|
Transitivity and Absolutive Extraction in Tagalog
Edith Aldridge SUNY Stony Brook
Many Austronesian languages exhibit a constraint under which only absolutives (or subjects) are able to undergo A’ movement, as shown in (1) for Tagalog. In the transitive clause in (1a), the theme wh-word can move to clause-initial position. In order to extract the agent, the clause must be antipassivized, as in (1b). But the theme cannot be extracted from the antipassive, as shown in (1c).
This paper proposes an account of this restriction in Tagalog based on transitivity and Case checking. I propose that absolutive Case in ergative languages is checked either by T or v, depending on the transitivity of the clause. In transitive clauses, v hosts an uninterpretable absolutive Case feature which is checked with a VP-internal DP goal. In intransitive clauses, like antipassives, the absolutive Case feature is located in T, which serves as the probe for the DP closest to it, the external argument in [Spec, v].
The A’ extraction restriction is accounted for in the following way. I propose that a [wh] feature is allowed to appear on v only when this functional head also hosts an absolutive Case feature, i.e. when the clause is transitive. The result is that a VP-internal wh-word in a transitive clause will be attracted to the outer specifier of v to check the [wh] feature. This DP will then be at the vP phase edge and can move further to [Spec, C], as shown in (2a). In intransitive clauses like antipassives, v does not have a [wh] feature and therefore cannot attract a VP-internal wh-word, accounting for the fact that VP-internal DPs cannot be A’ extracted in antipassives, as shown in (2b). However, the agent in an antipassive clause, which is merged in the vP phase edge, can be attracted to [Spec, C], as shown in (2c).
This paper argues against the Case agreement approach to the absolutive extraction restriction proposed for Austronesian languages by Georgopoulos (1985), Chung (1998), Rackowski (2002), and others. These analyses claim that A’ movement in Austronesian languages is allowed only when the verb agrees with the Case feature of the dislocated DP - nominative Case for external arguments and accusative Case for internal arguments. One argument against this approach is long distance wh-movement. The embedded clause in (3) is intransitive to allow dislocation of the agent. This might be construed as nominative Case agreement under the traditional account. However, the matrix verb must be transitive, which would not be expected for nominative Case agreement. This is accounted for straightforwardly by the approach taken in this paper. The DP moving from the embedded clause must pass through the vP phase of the matrix clause. This v must therefore have a [wh] feature, which is only allowed by transitive morphology on the matrix verb, according to the current proposal.
(1) a. Ano ang b-in-a-basa ni Maria? what Comp Red-Tr.Perf-read Erg Maria “What is Maria reading?” b. Sino ang b-um-abasa ng libro? who Comp Red-AP.Perf-read Obl book “Who is reading the book?” c. *Ano ang b-um-abasa si Maria? what Comp -AP.Perf-read Abs Maria “What is Maria reading?” (2) a. Ano ang b-in-a-basa [vP twhat [v’ ni Maria [v’ twhat ]]] [wh] what Comp Red-Tr.Perf-read Erg Maria “What is Maria reading?” b. *Ano ang b-um-abasa [vP [v’ si Maria twhat ]] what Comp -AP.Perf-read Abs Maria “What is Maria reading?” c. Sino ang b-um-abasa [vP twho ng libro] who Comp Red-AP.Perf-read Obl book “Who is reading the book?” (3) Sino ang s-in-abi [vP twho [v’ mo-ng [CP p-um-munta sa Maynila] [wh] who Comp -Tr.Perf-say 2sErg-Lk -Intr.Perf-go to Maynila “Who did you say went to Manila?”
References Chung, Sandra. 1998. The Design of Agreement. Chicago: University of Chicago Press Georgopoulos, Carol. 1985. Variables in Palauan Syntax. Natural Language and Linguistic Theory 3 Rackowski, Andrea. 2002. Voice and Configurational Case in Tagalog. paper presented at AFLA 9, Cornell University
|