Accusative and oblique subjects

in Ida’an-Begak transitive verbs

Nelleke Goudswaard

Vrije Universiteit Amsterdam

Ida’an-Begak is a Western Malayo-Polynesian language spoken on the east coast of Sabah, Malaysia, Borneo. This paper will concentrate on the Begak dialect and is based on spontaneous and elicited data gathered in the field.

Begak has two voices: Active Voice (AV) and Undergoer Voice (UV).  In the Active Voice, the agent is the subject, whereas in the Undergoer Voice, the patient is the subject. Full NPs are unmarked for case, but pronouns can appear in Nominative, Genitive and Accusative case.

Begak has two word orders. The verb initial word order is semantic: Verb - Agent - Patient and the verb medial word order is syntactic: Subject - Verb - Object.

The first phenomenon I will adress is the case marking of patients (objects) of AV verbs. Example (1) illustrates a sentence with an AV verb. Note that the object can appear either in the accusative (nakon) or in the oblique (nong nakon) without any difference in meaning. Example (2) shows that objects expressed by personal names are obligatorily marked by the oblique preposition nong. Oblique marking of other full NPs gives the NP a specific reading. Factors conditioning oblique marking of object NPs are humanness and specificity.

The second problem I will address is the case marking of patient-subjects of UV verbs. Examples (3) and (4) show the same UV verb, but in (3) the patient appears before the verb whereas in (4) it appears after the verb. The pronominal patient-subject of an UV verb receives nominative case only if it appears in pre-verbal position, as in (3). Post-verbal pronominal patient-subjects receive accusative or oblique case as in (4). 

Although the case marking of post-verbal UV patients suggests otherwise, tests such as relativization and control show that the patient of  UV verbs is the subject. Sentence (5) shows a relative clause of which the antecedent ulun ‘person’ is the agent, so the verb has to appear in the AV. Sentence (6) is ungrammatical because the verb appears in the UV so the gap wrongly refers to the patient. The controllee in (7) is the agent of the subordinate verb so the verb has to appear in the AV; the UV makes the sentence ungrammatical. The controllee in (8) is the patient of the subordinate verb so the verb has to appear in the UV; the AV makes the sentence ungrammatical.

I will argue that post-verbal patient-subjects appear in VP and pre-verbal patient-subjects in specIP. Only subjects that are known and definite may appear in specIP (cf. Diesing 1992, Woolford 2000). Patient-subjects appearing in VP need not be definite but may be specific, in which case they receive oblique case marking.

I  argue that nominative case marking depends on grammatical relations (UV or AV), the position in the sentence (VP or specIP) as well as on pragmatic factors such as the topic / focus status of an argument, and whether it is foregrounded or not.

Examples

Please note that @ represents schwa.

         The dog bit me (Acc/Obl).

         Sius is beating *Elvin / Elvin (Obl) / the dogs / this particular dog (Obl).

         I (Nom/*Acc) /Elvin have/has been bitten by a dog.

         Yesterday, I (Obl/Acc/*Nom) have been bitten by a dog.

         I do not know the person (who was) stealing (AV) the chicken.

         *not good for: I do not know the person (who was) stealing (AV) the chicken.

         (good for: I do not know the person (who was) stolen (UV) by the chicken.)

         The policei try to [ 0i catch (AV / *UV) Martin].

         Martini is afraid to [0i be caught (*AV / UV) by the police].

References:

Diesing, Molly (1992) Indefinites, Linguistic Inquiry Monograph 20, Cambridge,

Massachusetts: MIT Press

Woolford, Ellen (2000) “Object Agreement in Palauan: Specificity, Humanness, Economy and Optimality”, in I. Paul, V. Phillips and L.Travis (eds) Formal Issues in Austronesian Linguistics, pp 215-45, Kluwer, Dordrecht