Asia-Pacific Language Variation - Volume 5, Issue 2, 2019

The borrowing of words from one language into another is most likely as ancient as language itself. While ample linguistic attention has focused on various linguistic contact scenarios in which words from one language enter productive use into another, their aim has been largely restricted to documenting the words which are borrowed, their frequency, and other situation-specific information. In this paper, we propose new methods for studying loanwords, namely a combination of statistical testing techniques which can be used together to increase knowledge in this area. We illustrate these tools with a case-study of loanwords from an indigenous language (Māori) into a world dominant language (New Zealand English). Using a topic-constrained newspaper corpus in conjunction with quantitative methods, we explore the use of loanwords diachronically and analyse variation in loanword use across newspapers and across writers.

Referential kinship terms in Matukar Panau (Oceanic, Papua New Guinea) are obligatorily possessed. Traditionally, kinship terms are directly possessed in Oceanic languages (with an obligatory suffix on the root that agrees with the person and number of the possessor). In Matukar Panau, some kinship terms are also indirectly possessed (with a classifier that agrees with the person and number of the possessor). A third pattern shows double-marking of possessors with directly possessed terms co-occurring with a classifier. I present a multivariate analysis of the predictors that influence the choice of the direct, indirect or double-marked patterns. Older women and younger men are most likely to use the indirect pattern, particularly when discussing their own kin from their households, especially in conversational situations. The indirect possession pattern, then, is used for more integral relationships, what has previously been the semantic domain of direct possession in Oceanic.

With 20,000 speakers across Northern Australia, Australian Kriol is well known to exhibit geographic variation but this has never been systematically studied. This article stems from the first dialectological study of Kriol, focusing on the eastern portion of the Kriol-speaking area. It analyses variation in forms of the Kriol reflexive, which is derived from the English form ‘myself/meself’ but is invariant for person and number. The analysis utilises random forests modelling to analyse the importance of factors, a new method available to variation studies that is particularly useful when applied to small languages with small datasets. With geography confirmed as the major factor accounting for variation, areal patterns showing variation in lexical form of the reflexive, the medial consonant, the final vowel and the final consonant are considered. This study also documents new variants of the Kriol reflexive and incorporates perceptual dialectology, combining to better inform classifications of Kriol dialects.

The number interpretation of Noun Phrase structures (NPs) with different animacy values and variable overt marking of number is investigated in a corpus from the Oceanic language Vera’a. It is found that number marking is optional for all nouns and that number marking is essentially semantic, expressing collectivity and distributivity. While statistical patterns of greater likelihood of number marking comply with the animacy hierarchy, the central motivating factor for number marking is found to be likelihood of individuatedness. This study thus contributes to a functionalist account of cross-linguistically observed patterns of hierarchical number marking.