The purpose of the present study is to examine the use of cohesive devices, especially conjunctions, in oral narratives spoken by adult Korean EFL learners and compare the results with those by English native speakers in the same experimental setting. The data consist of a total of 83 narratives spoken by 43 Korean EFL learners and 40 English native speakers, who were each asked to tell a brief story based on a sequence of 12 pictures in accordance with a given title. Our hypothesis is that learners will have difficulty in using cohesive devices such as conjunctions appropriately; More specifically, learners will underuse and misuse conjunctions in discourse due to the lack of their language proficiency. The results show that except for differences in some of the conjunctions, the learners were strikingly similar to the native speakers in the inventory and distribution of conjunctions they used, although there were some misused errors. The learners had a great tendency to overuse coordinate conjunctions, especially but and so, whereas they tended to underuse temporal subordinate conjunctions such as while and as. However, the overuse of initial subordinate clauses by the learners suggests that they haven not quite acquired the uses of conjunctions in discourse.

This paper aims to give a preliminary proposal to DRT to represent the universal and existential readings of donkey sentences within the framework of Kamp and Reyle (1993). When only EVERY DRS construction rule is applied to donkey sentences without any universal effect, existential readings as basic and universal readings as additional can be obtained from them. Meanwhile, whenever universal effect comes out of the context, UNIVERSAL EFFECT DRS construction rule I propose in this paper should be applied to the donkey pronoun. The result is that only universal readings will be obtained. Though some donkey sentences have only existential readings and others only universal readings, the typical donkey sentences are ambiguous for two readings. In this respect, along with Chierchia (1995), I presume that existential readings are basic and universal readings are additional. I try to seek an adequate proposal in this paper, on the basis that universal readings are stronger, in the sense that they entail the corresponding existential readings.

Young-Soon Kim. 2001. Discourse Functional Interpretation of the Word Order Variation. Studies Modern Grammar 24, 137-159. In many linguistic studies there has been a dichotomy between formal and functional approaches. This paper will attempt to connect the two approaches because linguistic expressions always simultaneously have formal grammatical properties and functional pragmatic properties. In order to do it, this study explains the concept of the discourse-grammar, demonstrates the discourse-grammar relationship of Korean, German, and English, and deals with the motivation of word order variation in English. There are expressions of discourse functions in English, for example, the left attachment and the right attachment. The former has the interpretation of marked topicalization and the latter has the effect of creating a new focus of statement. The outline of results in this study is as follows: Korean and German are pragmatical languages, which make use of grammatical means to cope with discourse-functional requirements in quite a different way from English.

Lee, Bong-seon. 1998. A Semantic Analyisis of the Korean Discourse Marker `Keosigi. Studies in Modern Grammar 13, 197-212. The purpose of this paper is to give a pragmatic description of the Korean discourse marker `keosigi`, that is, how a speaker uses it for his purpose and how a hearer interprets it in discourse. The conclusion of this study is summarized as follows; First, the charateristics and the definition of the Korean discourse marker are given. Second, the relevance theory is suggested in which all the pragmatic uses of the discourse marker are inferred. Third, some discourse properties of `keosigi` are examined in accordance with relevance in the context where it is used.

Lee Bok-hee. 1998. Coherence and Relevance Principle in Discourse Interpretation. Studies in Modern Grammar 13, 181-196. Many coherence theorists set themselves two related goals: (a) to provide a theory of comprehension, explaining how discourses are understood; (b) to provide a theory of evaluation or appropriateness. Goal (a) is shared with relevance theory, which aims to provide a theory of comprehension. However, relevance theory has no explicit goal equivalent to (b). Giora`s interest is in goal (b). In particular, she suggests that relevance theory sheds no light on two main types of intuition: (a) local intuitions about when two adjacent discourse segments are related to each other or to a common topic; (b) global intuitions about when a text or discourse hangs together as a unit, rather than constituting a series of isolated remarks. This is the substantive issue in this paper I want to discuss.

Sun Ok Park. 1998. Differences in Discourse Functions of Modal Auxiliaries and Quasi-Modals. Studies in Modern Grammar 13, 107-131. The main concern of this study is to provide the English learners with the differences in discourse functions of will and be going to and suggest the differences can be applied to other modal auxiliaries and quasi-modals. Will has hypothetical condition and the speaker reaches his or her subjective prediction using the background experience and common sense. Be going to has present indication for the future event, so the speaker reaches his or her prediction based on the actual condition. The modal auxiliaries, must and can also have the speaker`s subjectivity while the corresponding quasi-modals, have to and be able to don`t.

Chung Moo-Joo. 1998. Personal Pronouns and Their Speech Roles. Studies in Modern Grammatical Theories 12, 169-186. The referents of the 1st and 2nd person pronouns are not fixed, but shift according to the situation, as participants take turns to speak. This paper was written to describe a wide variety of social and political roles and stances of these inter-personal pronouns. The actual discourse referents for use are seemingly limitless its precise interpretation is dependent on the particular context of use. `Inclusive we` ostensibly implies joint activity; but in different contexts there is actually a greater pull either towards egocentricity [+ego] or a vocative function [+voc]. The most striking example of egocentric reference is the so-called `royal We` and `editorial use`, both representing the authority of the speaker. The `inclusive we` with its addressee role is used when the speakers feel solidarity with the subordinate addressees. Impersonal pronoun one, recently, is used as a socio- linguisticaly marked form of I among the upper class and royalty and their hangers-on. To refer to the addressee, the third person forms can be used, implying deference, distance, or intimacy. On the evidence of medieval and Early Modern English texts the polite form you was used to address a singular social superior and also a social equal among the upper class, leaving thou for address to a social inferior and for reciprocal address among the lower classes. There was, however, considerable fluctuation between you and thou in address to the same individual, even within the same utterance. This code-switching of you and thou frequently signals expressive shifts of feeling that are not easily explained by a power semantic.