نقش ویژگی‌های رده‌شناختی ساخت موصولی در تعیین ترتیب سازه‌ای فارسی

Relative Clause (RC) as a complex syntactic structure in most languages has attracted the attention of many researchers. According to Lehman (1986), there are three reasons for this including Greenberg’s Universals (1963), Keenan & Comrie NP relativized hierarchy (1977) and the role of RC in studying the phases in Generative Grammar. Several linguists have studied RC and classified it into various types. In other words, there are some typological parameters for the classification of RC universally. RC at the same time has a vital role in the typological studies of word order. In the current study, we intend to examine Persian Relative Clause (PRC) based on these universal features, and afterwards to investigate the role of PRC in predicting Persian word order according to some famous typological hypotheses. In spite of several studies which analyzed syntactic or pragmatic structures of PRC, there is no comprehensive study about PRC in this respect. Therefore, this study is divided into three  main parts. The first part devotes to the studyof universal features of RC in pragmatic and syntactic aspects. Second part of this study is about the behavior of PRC according to the universal features. The role of PRC in the typological studies of Persian word order is the subject of the third part.
As mentioned above, the first part of this article is about universal pragmatic and syntactic features of RC. Based on the pragmatic aspect, RCs are universally divided into two main types of restrictive and non-restrictive. Based on the syntactic features, RCs are studied from five different aspects. First, RCs are divided into the headed and the headless. The frequency and importance of the headed RC is more than the headless one. The second and third syntactic aspects of RC are restricted to the headed RC. In other words, there are two general parameters for the headed relative clause; the first one concerns the place of the head in the structure (whether internally-headed or externally-headed) and the second one concerns the position of the modifying clause (whether joined or embedded). There are five universal kinds of the headed relative clause based on the second and third syntactic aspects of RC including proposed, circumnominal, postposed, prenominal and postnominal. Using the fourth aspects, different methods of producing RCs in all languages are studied. There are four universal methods including gap method, resumptive pronoun method, pronoun retention method and full noun phrase method. Typological studies of RC in this aspect show that the fourth method, full noun phrase, has three different approaches including correlative clause approach, internally headed relative clause approach, and paratactic relative clause approach. At last, the fifth syntactic aspect of RCs focuses on the grammatical function of the head. The best known study in this area is done by Keenan and Comeri (1977) entitled as NP relativized hierarchy. They introduced this hierarchy by studying the behavior of RC in 30 different languages. According to their findings there are a tendency in these languages for relativization of nouns by the following grammatical function:
«genitive< obliqueThe second part of this article investigates the behavior of PRC concerning the above universal parameters. For this reason, syntactic and pragmatic structures and features of PRC were studied separately. Before that, a list of related studies of PRC from different perspectives is presented as a comprehensive review of literature. From pragmatic perspective, there are both restrictive and non-restrictive RC in Persian. The results of pragmatic study of PRC show that Persian distinguishes between restrictive and non-restrictive relative clauses too. Three criteria- adding a suffix ‘-i- ‘to the head of restrictive relative clause, obligatory pause between head and modifying clause in non-restrictive clause, and optional usage of non-restrictive clause - in addition to different examples are presented for showing this distinction. After that, all syntactic aspects of PRC are studied too. Findings of the first syntactic aspect of PRC show that there are both headed and headless relative clauses in Persian, although – as most other languages- headed relative clauses are more frequent in Persian. By considering the second syntactic aspect, Persian has externally headed RCs because the results of this part show that head of this constituent in Persian is always out of modifying clause. Some examples are presented to show that Persian modifying clauses are embedded within main clause (not joined). Furthermore, since modifying clauses in Persian always follow the head, Persian has a post nominal Relative structure. The results of the forth universal syntactic aspect of PRC indicate that Persian uses the first and the second method of relativization among the four universal methods including gap, resumptive pronoun, pronoun retention and full noun phrase. Of course, using gap method is prohibited when the head has a genitive function and using resumptive pronoun is prohibited when the head has a subject function in Persian. For the last syntactic aspect, all NPs in Keenan & Comrie (1977) can undergo relativization in Persian.
Focus of the third part of this study is on the role of PRC on word order. For this sake, firstly, the results of some famous studies about Persian word order are presented. After that, the role of RC in the word order investigations of Greenberg (1963), Vennemann (1973), Lehmann (1974), Dryer (2005) and Hawkins (2014) are presented and discussed. In the last part, the behavior of PRC for predicting Persian word order is presented. Investigations of PRC’s characteristics show that Persian must be a SVO language according to Greenberg (1963), Vennemann (1973) and Lehmann (1974), but it could be SOV or SVO language according to Dryer (2005). In other words, although Persian, like the East-Asian languages, is a head-final language with SOV word order, it has post-nominal RC. Therefore, Persian is potentially an interesting language in this respect, because typologically it falls in between the European and East-Asian languages. The findings of this part confirm the results of some famous Iranian linguists such as Dabirmoghadam (1392) who believes Persian is changing from a OV language into a VO one.