On the lack of wh-movement in Malagasy

Ileana M. Paul

University of Western Ontario

Clause-typing is marked in particular ways cross-linguistically (Cheng 1991). One common form is through wh-movement to the periphery. Languages with wh in-situ, however, resort to clausal particles (often abstract) to type the clause as [+wh]. Malagasy appears to have both wh-movement and wh in-situ. Such optionality is difficult to account for under a feature-driven account of movement (e.g. Chomsky 1995, inter alia). I suggest, however, that there is no wh-movement (overt or covert) to the periphery in Malagasy and examine the consequence that this has for various types of wh-constructions.

            It has long been recognized that Malagasy wh-movement is a type of cleft (1). According to Dahl (1986), in the cleft, the focused element is the predicate and the remainder of the sentence is a headless relative in subject position. Such an analysis also applies to wh-clefts – the wh-word is a predicate and therefore not in a clause-peripheral position. Instead an abstract Q particle appears in the clause periphery to mark [+wh].  (The overt version of this particle (ve) occurs in yes-no questions (2).)

The Q particle also (unselectively) binds wh-elements in-situ (3).  No LF movement of wh-elements obtains – there are neither island (wh, complex NP or adjunct) nor weak cross-over effects (4). Similarly, in indirect questions, there is no wh-movement to the edge of the embedded CP. Instead, the matrix verb either selects for a DP headless relative (5), a CP headed by raha ‘if/whether’ or a cleft.  A range of examples thus shows that full XP movement to the periphery is not necessary to type a clause as [+wh]. In fact, the only true A-bar movement is operator movement, which obtains in relative clauses and clefts.

            This analysis of wh-constructions can be extended to account for the “partial” movement construction (6). The wh-word appears in an embedded position but the sentence is interpreted as a matrix wh-question. Unlike overtly moved wh-elements in English (Baker 1970), the scope of the clefted wh can be wide, bound by a matrix Q. Once again, there is no need to posit covert wh-movement to the matrix clause or even an indirect dependency à la Dayal (1994). Finally, I consider the formal semantics of wh-questions and how to accommodate the Malagasy data within current semantic analyses of questions.

I conclude by comparing Malagasy with other languages that form wh-questions with clefts (e.g. Irish) and by discussing how the full range of wh-question data fit with recent theories of V-initial languages (Oda 2001).

1.         Iza  no manasa lamba?

            who no wash    cloth

            ‘Who is washing clothes?’ (‘The one who is washing clothes is who?’)

2.         Manasa lamba ve i Be?

            wash     cloth     Q             Be

            ‘Is Be washing clothes?’

3.         Manasa inona i Be?

            wash     what      Be

            ‘What is Be washing?’

4.         a.             Namangy         ny lehilahy izay nanasa inona i Be?

                        met                   the man             rel washed what     Be

                        ‘What did Be meet the man who washed?’

            b.            Manaja an’iza   ny reniny?

                        respect who      the mother.his

                        ‘Who does his mother respect?’

5.         Tsy fantatry             Rakoto izay manasa lamba.

            neg know                Rakoto rel wash           cloth

            ‘Rakoto doesn’t know who is washing clothes.’

6.            Mihevitra i Be fa             iza no    nanasa lamba?

            think                 Be that who no wash    cloth

            ‘Who does Be think washed clothes?’

References

Baker, C.L. 1970. Notes on the description of English questions: The role of an absract question morpheme. Foundations of Language 6: 197-219.

Cheng, Lisa. 1991. On the typology of wh-questions.  PhD thesis, Cambridge, MA, MIT.

Chomsky, Noam. 1995. The minimalist program. Cambridge, MA: MIT Press.

Dahl, Otto Chr. 1986. Focus in Malagasy and Proto-Austronesian. In FOCAL I, Eds. P. Geraghty, L. Carrington and S.A. Wurm: 21-45. Pacific Linguistics.

Dayal, Veneeta. 1994. Scope marking as indirect wh-dependency. Natural Language Semantics 2: 137-170.

Oda, Kenji. 2002. Wh-questions in V-initial languages. MA thesis, University of Toronto.