화자부인(話者否認)자질과 가부정(假否定)간의 국소성관계에 대하여

최현숙(Choe Hyon Sook) 한국생성문법학회 2015 생성문법연구 Vol.25 No.4

Choe (2015a) suggests that while the negative marker -ci {an-h/ani-ha}- or an(i) can make a clause negative, it can also make a clause non-negative when the sentence-ending marker (= X) bears a lexical feature [Ds] (=[Speaker's Denial]), whose function is to deny the value of the feature [negative] of the negative marker. In Choe 2015a, when the negative marker loses the value of the feature [negative] that it has, it is called nNEG, while it does not, it is called NEG. In this paper, assuming the suggestion in Choe 2015a, I suggest the following: First, nNEG and NEG are lexically different (see also D-S. Kim 1981), and nNEG is defective in that its feature value should be denied by the feature [Ds] in order for it to be properly licensed. Second, X[Ds] can license nNEG across T and/or will, but not across (i) nNEG/NEG or focus markers such as "-to('even')" and "-(n)un" (contrastive focus marker), or (ii) aspectual elements or modal elements other than will (-keyss- or -(u)l-). The second suggestion implies that the licensing relation between X[Ds] and nNEG should be local in a certain sense, and as for the reason why the relation should be local, I suggest/speculate that the locality relation is derived because of the following reason: In Korean obeying the head movement constraint (the HMC), nNEG (a head) should move to the element with the feature [Ds] (i.e. X[Ds]) to be properly licensed, just as the nNEG element (= n't) in English should move to C via head movement.

On the (stacked) markers -{ka/lul}

Hyon Sook Choe 현대문법학회 2011 현대문법연구 Vol.65 No.-

In this paper, mainly based on the data with the constituent XP-man-i ('XP-only-F'), I discuss some distributional and functional properties of the markers -ka in Korean to suggest that the marker -ka that is stacked on the dative marker -eykey ('to') or on particles like -man ('only') (which is generally called stacked -ka) is not a case marker, but a focus marker with a special function, and that unstacked -ka may also function as a focus marker in certain contexts. I also discuss further data with the marker -lul to suggest the following: First, stacked -lul is a focus marker and unstacked -lul may function as a focus marker in certain contexts, like unstacked -ka. Second, there are two types of focus markers, which differ in their morphological realization patterns. Based on the discussions, I conclude that case duplication or case stacking is not allowed in Korean.

On the Ambiguity of (Negative) Interrogatives

Hyon Sook Choe 한국생성문법학회 2015 생성문법연구 Vol.25 No.1

In this paper, I discuss the following two semantic properties of interrogatives with NEG in Korean (and in English): First, they are interpreted either as (I) neutral negative interrogatives or as (II) non-neutral interrogatives where the negative meaning of NEG is considered to be lost. Second, type (II) non-neutral interrogatives can trigger either a question or a statement reading. By examining lexical/semantic properties of sentence-ending (SE) markers in Korean, which can be analyzed as MOOD (or as MOOD+COMP), I suggest that the loss of the negative meaning of NEG is triggered because some of the SE markers that are known as interrogative (INT) markers optionally bear the feature [Speaker's Denial] (=[Ds]), whose function is to deny the negative value of NEG in a way similar to how a NEG negates the negative value of another NEG in a sentence. By suggesting that SE markers that can have the feature [Ds] may be ambiguous between DEC and INT in a certain way, I attempt to explain the second semantic property. During the discussion, I suggest that some SE markers are many-way ambiguous and that there are at least three subclasses of interrogatives/declaratives: (neutral) negative interrogatives/declaratives (I), (non-neutral) interrogatives/declaratives with Speaker's Denial (II), and (non-neutral) interrogatives/declarative with Speaker's Assertion (III). Finally, I discuss empirical advantages of the present approach, and also examine some related issues discussed in the literature, which are properly explained under the present approach.

Reconstruction and Binding in the Minimalist Program: With Special Reference to Conditions A and C

Hyon Sook Choe 한국언어학회 2003 언어 Vol.28 No.3

On Some Distributional Differences between Shortened NEG and Unshortened NEG in Korean

Hyon Sook Choe(최현숙) 한국언어학회 2017 언어 Vol.42 No.4

In the Korean literature, it has been noted or assumed that so-called short-form(SF) and long-form(LF) negations cannot appear in imperatives and exhortatives. In this paper, I discuss the distributional properties of the four varieties of NEG in Korean (i.e., (i) ani-ha-, (ii) an-h-, (iii) ani, and (iv) an) by examining negative imperatives/exhortatives which employ more than one morphological variety of the imperative/exhortative markers. Based on empirical data, I suggest the following: Frist, all the four varieties can appear in imperatives/exhortatives in principle, contrary to what Han and Lee (2007) would suggest. Second, while the two unshortened varieties (= (i,iii)) may not be prohibited in imperatives/exhortatives, making sentences sound old-fashioned, the distribution of the two shortened varieties (= (ii,iv)) may be restricted because of some co-occurrence restrictions they should respect. Third, the distributional difference between (i,iii) and (ii,iv) can be attributed to the diachronic change from (i,iii) to (ii,iv), which partly triggers a tendency for negative imperatives/exhortatives to be replaced by mal-imperatives/exhortatives.

The (Non-)Specificity of the (Null) Degree Word in Korean Comparatives

Hyon Sook Choe 현대문법학회 2013 현대문법연구 Vol.71 No.-

In this paper, I discuss a semantic aspect of the null degree word/operator in Korean clausal and phrasal comparatives to suggest the following: First, the null degree word in Korean comparatives can be replaced with overt deictic degree words (cf. Choe (2012)). Second, -pota may license a generic or a specific phrase, but not a non-specific phrase. Third, while deictic degree words in Korean are ambiguous between specific and non-specific, the null (deictic) degree word in Korean comparatives is non-specific so that -pota may not come directly after the "stripped" compared constituent containing the (non-specific) degree word. During the discussion, to explain a certain cross-linguistic difference between Korean and English, I also suggest that the null degree word in English can be ambiguous between specific and non-specific, unlike the one in Korean, and that the null degree word of the "stripped" compared constituent should be specific in English, because than also cannot license a non-specific phrase.

On the Categorial Ambiguity of the Morpheme kes in Korean

Hyon Sook Choe 서울대학교 언어교육원 (구 서울대학교 어학연구소) 2007 語學硏究 Vol.43 No.2

In this paper, I discuss the ambiguous categorial status of the morpheme kes in Korean. The morpheme kes has been assumed to be a “dependent” or “defective” noun, because it cannot appear alone and its meaning is largely determined by discourse and/or syntactic contexts. Based on empirical data, I show that it may come either with a modifier or with a non-modifier, but exhibits different properties, depending on which it comes with: When it comes with a modifier, it has properties of a lexical N, but when it comes with a non-modifier, it does not have properties of a lexical category, but those of a functional category. As for the categorial status of the morpheme kes as a functional category, under the EP hypothesis in H S Choe (2006, 2007b), which suggests that a full realization of a nominal expression is a functional category projection above DP (called EP), I suggest that the morpheme kes is a realization of E, when it comes with a non-modifier. Under the present approach, I also attempt to characterize the nature of the syntactic and semantic dependency or defectiveness that the morpheme kes exhibits, and to explain some descriptive facts related to noun phrases in Korean.

Some Distributional Differences Between Adjectives and Verbs in Korean: A reply to Yeo (2004)

Choe, Hyon Sook 서울대학교 어학연구소 2005 語學硏究 Vol.41 No.2

'Restructuring' in Korean

Choe, Hyon Sook 서울대학교 어학연구소 1988 語學硏究 Vol.24 No.4

The Italian 'restructuring' phenomenon discussed by Rizzi (1982) and others raises an important theoretical problem on the status of categories in configurational structure, which is our main concern in this paper. We have two goals: First, based on Korean data containing a certain class of auxiliary verbs which trigger 'restructuring' in Italian, we show that 'restructuring' is universal and should be explained in terms of core grammar (UG). Second, we propose, under the principles-parameters theory of UG, that 'restructuring' effects are derived from V-to-V transformation we call RR, which is characterized as follows (under an assumption that categories and terminal strings are independent entities so that rules can affect either categories or terminal strings): (1) RR, which is triggered by a certain class of auxiliary predicates, affects categories. (2) V-to-V RR creats complex minimal/maximal projections in a certain sense: Minimal or maximal projections within an RR domain do not constitute inherent barriers or induce the Miimality Condition.

Agreement and Existential Sentences in English

Hyon Sook Choe 영남대학교 인문과학연구소 1994 人文硏究 Vol.16 No.1