A contrastive analysis of syntactic anaphora in Croatian and Italian

Batinić Angster, Mia. (2017). A contrastive analysis of syntactic anaphora in Croatian and Italian. PhD Thesis. Filozofski fakultet u Zagrebu, Department of Linguistics.
(Poslijediplomski doktorski studij lingvistike) [mentor Peša Matracki, Ivica and Kresić, Marijana].

Preview

PDF (Croatian) Download (3MB) | Preview

Abstract

This thesis analyses the syntactic anaphora in Croatian and Italian in a contrastive perspective. Anaphora can be generally defined as the reference of one expression to another one already mentioned in discourse (cfr. Safir 2004a: 4). Anaphora has been studied from the perspective of syntax and semantics, linguistic typology, pragmatics, textual linguistics and discourse analysis, computational linguistics, language processing and language acquisition. There are two basic types of anaphora: intrasentential and discourse anaphora. The aim of the present study is the analysis of intrasentential or syntactic anaphora, i.e. binding relations in Croatian and Italian language. A syntactic anaphor is a nominal expression that has no referential value on its own and as such its interpretation is determined by the interpretation of another element expressed in the same sentence (cfr. Huang 2000: 1, Safir 2013: 517). Syntactic anaphors are intrinsically dependent forms and must therefore have an antecedent in the sentence (cfr. Safir 2004a: 9, 2013: 517, Giorgi 109). Anaphora is the term indicating the relation between an anaphor and its antecedent. Syntactic anaphora attracted in the 1970s the attention of scholars (Chomsky 1973, Reinhart 1976). In the ’80s has become a central topic in generative grammar and the analysis of anaphoric relations has frequently influenced the development of syntactic theory. The thesis consists of six chapters. In the introductory chapter, the basic terminology related to the phenomenon of anaphora is explained allowing the author to situate her study in the context of existent theories. In the second chapter, the basic research questions concerning anaphoric dependencies are presented: the distinction between anaphors, pronominals and R-expressions, typology of anaphors and their characteristics. The following chapter presents a review of the theoretical background: after an extensive survey of standard binding theory, different contemporary approaches to anaphora within the generative framework are discussed. In the fourth chapter, we present the object and the aims of the present study, its theoretical approach and basic hypotheses as well as the methodology used. The object of the analysis are nominal reflexives divided in two classes according to their function: a) possessive syntactic anaphors: cro. svoj, vlastit vs. it. proprio, b) personal syntactic anaphors: cro. sebe, samog sebe vs. it. sé, se stesso. In this thesis Ken Safir’s (2004a, 2004b) approach is adopted: according to Safir the complementary distribution of anaphors and pronominals, although not universal, cannot be accidental for it is attested in a wide variety of languages. He assumes it is the result of the competition between anaphors and pronominals in taking dependent or bound interpretation. According to him different forms (anaphors, nominals and R-expressions) are ranked on a scale on which anaphors as the most dependent forms and they thus outcompete the others in being dependent on a given antecedent in a specific syntactic context. Pronouns are excluded if syntactic anaphors are possible, and this is regulated by the Form to Interpretation Principle (FTIP). In this thesis the existence of another principle is also assumed: the principle of interpretation of the syntactic context, which makes specific expressions available for competition. It is argued that possessive expressions are available in the possessive context, and the context of personal retrieval is necessary for the availability of personal expressions alone, while all other nominal expressions are excluded in that context. This principle assures that possessive and personal anaphors cannot compete with each other. In the analysis of these expressions the author hypothesises that the distributional differences between the correspondent elements in the two languages are due mostly to the differences in their morphosyntactic or φ-features (cfr. Burzio 1989, 1991, 1994, in press), to the fact that they belong to different types of syntactic anaphors (pronominal or relational anaphor) (cfr. Safir 2004a), as well as to the fact that Croatian and Italian inventory of syntactic anaphors are not entirely superimposable. Since these two languages have different expressions available for a dependent reading in the same context (while in Croatian both svoj and vlastit are available in a possessive context, Italian has only one possessive anaphor – proprio), their most dependent scales differ as well. The study presented in this thesis is contrastive and the analysis of the syntactic distribution of anaphors is corpus-based. Two comparable Web corpora has been used: hrWaC v2.2 (Ljubešić & Klubička 2014) for Croatian and itWaC (Baroni et al. 2009) for Italian. The next two chapters constitute the central part of the thesis: the fifth chapter focusses on possessive anaphors while in the sixth chapter the analysis of personal anaphors is presented. Both chapters are structured in the same way. In the introductory part the concept of possession or the term “personal anaphor” is explained and the relevant possessive or personal anaphors are situated in the context of other possessive or personal expressions. In the next part the author contrasts reflexive and non-reflexive elements in both languages and highlights the particular behaviour of the reflexive ones. Subsequently the author analyses the status of anaphors and of reflexive elements on the basis of their intrinsic nature, their morphological and lexical properties, as well as the type of syntactic anaphor to which the specific expression belong, i.e. whether it is a pronominal (i.g. svoj) or relational anaphor (i.g. cro. vlastit, it. proprio). Once their anaphoric status is determined, the competitions into which these forms can enter can be predicted. The author presents the scales and some further hypotheses on the distributional properties of the anaphors taken into consideration. For example an expression can be available for the competition only if the syntactic context corresponds to its function or that in a competition between anaphors svoj and vlastit it is dependent on the atypical antecedent (e.g. in accusative or dative case), the relational anaphor will outcompete the pronominal one. The corpus-based analysis of the syntactic distribution of the anaphors in both languages is thus presented. The author considers their possible antecedents, their complementary distribution with pronouns, their orientation, domain and some distributional particularities. Theoretical explanations are proposed throughout the analysis. This thesis is therefore both descriptive and theoretical. Our findings show that the competitive theoretical approach applied in the thesis is adequate for the explanation of Croatian anaphoric dependencies, whilst not satisfying for the description of Italian anaphoric system. The problem arises when it seems that the complementarity between Italian anaphors and pronouns lacks and considering Italian proprio and sé are longdistance anaphors which therefore do not confirm the Local Antecedent licensing (LAL). Given that also Croatian svoj in contemporary use does not always necessarily outcompete nonreflexive possessive pronouns, at first sight it seemed that also Croatian would present a challenge for theoretical assumptions. In the discussion of Croatian examples, it turned out to be a phenomenon affected by the pragmatically conditioned shift of the point of view. In this thesis Croatian vlastit and samog sebe are established as syntactic anaphors and, along the other anaphors of Croatian, are described for the first time in a systematic way. This is also the first contrastive analysis of this phenomenon in Croatian and Italian languages that applies the same theoretical approach on both of them. Although Italian anaphors have been treated before in a systematic way (Giorgi 1982, 1984, 1990, 2004, 2006, Burzio 1989, 1991, 1994, in press), those studies were always based on the researchers’ intuitions, while the corpus-based analysis made us possible to observe some particularities that have not been considered before. In conclusion this thesis paved the way for the description of the full pattern of anaphora in two languages.

Actions (login required)

View Item