تحليل يك نوع فعل مركب فارسي بر اساس نظريه اشتقاق فاز

Phase Derivation of a Class of Persian Compound Predicates

فعل‌هاي مركب در زبان فارسي از زايايي بالايي برخوردارند. اين فعل ها، از يك طرف ساختاري تركيبي و نحو بنياد دارند، اما از طرف ديگر، عملكرد آن‌ها در نحو يكپارچه و به صورت يك هستة واژگاني منسجم است و هم چنين درونداد فرايندهاي ساختواژي قرار مي گيرند. در اين پژوهش، فرض بر اين است كه فعل‌هاي مركب عمل كرد دوگانه صرفي-نحوي دارند. هدف پژوهش حاضر اين است كه نشان دهد فعل‌هاي مركب ساخته شده با فعل سبك "كرد" در زبان فارسي داراي ساخت اشتقاقي فازگونه هستند. داده هاي تحقيق، نمونه هايي از فعل هاي مركب هستند كه از فرهنگ دو جلدي سخن استخراج شده اند. مبناي نظري اين تحقيق با روش هاي اشتقاقي مطرح شده در آثار چامسكي (2000؛ 2001؛ 2008) ، مرنتز (2001؛ 2007) هم سويي دارد، هرچند از بعضي مفاهيم نظري در نظريه هاي ساختواژي همچون ليبر (1980) هم استفاده شده است. نتايج نشان مي دهند كه فعل هاي مركب داراي اشتقاقي فازگونه اند. در زبان فارسي، ريشه فعل و جزء غيرفعلي در دامنة يك فاز ساختواژي ادغام مي شوند سپس آن ريشه ها براي رفع نيازهاي اشتقاقي يا محاسباتي به پوستة آن فاز ساختواژي حركت مي كنند كه در آن جايگاه تحت تأثير عمليات هاي نحوي مي توانند به جايگاه هاي نحوي حركت كنند و گسترش يابند.

1. Introduction Persian compound verbs have been the topic of much research and have been investigated on the basis of various approaches. A clear fact about complex predicates is that they are constructions formed by more than one lexeme. The light verb in a complex predicate is the head and the other segments are considered as nonverbal. This paper tries to describe how complex predicates formed by the light verb "Kard-an = to do" are derived. On the one hand, complex predicates are regarded as lexical units as they are input to morphological rules; on the other hand, they have been supposed to be syntactic because they can split in syntax like independent lexical items. This paper describes morphological derivation of Persian complex predicates and their syntactic separability on the basis of Phase Derivation Theory which has recently gained much attention in the minimalist tradition. The data were extracted out from Sokhan Dictionary and the online version of the Persian Linguistic Database. Tree diagrams are usually used in this theory to show the details of the structural derivations. Phase derivation theory (Chomsky, 2000, 2001, 2008) is the newest version of the generative grammar. Marantz (2001, 2007), Di Sciullo (2003) referred to the morphological phases inside word structure and explained how derivation is performed morphologically. Megerdoomian (2002) following Marantz (1997) and Chomsky (2000) claimed that there is a phase head in Persian complex predicates. In studies by VahediLangrudi (1996), Karimi (1997), Megerdoomian (2002), Folli, Harley, and Karimi (2005), and Pantcheva (2008), Persian complex predicates are syntactic unites as output of the syntactic operations. Karimi Doostan (1997) regarded complex predicates as morphosyntactic units. 2. Methodology This paper deals with two theoretical problems: first, how is the complex predicate derived, which leads to the interaction of morphology and syntax; second, why does the interaction between morphology and syntax take place during the derivation of complex predicates. In this paper, complex predicates are the output of the derivational operations in morphology that are in turn the input of the syntactic operations, so the interaction between morphology and syntax emerges. The derivation method suggested here is that roots and categories are merged first. Roots are bare and have no category and inflection. Then, a phase head is merged which carries inflectional or functional information. Phase head in complex predicates is a light verb head (v) which is merged after the merge of all the roots. The sister of the phase head is called the phase domain. This domain is impenetrable. By phase impenetrability condition, phase domain is transmitted to the interface levels and remains out of the access of syntactic operations; while, phase head plus Spec-phase is at the phase edge which is accessible to the syntactic operations. Before the spell-out, every segment should satisfy its computational needs. 3. Results and Discussion: In Persian grammar, verb roots represent out as a tensed stem. In other words, verb root is [u-tense] which should be checked during the derivation. It is possible for a phase head to have [tense] feature (Chomsky, 2008; 2013); so, it is assumed that phase head (v) carries feature [tense]. Before spelling out, the verb root "Kon = do" in the phase domain is adjoined to (v) to check [u-tense]. By default, in Persian the nonverbal segment is represented before the verb head. Based on the correspondence axiom principle (Chomsky, 1995) the linear order is dependent on the c-command in the derivation; that is, the nonverbal element is located at the phase edge and c-commands the light verb. The nonverbal element moves in order to omit [EPP] feature of the phase head and to remain in a common derivational domain near the light verb, so that they spell out in a common step of the derivation and hence, the interpretation of the complex predicate is performed in a common minimal domain. Therefore, the verb root adjoined to the phase head (v) operates as a light verb head, and the nonverbal is adjoined to the phase edge, which is accessible to syntactic operations and can split in syntax. In some cases, a lexical root which moves to the phase edge is adjoined to an affix head. According to Lieber's (1980) morphological theory, every affix has a subcategorization frame containing the category information and the selectional restrictions. When a lexical root moves to edge of the morphological phase; if adjoined to an affix, it is inserted into the subcategorization frame. 4. Conclusion It is concluded that the complex predicates are derived by a morphological phase. The domain of the phase remains out of access to syntactic operations. But the nonverbal element in the phase edge is accessible to the syntactic operations and becomes separable in syntax. Then, the complex predicates are derived by a phase the edge of which can split in syntax and remain separable during the syntactic derivation.