hide
Free keywords:
-
Abstract:
In this paper, I investigate clause linkages in the Oceanic language Whitesands (ISO: TNP). Whitesands, like its sister languages of the southern Vanuatu sub-group, uses a switch reference system called the Echo Subject (ES) (Lynch 2001:177, Crowley 2002:201). I present results from a response time-based experiment that targets the comprehension of typical and non-typical forms of the switch reference system.
The m- ‘ES’ inflection is typically used when two adjacent predicates have identical reference for the subject. The m- replaces the person agreement and tense operators in the second clause. In (1) the m- indicates co-reference of the subject of the predicate with the subject of the preceding predicate with no other tense or person operators present. If the subjects are different in each clause, such as (2), full inflection — both person agreement and tense — is required on both predicates (even if there is no overt nominal reference).
(1) t-am-ø-uven apaha iVila kani m-ø-uven apaha itehi
3-PST-SG-go LOC Port.Vila and ES-SG-go LOC saltwater
He went to Port Vila and went to the beach.fn2_29
(2) t-am-ø-ek kapiel apiapwei kani t-us nelma-n
3-PST-SG-touch stone hot and 3SG.NPST-bite hand-3SG
He touched a hot stone and it burnt him (lit. it bit his hand).fn2_49
However, in the Whitesands corpus it is clear that a simple “antecedent equals subject” rule does not always hold for Echo Subject clauses and that a notion of discourse topic might be a potential antecedent alternative (see Reesink (1983) for a similar claim in non-Austronesian languages). For example, there are topic chains that use the Echo Subject for continual reference whilst skipping immediately adjacent non-topical subjects. Further, there are forms where the Echo Subject prefix combines previously distinct arguments into a single argument slot.
The experiment presented here was designed to test hearer comprehension of the Echo Subject system. In particular, it aims to answer two questions. Firstly, what is the relative ordering of canonicity of the different Echo Subject forms? That is, are some antecedent types more or less default OR is there evidence to posit a single syntactic or pragmatic class that is the antecedent for the Echo Subject clauses? It
seems there are differences in response times and accuracy within the different conditions suggesting a hierarchy of preferred antecedents for Echo Subject clauses.
Secondly, is the Echo Subject system more or less inferential than alternative referential systems? In particular, what is the pragmatic loading compared to when there is fully specified verb agreement. We see that in same subject contexts a fully inflected predicate (which is grammatical) results in longer response times. This suggests that additional pragmatic inference is required for non-Echo Subject clauses in these same subject configurations.
