بررسی کارایی آزمون‌های شناسایی سازه در نحو زبان‌ فارسی؛ رویکردی کمینه‌گرا

شناسایی سازه، به عنوان یک واحد مستقل نحوی و معنایی که شالوده اصلی جمله های زبان است، هم به منظور اجتناب از ابهام و هم از لحاظ درک و تحلیل صحیح فرایندهای نحوی از اهمیت بسیاری برخوردار است. بدین منظور آزمون هایی در نحو معرفی شده‌اند که زبان‌شناس را در امر شناسایی مرز بین سازه‌ها یاری می‌کنند. در تحقیق حاضر، در چارچوب برنامه کمینه گرا هشت آزمون مهم همپایگی، حذف، پرسش‌سازی، پاره‌جمله، جایگزینی، اسنادی‌سازی، شبه اسنادی‌سازی و مبتداسازی را که دارای بیشترین بسامد بودند، بر روی هفت سازه منتخب و مهم جمله یعنی گروه فعلی، گروه حرف تعیین، گروه حرف اضافه‌ای، گروه صفتی، گروه قیدی، گروه متمم‌نما و گروه زمان در زبان فارسی به کار بردیم. از مقایسه ی نتایج آزمون های یادشده به این نتیجه رسیدیم که آزمون های شبه اسنادی‌سازی و مبتداسازی در زبان فارسی کارایی کمتری دارند و قادر به شناسایی سازه‌های کمتری هستند؛ به علاوه در زبان فارسی نمی توان از همپایگی زنجیره مشترک به‌عنوان آزمونی مجزا بهره برد. از مجموع مباحث تحقیق نیز دریافتیم که با توجه به وجود مثالهای نقض و استثنائات فراوان، نمی‌توان آزمونهای فوق را ابزاری مطلق برای شناسایی سازه‌ها در نظر گرفت.

An Analysis of the Efficiency of Constituency Tests in Persian; a Minimalist Analysis

Detecting constituents as independent syntactic units having syntactic and semantic coherence on which the structure of the sentences is based is vital both for the prevention of ambiguity and for a correct reading and analysis of syntactic processes. To this end, some tests have been introduced in syntax which help the linguist in diagnosing the boundaries of constituents. Within a Minimalist framework, this study aimed to analyze the efficiency of eight constituency tests, namely: coordination, ellipsis, question formation, sentence fragment, replacement, clefting, pseudoclefting and topicalization. These had the highest frequency in important syntactic sources as means for diagnosing seven selected important constituents of the sentences, i.e., VP, DP, PP, AP, AdvP, CP and TP. We applied the tests to these categories in Persian to see which ones are most efficient and can cover a wider range of constituents. The results suggested that pseudoclefting and topicalization are more effective and have the capability of identifying more constituents in English than in Persian. It was also concluded that, given the large number of exceptions, in general, the tests mentioned above cannot be considered as absolute tools in identifying constituents.1. IntroductionDistinguishing the borderline between constituents in sentences can be a great help in preventing syntactic ambiguity. But this can rsquo;t be achieved without specific discovery instruments for the detection of constituents. This instrument is the very constituency test, which is the focus of the present study. Although these tests have been presented in most sources in the literature, their efficiency for the detection of constituents in Persian has rarely been the topic of a comprehensive study. Since in Minimalism only constituents can participate in syntactic operations, it is necessary to define constituents precisely to be able analyze syntactic structures and operations. Chomsky (1971, p 30) states that all syntactic operations are structuredependent. Byram (2004, p 647) further explains that the Structuredependency Principle makes all languages move sentence elements based on their structures and not merely the linear order of words. Minimalism, as Chomsky (2001) defines it, has two basic processes, namely Merge and Move (cited in Cook Newson, 2007, pp 2723) and the tree diagram of sentence structures is built through the bottomup formation of structures by means of merge. It is the constituents which merge or move.Research Question(s)The main questions in the present study are the following:Which of the constituency tests successfully detect syntactic constituents in Persian?Which lexical and functional categories in Persian can be detected by means of each constituency test? 2. Literature ReviewThe works in the literature dealing with constituency tests in Persian can be roughly divided into two groups. First are those which mention the tests superficially among other topics in syntactic analysis. Kavoosinezhad (1997), Dabirmoghaddam (2005), Rasekh Mahand (2006 2011), Toosarvandani (2007), Mahootian (2008) and Karimi Azmoudeh (2012 2015), among others, belong to this group. The second group are those studies which reviewed the tests in more detail although not thoroughly enough. So far, we know of only two of such studies in Persian. Gholamalizadeh (1995) described five tests in Persian, namely: ellipsis, substitution, whquestion formation, wordchain movement (or topicalization), clefting and psudoclefting. Also, Golfam (2012) presents a summary of the main constituency tests and names three of them: substitution, movement and coordination.On the other hand, English language is very rich in constituency test sources. In his successful series (1997, 2006, 2009a b), Radford meticulously lists, describes and exemplifies most of the tests in English in detail. Adger (2002), Kim Sells (2007) Carnie (2001 2010) and Tallerman (2011) also present different constituency tests in the English syntax. 2.1. Constituency Tests in PersianThe following test were used for the detection of different lexical and functional categories in Persian.2.1.1. Coordination2.1.2. Ellipsis2.1.3. Question formation2.1.4. Sentence fragment/Stand alone2.1.5. replacement/proform/substitution2.1.6. Clefting2.1.7. Psudoclefting2.1.1. fronting/topicalization/ preposing3. MethodologyEach of the above tests were used for the detection of the lexical categories DP, VP, AP, PP, AdvP and the functional categories CP and TP in many Persian example sentences to see which tests are efficient in the detection of which categories. The examples were analyzed in terms of acceptability and wellformedness based on the intuition of the researcher, who is a native speaker of Persian. 4. ResultsTables 1 and 2 below represent the efficiency of these tests in the detection of different constituents in English and Persian. Table 1 Constituency tests and target structures in English Target structure TestDP, PP, VP, AP, AdvP, TP, CPCoordinationAP, AdvPEllipsisDP, PP, VP, AP, AdvP, TP, CPQuestion formationDP, PP, VP, AP, AdvP, TP, CPSentence fragmentDP, PP, VP, AP, AdvP, TP, CPSubstitutionDP, PPCleftingDP, PP, VP, AP, CPPsudocleftingDP, PP, VP, AP, AdvP, CPTopicalization Table 2 Constituency tests and target structures in Persian Target structureTestDP, PP, VP, AP, AdvP, TP, CPCoordinationAP, AdvPEllipsisDP, PP, VP, AP, AdvP, TP, CPQuestion formationDP, PP, VP, AP, AdvP, TP, CPSentence fragmentDP, PP, VP, AP, AdvP, CPSubstitutionDP, PPCleftingDP, PPPsudocleftingDP, PP, AdvPTopicalizationAs the tables suggest, constituency tests act more successfully in English than in Persian; they can detect more various constituents. The different efficiency of the tests in two languages lies in the particular syntactic features of their structure. The exploration of these features can be a topic for further studies.The results indicate that some tests like clefting or ellipsis fail in detecting most constituents and are not good options as such. On the other hand, although tests like coordination and sentence fragment are more successful, they face a lot of exceptions, some of which were presented in this paper. Overall, the tests are not always reliable and researchers must use various instruments in their analysis, one of which could be a test. The results of this research can remind the students of syntax of the limitations in the application of constituency tests as an analysis instrument.