resumptive pronouns

[Lars Hellan]

Hello,

Responding to Mehran’s question, Edo, a Niger-Congo language related to the 
Kwa languages of West Africa, is a language that allows for two distinct 
ways of marking an ‘extraction site’, used in wh-constituent questions, 
relativization, and under focus extraction. One of these strategies is 
pronominal and could be seen as a resumptive pronoun strategy if it were 
not for the lack of agreement.

Here are the two strategies that Edo uses to mark extraction sites:

         A pronominal item, with a constant form (normally the form of a 
3rd person sg. pronoun) is inserted at the ‘extraction-site’. In a 
manuscript we wrote we refer to it as a plug rather than a resumptive 
pronoun because of the lack of agreement. Plugs mark the position of 
extracted subjects, and of both indirect and direct objects in double 
object constructions. The use of plugs suggests the notion of an openly 
spelled-out trace, but again what might count against this is a general 
lack of agreement between the ‘filler’ and the plug.

·       The ‘extraction site’ is empty, and the verb subcategorizing for 
the extracted argument, signals a reduction in valence by taking the tone 
of an intransitive verb.

Notice that both strategies respect the standard island constraints.

The above information is also part of a doctoral thesis that is at present 
under preparation at the Linguistics Department in Trondheim by Ota Orgie, 
who is a native speaker of Edo.

A summary of her findings concerning extractions and a discussion of the 
data from an HPSG perspective can be found in a manuscript written by Lars 
Hellan, Ota Ogie and me, called:

“Three kinds of Extractions in Edo”

For people interested in this paper, please send an e-mail to:
dorothee.beermann @ hf.ntnu.no.




-------------- next part --------------
A non-text attachment was scrubbed...
Name: resumptive pronoun.doc
Type: application/msword
Size: 22016 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/hpsg-l/attachments/20021001/1852e9c3/attachment.doc>