Possession in Vamale (Southern Oceanic, New Caledonia)

Rohleder, Jean Karl; Zúñiga, Fernando (2018). Possession in Vamale (Southern Oceanic, New Caledonia) (Unpublished). In: 51st Annual Meeting of the Societas Linguistica Europaea. Tallinn, Estonia. 31.08.2018.

Many non-Polynesian Oceanic languages are known to have two morphological strategies to express attributive possession, namely direct and indirect possession: the host of the affixes indicating person and number of the possessor is the nominal with the former and a possessive/relational classifier with the latter (Ross 2004, Lichtenberk 2009). Direct possession is typically found with a distinct set of possessums: kinship terms, body parts, and things done to or used on the possessor. Indirect possession is typically used with items related to food, items related to potables, and general items; Lynch (2000) reconstructs *ka, *m(w)a-, and *(n)a- as the classifiers for these three groups, respectively.
Based on original fieldwork, the present paper outlines how possession marking works in Vamale, which shows a picture that is both similar and different from the above, in several intriguing respects. First, there are two sets of possessive suffixes:

(1) Set I: xhetham ‘plate’ - xhetham-an ‘his plate’
Set II: vwen ‘turtle’ - vwen-ea ‘his turtle’

Second, there is an inalienable-alienable opposition that follows the pan-Oceanic distribution rather closely. Inalienable possessums include most kinship terms (but not forms of address), body parts (except blood), and characterizing items (e.g., spirit, appearance, strength). Alienable possessums include animals and plants, as well as general items. Nevertheless, while alienable items can take suffixes of either set (those of Set II typically for contrastive focus), inalienable items invariably take Set I suffixes. Third, possession can be expressed via juxtaposition if the possessor is lexical rather than pronominal; in such cases, vowel-final possessum stems take a marker -n (< POc *-ɲa ‘3SG.PSR’; Ozanne-Rivierre 1982: 43), e.g. mwa-n daahma [house-N chief] ‘the chief’s house’. Fourth, when taking Set II suffixes of any person, a semantically bleached version of this marker -n appears under (apparently non-semantic) conditions not fully understood yet.
Lastly, and most notably, some alienable kinship terms and inalienable possessums, as well as nominalizations expressing ways of acting, additionally take an obligatory marker ka when possessed:

(2) a. bifidu-ka-n [twin-KA-3SG.PSR] ‘his twin’
b. vwasee-ka-n [sadness-KA-3SG.PSR] ‘his/somebody’s sadness’
c. hun-vwa-ka-n [NMLZ-do-KA-3SG.PSR] ‘his/somebody’s way to do (sthg.)’

Unlike Rivierre & Ehrhart’s (2006) account of closely related Bwatoo, the analysis espoused here casts some doubt on seeing ka either (i) as the host of an element formally detached from the possessum nominal or (ii) as a relational classifier (e.g., as the present-day reflex of the POc food classifier *ka). The present paper argues on semantic and morphological grounds in favor of considering the possibility of (synchronically) regarding -ka as a suffix that allows selected items to become possessed and, therefore, in favor of regarding the morphology-based direct-indirect distinction as inadequate for Vamale. We claim that the latter analysis is not only explanatorily and descriptively more adequate but that it also does more justice to the quite simple and well-behaved morphophonemic regularities found in the language, to which a purported classifier ka- would be an exception.

Item Type:

Conference or Workshop Item (Speech)


06 Faculty of Humanities > Department of Linguistics and Literary Studies > Institute of Linguistics

UniBE Contributor:

Rohleder, Jean Karl and Zúñiga, Fernando


400 Language > 410 Linguistics
400 Language > 490 Other languages




Fernando Zúñiga

Date Deposited:

10 Sep 2018 13:51

Last Modified:

10 Sep 2018 13:51



Actions (login required)

Edit item Edit item
Provide Feedback