دستوری‌شدگی حرف تعریف نکره در فارسی باستان از‌منظر دستور ساخت‌مدار

در این مقاله، به شناسایی ساخت‌های دستوری بیانگر »مفهوم نکره « در زبان فارسی باستان می‌پردازیم. براساس داده‌های موجود از این دوره زبان فارسی، می‌توان بیان داشت قبل از فارسی باستان عنصر دستوری »حرف تعریف نکره « وجود نداشته و پیدایش »حرف تعریف نکره « از رهگذر دستوری‌شدگی عدد aiva »یک « به‌وسیله بازتحلیل این عنصر زبانی براساس روابط مجازی و سپس بسط استعاری به »حرف تعریف نکره « رخ داده است. این دستوری‌شدگی در ساخت مشخصی (تنها یک »x « وجود دارد، این »x « معرف انسان است) رخ داده که نتیجه آن تغییر حوزۀ مفهومی به بیان »مفهوم نکره « در فارسی باستان است، به‌منظور تبیین تکوین »حرف تعریف نکره «، الگوی دستور ساخت‌مدار برگزیده شده است. داده‌های این پژوهش، مستخرج از پیکرۀ جمع‌آوری‌شدۀ کنت (1953) است. بسامد وقوع »حرف تعریف نکره « در بین 4130 گروه اسمی موجود در پیکره 0.5 درصد است. این مقدار بسامد وقوع که چندان زیاد نیست، مبین جدید بودن مقولۀ »حرف تعریف نکره « یا به ‌عبارتی مراحل اولیۀ دستوری‌شدگی این مقولۀ دستوری در فارسی باستان است.

The Grammaticalization of the Indefinite Article in Old Persian:A Construction Grammar Account

This paper identifies the grammatical constructions expressing indefiniteness in Old Persian. Based on the extant data from this stage of Persian language, There had been no indefinite article before Old Persian era and its grammaticalization is the result of the bridging context chr('39')There is only one X, the X represents a human beingchr('39'). In fact, the metonymic relations triggered reanalysis of aiva chr('39')onechr('39') to an chr('39')indefinite articlechr('39') which in turn paved the way for its metaphoric extension. To explain this path of development, a Construction Grammar model is adopted. This research is based on the corpus of OP gathered by Kent (1953). The investigation illustrates that the token frequency of the aforementioned indefinite article in the total NPs (4130) amounts to 0.5%. This percentage is not much high; however, it implies that it is a newly formed grammatical construction in Old Persian. In other words, it illustrates the early stages of the indefinite article grammaticalization. 1. IntroductionThe terms Definiteness and Indefiniteness are usually applied to noun phrases (for example Abbott, 2004; 2006) and encoded using different linguistic means. The most common diachronic source for the indefinite marker in the languages of the world is the numeral chr('39')onechr('39') (Giv oacute;n, 1981; Lyons, 1999; Heine Kuteva, 2002).Only a minority of the languages of the world have grammaticalized indefinite markers (Heine, 1997:68). In other words, languages of the world tend to grammaticalize the definite marker prior to the indefinite one (Moravcsik, 1969; Croft, 2003). In Moravcsikchr('39')s (1969) sample of 108 languages, in 61 languages (56%) only definite markers were identified, whereas 5 languages (5%) contained only indefinite markers and there was no information on the existence of the definite ones; Those 5 languages are Bambara, Gipsy, Rotuman, Sundanese, and Aztec. As a result Persian deserves studying in this regard. In this research, it is investigated whether the indefinite article existed in Old Persian grammar. If the response to this question is positive, the mechanism of its grammaticalization will be investigated. The data of this research are extracted from the corpus gathered by Kent (1953). 2. Literature ReviewThe most related work in the related literature is the work of Paul (2008). Paul describes indefiniteness marking in three stages established for Persian language. His findings are represented as follows (figure 1) (Paul, 2008:310): Figure 1 illustration of the history of i and yek In his study, there is no grammatical element expressing indefiniteness in Old Persian and little attention is paid to the contexts and mechanisms of this development. In contrary to Paul rsquo;s work, in this research it has been illustrated that in Old Persian the indefinite article is grammaticalized and a part of the grammar. 3. MethodologyIn Construction Grammar, all levels of grammar involve constructions. By construction, we mean the pairing of form and meaning which are conventionalized (Hoffmann Trousdale, 2011, p. 2). The constructions are not just listed but instead they form a structured inventory in the speakerschr('39') mind (Croft Cruse, 2004, p. 262265). Due to the structured inventory of constructions which form a network, in CG model there is no distinction between lexicon and syntax; therefore, all constructions can be located on a lexiconsyntax continuum (Fillmore, 1988). This allows for constructions to possess more or less lexical/grammatical characteristics. Consequently, no mere distinction between synchronic and diachronic phenomena is observable. In this process a change in the network of constructions is said to be observable in the course of time. In CG model, the process does not operate on a single word or morpheme but on the whole construction by syntagmatic relations (Lehmann, 1992, p. 406).The reanalysis based on metonymic relations in the context of change which in turn paves the way for metaphoric extension results in grammaticalization (Davari NaghzguyKohan, 2017), being a construction creation in CG. This model and its envisioned chain of changes can be summarized as follows:(2) Metonymic relation > reanalysis > metaphoric relation > new construction > analogy In this research, the distribution of aiva lsquo;one rsquo; is provided in diferent contexts. After that, the aforemtioned mechanism is investigated for the indefinite article in Old Persian. In adittion, the token frequency of the indefinite article is presented in the corpus. 4. ResultsThe frequent use of aiva chr('39')onechr('39') before human NPs in OP has triggered its reanalysis as chr('39')indefinite articlechr('39') in mytonymic relations. Here, one can observe a linkage between numeral system (one as a numeral) and indefiniteness concept as a more abstract conceptual space. This can be illustrated as figure (2): B Figure 2: the conceptual shift from the numeral to indefiniteness The mentioned mechanisms are responsible for the emergence of the new grammatical category filling the determination slot; the schema is mentioned again as follows:(1) [aiva[X]NPi]j [introducing a participant into discourse a X ]j, X=humanIn spite of this usage of aiva as chr('39')indefinite articlechr('39'), its token frequency is not much high in OP. in our corpus of 8077 words and 4130 NPs, there are 21 NPs containing indefinite articles. This can be illustrated in the following figure (3): Figure 3: The token frequency of indefinite article in the corpus As it is obvious, the frequency of use (0.5%) of the indefinite article is not much high; however, as Denison (2006:290291) notes, rare patterns of low frequency should be taken seriously. 5. DiscussionThe presence of an indefinite article in the grammar entails the existence of a determination slot. Finding determinatives in a particular position (preceding modifiers) most of the time, speakers tend to regard that location as a place where determinatives occur and by the increase of the frequency, the speakers abstract a prehead pattern for determinatives (Sommerer, 2012:204205). Following the establishment of the determinative slot, it turns into a chr('39')gravitational polechr('39') which attracts items (Krug, 2000). Actually, a syntactic slot is regarded as a position which invites linguistic elements to grammaticalize there. This is also true for the indefinite article in OP; the emergence of the determination slot triggered the grammaticalization of the indefinite article. Therefore, chr('39')determination slotchr('39') and chr('39')indefinite articlechr('39') are inseparable notions. 6. ConclusionIn this research, the grammaticalization of the indefinite article in Old Persian is studied. The first bridging context which leads to the grammaticalization of indefinite article is the occurrence of numeral aiva chr('39')onechr('39') before human NPs. In simple words, the context is chr('39')There is only one Xchr('39'), here the X represents a human being. In this context, two readings of aiva are possible: a. there is one man, not more; b. the indefinite article reading, to introduce one person into discourse. In the latter reading the hearer reanalyzed aiva as an indefinite article based on metonymic relations. The reanalysis paves the way for the metaphoric extension from a numeral to a more grammaticalized item, namely an indefinite article. This conceptual shift/metaphoric extension becomes possible due to the fact that both chr('39')numeral onechr('39') and chr('39')indefinite articlechr('39') mean chr('39')one entitychr('39'). The difference is that the former means chr('39')onechr('39') and the latter chr('39')one out of the typechr('39'). The path of this grammaticalization is represented in (2):(2) Numeral one rarr; numeral one, indefinite article rarr; indefinite article