Evidentials in the Shuri Dialect of Luchuan (Ryukyuan)

Abstract

This dissertation attempts to demonstrate that Luchuan is a language which has a grammatical category of evidentiality. Luchuan is the only sister language of Japanese; however, since they are considered to have diverged as early as the beginning of the 8th century, their vocabularies and grammars have developed in somewhat different directions. The grammatical category of evidentiality is one of the categories which Luchuan has developed in different ways from Japanese. Evidentiality is a linguistic category which marks source of information. Evidentials have often been overlooked in the study of Luchuan, and the morphology has often been misanalysed as belonging to other grammatical categories because of failure to take account of the concept of evidentiality. On the one hand, some existing studies classify these evidential markers as temporal categories, such as tense or aspect, and on the other hand, some studies claim that these evidential markers should belong to the category of modality. I argue that both approaches have failed to describe Luchuan grammar accurately, and that the concept of evidentiality can resolve problems which other existing approaches could not deal with adequately. The main purpose of this study is to analyze the evidentials in Luchuan systematically and to formulate a model of the evidential system of this language. The proposal I make and defend in this dissertation is that Luchuan has a grammatical evidential system which contains one direct evidential and three indirect evidentials (Inference, Assumed, and Reportative). I argue that these four evidential markers should be considered to belong to a single grammatical category. Further, I discuss the relation between evidentials and other grammatical categories such as tense, aspect, and modality. It is obvious that these categories are closely related in Luchuan, but I attempt to tease apart their functions as clearly as possible. The Direct evidential -N is used when the speaker has 'direct evidence‘, such as direct experience or direct perceptions. When direct evidence is unavailable but the best possible source of information‘ is available, such as a report from a participant in the event, the use of -N is licensed. The Inferential evidential tee is used when the speaker makes an inference based on direct evidence. The Assumed evidential hazi is different from the Inferential evidential in that it does not require direct evidence, but the speaker‘s assumption has to be based on knowledge of habits or general knowledge. Finally, Reportative evidential Ndi indicates that the speaker acquired information from someone else, mainly orally report but in any case through the use of language. Luchuan is an endangered or moribund language which has very limited native speakers. Therefore, the principal focus of this dissertation is a descriptive study of verb forms whose syntactic features have not been fully described: for example, I set out whether or not each evidential can appear in negative or interrogative sentences, whether or not each evidential can have a non-past or a past form, whether or not each evidential markers can co-occur a subject of any person. Although my primary focus is a description of the evidential system of this language; at the same time, I relate my discussion to cross-linguistic issues such as how evidentiality is related to epistemic modality, with the intention that this work should constitute a contribution to the typological and theoretical study of evidentiality. I propose that evidentiality should be distinguished from the category of modality because in Luchuan the Direct evidential and the Reportative evidential belong to the category of evidentiality, though the other two indirect evidentials ― the Inferential and Assumed― might be regarded as an overlap category between evidentiality and modality.