تحلیل نحوی حالت دهی کنائی در منتخبی از زبانهای ایرانی: رویکردی کمینه گرا

Tatic-type languages are among west Iranian languages divided into four main groups: northern Tatic, central Tatic, southern Tatic and Taleshi group (Stilo, 1981, p. 139). Among these languages, we focused on three southern Tatic languages namely Chali, Taleshi (Anbarani) and Vafsi. Theses Tatic languages present the most complex kind of split alignments and this phenomenon follows the universal tendency seen in tense/aspectual split ergative alignments in which the ergative alignment only appears in a specific tense/aspect generally past/perfect tense. In these Tatic languages, the oblique subject only appears in past transitive clauses where in there is no verbal agreement, but in other environments, the subject is direct and the verb agrees with direct subjects: In present tense sentences, the subject is direct and the verb shows full agreement with the subject. In contrast, in past tense sentences the subject of intransitive clause is direct and the verb shows full agreement with the direct subject. In transitive clauses, the subject bears marked case oblique and the agreement in the verb would be default 3s. In all three languages, a pronominal mobile clitic optionally cross-references the subject. The split ergative alignment of these Tatic languages is of potential theoretical interest mainly for two reasons. First: In one of the most influential views in the current literature on ergativity, ergative case is an inherent case (Nash, 1996, 2015; Woolford, 1997, 2006; Aldrige, 2004, 2008, 2012; Laka, 2006; Anand & Nevins, 2006; Legate, 2006, 2008, 2012; and Massam, 2006). On this view, ergative case is attributed to the lexical properties of the agentive v head and theta marks the subject, not to the subject’s surface structural position or to the agreement with non-theta marking heads (Baker 2015, p. 54). In other words, the main assumption for considering ergative as a kind of inherent case is the relationship between agent theta role and ergative case and in languages that ergative is inherent case, we should see a close correspondence between agent theta role and ergative subjects and also the presence of active alignment. However, it does not seem right for Tatic languages since what thematic roles an NP has is not a primary determinant of its case in these languages; all past transitive subjects are marked oblique regardless of their theta roles and no active alignment can be seen in these languages. Additionally, the restrictedness of the appearance of ergative/oblique case in past transitive sentences suggests that it is a kind of structural case and should be accounted for by an structural case mechanism. Secondly, the split ergative alignment in these Tatic languages is different from other aspectual split alignments in spite of the fact that the restriction of ergative to past transitive clauses conforms to the well-known universal tendency in this regard. But in these languages, it is not aspect that conditions splitness. So generally, it can be argued that these Tatic languages cannot be classified as aspectual based ergative languages and none of the analytical studies (Laka, 2006; Coon & Preminger, 2014; Nash, 2015; Ura, 2006; Baker, 2015) which derive their analyses based on aspect can account for these Tatic languages. As it can be seen, none of the analyses proposed so far can be readily used for explaining split ergative alignment in these Tatic languages. Thus, the following questions and hypotheses are addressed here:  
1-      What is the source of Ergative case on past transitive subjects?
2-      What is the source of direct case on present transitive subjects and intransitive ones?
In an alternative view (Baker, 2015), ergative case is a structural case. Baker (2010, 2015) believes that the inherent view of ergative case has advantages for non-strict ergative languages like Hindi and Georgian and he tentatively accepts it for those languages; however, he argues that in stricter ergative languages, ergative is a structural case not an inherent one (see Baker 2015, p. 54 for more discussion). In this regard, he invokes the idea that in addition to the agreement-based theory of case (Agree), case can be assigned by a rule of dependent case assignment in the sense of Marantz (1991). He also invokes the possibility that one language may use a combination of case assigning mechanisms; that is to say, both Agree and Marantzian ones.
We have adopted Baker’s (2014) argument regarding the conditioning factor in splitness in Kurmanji and claim that the fundamental difference between clauses with past and present verb stems, which drives the split ergative pattern in these Tatic languages, is in the phrasal status of the v node. We claim that subject direct case is related to agreement on T in the familiar way, but oblique case on past transitive subjects is not related to agreement with a functional head and instead, we claim that the rule for oblique case assignment can be formalized in terms of a dependent rule.