out further fieldwork on the undocumented RASH languages spoken in and near Raja Ampat.
The data I collected in this project provide a basic documentary record of the languages, and allow us to identify priorities for future major documentation projects. All data collected in the project are archived here: datashare.ed.ac.uk/handle/10283/8573.
Besides this, the project had three aims:
You can find a paper on the fundamental frequency of vowels (in collaboration with Jiayin Gao and James Kirby) here, slides from a presentation on unusual tone changes in Raja Ampat here, and slides from a presentation on the subclassification of the Raja Ampat languages here. For more on the languages I collected data from in this project, take a look at the Fieldwork page.
In this project, I look at the morphosyntax of possessive constructions in a linguistic area called Wallacea, which spans east insular Southeast Asia and west New Guinea. Many languages spoken in Wallacea have a basic two-way morphosyntactic split in their adnominal possessive constructions: one construction is typically used to express relationships of inalienable possession (such as body parts and kin terms), and the other construction is used elsewhere. Here's an example from Ambel. In this example, the 1SG possessor of the inalienably possessed 'eye' is marked with the suffix -k, which is attached directly to the possessed noun. For the alienably possessed 'canoe', the marker of the 1SG possessor is attached to a prenominal possessive particle.
| taji-k | ne |
| eye-1SG | DEM.PROX |
| 'my eye' | |
| ni-k | wan | ne |
| poss-1SG | canoe | DEM.PROX |
| 'my canoe' | ||
In some languages with this distinction, there's a further split in the way 'inalienable' constructions are expressed – this is what I'm calling 'split inalienable coding' (or 'SIC' for short). In most cases, the split is found in the morphological paradigms. So, for example, Ambel has one paradigm marking kin terms, and another marking body parts. This difference is shown in example 2: a 3SG possessor is marked with the prefix i- for kin terms like 'mother', but is not marked at all for body parts like 'eye'.
| i-nya | pa |
| 3SG-mother | DEM.MID |
| 'his/her mother' | |
| taji | pa |
| eye.3SG | DEM.MID |
| 'his/her eye' | |
Languages with SIC are rare worldwide, but comparatively common in Wallacea. There were two factors influencing the development of SIC in Wallacea. First, languages belonging to the Austronesian family are predisposed to develop SIC: kin terms in these languages are often marked in special ways, and SIC is another example of this. Second, in Northwest New Guinea, contact between languages helped facilitate the development.
You can read much more about the distribution and development of SIC in Wallacea in a paper published in STUF, which is available at doi.org/10.1515/stuf-2023-2013.
In this project, I'm collaborating with language acquisition specialist George Saad (Palacký University) and my former MA dissertation supervisee Emma Peddie to investigate an underdocumented pattern of language acquisition attested across Island Southeast Asia and the Pacific: Late Vernacular Production (LVP). People with LVP grow up bilingual in both the local vernacular (the language of their parents) and the regionally dominant lingua franca. In childhood these speakers have only passive competence in the vernacular, becoming active users upon reaching adulthood. The result is quite striking, with speakers often appearing to go from almost no active competence to full fluency almost overnight.
We're investigating this phenomenon from several perspectives: at the moment, we're particularly thinking about the implications of LVP for models of language endangerment, bilingual acquisition, and language change. George is also carrying out real time studies to investigate the morphosyntactic changes that occur in vernacular languages acquired in situations of LVP.
My PhD project was a grammatical description of Ambel, another RASH language, which was based on a major audio-visual documentation of the language. Some particularly cool features in Ambel include lexical tone, differential marking in inalienable possessive constructions, and a super-complex system of spatial deixis.
You can find my description of Ambel here, and some highlights from the documentation here. Martinus Wakaf and I compiled a trilingual (Ambel-Malay-English) dictionary for the Ambel community; you can find a pdf copy here. The full documentation of Ambel is available at the Endangered Languages Archive – you'll need an account (it's free) to access it.