. Toba Batak displays a variety of properties that are typologically rare: (1) The predominant word order is V-O-S-IO and V-O-S-Adv. (2) Extraction of direct objects/passive agents (by relativization, etc.) is ungrammatical, but subjects, IOs, and adverbials can be extracted. (3) In Toba Batak anaphoric binding, the active subject binds the DO and active DOs can bind the IO, but, additionally, the “passive agent” can bind a passive subject. We argue that these facts are explained by an analysis in which V-O-S-IO order is derived by VP-raising and in which the passive agent is an argument, generated as the specifier of vP rather than as an adjunct. Furthermore, the passive subject undergoes optional reconstruction to its base-generated position in VP. We also argue that, although c-command relations among elements of vP are critical, linear order plays no role whatsoever in the base structure. That is, the output of Merge can be unordered linearly, and it can be shown that all surface word orders in Toba Batak can be derived from various movement rules out of an unordered base. Linear order comes into play only when constituents move out of the vP to higher functional projections, linear order being simply a by-product of Move.
[1]
Peter Cole,et al.
Voice in Malay/Indonesian
,
2008
.
[2]
N. Richards,et al.
Phase Edge and Extraction: A Tagalog Case Study
,
2005,
Linguistic Inquiry.
[3]
Chris Collins,et al.
A smuggling approach to the passive in english
,
2005
.
[4]
Joachim Sabel,et al.
Clause structure and adjuncts in Austronesian languages
,
2006
.
[5]
Peter Cole,et al.
Is There Pasif Semu in Jakarta Indonesian?
,
2006
.
[6]
Gabriella Hermon,et al.
Subject And Non-subject Relativization in Indonesian
,
2005
.
[7]
Matthew Pearson,et al.
The Malagasy Subject/Topic as an A′-Element
,
2005
.
[8]
R. Larson.
On the double object construction
,
1988
.
[9]
Lisa deMena Travis,et al.
Spec of IP and spec of VP: Two subjects in Austronesian languages
,
1992
.