4325 - Morphological characteristics and the geographical distribution of Arawakan features

Comparing all Arawakan languages on the basis of general grammatical features, only a few clear clusters can be distinguished (cf. Danielsen et al. in press). In contrast to this finding, Aikhenvald (1999) claimed for a general split between northern and southern Arawakan languages. Even though I claim that this split cannot be confirmed in general, it can be observed with respect to one particular group of features. There is a tendency towards less morphologically complex verbs in the northern part of the language family (cf. Aikhenvald 2001) . One reason for this is the fact that semantic roles of (oblique) arguments are generally marked by adpositions in the North (dependent-marking), where southern Arawakan languages apply a wide range of different verbal applicative affixes (head-marking), cf. examples (1) through (4):

(1) suvi inà Maipure (Northern Arawakan)
‘with a stick’ (Zamponi 2003: 35)

(2) li-sika-bo to polata da- myn Lokono (Northern Arawakan)
3SGm-give-CONT ART money 1SG- BEN
‘He gave me the money.’ (Pet 1987: 65)

(3) yu- ma nu-íjluami Yavitero (Northern Arawakan)
3SGf- COM 1SG-mother
‘with my mother’ (Mosonyi & Mosonyi 2000: 618)

(4) ni=wo’ik- ino -wo=pi=ro Baure (Southern Arawakan)
1SG=butcher- BEN -COP=2SG=3SGm
‘I am butchering it for you.’ (Danielsen 2007: 252)

In this presentation the existence of adpositions in the Arawakan language family is calculated for a sample of 29 Arawakan languages (phylogenetic analysis). It shows the typical split between dependent-marking and head-marking strategies, partly reflected by the geographical distribution ((1) through (3) versus (4)).

The outcomes of these features are compared to the distribution of the personal cross-reference markers and the patterns the paradigms show. The person markers of Arawakan languages are usually the first features that are taken in order to determine the genetic relation. The question is here if the distribution of the personal forms is similar to other clusters, and in particular that of split between dependent-marking versus head-marking (the so-called “valency split, Danielsen 2011).

Keywords: Arawakan, morphological features, phylogenetics

Author: Danielsen, swintha (University of Leipzig, Germany / Deutschland)


University of Vienna | Dr.-Karl-Lueger-Ring 1 | 1010 Vienna | T +43 1 4277 17575