Abstract: ​To what extent can the results of variation studies in large-scale speech communities be extended to small-scale speech communities? Many sociolinguistic differences would lead us to reject the "uniformitarian principle" that allows us to extrapolate across space and time (Labov 1972), for example: rates of linguistic change, rates and types of multilingualism, and conditions of language acquisition are all different between WEIRD and small-scale societies like in southern New Guinea. In this talk, I will discuss how I planned, collected, analyzed, and interpreted a sociolinguistically annotated corpus of Ende, a Pahoturi River language of southern New Guinea, taking into account the unique characteristics of this community. Together, we can discuss the advantages and difficulties for engaging in this type of research in small-scale speech communities.

Phonological Variation and Gradient Representation
This Friday, I would like to discuss two topics with the P-interest group. For the first half of our meeting, I will discuss a phonological variable relevant to my dissertation: verb-final /n/-realization in Ende. I will discuss my methods and preliminary findings. Then, I will present a recent paper (Smolensky & Goldrick 2016) on the use of Gradient Symbolic Representations (GSR) to model French liaison consonant realization. The big take-away from this paper is that a symbol in the input can be gradient in its degree of presence in the input (i.e., partially present), such that it incurs a partial violation of Dep to be realized and a partial violation of Max to not be realized. My hope is that we can discuss the advantages and disadvantages of introducing gradience in the input of phonological grammars in order to account for idiosyncratic patterns in languages like French and Ende.