用户名: 密码: 验证码:
《先祖阔尔库特书》形态句法研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
《先祖阔尔库特书(德累斯顿本)》是15世纪末16世纪初成书的一部突厥语口头文学作品。本文是对该文献语言的形态句法描写,主要包括词类、词组和短语、命题结构、限定标记和句子类型等几个部分。
     我们的描写以句子为基本单位,较少关注句子之间的连贯和衔接。句子按其结构特征可以分为简单句、并列句和主从句几种。主从句为讨论的重点。我们认为,只有以借词ki(m)引导的句子才属于从句,标准有二:第一,ki(m)从句具有自己的限定标记;第二,ki(m)从句是主句的一个句法成份。由形动词、动名词和副动词构成的主谓结构虽然也属于句子成份,但不能独立成句,故本文视之为句子降级结构而排除在从句之外。包括直接引语的句子是比较特殊的一种句型,它以独立句作为言语动词的补足成份嵌于句内,与以上几种句型均有区别。
     句子按功能分有陈述句和祈愿句两个主要类别。区别于一般的主谓二分,我们提出陈述句二分为命题和限定标记。其中,命题二分为主语和表语两个部分;限定标记则包括极性、直陈、语气、人称/数和情态五个部分。对静词句而言,极性范畴包括无标记的肯定和以degul表示的否定这两个成员;于动词句而言,极性范畴嵌于词内(因而不属于限定标记),包括无标记的肯定和以-mA-表示的否定两个成员。我们以“直陈”表示陈述与疑问的对立,前者为无标记形式,后者对是非疑问句而言以附缀mI标记。语气统指系词i-(<古突厥语ar-)的四种变化形式构成的聚合序列,另有无标记形式与其中的idi构成对立,分别表示现在时和过去时,其余三种形式(imis、ise和iken)分别表示传信、条件和时间。人称/数标记源于人称代词或领属词缀。句子最外层有无标记形式与附缀-dUr构成的对立,表示认知情态意义。
     命题结构是从句子成分关系的角度对句子进行的分析。以形动词短语作表语者为动词句,余为静词句。动词有不同题元结构,动词的题元为其主语和补语。补语一般被视为动词短语的内部结构,故在命题结构中未将补语列为与主、表并列的一种成分。状语无论从意义和形式上都更为多样。从意义上看,状语可以分为与命题相关的环境状语和与限定标记所表意义密切相关的语气状语两类。
     基于层级的观念,词组和短语是构成句子的直接成分。二者的区别在于,词组可以只有中心语而没有任何修饰限定成份,短语则必须有补足成分。作为句子成分者如果是一个词,我们视之为没有扩展成分的词组。与句子结构对应,词组/短语结构也是本文关注的内容之一。
     词不是句子的直接成分,但对词进行分门别类是对更高层级语言单位进行描写的必要条件。除了一般语法描写中必然提及的名词、动词、形容词等等主要词类外,本文特别将附缀列为词类的中的一项。附缀是介于自由语素(词)和粘着语素(后缀)中间的一种语言形式,它是很多语言都存在的一种现象,但是在过去的突厥语言描写中对此并未特别注意。事实上,附缀有很多不同于后缀的表现。附缀既有附着形式,也通常会有独立形式。作为附着形式,附缀和后缀类似,也会与其词基发生语音和谐。附缀与后缀的另一个不同之处在于它对词基的类别较少限制。此外,附缀不带重音——从现代语言的这一特点可以推断《先祖阔尔库特书》也应具有这一特点。
     在尝试对《先祖阔尔库特书》语言进行系统性描写的同时,本文特别突出对时、体范畴的讨论。与多数突厥语语法描写不同,我们认为动词只有体范畴而没有时范畴,时的辖域是整个命题而非动词。一般描写中的简单时形式是动词的不同体,属于形动词,换言之,是非限定动词形式。体分为完整体、非完整体和未然体三个大类。是为语法体。动词短语在其词汇-语义层面不同的时间特征属于词汇体的范畴,包括成就、达成、活动、瞬时、状态等成员。
     共时描写是本文的主要目的,但其潜在的用意是历时的。对不同时代语言的共时描写是进行历时观察的前提。基于这样的考虑,我们在讨论某些语言现象时偶尔会以现代语言(特别是土耳其语)的相关现象作为参照,试图揭示语言演化的趋势或规律。这样的比较涉及到文献语言与现代语言的亲疏关系问题,本文无意对此发表意见,但是希望通过尽量客观、准确、系统的描写,供感兴趣者结合特定的现代语言、通过比较作出自己的判断。
     结构-功能主义是本文的基本理论路向,涉及索绪尔、布拉格学派、语言类型学和认知语言学的一些基本概念,可以概括为:语言是一个形义结合的层级系统,同一范畴具有相同的组合-聚集关系(分布),范畴成员可以是无标记的,很多语法范畴属于典型范畴,一些语言现象具有跨语言共性,可以从认知的角度加以解释(如通过像似性原则)。对文献语言描写的过程中充分利用现代语言学的成果是本文的一个特点。
     本文的另外一个特点是借助计算语言学手段实现定性与定量的并重。我们提出并以Python语言实现了对突厥语具有较好适用性的“削尾与生成相结合”的自动形态分析算法,不仅大大提高了本项研究的效率,也可以为相关研究提供得力的工具。
This study is a morpho-syntactic description of The Book of Dede Korkut (Dresten version), a collection of epic tales passing from generation to generation in an oral form for centuries before the first written copies emerged (in the15th or16th centuries). The paper consists, among its major parts, of word classes, word groups and phrases, propositional structure, finiteness markers and sentence patterns.
     We take the sentence as the basic unit of our description, with much less attention paid to structures above sentences, i.e. coherence and cohesion. A sentence can be classified as simple, compound or complex according to its structural complexity. Much emphasis is placed on complex sentences, which include only those introduced by the loan word ki(m)-two criteria apply:such a sentence has its own finiteness markers, and is a component of the main clause. A subject-predicate structure headed by a verbal, nominal or adverbial adjective is not independent, though it also constitutes a part of the main clause-it is nothing but a downgraded sentence and so should be excluded from dependent clauses. Direct speech differs from the aforementioned sentence patterns in that it is independent but embedded inside a sentence as the complement of a speech verb.
     Declaratives and optatives are the major sentence types from a functional point of view. We argue for a two level bipartition of declarative sentences, between a proposition and a set of finiteness markers on the sentential level, and a subject and a predicate on the propositional level, as opposed to the traditional straightforward subject-predicate bipartition of sentence patterns. Finiteness is expressed by a syntagmatic sequence of polarity, indicative, mood, agreement and modality markers. For nominal sentences, polarity includes an unmarked member for positive and degul for negative meaning; for verbal sentences, this is embedded inside the verb (and therefore is not a part of the finiteness) with, again, an unmarked member for positives and-mA-for negatives.'Indicative', though not a satisfactory term, is here used to cover both declarative and interrogative meanings, with the former implicit and the latter explicit through the clitic ml where a yes-no type interrogative is in question. Mood is used here as a cover term for the paradigm of the four inflectional forms of copular i-(     Subjects and predicates reflect the relations between sentence components. Sentences with verbal adjective (i.e. participle) phrases as their predicates are called verbal, whereas others are nominal. A verb has a subject and zero or more complements in its argument structure. Complements are not listed in this study as distinct components on the same rank of subjects and predicates, as they are viewed as part of verbal phrases. Adjuncts show much diversity both in form and in meaning and can be divided roughly into two groups:circumstantial and modal. Circumstantial adjuncts are those closely related in meaning to the proposition, modal adjuncts to the finiteness.
     Word groups and phrases make the immediate constituents of sentences-a well established approach within the structuralist tradition. Describing a sentence as a construction of words is rather like describing a house as a construction of bricks, without recognizing the walls and the rooms as intermediate structural units.
     Classification of words is necessary for the description of structures on a higher rank within the language hierarchy. Beside nouns, verbs, adjectives etc., clitics are also discussed as an item of special importance on the traditional list. A clitic is a linguistic form half-way between autonomous words and affixes. Though quite common in a wide range of genealogically unrelated languages, clitics have been neglected to a certain extent in most descriptive studies of Turkic languages.
     We have placed much emphasis on the discussion of tense and aspect categories. We argue that only aspect is a verbal category, while tense takes the proposition as a whole in its scope. The so called'simple tense'forms are believed to be verbal adjectives of different aspects,; in other words, they are not finite. Temporal meanings on the lexical-semantic level of verb phrases can be classified as lexical aspects (or, situation types), with such members as ACComplishments, ACHievements, ACTivities, SEMELfactives and STAtes.
     The underlying goal of this synchronic description is, not surprisingly, diachronic. Related forms of modern languages (especially those of Turkish) are taken into consideration with the aim of bringing to light the mechanism of change in relation to at least that part of language. This sort of observations necessarily involves judgment on the genealogical relations between the language of Dede Korkut and its modern relatives, a topic beyond the scope of this study. We believe, however, an objective, thorough and systematic description may help those interested draw their own conclusions in this respect.
     Underlying the presentation of the language system is a structuralist-funtionalist approach. Scattered throughout this dissertation are basic ideas and concepts of F. de Saussure, the Prague School, M.A.K. Halliday, linguistic typology and cognitive grammar-i.e. language is a hierarchical system of form-meaning combinations; a grammatical category is a set of linguistic elements of the same syntagmatic and paradigmatic distribution; a member of a category can be unmarked; many grammatical categories are prototypes; some phenomena are cross-linguistically universal and can be explained from a cognitive point of view (e.g. iconicity).
     Use of NLP (Natural Language Processing) techniques not only improves the efficiency of this study, but also make quantitative observations easier. In addition to qualitative presentations, such observations are necessary for any thorough descriptions, either synchronic, diachronic or both. A'stripping plus generating'algorithm realized in Python is applied for morphological tagging and a KIC search program is developed. We believe this algorithm is also applicable to other Turkic languages and thus makes a promising by-product of this thesis.
引文
[1]Adamovic, Milan. (2001) Zur Herkunft der Partikel da/da. Materialia Turcica,22, s.115-120.
    [2]Akalin, Sukru Haluk. Dede Korkut Kitabinde Gecen bir Tumce Turu Uzerine. TDAY-B 1998/Ⅰ-Ⅱ.
    [3]Akar, Ali (2008) Dede Korkut Kitabiinda tur- Fiili. Turkish Studies, vol.3.1.
    [4]Ata, Aysu. Dede Korkut Hikayelerinde Supheli Birkac Kelime Uzerine Dusunceler. TDAY-B 1998/Ⅰ-Ⅱ.
    [5]Basdas, Cahit (2008) Turkcede iyelik-Yukleme Sorunu Dede Korkut Ornegi. Turkish Studies, vol. 3/1.
    [6]Blake, Barry J. (2005[2001]) Case. Beijing: Pekin University Press.
    [7]Buran, Ahmet. (1994) -Ik Ekinin Anadolu Agizlannda Kullanihsi.
    [8]Cemiloglu, Ismet. (2001) Dede Korkut Hikayeleri Uzerinde Soz Dizimi Bakimindan Bir inceleme. Ankara: Turk Dil Kurumu Yayinlan.
    [9]Chafe, W. (1976) Givenness, contrastiveness, definiteness, subjects, and topics. In Li (ed.) Subject and topic. New York: Academic Press., 25-56.
    [10]Comri, B. (2005[1976]) Aspect. Beijing: Beijing University Press.
    [11]Croft, William. (2000[1990]). Typology and Universals. Beijing: FLTRP.
    [12]Dasdemir, Muharrem. (2000) Dede Korkut Kitabinin Soz Dizimi. Yayinlanmamis, Doktora Tezi. Ataturk Universitesi Sosyal Bilimler Enstitusu Turk Dili ve Edebiyati Anabilim Dali.
    [13]Demirtas, Ahmet (2008) Dede Korkut Hikayelerinde "Gerek" Kelimesiyle Kurulu Cumleler. Turkish Studies, Vol. 3/1.
    [14]DLT = Atalay, Besim (1985) Divanu Lugat-it-Turk Tercumesi,Ⅰ,Ⅱ,Ⅲ. Ankara: TDK Yayinlan.
    [15](?)lizad(?), Sam(?)t (2004) Kitabi-D(?)d(?) Qorqud, (?)sil v(?) sad(?)l(?)sdirilmis m(?)tnl(?)r. Baki: Ond(?)r N(?)sriyyat.
    [16](?)lizade, Sam(?)t. Azerbaycanda Kitabi Dede Gorgud'un Elmi Metninin Nesri Problemlerine Dair. TDAY-B 1998/Ⅰ-Ⅱ.
    [17]Ercilasun, Ahmet B. (1984) Kutadgu Bilig Grameri -Fiil-. Ankara: Gazi Universitesi Basm Yayin Yuksekokulu Basimevi.
    [18]Ercilasun, Ahmet B. 1994. Dede Korkut Kitabi ile Oguz Destani Arasmdaki Munasebetler. TDAY-B 1988, s.69-89.
    [19]Ercilasun, Ahmet B. ed. (2007) Turk Lehceleri Grameri. Ankara: Akcag.
    [20]Ercilasun, Ahmet B. Zikredilen Sahislara Dayamlarak Dede Korkut Boylarinin Kronolojik Siralanmasi. TDAY-B 1998/Ⅰ-Ⅱ.
    [21]Ercilasun, Ahmet B., 2007. Baslangigtan Yirminci Ytizyila Turk Dili Tarihi, 4.haskt. Ankara: Akcag Yayinlan.
    [22]Erdal, Marcel. (2004)A Grammar of Old Turkic. Leiden-Boston: Brill.
    [23]Ergin, Muharrem. (2008[1958]) Dede Korkut Kitabi - 1. Ankara: TDK Yayinlan.
    [24]Ergin, Muharrem (2009[1963])Dede Korkut Kitabi - 2. Ankara: TDK Yayinlan.
    [25]Eryigit, Gulsen & Adali, E§ref. 2010. An affix stripping morphological analyzer for Turkish, Proceedings of the LASTED International Conference ARTIFICIAL INTELLIGENE AND APPLICATIONS, February 16-18, 2004, Innsbruck, Austria.
    [26]Goksel, Ash & Celia Kerslake. 2005. Turkish: A Comprehensive Grammar. London: Routledge.
    [27]Goksel, Asli & Kerslake, Celia (2005) Turkish: A Comprehensive Grammar. Oxon: Routledge.
    [28]Gokyay, Orhan Saik. (1985[1976]). Dede Korkut Hikayeleri, 3. baski. Istanbul: Dergah Yaymlan.
    [29]Gokyay, Orhan Saik. (1994) Dede Korkut Kitabi Uzerine. TDAY-B 1988, s.91-96.
    [30]Haciyev, Asif. (2007)D(?)d(?) Qorqud Kitabt: Oxunuslar, Actmlar. Baki: Elm.
    [31]Haciyev, Tofiq ed. (2004a) Dddd Qorqud Dunyasi, Ensiklopedik Luget. Baki: Ond(?)r N(?)sriyyat.
    [32]Haciyev, Tofiq ed. (2004b) D(?)d(?) Qorqud Dtinyasi, Mdqal(?)l(?)r. Baki: Ond(?)r N(?)sriyyat.
    [33]Haciyev, Tofiq. (1994) "Dede Gorgud Kitabi" Dilinin Ikin Sintaktik Hususiyyeti Hagginda. TDAY-B 1988, s. 23-33.
    [34]Haiman, John (1978) Conditional Are Topics. Language, Vol. 54, No. 3 (Sep., 1978), pp. 564-589.
    [35]Haiman, John (2009[1985]) Natural Syntax:Iconicity and Erosion. Beijing:Word Publishing Corporation.
    [36]Halliday, M.A.K. (2000[1994]). An Introduction to Functional Grammar. Beijing: FLTRP.
    [37]Hankamer, J. Finite state morphology and left to right phonology, Proceedings of the Fifth West Coast Conference on Formal Linguistics, Stanford, CA,1986,29-34.
    [38]Hopper, Paul J.&Traugott, E.C. (2001 [1993]). Grammaticalization. Beijing:FLTRP.
    [39]Kacalin, Mustafa S. (2006) Oguzlarm Diliyle Dedem Korkudun Kitabi. Istanbul: Kitabevi.
    [40]Kacalin, Mustafa S. Dede Korkut Kitabi'nda Okuma Onerileri. TDAY-B 1998/Ⅰ-Ⅱ.
    [41]Kara, Mehmet.2005. Turkmen Turkcesi Grameri. Ankara:Gazi Kitabevi.
    [42]Keenan, Edward and Bernard Comrie (1977) Noun phrase accessibility and universal grammar. Linguistic Inquiry.8:36-99.
    [43]Korkmaz, Zeynep.1995. Turk Dili Uzerine Arastirmalart. Ankara:TDK Yayinlan.
    [44]Kornfilt, Jaklin. (1996) On Some Copular Clitics in Turkish. In:A.Alexiadou et al. (eds.) ZAS Papers in Linguistics, vol.6, Berlin, Zentrum fur Allgemeine Sprachwissenschaft.
    [45]Langacker, Ronald W. (2008) Cognitive Grammar:A Basic Introduction. Oxford: Oxford University Press.
    [45]Lewis, Geoffrey. (1974) The Book of Dede Korkut. Maryland:Penguin Books.
    [47]Liu Zhao (2008) Turkce Bildirim Tumcelerinde Yuklem Yapilanist. Yayinlanmamis. Yukseklisan Tezi. Ankara Universitesi.
    [48]Lyons, John (2000[1995]) Linguistic Semantics:An Introduction. Beijing:FLTRP.
    [49]Mansuroglu, Mecdut. (1998) Eski Osmanhca. In:Mehmet Aklin ed. Tarihi Turk Siveleri. Ankara:Ankara Universitesi Basimevi.
    [50]Oflazer, K. Two-level Description of Turkish Morphology, Literary and Linguistic Computing,9(2),1994,137-148.
    [51]OTWF= Erdal Marcel (1991) Old Turkic Word Formation. Wiesbaden:Otto Harrassowitz.
    [52]Ozcelik, Saadettin.2009. Dede Korkut Metinlerin Uzerine Tespitler (1):Dede Korkut'ta Cozulememis bir Ibare. Turk Dunyasi incelemeleri Dergisi, Cilt:Ⅸ, Sayi:1, Sayfa:145-51. Izmir.
    [53]Rentzsch, Julian. (2010). Why Turkic DI is not [+PAST]. In:Matthias Kappler, Mark Kirchner and Peter Zieme eds. Trans-Turkic Studis:Festschrift in Honour of Marcel Erdal. Istanbul:Urun Yaymlan. pp.267-279.
    [54]Sertkaya, Osman Fikri.1994. Dede Korkut Kitabi'nin Bazi Manzum Parcalarm Hece Vezni ve Manzume Turu Acisindan Degerlendirilerek Yeniden Okunmasi. TDAY-B 1988, s.141-156.
    [55]Sertkaya, Osman Fikri.1994. Dede Korkut Kitabi'nin Dresden Yazmasindaki Bazi Kelimeler ve Sekillerin Imla Ozelliklerine Dayanilarak Yeniden Okunusu ve Anlamlandirilisi Uzerine. TDAY-B 1998/Ⅰ-Ⅱ.
    [56]Solak, A.& Oflazer, K. Design and Implementation of a Spelling Checker for Turkish, Literary and Linguistic Computing,8(3),1993,113-130.
    [57]Sumer, Faruk and Ahmet E. Uysal and Warren S. Walker (1972) The Book of Dede Korkut, A Turkish Epic. Austin and London:University of Texas Press.
    [58]Taylor, John R. (2001[1995]) Linguistic Categorizaion:Prototypes in Linguistic Theory. Beijing:FLTRP.
    [59]TDK(Turk Dil Kurumu) (2006) Karsilastirmah Turk Lehceleri Grameri, Ⅰ,-Fiil-. Ankara: TDK Yayinlan.
    [60]Tekin, Talat (1986) Dede Korkut Hikayelerinde Bazi Duzeltmeler. TDAY-B 1982-1983, s.141-161.
    [61]Tezcan, Semih & Boeschoten, Hendrik. (2001) Dede Korkut Oguznameleri. Istanbul: Yapi Kredi Yaymlan.
    [62]Tezcan, Semih. (2001) Dede Korkut Oguznameleri Uzerine Notlar. Istanbul:Yapi Kredi Yayinlan.
    [63]Tural, Giizin (2000) Dede Korkut Hikayelerinde Zarf-Fiiller. Alev Kahya-Birgul & Aysu Simsik-Canpolat (eds.) Uluslararasi Dede Korkut Bilgi Soleni, Ankara,19-21 Ekim, 1999. Ankara:Ataturk Kultur Merkezi Baskanligi.
    [64]Ustunova, Kerime (1998) Dede Korkut Destanlan ve Sentaktik Paralelizmde Eksiltili Yapilann Rolii. TDAY-B 1998/Ⅰ-Ⅲ
    [65]奥托·叶斯柏森(2009[1924])语法哲学.北京:商务印书馆.
    [66]巴斯卡科夫,H.A.(2004[1981]).阿尔泰语系语言及其研究.陈伟周建奇译.呼和浩特:内蒙古教育出版社.
    [67]布龙菲尔德(2004[1955])语言论.袁家骅,赵世开,甘世福译.北京:商务印书馆.
    [68]戴维·克里斯特尔(2007[1997]).现代语言学词典.沈家煊译.北京:商务印书馆.
    [69]冯.加班(2004[1950])古代突厥语语法.耿世民译.呼和浩特:内蒙古教育出版社.
    [70]冯志伟(2010)自然语言处理的形式模型.合肥:中国科技大学出版社.
    [71]陆丙甫(2001)从宾语标记的分布看语言类型学的功能分析.载:《当代语言学》第三卷2001年第四期253-263页,北京.
    [72]陆俭明(2005)现代汉语语法研究教程(第三版).北京:北京大学出版社.
    [73]尚新(2007)英汉体范畴对比研究——语法体的内部对立与中立化.上海:上海人民出版社.
    [74]石毓智(2004)汉语语法化的历程——形态句法发展的动因和机制.北京:北京大学出版社.
    [75]张定京(2004)现代哈萨克语实用语法.北京:中央民族大学出版社.
    [76]赵明鸣(2001)《突厥语词典》语言研究.北京:中央民族大学出版社.
    [77]赵相如,朱志宁(2008)维吾尔语简志.《中国少数民族语言简志丛书(修订本卷伍)》.北京:民族出版社.
    [78]朱拉斯凯和马丁.2010.语音和语言处理(英文版第2版).北京:人民邮电出版社.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700