Can Korean -(n)un mark (contrastive) focus?

June 3, 2017 | Autor: Ilkyu Kim | Categoría: Korean linguistics, Information Structure
Share Embed


Descripción

Language Sciences 56 (2016) 105–117

Contents lists available at ScienceDirect

Language Sciences journal homepage: www.elsevier.com/locate/langsci

Can Korean -(n)un mark (contrastive) focus? Ilkyu Kim Department of English, Kangwon National University, Joongang-ro 346, Samcheok-si, Gangwon-do, 245-711, South Korea

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 March 2015 Received in revised form 12 April 2016 Accepted 15 April 2016

This paper investigates the relation between the grammar and the basic notions of information structure, focusing on Korean -(n)un and (contrastive) focus. It has been claimed that -(n)un marks contrastive focus (CF) in certain contexts. However, no consensus exists on in which context(s) it can mark CF. Moreover, the very claim that -(n)un is used to mark CF itself has been sometimes questioned and refuted. The purpose of this paper is to shed light on this issue by providing a new analysis of the relation between -(n)un and (contrastive) focus. Based on a language- and theory-independent definitions of focus, contrast, and CF, and a detailed investigation of the information structure of the relevant phenomena at issue, the main claim of this paper is that -(n)un marks relationally old information and thus cannot mark (contrastive) focus, which is necessarily relationally new by definition. How it is possible for -(n)un to mark relational givenness will be discussed based on the comparison of its inherent meaning and that of the Korean nominative marker -i/ka. Also, the standard understanding of the notion of relational givennessnewness will be critically examined in the course of unveiling the nature of the relation between -(n)un and (contrastive) focus. Ó 2016 Elsevier Ltd. All rights reserved.

Keywords: Information structure (Contrastive) focus (Contrastive) topic Relational oldness/newness Korean -(n)un

1. Introduction In information structure theory, one of the central issues is which basic category of information structure (e.g. topic, focus, contrast) is marked by which linguistic phenomenon (or device) in a certain language. For instance, Hungarian preverbal position, Korean -(n)un, and the so-called English B-accent (or rise-fall-rise contour) have often been claimed to mark contrastive focus (CF) (or identification focus), topic and contrastive topic (CT), and CT, respectively (e.g. É Kiss, 1998; C. Lee, 2007; Büring, 2003). What is often missing in the works dealing with this issue is consideration of what it really means for some linguistic device to “mark” some category of information structure. For example, when the English B-accent is argued to mark CT, it is not clear whether it has CT as its inherent denotation, or the meaning of CT is derived from some independent meaning of the contour which is not the CT meaning itself. In other words, it needs to be clarified whether the contour is directly related to CT or the relation is only indirect. Without a clear understanding of the exact relation between the B-accent and CT, it would be hard to say that we have understood the exact nature of the English B-accent in terms of its relation to information structure. In Korean, the particle -(n)un has been studied by numerous researchers with respect to its relation to the informationstructural notions of topic and contrast (e.g. H.-W. Choi, 1996; C.-h. Han, 1998; S. Choi, 2000; C. Lee, 2003, 2007; C. Oh, 2007; J. E. Kim, 2010). However, whether the relation between -(n)un and the two notions is direct or indirect has been rarely

E-mail address: [email protected]. http://dx.doi.org/10.1016/j.langsci.2016.04.002 0388-0001/Ó 2016 Elsevier Ltd. All rights reserved.

106

I. Kim / Language Sciences 56 (2016) 105–117

discussed, which resulted in only superficial understanding of the nature of the meaning of -(n)un and its relation to information structure.1 Recently, Kim (2015) has tried to solve this problem by delving into the nature of the relation between -(n)un and the two notions, topic and contrast. According to Author, the inherent meaning of -(n)un, contrary to the standard view, is not topichood or contrast; rather, its meaning is performative in nature, and what the speaker does with -(n)un is to impose salience on the referent of a -(n)un-marked element. From this perspective, topicality and contrast expressed by -(n)un are just interpretive effects derived from the interaction of the meaning of -(n)un and various syntactic, semantic, and pragmatic factors. Although not studied as much as the relation between -(n)un and (contrastive) topic, the relation between -(n)un and (contrastive) focus has also been discussed by previous studies, most of them arguing that the particle can mark CF (e.g. H.-W. Choi, 1996; C.-h. Han, 1998; K.-S. Choi, 2004; Y.-B. Kim, 2004; Y. Jun 2005, 2006). However, unlike the case of the relation between -(n)un and (contrastive) topic, whose existence, whether direct or indirect, is taken for granted by most, if not all, Korean linguists, it is controversial whether -(n)un is related to CF, that is, whether it can really “mark” CF. For instance, Lee (2003) explicitly opposes the view that -(n)un can mark (contrastive) focus and limits its function to marking (contrastive) topic. There could be many reasons for the disagreement, among which I think the following two problems are the main causes for it: first, no consensus exists on what (contrastive) focus is and the notion is understood and used differently from scholar to scholar, and second, there is a deeply rooted (but never justified) assumption that the particle -(n)un is and must be directly related to (or inherently expresses) some notion(s) of information structure such as topic, focus, CT, and CF. In this paper, I try to unveil the nature of the relation between -(n)un and (contrastive) focus by dealing with the two problems just mentioned above. Based on a language- and theory-independent definitions of focus, contrast, and CF, and a detailed investigation of the information structure of the relevant data at issue, the main claim will be that -(n)un marks relationally old information and thus cannot mark (contrastive) focus per se, which is necessarily relationally new by definition. In discussing the relation between -(n)un and (contrastive) focus, the paper also deals with two important theoretical issues: 1) the relation between relational givenness and salience imposed by -(n)un and 2) the limits of the standard view of relational givenness-newness, with which one cannot fully capture the information structure of a number of sentences. First, considering the inherent function of -(n)un, which is argued to impose salience on the referent of a -(n)un-marked element (Kim, 2015), the conclusion that -(n)un marks relationally given information leads to a theoretically interesting question: how is the function of imposing salience related to marking relational givenness? As far as I know, no previous studies have dealt with the connection between relational givenness and saliency, and this paper builds a bridge connecting the two concepts, thus further supporting the analyses proposed by Kim (2015) and this paper. Second, it will be pointed out that the widely accepted view of the notion of relational givenness-newness is not adequate to explain various linguistic data. The standard understanding of relational givenness-newness corresponds to the traditional notion of topic-comment, theme-rheme, etc., which divides a sentence into relationally given part (or topic or theme) and relationally new part (or comment or rheme). This limitation of allowing only one partitioning of a sentence in terms of its information structure will be shown to make it hard to capture the exact information structure of a number of sentences with or without -(n)un. The structure of the paper is as follows. In Section 2, I will introduce previous studies on the relation between -(n)un and CF and discuss their problems, especially focusing on Y. Jun’s (2005) analysis, which provides the most recent and detailed analysis of the relation between -(n)un and CF. In Section 3, I will provide definitions of focus, contrast, and CF, which are language- and theory-independent. In Section 4, I will propose the main claim of the paper that -(n)un divides information into relationally old and new parts by signaling that the element to which it is attached is relationally old. In doing so, I will discuss the two theoretical issues introduced above in detail. Finally, Section 5 concludes the paper. 2. Previous studies It has been often claimed that -(n)un can be used as a CF marker in Korean (e.g. H.-W. Choi, 1996; C.-h. Han, 1998; K.-S. Choi, 2004; Y.-B. Kim, 2004; Y. Jun 2005, 2006). However, as Y. Jun (2005) correctly points out, those who make this claim are not in agreement with respect to what CF is. For instance, H. W. Choi (1996), C.-h. Han (1998), and K.-S. Choi (2004) argue for the status of -(n)un as a CF marker based on the (assumed) exhaustivity or exclusivity conveyed by the particle, whereas Y.B. Kim (2004) characterizes CF in more detail with respect to its phonetic property, the type of implicature it generates, and its cognitive status. To be more specific, Y.-B. Kim claims that CFs marked by -(n)un are phonetically accented (with high pitch), generate indifference or uncertainty implicature, and are always cognitively active (or prominent). Pointing out the problems and limits of the definitions of CF proposed by these previous studies, Y. Jun (2005) argues that CF should be defined as what conveys focus-hood, exhaustivity, and scalar implicature. According to Jun’s analysis, -(n)un that is used contrastively can (and must) mark either CT or CF depending on whether the -(n)un-marked expression is topic or focus. The examples in (1) and (2) are the examples of CT and CF provided by Y. Jun in his (4) and (5).

1

See Kim (2015) for a comprehensive review of the previous studies on the relation between -(n)un and topic and contrast.

I. Kim / Language Sciences 56 (2016) 105–117

(1)

A:

B:

Bʹ:

(2)

A:

B:

Bʹ:

107

Bill-uy yetongsayng-tul-un mwe hay-ss-ni Bill-Gen sister-Plu-Nun what do-Past-Int ‘What did Bill’s sisters do?’ Bill-uy khun yetongsayng-un John-eykey khisu hay-ss-e Bill-Gen big sister-Nun John-Dat kiss do-Past-Dec ‘Bill’s big sister kissed John.’ Bill-uy khun yetongsayng-kwa cakun yetongsayng-un Bill-Gen big sister-and small sister-Nun John-eykey khisu hay-ss-e John-Dat kiss do-Past-Dec ‘Bill’s big sister and small sister kissed John.’ ne ton iss-ni? you money exist-Int ‘Do you have money?’ na tongcen-un iss-e I coin-Nun exist-Dec ‘I have coins.’ na tongcen-kwa cicen-un iss-e I coin-and bill-Nun exist-Dec ‘I have coins and bills.’

The topic of the sentence in (1B) is the referent of the -(n)un-marked element, that is, Bill’s big sister, whereas it is I in (2B). Their topicality is guaranteed by the previous contexts, namely (1A) and (2A).2 What is important here is that Y. Jun claims that the -(n)un-marked element in (2B) is focus because it is what is relationally new in the given context. According to him, what is relationally new in (2B) is that she has coins but not bills and this new information is conveyed by the NP, tongcen-un ‘coin-NUN’. As for the information-structural status of the predicate iss-e ‘exist-Dec’, he argues that it is relationally old since it has been already introduced in the previous utterance, that is, (2A) (Y. Jun (p. c.)). Why his analysis of (2B) is problematic will be discussed in detail in Section 4. Also, it will be ultimately claimed that the relationally new information in (2B) is not just the -(n)un-marked NP but the combination of the NP and the predicate iss-e ‘exist-Dec’. Regarding the exhaustivity and scalar implicature of CF, Y. Jun claims that they exist in (2B) but not in (1B). First, following Szabolci (1981), he bases his argument for the existence of exhaustivity in (2B) and its non-existence in (1B) on the assumption that (2B0 ) does not entail (2B) whereas (1B0 ) entails (1B). Putting aside the validity of the criterion based on the notion of entailment, it is hard to understand why only (1B0 ) but not (2B0 ) entails its counterpart. According to Y. Jun, (2B0 ) does not entail (2B) because (2B) means that I have only coins (but not bills). Note that if we apply the same logic to the relation between (1B) and (1B0 ), it turns out that (1B0 ) does not entail (1B) either. That is, when the speaker utters (1B) instead of (1B0 ), it is because either she believes that Bill’s little sister did not kiss John or she is not sure about whether Bill’s little sister kissed John or not. This contrastive/uncertainty implicature is exactly the same as the implicature generated by (2B), both of which are caused by the use of -(n)un.3,4 Thus, it is problematic to argue that only (1B0 ) entails (1B). Semantically, both (1B0 ) and (2B0 ) entail (1B) and (2B) respectively; and pragmatically, both (1B) and (2B) convey the same type of contrast/uncertainty. With regard to scalar implicature, Y. Jun claims, based on (1) and (2), that only CFs (but not CTs) generate scalar implicatures. That is, he attributes the alleged lack of scalar implicature in (1B) to the topichood of the -(n)un-marked phrase, while relating the scalar implicature generated by (2B) to the focus-hood of the -(n)un-marked element. However, it is not hard to find counterexamples to this claim. (3)

(In a situation where Tom is more important than John in that a fashion show cannot start without him,) A: What about Tom and John? Have they arrived yet? B: John-un tochak ha-yess-e John-Nun arrival do-Past-Dec ‘John arrived.’

2 With regard to what topic is, I follow the standard aboutness-based view of the notion, that is, what the (rest of the) sentence is about (e.g. Hockett, 1958; Reinhart, 1981; Lambrecht, 1994). Y. Jun (2005) also adopts this definition of topic. 3 Although Y. Jun states that (2B) means ‘I have only coins (but not bills)’, it is also possible that the sentence can generate an uncertainty implicature (i.e. ‘but I don’t know whether I have bills (in my wallet)’) depending on context. 4 One reviewer questioned whether it is possible to get the uncertainty/contrastive implicatures in (1B) and (2B) without -(n)un. According to my intuition, the acceptability of (1B) and (2B) (and (3B)) decreases significantly if -(n)un is left out. Putting aside the decreased acceptability, however, the implicature that would be generated is the same uncertainty/contrastive implicature as that generated when -(n)un is used, which supports the view that the uncertainty/contrastive implicature (argued to be) induced by -(n)un is conversational but not conventional in nature.

108

I. Kim / Language Sciences 56 (2016) 105–117

In the given context, (3B) induces a scalar implicature (i.e. ‘but not Tom’), the scalarity of which is made possible by the pragmatically established scale between Tom and John, even though the -(n)un-marked phrase, that is, John, is guaranteed to be the topic by the first sentence of (3A). Thus, whether or not scalar implicature is generated has nothing to do with whether a -(n)un-marked element is topic or focus. To sum up, exhaustivity and scalar implicature, contra Y. Jun, are not determined by focus-/topic-hood of a -(n)un-marked element. Rather, their (non-)existence must be explained by various pragmatic factors, to find out which is beyond the scope of this paper. 3. What are focus, contrast, and CF? Given that the definitions of CF provided by the previous studies are problematic, it is necessary to provide a more stringent definition of CF before discussing the issue of whether -(n)un can mark (contrastive) focus or not. In this section, I will provide a definition of CF based on language- and theory-independent definitions of focus and contrast. 3.1. What is focus? Focus has been understood in many different ways, but the two most influential approaches are to see it as what is relationally new and what indicates alternatives. First, within a group of functionally motivated theorists, focus is defined as what is relationally new (e.g. Halliday, 1967; Lambrecht, 1994; Vallduví and Engdahl, 1996; Gundel and Fretheim, 2004; Y. Jun 2005, 2006). Second, in the alternative semantics theory, one of the most popular formal semantic theories on focus, it is characterized as what indicates alternatives (e.g. Rooth, 1985, 1992; Krifka, 2007). In this paper, I will adopt the definition of focus proposed by the functionalists. There are several reasons for this, and let me briefly explain why I prefer the functionalist definition to the one proposed by the alternative semantics.5 First, the definition from the functionalist approach better captures our intuition than that from the alternative semantics. As pointed out by Umbach (2004) and Mati c (2009), the definition based on relational newness seems more intuitively appealing.6 The reason is that a typical example of focus, an answer to a wh-question, can best be captured by the pragmatic notion of relational givenness-newness. The notion of indicating alternatives is intuitively less attractive, because, by answering a typical wh-question, what the speaker tries to do is to propose one of the alternatives as the correct answer but not to indicate (possibly) wrong “alternative answers”. Second, the meaning of focus proposed by the alternative semantics is, unlike the functionalist definition, motivated theory-internally. Notice that it is very closely related to the meaning of a question proposed by Hamblin (1973), according to whom the meaning of a question is a set of alternative answers. Adopting this view of questions and also assuming that every focus-containing utterance is an answer to an (implicit or explicit) question, Rooth (1985: 14) claims that “the function of the focus . is to signal that alternatives [the meaning of a question to which the focus-containing sentence is an answer] are indeed under consideration”. To put it simply, the meaning of focus is understood as to indicate a counterpart question. For instance, the focus in (4a) indicates that the utterance serves as an answer to the question in (4b). (4)

a. [John]F likes Kim. b. Who likes Kim?

This understanding of focus is motivated only when Hamblin’s theory of the meaning of questions is adopted. However, as noted by Beaver and Clark (2008: 27), “questions do not have to be analyzed using alternatives at all”. In fact, another standard way of dealing with questions in formal semantics and pragmatics is to treat them as properties, which is, according to Beaver & Clark, intuitively more direct (e.g. Scha, 1983; Hausser, 1983; Krifka, 2001, 2004; Ginzburg, 1995a,b; Ginzburg and Sag, 2000). Moreover, despite the fact that the theory-internally motivated definition of focus can be useful in accounting for various linguistic phenomena such as question–answer congruence, reconstruction of elided VPs, focus on contrastive pairs, focustriggered implicature, and the focus sensitivity of expressions like only and even, it is not without theoretical problems (Krifka, 2001; Beaver and Clark, 2008; Wee, 2010). Rather than re-introducing the problems discussed by previous studies, I will introduce a problem that has been ignored by most researchers. The problem is that, in principle, not only focus but topic can also be understood to indicate alternatives. Under the assumption that an assertion is an answer to an implicit or explicit question (Beaver and Clark, 2008; Roberts, 2012), the “topical meaning” (as opposed to the ordinary meaning) of a topic-containing sentence can be taken to indicate a set of propositions each of which contains an alternative topic. Logically, this makes perfect sense, because it is totally

5 As one reviewer correctly points out, “the issue is too complex to be done away with on a page and a half: one would need a whole paper to analyze the problems of the alternative semantics in sufficient depth”. In fact, it is not important whether one finds my argument against the alternative semantics convincing or not. My reason for discussing the problems of the alternative semantics here is to make the reader understand why I prefer the functionalist definition of focus to the alternative semantics version. 6 For instance, Umbach (2004) claims that “[i]nterpreting focus as evoking a set of alternatives raises the question of how to account for the intuitive idea that a focus in most cases presents new information” (Umbach, 2004: 160).

I. Kim / Language Sciences 56 (2016) 105–117

109

plausible to assume that a set of potential topics exists in discourse and the speaker selects one of them in her topiccontaining utterance. In fact, this kind of approach has been pursued by Büring (2003) for what he calls CT.7 From this point of view, no distinction can be made between focus and topic at least in their inherent nature of indicating alternatives. Lastly, the alternative semantics approach to focus causes serious confusion in distinguishing focus from contrast, another basic notion in information structure. Basically, contrast is established between one element and its alternative(s). Thus, no matter what the exact definition of contrast is, it necessarily involves alternative-indication as the central part of its definition. Hence, the problem is that it is hard, if not impossible, to make a clear distinction between focus and contrast. Indeed, it is generally believed among alternative semanticists that every focus is inherently contrastive (Umbach, 2004). Given the weaknesses of the definition of focus as indicating alternatives, I propose that a better and more useful definition (at least for the purposes of this paper) is the one based on relational newness. Note that focus understood as what is relationally new does not have the problems mentioned above, for it is more intuitive, not theory-internally motivated, and does not overlap with the notion of contrast in meaning. According to Gundel and Fretheim (2004), there are two dimensions of newness-givenness: referential and relational. First, referential givenness-newness has to do with a relation between a linguistic expression and its meaning in the speaker/ hearer’s mind, the discourse (model), or some real or possible world, depending on where the meaning of the expression is assumed to reside. For instance, in (3), Tom and John are referentially given or “familiar” by the time when (3B) is uttered, due to their prior introduction in (3A). On the other hand, relational givenness-newness reflects how the informational content of a particular event or state of affairs expressed by a sentence is represented. To be more specific, relational givenness-newness involves a partition of the propositional content of a sentence into two complementary parts, X and Y, where X is a noninformative, known, or expected part and Y is an informative, “newsy”, dominant, or contrary-to-expectation part. In other words, “X is given in relation to Y in the sense that it is independent of, and outside the scope of, what is predicated in Y. Y is new in relation to X in the sense that it is new information that is asserted, questioned, etc. about X” (Gundel and Fretheim, 2004: 177). Assuming that the purpose of uttering something is to give the addressee some new information about the world (Roberts, 2012), every utterance must have relationally new information. That is, no utterance can contain only relationally old information, for that would make the utterance totally uninformative and thus ultimately useless in meeting the purpose of participating in the discourse. Let us look at an example to understand what it means for focus to be relationally new information. (5)

A: B:

Among John, Tom, and Mary, who went to the party? John (went to the party).

In (5), John is referentially given by the time (5B) is uttered, due to the fact that it has already been mentioned in (5A). However, he is relationally new in that he is the answer to the question and thus being not predictable or recoverable from the given context. In contrast, the meaning of went to the party is the relationally given information in that it is expected from the prior discourse and thus being old/given in relation to John. Therefore, under the view that sees focus as what is relationally new, the focus of the sentence in (5B) is John. The dichotomy of relational givenness-newness has been captured by different terms by different researchers, including presupposition-focus (e.g. Jackendoff, 1972), topic-comment (e.g. Gundel, 1974), theme-rheme (e.g. Vallduví, 1990), and topicpredicate (Erteschik-Shir, 1997). Here, the terms focus, comment, rheme, and predicate all have the same meaning as that of focus used in this paper, that is, what is relationally new. 3.2. What is contrast? Unlike focus, there seems to be no standard view of contrast accepted by the majority of linguists (Repp, 2010), the reasons for which are (at least partly) that in defining the notion, 1) a number of scholars base their analysis on a specific language that they work on (e.g. É Kiss, 1998), 2) they tend to relate contrast to either topic or focus rather than treating it as an independent concept of its own (e.g. Lee, 2003, 2007), and 3) they focus more on making the notion fit into the theory under which they work than on adhering to our intuition about the notion (e.g. Rooth, 1985, 1992; Krifka, 2007).8 As suggested above, in this paper I will focus on providing a definition of contrast that is language- and theoryindependent and at the same time well captures our intuition about it. In doing so, let us first consider what our intuition says about the notion. Intuitively, contrast can be understood as unlikeness or dissimilarity between things. This intuition is well captured by the Oxford English Dictionary (OED), according to which contrast is defined as “comparison of objects of like kind whereby the difference of their qualities or characteristics is strikingly brought out; manifest exhibition of opposing qualities”. Given this intuitive sense of contrast, the remaining job is to make it as much explicit as possible.

7 The only difference is that he deals only with topics that have a contextually salient set of alternatives. But nothing in principle prevents us from extending his analysis to “plain” non-contrastive topics. 8 As a result, in the alternative semantics approach, no distinction can be made, for instance, between addition and contrast, which are intuitively opposite meanings (Krifka, 2007).

110

I. Kim / Language Sciences 56 (2016) 105–117

To the best of my knowledge, Mulkern (2007) provides one of the best formalizations of the meaning of contrast understood as above, arguing that “[t]he purpose of contrast is to partition a set of discourse entities with respect to some semantic property P such that it is established (either via assertion or implicature) that the value ‘true’ results when P is applied to one part of the set and ‘false’ or ‘unknown’ when applied to the other” (Mulkern, 2007: 121). Here, the meaning of “difference” or “opposing qualities” in the OED is understood in terms of the difference in truth value with respect to some semantic property P, and the meaning of “comparison” is captured by the notion of partitioning a set of discourse referents. Based on the purpose of contrast proposed by Mulkern, its definition is provided in (6). (6)

Definition of Contrast Contrast is a relation between discourse referents that are partitioned with respect to some semantic property P such that it is established (either via assertion or implicature) that the value ‘true’ results when P is applied to one part of the set and ‘false’ or ‘unknown’ when applied to the other.

It is important to note that contrast, unlike focus (and topic), is a relational notion. That is, as Repp (2010: 1335) correctly points out, “we have focus ON an item but contrast BETWEEN items”. This relational nature of contrast is also well captured in the above definition. It is also important to note that contrast can be further divided into subtypes. Let us just look at two subtypes, which are relevant to the topic of the paper. The first type of contrast is correction, which occurs when one denies what the other interlocutor said and provides what she believes to be correct information. (7)

A: John bought the book. B: No. Mary bought the book.

In (7), the speaker B denies what is expressed by the speaker A and provides new information which she believes to be true. This utterance is contrastive since a set of alternatives is evoked, John and Mary in this case, and different truth values come out when the semantic property of buying the book is applied to the alternatives. The second type of contrast is selection, which is basically done by choosing one of several contextually evoked alternatives. Note that what is selected is necessarily distinguished from the rest of the alternatives with respect to some semantic property that motivates the selection; hence, it is contrastive. It is different from correction, however, because one who performs selection does not deny the other interlocutor’s utterance, as shown in (8). (8)

A: Among John, Mary, and Kim, who is the oldest? B: John is the oldest.

In (8), the speaker A provides a set of alternatives to be selected by the speaker B. Then, what the speaker B does is to partition the discourse referents into two groups with respect to their age, and apply the property of being the oldest to one group, that is, John. 3.3. CF ¼ contrast þ focus With the definitions of focus and contrast provided above, we are now ready to understand what CF is. Before going on any further, however, it has to be noted that most, if not all, previous studies, when providing a definition of CF, have taken the notion to be independent of the meanings of contrast and focus. For instance, É Kiss (1998), who provides one of the most influential studies on CF, takes CF (her identificational focus) to be focus that conveys exhaustivity. The problem of this definition of CF is that no independent motivation is provided for treating exhaustivity as a crucial characteristic of CF. The claim is only based on the assumption that the meaning of a particular linguistic phenomenon of a specific language, the preverbal position in Hungarian in her case (É Kiss, 1998), is the meaning of CF. Regardless of whether the preverbal position in Hungarian is really directly related to exhaustivity, the definition of CF defined in this way is problematic because there is no independent motivation for relying on the meaning of a specific linguistic phenomenon of a specific language for defining CF. Rather, as pointed out above, the meaning of CF should be understood languageindependently.9 Intuitively, the notion of CF must be dependent on the meanings of contrast and focus, for it is contrast(ive) and focus. To be more specific, the most natural and reasonable understanding of CF should be to think of it as a simple combination of the meanings of contrast and focus. According to this view, CF can be defined as in (9). (9)

9

Definition of Contrastive Focus (CF) Contrastive focus is a focus whose alternatives are evoked in the context and which is distinguished from the alternatives with respect to some semantic property P in that some truth value results when P is applied to the focus and the opposite truth value or ‘unknown’ when applied to the alternatives.

The same problem is also shown in Y. Jun’s (2005) analysis, which tries to define CF based on the specific lexical item -(n)un.

I. Kim / Language Sciences 56 (2016) 105–117

111

Note that the meaning of CF provided in (9) is a simple unification of the meaning of focus and that of contrast proposed in this paper. That is, a CF is a focus that is contrastive, with the meaning of contrast being exactly the same as defined in (6) above. Again, the reason that this intuitive and explicit definition has not been adopted by most previous studies mainly seems to be the unjustified assumption that any notion of information structure must be directly realized (or manifested) by a certain linguistic device in every language, which has been convincingly argued to be invalid based on a detailed investigation of linguistic expressions in various languages that have been known to be “markers” of the categories of information structure (cf. Mati c, 2009; Mati c and Wedgwood, 2013; Kim, 2015). And because of this assumption, a number of previous studies on CF, I believe, are involved in consuming debates on non-essential aspects such as whether CF is exhaustive or not. Lastly, note that in the view of CF proposed here, correction and selection require CFs in their context. That is, in correction or selection, the correcting/selected information is always focus, for it is new in relation to the rest of the sentence by being unexpected from the previous discourse (whereas the other part is always expected). Also, it is contrastively related to its alternatives with respect to some semantic property that distinguishes it from them. Thus, the correct information in correction and the selected information in selection are necessarily CFs. 4. Proposal 4.1. -(N)un “marks” relationally old information There is a long standing observation that -(n)un marks referentially old (or given) but not new information (e.g. I. Yang, 1973; W. Chay, 1983; S.-Y. Bak, 1981; I. Kim, 2013). Its oldness/newness status from the relational viewpoint, however, has been little discussed. A crucial point to take into account in determining the relational givenness/newness of -(n)un-marked elements is the pragmatic effects derived from the meaning of -(n)un. According to Kim (2015), there are four interpretive effects of -(n)un: topic and frame10 introduction, contrast indication, and simple emphasis. First, topic/frame introduction is “an act of proposing a new topic/frame into the discourse for the purpose of either beginning a discourse or shifting/contrasting an existing topic/frame” (Kim, 2015: 97). It is well known that in the topiccomment structure, a topic is given in relation to its comment “in the sense that it [the topic] is independent of, and outside the scope of, what is predicated in Y [the comment]” (Gundel and Fretheim, 2004: 177), and a comment is new in relation to its topic “in the sense that it is new information that is asserted, questioned, etc. about X [the topic]” (Gundel and Fretheim, 2004: 177). For instance, the sentence in (10) is divided into two parts with regard to its information structure. (10)

John-un hakkyo-ey John-Nun school-Dat ‘John went to school.’

ka-ss-e go-Past-Dec

That is, by the use of -(n)un, the sentence bifurcates into John, the topic of the sentence, and the rest of the sentence (i.e. the comment about the topic). That -(n)un divides a sentence into relationally old and new parts is strongly supported by the fact that (10) cannot felicitously serve as an answer to a question requiring its answer to have the “sentence-focus” structure (Lambrecht, 1994), that is, a question that calls for an answer with only relationally new information (e.g. What happened?). Thus, when -(n)un is used to introduce a topic, it marks relationally old information by the nature of topic itself. The same logic applies to the relation between a frame and the proposition whose truth value is restricted to it. The frame marked by -(n)un is independent of the proposition that follows it. An example of frame introduction is shown in (11) below. (11)

ecey-nun pi-ka wa-ss-e yesterday-Nun rain-Nom come-Past-Dec ‘Yesterday, it rained.’

Here, the speaker introduces the temporal frame, which is ecey ‘yesterday’, and then utter a clause whose truth is restricted to the frame. Note that, by the use of -(n)un, the whole sentence is decomposed into relationally given and new parts, which is shown by the fact that when the utterance is denied by a lexical item denoting negation such as ani ‘no’, it is not the frame but only the proposition following the frame that is denied. As for contrast-indicating function of -(n)un, it is important to note that -(n)un induces contrast by partitioning a set of alternatives into two groups and making only one group salient (or prominent) so that contrastiveness arises between what is highlighted and what is not. In order to understand more clearly what it means for -(n)un to induce contrast, let us look at (2) again, repeated below. (2)

10

A:

ne ton iss-ni? you money exist-Int ‘Do you have money?’

“In (X Y), X is the frame for Y iff X specifies a domain of (possible) reality to which the proposition expressed by Y is restricted” (Jacobs, 2001: 656).

112

I. Kim / Language Sciences 56 (2016) 105–117

B:

na tongcen-un iss-e I coin-Nun exist-Dec ‘I have coins.’

By saying tongcen-un ‘coin-Nun’ in (2), the speaker of (2B) induces contrast between coins and bills. Then, she provides information about what she has just highlighted, that is, coins, by saying iss-e ‘exist-Dec’. Two crucial points must be made here. First, the type of contrast expressed in (2) is fundamentally different from the type of contrast shown in (7) and (8) above, each of which is about correction and selection. The difference comes from the source of contrast. In (2), contrast is induced by the lexical item -(n)un itself, whereas in (7) and (8), it is made possible by the context rather than by some specific lexical item. Second, the sentence with contrast expressed by the use of -(n)un is also divided into relationally old and new parts in terms of its information structure. That is, what -(n)un does here is not only to induce contrast by making the selected set of alternatives salient, but also makes the hearer expect the highlighted part to be commented upon by what follows it. In this sense, the sentence is made up of the relationally old (i.e. what is highlighted for contrast) and new parts (i.e. the information about what is highlighted), and -(n)un marks the relationally old part. That -(n)un, when used to induce contrast, marks relationally old information, is made clear when (12), which is a slight modification of (2), is compared to the contexts of (contrastive) focus in (13) and (14), each of which is the context of selection and correction involving CF. (12)

A: B:

Do you have money? ??tongcen-un coin-Nun ‘I (don’t) have coins.’

(13)

A: B:

(14)

A: B:

Among John, Mary, and Tom, who went to the party? Mary-ka (ka-ss-e) Mary-Nom go-Past-Dec ‘Mary (did).’ John bought the book. aniya. Mary-ya. no Mary-Cop-Dec ‘No. Mary (did).’

As shown in (13B) and (14B), in Korean it is possible and even most natural to omit the relationally old part of the sentence. In contrast, it is much less felicitous to omit the predicate in (12B), which would be hard to explain if it is relationally old (and thus expected). It is also crucial to note that in (12B), the speaker can felicitously say either that she has coins or that she does not, which clearly shows that the predicate is not given/old at all in relation to the -(n)un-marked NP. For instance, one could provide (15) as the answer to the question in (12A). (15)

tongcen-un eps-ko ciphyey-man iss-e coin-Nun not exist-and bill-only exist-Dec ‘(I) have only bills but not coins.’

That is, the hearer does not know until the speaker utters the predicate whether or not she has coins. This is clearly different from cases where (contrastive) focus is involved (e.g. (13) and (14)), where the relationally old information is definitely fixed and no possibility exists for its change. One might argue that in (2) and (10) above, the -(n)un-marked NP can be assumed to be relationally new because it is unexpected from the previous discourse. For instance, from the hearer’s point of view, it is totally unexpected that the speaker uttered tongcen-un ‘coins-Nun’ in (2B). That is, the hearer cannot know whether the speaker would choose coins or bills as the subject argument of the sentence until she actually utters one or the other. What is missing in this counterargument, though, is that the distinction between relational givennness and newness must be made within a sentence. In other words, what is relationally old/new must be old/new in relation to the other part of the same sentence, not to some other part of the prior discourse outside the sentence. Of course, there is a correlation to a certain extent between relational oldness-newness within a sentence and the possibility of expectation from prior discourse. For instance, if some information conveyed in a sentence is expected from prior discourse, it is also necessarily relationally old in the sentence. The predicates of the examples in (5), (7), and (8) are the cases clearly showing this point. That is, the expected nature of the predicates from the prior discourse guarantees their relational givenness in the sentence in which they occur. However, if something unexpected from prior discourse is uttered, it does not necessarily lead to relational newness within a sentence; rather, it can be either relationally old or new depending on its relation to the rest of the sentence. Gundel and Fretheim (2004: 178) also points out that “relational givenness notions like topic . may be constrained or influenced by the discourse context . but they are not uniquely determined by it [emphasis added]”. How to “package” unexpected

I. Kim / Language Sciences 56 (2016) 105–117

113

information is entirely determined by the interests and perspective of the speaker. In fact, we have just seen two of the most representative motivations for making unexpected information relationally old, that is, topic introduction and contrast indication using -(n)un. Lastly, the function of simple emphasis expressed by -(n)un also supports the claim that -(n)un marks relational givenness. Simple emphasis is defined as “a non-information-structural effect of attracting attention, which is caused by the difference between the actual (and unexpected) salience of an item and its expected salience” (Kim, 2015: 97). (16)

Q: A:

How much is the book? I think I heard it is 5 dollars. 5 talle-nun ani-ya 5 dollar-Nun not-Dec ‘(It) is not 5 dollars.’

In (16A), the -(n)un-marked element, 5 talle ‘5 dollars’, is neither topic nor frame. There is no frame in the sentence and the topic is the omitted subject, that is, the price of the book, in that the sentence is about the price of the book. Also, in the given context, in which the issue is not about comparing different prices but about whether the price of the book is 5 dollars or not, contrast is hardly induced by -(n)un either. Rather, the speaker can best be assumed to emphasize 5 dollars by imposing unexpected salience on it. Why would she make 5 dollars the center of attention by making it more prominent than expected? In the situation provided in (16), the most plausible account would be that the speaker wants to emphasize that what the hearer heard about the price of the book is incorrect. In other words, the speaker tries to emphasize that the hearer is misinformed on how much the book costs by making the core of the misinformation, 5 dollars in this case, the center of attention and commenting about it. For our purposes, it is important to note that simple emphasis is necessarily followed by comments about what is emphasized. For instance, in (16A), after emphasizing 5 dollars, the speaker provides information about the relation between 5 dollars and the price of the book (i.e. whether or not the book is 5 dollars). The relation between what is emphasized and the following comment can be captured by relational givenness-newness, and again, this is done by the use of -(n)un. 4.2. The connection between relational givenness and saliency Given that -(n)un always marks relationally old information, I would like to ask a more fundamental question: how is it possible for -(n)un to mark relational oldness? This issue, just like the issue of whether -(n)un is directly or indirectly related to topic and contrast, is essential for a deeper and systematic understanding of the nature of -(n)un and its relation to information structure. In order to solve this problem, let us first recall the difference between relationally old information determined by discourse context and that determined by -(n)un. The crucial difference is that, unlike the former, the latter is always highlighted by the speaker, which means that it is more prominent than expected from the prior discourse context. The kind of highlighting done by -(n)un is special in that the hearer expects more to follow after highlighting the referent of a-(n)un-marked element, and this intuitive expectation of the comment about the highlighted part is the very source of the theoretical notion of relational oldness. What makes this special type of highlighting possible? The answer must be found in the common property shared by the four interpretive effects of -(n)un discussed above, and it is the act of imposing salience using -(n)un. First, the highlighting itself is easily explained by the function of imposing salience. Note that highlighting something is equivalent to making it prominent, and making something prominent can be done by imposing salience on it, with salience being understood as “cognitive prominence of the meaning of any part of an utterance made by discourse participants, the degree of which is determined by the amount of attention allotted to it” (Kim, 2015: 93). Second and more importantly, the property that makes the hearer expect more to follow (after highlighting) can be understood better by comparing the meaning of -(n)un with that of -i/ka, which is known as the nominative marker in Korean. A number of previous studies have noted that -i/ka has more than its nominative-marking function, and the additional meaning/function has been explained based on the notion of specification (e.g. C. S. Shin, 1975; S.-J. Ko, 2002; Kim, 2013, 2014). For instance, Kim (2013, 2014) has provided an analysis of the meaning of -i/ka based on this notion, according to whom the meaning of -i/ka is, just like -(n)un, performative in nature and the act conducted by the use of -i/ ka is to uniquely specify the meaning of an -i/ka-marked element. Informally, the act of unique specification can be understood as specifically pointing at (or pinpointing) something. In other words, just like one can use one’s finger to point at a physical object, the same thing can be done to the meaning of linguistic expressions by using -i/ka as a “linguistic pointer”.11 Based on various kinds of contexts in which -i/ka and -(n)un are (in)appropriate, Kim (2013) summarizes the difference between -i/ka and -(n)un as shown in Table 1.12

11 12

For more on the meaning of -i/ka, see Kim (2014). For a detailed discussion on the pragmatic difference between the two, see Kim (2013).

114

I. Kim / Language Sciences 56 (2016) 105–117 Table 1 Difference between -i/ka and -(n)un.

-(n)un -i/ka

Imposing salience

Unique specification

Yes No

No Yes

That is, -(n)un has the function of imposing salience but not unique specification, whereas -i/ka is used for unique specification but not for imposing salience. Note that in principle there is no reason for a single linguistic device not to have both (or more) functions.13 The distinction between the two particles provides a clue to the puzzle at hand. That -(n)un lacks the function of unique specification is what makes the hearer feel that the use of -(n)un cannot be the end of the speaker’s utterance but there is more to come. A good example showing this point is (12), which is repeated below. (13)

A: B:

Among John, Mary, and Tom, who went to the party? Mary-ka (ka-ss-e) Mary-Nom go-Past-Dec ‘Mary (did).’

In (13B), by using -ka, the speaker uniquely specifies (or “pinpoints”) one of the alternatives so that the hearer can identify who went to the party among the three. Once the focus referent is identified by unique specification, the predicate need not be overtly uttered as shown in the translation. If -ka is replaced by -(n)un, however, the specifying function is not intuitively felt anymore and the answer becomes infelicitous in the given context. With -(n)un, Mary becomes CT of the sentence and one has a strong feeling that the subject argument must be followed by a comment on it. This intuition is well captured by the pragmatic difference between -i/ka and -(n)un shown in Table 1 above. That is, the intuition that more to be uttered after a -(n)un-marked phrase is explained by a referent’s being highlighted without being uniquely specified. To sum up this subsection, relational givenness marked by -(n)un is made possible by its inherent function of imposing salience and its lacking the function of unique specification. Notice that the function of marking relational oldness is not mere interpretive effects like topicality and contrast, because it is not derived from the interaction of various factors. Rather, relational givenness is a direct result of the inherent nature of -(n)un. 4.3. Relation between -(n)un and (contrastive) focus If the claim that -(n)un marks relational oldness is on the right track, it becomes clear that -(n)un cannot mark (contrastive) focus, because (contrastive) foci are always relationally new by nature and -(n)un marks only relationally old information. In this subsection, I will provide further support for this claim by analyzing the information structure of (2), repeated below, and other related linguistic data. (2)

A:

B:

ne ton iss-ni? you money exist-Int ‘Do you have money?’ na tongcen-un iss-e I coin-Nun exist-Dec ‘I have coins.’

According to Y. Jun (2005), the -(n)un-marked NP is CF and the predicate is relationally old in (2B). For him, the reason for treating the NP as CF is that it is (part of) relationally new information. As for the predicate in (2B), that is, iss-e ‘exist-Dec’, he takes it to be relationally old because he believes that it is expected from the prior utterance. The crucial problem with Y. Jun’s analysis is that it is not explicit about what is relationally new and old in the sentence. According to him, the relationally new information in (2B) is that the speaker has only coins but not bills. However, that information is in fact the meaning of the whole sentence (plus the contrastive implicature generated by the use of -(n)un), which would ultimately lead to the conclusion that the sentence consists of only relationally new information. Thus, his position on what is relationally new in the sentence and his claim that only the -(n)un-marked NP (but not the subject argument (i.e. na ‘I’) and the predicate) conveys relationally new information are contradictory to each other. In order to solve this problem, I provide the information structure of (2B), which is shown in (17) below. (R-Old/New means relationally old/new.)

13

I suspect that English it-clefts have both the function of imposing salience and unique specification.

I. Kim / Language Sciences 56 (2016) 105–117

(17)

[na]Topic [[tongcen-un]R-Old/Contrast I coin-Nun ‘I have coins.’

115

[iss-e]R-New]Focus exist-Dec

First, the subject argument, that is, na ‘I’, is the topic of the sentence in that the (rest of the) sentence is about the subject, i.e. as expressing information which is relevant to and which increases the addressee’s knowledge of the subject referent. The topical status of the subject is guaranteed by (2A), in which ton iss-ni ‘money exist-Int’ is new in relation to ne ‘you’ in the sense that it is questioned about ne ‘you’. Since the subject argument is the topic, the rest of the sentence automatically gets focus (or comment) as its informationstructural status. Here, it is important to note that the focus of the sentence is further divided into relationally old and new parts as shown in (17). That is, iss-e ‘exist-Dec’ is new in relation to tongcen-un ‘coin-Nun’ as discussed in 4.1.14 According to this analysis, the -(n)un-marked NP is neither focus nor CF. Rather, it is part of the focus, and what -(n)un does here is to induce contrast and make the element to which it is attached the relationally old part against what follows it. In other words, -(n)un is not directly related to the notion of focus.15 One might raise two issues concerning the claim that -(n)un can mark relational old information and induce contrast inside focus. The first issue is how focus, which is relationally new by definition, can have some part that is relationally old. The second issue is why an element that is in focus but relationally old is supposed to induce contrast. The first issue can be resolved by carefully taking into account the notion of relational givenness/newness. As discussed in 3.1, the usual understanding of relational givenness/newness is provided by Gundel and Fretheim (2004). For our purposes, it is important to note that “relational givenness-newness involves a partition of the semantic/conceptual representation of a sentence into two complementary parts [emphasis added]” (Gundel and Fretheim, 2004: 177). This understanding of relational givenness-newness is unsatisfactory because it overlooks one important point that, in principle, there is no reason for what is relationally new/old not to allow more divisions within it in terms of relational givenness-newness. For instance, in (17), the focus of the sentence (i.e. tongcen-un iss-e ‘have coins’) is new information in relation to the topic na ‘I’. However, this does not mean that the relation between tongcen-un and iss-e is undefinable in terms of relational givenness-newnes. In other words, the relation between na and tongcen-un iss-e is independent of the relation between tongcen-un and iss-e is. Regarding the second issue, note that topic introduction is only possible when -(n)un is attached to a sentence-initial element, due to the syntactic constraint on topic introduction. (18)

Syntactic constraint on topic introduction (Kim, 2015: 106) [S [NP X]Topic [ VP/S Y]Comment]

What the constraint in (18) means is that when a new sentence topic is newly introduced, it must be sentence-initial and separated from the rest of the sentence (i.e. the comment), thus forming the typical [topic-comment] structure. This structural constraint for topic introduction is functionally well motivated. For efficient and coherent communication, it would be undesirable to start one’s utterance with commenting on something before even introducing it. If a -(n)un -marked phrase is not sentence-initial, it cannot be topic and must be used for contrast or simple emphasis depending on the existence of alternatives in discourse context. That is, if alternatives to the referent of a -(n)un-marked element already exist or are easily inferred to exist, the use of -(n)un strongly suggests contrast; however, if no alternative exists and is not easily inferable, -(n)un tends to convey simple emphasis. In case of (17) above, -(n)un used in focus is supposed to induce contrast because it cannot be used to introduce a new topic due to the syntactic position of the -(n)unmarked phrase and a set of alternatives to the referent of the -(n)un-marked phrase is easily inferable, that is, a set of bills. Based on the discussion so far, I provide two ways of realizing (non-contrastive) focus as in (19). (19)

Two ways of realizing non-contrastive focus a. [R-new]Focus b. [[R-old] [R-new]]Focus

First, focus can consist only of what is relationally new as in (19a). Also, it can bifurcate into relationally old and new parts as in (19b). When -(n)un is used within focus, it can only attach to an element that conveys relationally old information. Its function there must be contrast indication or simple emphasis but not topic introduction, which is understandable considering that topic introduction in principle cannot be done within focus, that is, the comment about the topic. Note that in (19b) what is relationally old can be divided into two types depending on whether or not -(n)un is attached to it. When -(n)un marks the relationally old information within focus, it always induces contrast or emphasizes the relationally old information. On the other hand, if -(n)un does not appear in what is relationally old in focus, the relationally old part corresponds to Vallduví’s (1990) “tail”, an example of which will be shown below.

14

Also, the fact that -(n)un here induces contrast is represented in the information structure with the subscript Contrast. According to C. Lee (2003, 2007), every -(n)un-marked phrase is (contrastive) topic, which predicts that the -(n)un-marked NP in (2B) is CT. In order for this view to be valid, however, the definition of topic must be broadened so that it can allow what the (rest of the) sentence is not about. Unless we want to stick to the idea that -(n)un inherently expresses topicality, for which no independent (or empirical) motivation exists, it is not necessary (and even undesirable) to accept his view with the counterintuitive definition of topic. 15

116

I. Kim / Language Sciences 56 (2016) 105–117

Now, let us investigate the relation between -(n)un and CF. Note that correction can be done in more than one way. For instance, (14A) can be corrected as in (20). (20)

(aniya.) [John-un]CT [[(ku chayk)]R-Old/Background no John-Nom the book ‘(No.) John didn’t buy the book.’

[an not

sa-ss-e]R-New]Focus buy-Past-Dec

Although it does not give the correct information about the question under discussion (QUD), Who bought the book? in this case, the utterance in (20) can be safely considered to be correction, for it denies what has been previously uttered and the denial itself can be considered as correcting information, albeit partially. What is important here is that the correcting information in (20) is “packaged” differently from that in (14B). In (14B), correction is done by using CF, that is, the correct answer to the QUD. However, in (20), the correcting information is an explicit denial of the previous utterance but not the correct answer to the QUD. The information structure of the denial in (20) is CT-focus, in which the focus is further divided into relationally old and new parts. First, the subject argument John is CT because -(n)un that is attached to it makes it relationally old and gives rise to aboutness and contrast at the same time. Topicality of the subject makes the rest of the sentence focus, which is further decomposed into relationally old and new information. First, the object argument ku chayk ‘the book’ is relationally old in the sense that it is expected from the prior discourse. Note that it is marked with the subscript Background, which simply means that it is relationally old information with no contrast involved. In contrast, the predicate an sa-ss-e ‘didn’t buy’ is relationally new since it is not predicted to be uttered by the previous utterance. So far in this section, we have seen that -(n)un can be used to induce contrast within focus (of a sentence) by making the referent of the element to which it is attached relationally given/old. The possibility for focus to have relationally old part within it is neither new nor surprising at all given the data like (20).16 Also, the idea that -(n)un-marked element is old in relation to the part that follows it is strongly supported by the fact 1) that -(n)un-marked element cannot stand alone but requires following part which is the comment with regard to the referent of the -(n)un-marked element and 2) that the exact content of the comment part is not expected or predicted from the hearer’s perspective. 5. Conclusion In information structure theory, basic notions like topic and focus are often treated as universal primitives that have a direct reflection on some linguistic device in every language. As for focus, for instance, “there is a popular perception among linguists . that there is a single notion of focus that can be associated with a variety of grammatical phenomena in different languages” (Wedgwood, 2009: 101). This widespread but never empirically justified assumption has had deep and often undesirable effects on the study of various linguistic phenomena in different languages. For example, most previous works on the meaning/function of Korean -(n)un have taken it for granted that the particle is directly related to either topic, CT, and/or CF, thus hindering us from getting deeper insights into important questions such as how exactly -(n)un is related to CT/CF (if it is). By not blindly granting the longstanding assumption, this paper has shown that what has been claimed as the CF-marking function of -(n)un can best be captured by the notion of relational givenness. That is, -(n)un itself does not lexically (or inherently) express CF; rather, it makes the element to which it is attached relationally given/old by imposing salience to that element. This has an important implication on the relation between grammar and information structure. Korean is a language in which the notion of relational givenness is more primitive than CF (or CT) in the sense that it (but not CT/CF) is not a mere interpretive effect but the direct effect of the function of imposing salience. Lastly, I would like to reemphasize the importance of allowing the partitioning of a sentence in terms of relational givenness-newness more than once, in order to capture the correct information structure of the sentence. Particularly, when a -(n)un-marked item is not in the sentence-initial position, it is likely to be used as the relationally given part within the relationally new part of a sentence, the function of which is either contrast indication or simple emphasis. References Bak, Sung-Yun, 1981. Studies in Korean Syntax. Ph.D. dissertation. University of Hawaii. Beaver, David I., Clark, Brady Z., 2008. Sense and Sensitivity: How Focus Determines Meaning. Wiley-Blackwell. Büring, Daniel, 2003. On D-trees, beans, and B-accents. Linguist. Philos. 26, 511–545. Chay, Wan, 1983. Meaning of the particle nun (written in Korean). In: Koh, Y.K. (Ed.), Syntax and Semantics of Korean. Thap Press, Seoul. Choi, Hye-Won, 1996. Optimizing Structure in Context: Scrambling and Information Structure. Ph.D. dissertation. Stanford University. Choi, Kyu-Soo, 2004. Topic and contrastive focus: focusing on word order and phonetic accent (written in Korean). Wulimalyenkwu 15, 149–172. Choi, Seungja, 2000. Topicality, Genericity, and Logophoricity: the Postpositional Markers nun in Korean and wa in Japanese from an Argument Perspective. Ph.D. dissertation. Yale University. Erteschik-Shir, Naomi, 1997. The Dynamics of Focus Structure. Cambridge University Press, Cambridge. É Kiss, Katalin, 1998. Identificational focus versus information focus. Language 74 (2), 245–273.

16 I am grateful to one reviewer for informing me of Kruijff-Korbayová and Steedman (2003), who also introduces a similar system of informationstructural categories to the system proposed here.

I. Kim / Language Sciences 56 (2016) 105–117

117

Ginzburg, Jonathan, 1995a. Resolving questions I. Linguist. Philos. 18, 459–527. Ginzburg, Jonathan, 1995b. Resolving questions II. Linguist. Philos. 18, 567–609. Ginzburg, Jonathan, Sag, Ivan A., 2000. Interrogative Investigations. CSLI Publications, Stanford, CA. Gundel, Jeanette K., 1974. The Role of Topic and Comment in Linguistic Theory. Ph.D. dissertation. University of Texas at Austin. Gundel, Jeanette K., Fretheim, Thorstein, 2004. Topic and focus. In: Horn, Laurence, Ward, Gregory (Eds.), The Handbook of Pragmatics. Blackwell, pp. 175– 196. Halliday, Michael A.K., 1967. Intonation and Grammar in British English. The Hague, Mouton. Hamblin, Charles L., 1973. Questions in Montague English. Found. Lang. 10, 41–53. Han, Chung-hye, 1998. Asymmetry in the interpretation of -(n)un in Korean. In: Akatsuka, Noriko, Hoji, Hajime, Iwasaki, Shoichi, Sohn, Sung-Ock (Eds.), Japanese/Korean Linguistics, vol. 7. CSLI, pp. 1–15. Hausser, Roland, 1983. The Syntax and Semantics of English Mood. Questions and Answers. Reidel, Dordrecht. Hockett, Charles F., 1958. A Course in Modern Linguistics. Macmillan, New York. Jacobs, Joachim, 2001. The dimensions of topic-comment. Linguistics 39 (4), 641–681. Jackendoff, Ray, 1972. Semantic Interpretation in Generative Grammar. MIT Press, Cambridge. Jun, Youngchul, 2005. Contrastive focus in Korean. Eonehak 43, 215–237. Jun, Youngchul, 2006. Korean marker ‘NUN’ for contrastive topic and contrastive focus. Hankul 274, 171–200. Kim, Ilkyu, 2013. Referential status of -(n)un-marked NPs. Korean J. Linguist. 38 (4), 855–874. Kim, Ilkyu, 2014. On the meaning of Korean -i/ka. Lang. Linguist. 63, 1–26. Kim, Ilkyu, 2015. Is Korean -(n)un a topic marker? On the nature of -(n)un and its relation to information structure. Lingua 154, 87–109. Kim, Ji Eun, 2010. The Generation of Implicit Propositions in “alleged” Korean Topics. Ph.D. dissertation. UCLA. Kim, Yong-Beom, 2004. Focus, topic and their phonetic relevance. Lang. Inf. 8 (1), 27–52. Ko, Seok-Joo, 2002. On the meaning of ‘ka’ (written in Korean). Kwukehak (Korean Linguist. 40, 221–246. Krifka, Manfred, 2001. For a Structured Meaning Account of Questions and Answers. In: Audiatur Vox Sapientia. A festschrift Arnim von Stechow, vol. 52, pp. 287–319. Krifka, Manfred, 2004. The Semantics of Questions and the Focusation of Answers. Topic and Focus: a Cross-linguistic Perspective. Kluwer Academic Publishers, Dordrecht, pp. 139–151. Krifka, Manfred, 2007. Basic notions of information structure. In: Féry, C., Fanselow, G., Krifka, M. (Eds.), The notions of information structure: interdisciplinary studies on information structure, vol. 6, pp. 13–55. Kruijff-Korbayová, Ivana, Steedman, Mark, 2003. Discourse and information structure. J. Log. Lang. Inf. 12 (3), 249–259. Lambrecht, Knud, 1994. Information Structure and Sentence Form: Topic, Focus, and the Mental Representation of Discourse Referents. Cambridge University Press, Cambridge. Lee, Chungmin, 2003. Contrastive topic and/or contrastive focus. In: McClure, B. (Ed.), Japanese/Korean Linguistics, vol. 12. CSLI. Lee, Chungmin, 2007. Contrastive (predicate) topic, intonation, and scalar meanings. In: Lee, Chungmin, Gordon, Matthew (Eds.), Topic and Focus: Crosslinguistic Perspectives on Meaning and Intonation. Springer, pp. 151–175. Mati c, Dejan, 2009. On the variability of focus meanings. In: Current Issues in Unity and Diversity of Languages (Collection of the Papers Selected from the 18th International Congress of Linguistics (CIL 18)). Korea University, pp. 176–191. Mati c, Dejan, Wedgwood, Daniel, 2013. The meanings of focus: the significance of an interpretation-driven category in cross-linguistic analysis. J. Linguist. 49 (1), 127–163. Mulkern, Ann, 2007. Knowing who’s important: relative discourse salience and Irish pronominal forms. In: Hedberg, Nancy A., Zacharski, Ron (Eds.), The Grammar-pragmatics Interface: Essays in Honor of Jeanette K. Gundel. John Benjamins Publishing Company, Amsterdam, pp. 113–142. Oh, Chisung, 2007. Topic and Focus Constructions in Spoken Korean. Ph.D. dissertation. University of Texas at Austin. Reinhart, Tanya, 1981. Pragmatics and linguistics: an analysis of sentence topics. Philosophica 27 (1), 53–94. Repp, Sophie, 2010. Defining ‘contrast‘ as an information-structural notion in grammar. Lingua 120, 1333–1345. Roberts, Craige, 2012. Information structure: towards an integrated formal theory of pragmatics. Semant. Pragmat. 5 (6), 1–69. Rooth, Math, 1985. Association with Focus. Ph.D. dissertation. University of Massachusetts, Amherst. Rooth, Math, 1992. A theory of focus interpretation. Nat. Lang. Semant. 1, 75–116. Scha, Remko J.H., 1983. Logical Foundations for Question Answering. Ph.D. dissertation. Rijksuniversiteit te Groningen. Shin, C.S., 1975. A study on the issue of subject in Korean (written in Korean). Mwunpep Yenkwu – Gramm. Res. 2, 131–170. Szabolci, A., 1981. The pragmatics of topic-focus construction. In: Groendijk, J., Jansen, T., Stokhof, M. (Eds.), Formal Methods in the Study of Language. Matemaish Centrum, Amsterdam, pp. 513–541. Umbach, Carla, 2004. On the notion of contrast in information structure and discourse structure. J. Semant. 21, 155–175. Vallduví, Enric, 1990. The Informational Component. Ph.D. dissertation. University of Pennsylvania. Vallduví, Enric, Engdahl, Elisabet, 1996. The linguistic realization of information packaging. Linguistics 34, 459–519. Wedgwood, Daniel, 2009. Variation in focus. In: Riester, Arndt, Onea, Edgar (Eds.), Working Papers of the SFB732, Focus at the Syntax-semantics Interface, vol. 3. University of Stuttgart, pp. 101–119. Wee, Hae-Kyung, 2010. A critical review of semantic theories of focus and an alternative approach. Lang. Inf. 14 (1), 197–227. Yang, In-Seok, 1973. Semantics of delimiters in Korean. Lang. Res. 9 (2).

Lihat lebih banyak...

Comentarios

Copyright © 2017 DATOSPDF Inc.