Pre-verbal dislocations in Persian

Fri Mar 1 00:38:04 UTC 2013

Dear LINGTYP members,

 Persian has generally been considered throughout the literature as basically a SXV language with a variety of word order rearrangements, which situates it in the category of scrambling languages. Scrambling, as far as I know, is a linguistic term with a paramount importance in the generative literature since Ross' (1967) PhD dissertation on. However, several Iranian and non-Iranian linguists claim that scrambling is a discourse-motivated phenomenon particularly intended to elicit topic or focus interpretations (Karimi 2005 among others). Given that Persian is a canonically a SXV language, the abundance of linguistic terms exerted to describe the syntactic dislocations in the pre-verbal domain of the sentence has lead to a huge baffling puzzle in my mind. For instance, Birner and Mahootian (1996) claim that there is an inversion process in Persian, functionally akin to English inversion, and  English topicalization in terms of the linear order of
 dislocated constituents, according to which the canonical SXV order is modified to XSV monitored pragmatically by this constraint that the preposed constituent is required to represent information at least as familiar within the discourse as that represented by the postposed constituent, which is in line with Birner (1994). The functional overlap between inversion despite the different constituent order in the two languages  gives rise to the claim that inversion in English and Persian constitutes identical constructions on their own rights. 

Inversion in Persian: X (PP) S V
Inversion in English : X (PP) V S
Topicalization in English : X S V 

Ex.1: ru-ye   hær   qali     donyayi  hast.                          preposed PP:  ru-ye   hær   qali (Discourse-old)
        on-EZ every carpet  world     be.PRS.3SG                postposed S: donyayi (Discourse-new)
        'In every carper is a world.'                                                                                       (Birner and Mahootian ibid: 130)

Another functional contribution by Shahidi (2000) reveals that scrambling is a distinct process in the syntax of Persian in which the clausal constituents undergo leftward movements that is conditional to the preservation of their grammatical marking; for example the postposition "ra" and prepositions in case of direct object NPs and PPs respectively. Moreover, the moved elements in leftward scrambling are not bound by their information statuses; e.g. a discourse-new PP can be scrambled to the beginning of the clause in presence of a discourse-old subject, whereas the same alteration  in inversion along the lines proposed by Birner and Mahootian (ibid) would be in felicitous on the grounds that the preposed constituent is not supposed to represent newer information than the postposed one.   

Ex 2: čeraq-ha-ra    man  xamuš  mi-kon-am. 
         light-PL-RA   I      dark     IPFV-do.PRS-1SG
         'lights I'll turn off.'                                                         (DO scrambling)

Ex. 3: ba     yeki æz  qali-ha-š                        aqa      tæmam-e     mædresæ-ro       mi-xærid.                     
          with  one  of   carpet-PL-CL.3SG       master  entire-EZ     school-RA           IPFV-buy.PST-3SG
          'With one of his carpets the master would buy the entire school.'                                            (PP scrambling)                                          (Shahidi 2000: 206)

My query is concerned first and foremost with this issue as to whether the discourse-status of dislocated elements could be a valid criterion for the distinction between inversion and scrambling in a SXV language such as Persian with this in mind that both movements share the leftward shift of constituents; e.g. a PP. Perhaps, the question does not pose an interesting issue for the functionalist or generativist models of language; however, I would like to know how linguistic typology can approach this complication. 

My second question deals with this challenge whether or not it is possible to claim that a SXV language such as Persian can represent an inversion construction, i.e. the so-called XSV in light of merely discourse-functional similarities with the XVS inversion in English; namely the pragmatic constraint governing the distribution of given and new information in the preposed and postposed constituents, regardless of the fact that the so-called XSV order in Persian does not syntactically exhibit postposing of the X, say PP, contrary to the constituent reversal in the English counterpart. 

I would appreciate any comments, hints or any cross-linguistic evidence posing the same complication. 

Thank you.

Farhad

References

Birner, Betty J. 1994. Information status and word order: An analysis of English inversion. Language 70:233-59.
Birner, Betty J., and Shahrzad Mahootian. 1996. Functional constraints on inversion in. English and Farsi. Language Sciences 18:127-138.
Karimi, Simin (2005) A Minimalist Approach to Scrambling: Evidence from Persian, Mouton de Gruyter.
Shahidi, Noushin. 2000. Topicalization in Persian: A Functional Perspective. MA diss., Allameh Tabataba'i University, Tehran.  
Ross, John R. 1967. Constraints on Variables in Syntax. Doctoral dissertation,. MIT.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20130228/7135a571/attachment.htm>