Abstract
Urdu/Hindi has been characterized as a wh-in situ language. That is, the wh-word need not necessarily appear in a position other than the base position. However, the wh-word can in fact appear in various places in a clause. Most discussed have been topicalized structures, in which a wh word can also be extracted from embedded sentences, often in interaction with what has come to be known as the scope-marking construction (e.g., Mahajan 1990, Dayal 1994, Dayal 1996). However, the wh-word can also appear at the right periphery for reasons not completely understood so far. Bhatt and Dayal (2007) propose a rightward remnant movement account, Manetta (2012) has recently challenged this account, instead proposing a scramblingaccount in which the rightward scrambling is licensed by a an EPP feature. However, these syntactic accounts cannot do justice to theinterpretational differences that arise in conjunction with thedifferent positioning of the wh-word. Consider the contrasts in (1), whereby (1a) represents the canonical word order, (1b) is only good asan echo question and (1c) as a rhetorical question (data based on Bhatt and Dayal 2007).
| (1) a. | siitaa=ne | dhyaan=se | kis=ko | dekh-aa | thaa? |
| Sita.F=Erg | carefully | who.Obl=Acc | see-Perf.M.Sg | be.Past.M.Sg | |
| 'Who had Sita looked at carefully?' | |||||
| b. | siitaa=ne | dhyaan=se | dekh-aa | thaa | kis=ko? |
| Sita.F=Erg | carefully | see-Perf.M.Sg | be.Past.M.Sg | who.Obl=Acc | |
| 'Sita had looked at carefully at who?' | |||||
| c. | siitaa=ne | dhyaan=se | dekh-aa | kis=ko | thaa? |
| Sita.F=Erg | carefully | see-Perf.M.Sg | who.Obl=Acc | be.Past.M.Sg | |
| 'Who had Sita really looked at carefully?' (i.e., she had not looked at anybody carefully) | |||||
I argue that in order to undestand question formation in Hindi/Urdu, information-structural factors must crucially be taken into account. I have suggested in previous work (Butt and King 1996) that Hindi/Urdu should be seen as a discourse-configurational language (cf. Kiss 1994) and that position in a clause is associated with information-structure in the following way: topics [-new,+prominent] are clause initial, focus [+new,+prominent] is immediately preverbal and background information [-new,-prominent] is postverbal (Butt and King 1997). This analysis immediately provides an explanation for the interpretation of (1b) as an echo question, example (1c) is more interesting in that the wh-word here breaks up the verbal complex. I suggest that a further type of focus, namely emphatic focus, is signaled in a postverbal position within the verbal complex and that this ultimately accounts for the interpretation of (1c) as a rhetorical question.