Skip to main content Accessibility help

We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.

Close cookie message
×
  1. Home
  2. Browse subjects
  3. Language and Linguistics
  4. Asian language and linguistics
  5. Content listing

Refine search

Refine search

Access:
Content type:
Publication date:
Apply   From and To filters
Subject: Show more
Journals Show more
Publishers: Show more
Societies Show more
Series: Show more
Collections: Show more

Actions for selected content:

Select all | Deselect all
  • View selected items
  • Export citations
  • Download PDF (zip)
  • Save to Kindle
  • Save to Dropbox
  • Save to Google Drive

Save Search

You can save your searches here and later view and run them again in "My saved searches".

Please provide a title, maximum of 40 characters.
Cancel
×

2280 results in Asian language and linguistics

Page 29 of 114

  • 9 - Coordination

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 204-225

  • Summary

    Every language employs its own coordination strategies, according to the type of coordinating marking, the pattern of marking, the position of the marker, and the phrase types coordinated. Korean is also intriguing in the sense that it displays almost all the possibilities of these dimensions. This chapter shows how a construction-based HPSG can provide a robust analysis for nominal and verbal coordination in Korean.

    Two main types of nominal and verbal coordination

    Korean employs two kinds of coordination marking: morphological and lexical marking. In the morphological marking system, the language distinguishes nominal and verbal coordination. As seen in the corpus example (1a), nominal coordination uses suffixal markers (usually called particles in the traditional literature) like -(k)wa, -hako, -(i)lang ‘and’ for conjunctive and -(i)na ‘or’ for disjunctive coordination. Meanwhile, as in (1b), verbal coordination uses the suffixal marker -ko ‘and’ for conjunctive and -kena ‘or’ for disjunctive coordination:

    (1) a. [khemphyuthe-wa/hako/lang intheneys-ul] payw-ess-ta

    computer-and internet-ACC learn-PST-DECL

    ‘(He) learned computer and internet.’

    b. pelley-ey [mwulli-ko/kena sso-i-yess-ta]

    insect-DAT bite-and/or sting-PASS-PST-DECL

    ‘(He) was bitten and/or stung by an insect.’

    In addition to these morphological markers, the language has words like kuliko ‘and’, ttonun ‘or’ as lexical coordinators. Unlike the morphological coordinators, these coordinators can be used for both nominal and verbal coordination:

    (2) a. hay-wa tal kuliko sem-i hamkkey ha-nun kos

    sun-and moon and island-NOM together do-MOD place

    ‘the place where sun, moon, and island exist together’

    b. mak-kena ttonun phiha-l swu iss-ta

    block-or or avoid-MOD possibility exist-DECL

    ‘(You) can block or avoid it.’

  • 4 - Auxiliary and complex predicate constructions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 84-103

  • Summary

    The typical complex predicate construction is composed of more than one grammatical element, each of which contributes nontrivial syntactic and semantic information to it. Like other SOV languages, Korean employs a variety of complex predicate constructions which we can frequently find in daily language use. Complex predicate constructions in Korean can be classified into verbal and nominal ones. The former group consists of only verbal expressions, and auxiliary and serial verb constructions belong to this group. The latter has a predicative nominal and a verb. Light verb constructions and bound noun constructions belong to this group. This chapter discusses the grammatical properties of the auxiliary constructions in Korean as complex predicates.

    Types of auxiliary verbs and morphosyntactic properties

    Auxiliary verbs in Korean are traditionally subclassified according to their semantic contributions (see Suh 1996, Chang 1996, Sohn 1999, Kim 2004, among others). Consider one typical classification with examples:

    (1) a. Continuation

    phwul-e ka-ta

    solve-CONN gradually-DECL

    b. Service

    nol-a cwu-ta / ilk-e cwu-ta

    play-CONN for-DECL / read-CONN for-DECL

    c. Attempt

    mek-e po-ta / ilk-e po-ta

    eat-CONN try-DECL / read-CONN try-DECL

    d. Conjecture

    mek-na po-ta / o-na siph-ta

    eat-CONN maybe-DECL / come-CONN maybe-DECL

    e. Repetition

    wus-e tay-ta / mek-e tay-ta

    smile-CONN on-DECL eat-CONN on-DECL

    f. Hold

    mak-a twu-ta / kam-a noh-ta

    block-CONN keep-DECL / wrap-CONN keep-DECL

    g. Expectation

    mek-ko siph-ta

    eat-CONN wish-DECL

    h. Existence

    kelly-e iss-ta

    hang-CONN exist-DECL

    i. Obligation

    mek-eya ha-n-ta

    eat-CONN must-PRES-DECL

    j. Habit

    mek-kon ha-ta

    eat-CONN used.to-DECL

    Most of the auxiliary verbs carry aspectual, modal, or volitional interpretations, but they are also used as main verbs with different senses. Consider the following examples:

    (2) a. Mimi-ka hakkyo-ey ka-ss-ta

    Mimi-NOM school-LOC go-PST-DECL

    ‘Mimi went to school.’

    b. elum-i chachum nok-a ka-ss-ta

    ice-NOM gradually melt-CONN gradually-PST-DECL

    ‘The ice gradually melted away.’

    The main verb ka- in (2a) just describes the (motion/transition) event of Mimi's going to school, while its auxiliary use in (2b) denotes the ice's continuous melting process. Note that the auxiliary verb does not have a typical argument structure.

  • 10 - Passive constructions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 226-241

  • Summary

    Korean employs both lexical and syntactic passive constructions. The lexical passive is quite idiosyncratic in the sense that it is applicable for only a limited number of transitive verbs. Meanwhile, syntactic passive, formed with the passive auxiliary verb ci-, is productive in the sense that most of the transitive verbs have a corresponding syntactic passive. This chapter discusses key properties of the syntactic passive in Korean and then offers a construction-based HPSG analysis.

    Lexical vs. syntactic passive

    Korean passive constructions include both lexical and syntactic passives. Consider the following two pairs of active and passive sentences:

    (1) a. Active

    kyengchal-i pemin-ul cap-ass-ta

    police-NOM criminal-ACC arrest-PST-DECL

    ‘The police arrested the criminal.’

    b. Lexical passive:

    pemin-i kyengchal-eykey cap-hi-ess-ta

    criminal-NOM police-DAT catch-PASS-PST-DECL

    ‘The criminal was arrested by the police.’

    (2) a. Active

    ai-ka os-ul simhakey ccic-ess-ta

    child-NOM clothes-ACC severely tear-PST-DECL

    ‘The child tore the clothes severely.’

    b. Syntactic passive:

    os-i simhakey ccic-e ci-ess-ta

    clothes-NOM severely tear-CONN become-PST-DECL

    ‘The clothes were severely torn.’

    In the lexical passive (1b), the addition of the passive morpheme -hi to the predicate promotes the active object into the passive subject while demoting the active subject into the passive oblique argument. Meanwhile, in the syntactic passive (2b), the promotion and demotion of the arguments are triggered by the passive auxiliary ci-ess-ta ‘become-PST-DECL’ even though there is no morphophonemic change of the main verb ccic- ‘tear’.2 That is, the form value of the main verb is intact other than the attachment of the connective marker or linker -e.

    Lexical passive, realized with one of the passive morphemes, -i, -hi, -li, and -ki, is quite restricted and unproductive. For example, transitive verbs like the following cannot be lexically passivized:

    (3) a. ditransitive: cwu- ‘give’, tuli- ‘give’, pachi- ‘give’, neh- ‘put’

    b. benefactive: et- ‘obtain’, mantul- ‘make’, ilh- ‘lose’, top- ‘help’

    c. experience: al- ‘know’, paywu- ‘learn’, pala- ‘wish’

    d. symmetric: manna- ‘meet’, talm- ‘resemble’

    There is also a difference even among synonymous transitive verbs. For example, the verb tam-ta ‘fill-DECL’ has its lexical passive tam-ki-ta ‘put- PASS-DECL’, but its synonymous verb neh-ta ‘put-DECL’ does not have any lexical passive.

  • 1 - Theoretical foundations

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 1-25

  • Summary

    This book, encompassing both empirical and theoretical issues of the Korean grammar, develops a constraint- and construction-based HPSG (Head-driven Phrase Structure) analysis. The constraint view assumes that grammar is a system of constraints governing the relation between form and meaning. This view eventually is combined with the philosophy of Construction Grammar in which functional knowledge of language is being based on an individual's systematic collection of “form and function pairings.” The book tries to merge these two perspectives of the grammar while seeking the greatest possible precision.

    Derivational vs. constraint-based views

    The central tasks of current linguistic theory are twofold. The first task is to explain how children are able to acquire grammatical competence even though they do not receive explicit instruction or have limited access to the data. This task is thus to account for the gap between children's knowledge of language and the apparent lack of substantive input for language, also known as Plato's problem. For example, assume that Korean speakers may hear both of the following Korean sentences including a relative clause:

    (1) a. [_ppang-ul manhi sa-n] ai-ka wus-ess-ta

    bread-ACC much buy-MOD child-NOM smile-PST-DECL

    ‘The child who bought a lot of bread smiled.’

    b. [[_ppang-ul manhi sa-ss-ta-ko] Mimi-ka malha-n]

    bread-ACC much buy-PST-DECL-COMP Mimi-NOM say-Mod

    ai-ka wus-ess-ta

    child-NOM smile-PST-DECL

    ‘The child who Mimi said bought a lot of bread smiled.’

    From these two examples, speakers may infer that a subject from a matrix clause as well as from an embedded clause can be relativized. However, note that Korean speakers may not utter sentences like the following:

    (2) *[[_ppang-ul manhi sa-n] kakey-lul] Mimi-ka malha-n

    bread-ACC much buy-Mod store-ACC Mimi-NOM say-Mod

    ai-ka wus-ess-ta

    child-NOM smile-PST-DECL

    ‘(int.) The child who bought a lot of bread from the store which Mimi

    mentioned smiled.’

    Note that just like those in (1), the sentence in (2) also relativizes the subject of the embedded clause, but Korean speakers would not utter such a sentence. The question that follows is then how and why the language users, with no explicit instruction, can avoid generating such an unacceptable sentence.

    The next and related task is to explain syntactic competence in language. Native speakers have the ability to combine structures to create simple as well as complex sentences.

  • 11 - Wh-questions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 242-255

  • Summary

    Korean is a wh-in-situ language in the sense that the wh-expression stays in situ with an obligatory Q-particle marking its interrogative scope. This chapter reviews some basic properties of the wh-question construction in Korean and shows how a constraint- and construction-based HPSG can provide a robust, precise analysis of the wh-construction in the language.

    Dependency between wh-question and Q-particle

    Unlike English, Korean is a wh-in-situ language where the wh-word stays in the original position (see, among others, Hagstrom 1998, Chung 2000, Kim 2000, and Hong 2005).

    (1) a. ecey Mimi-ka nwukwu-lul manna-ss-ni?

    yesterday Mimi-NOM who-ACC meet-PST-QUE

    ‘Who(m) did Mimi meet yesterday?’

    b. ecey nwu-ka Mimi-lul manna-ss-ni?

    yeterday who-NOM Mimi-ACC meet-PST-QUE

    ‘Who met Mimi yesterday?’

    As seen from these examples, unlike English, there is no displacement (or movement) of the wh-phrase in forming the wh-questions. The wh-expression in Korean need not move to the clausal scope (clause-initial) position, though it can be optionally displaced to the position through a scrambling process:

    (2) nwukwu-lul ecey Mimi-ka manna-ss-ni?

    who-ACC yesterday Mimi-NOM meet-PST-QUE

    ‘Who(m) did Mimi meet yesterday?’

    In forming a wh-question in the language, all that is required is the presence of a Q-particle. That is, the wh-expression requires a question mood marking (Q-particle) like -ni on the matrix verb to have an interrogative meaning.Without such a Q-particle, the wh-expression can be interpreted as an indefinite pronoun (see Chung 1996, Kim 2000, Hong 2005):

    (3) a. Mimi-ka mwues-ul mek-ess-ta

    Mimi-NOM what-ACC eat-PST-DECL

    ‘Mimi ate something.’

    b. nwu-ka Mimi-lul manna-ss-ta

    who-NOM Mimi-ACC meet-PST-DECL

    ‘Someone met Mimi.’

    The wh-expression is thus ambiguous between an interrogative reading and an indefinite reading, depending on the presence of the Q-particle.

    The Q-particle can be attached to the final verb of a clause located to the right of the corresponding wh-element, marking its scope, as illustrated by the following:

    (4) a. Embedded clause wh-question

    Haha-nun [Mia-ka mwues-ul sa-ss-nunci] mwul-ess-ni?

    Haha-TOP Mia-NOM what-ACC buy-PST-QUE ask-PST-QUE

    ‘Did Haha ask what Mia bought?’

    #‘What did Haha ask if Mia bought?’

    b. Matrix clause wh-question

    Haha-nun [Mia-ka mwues-ul sa-ss-ta-ko]

    Haha-TOP Mia-NOM what-ACC buy-PST-DECL-CONN

    malha-yess-ni?

    say-PST-QUE

    ‘What did Haha say Mia bought?’

    In both cases, the wh-question is in the embedded clause but the position of the Q-particle differs.

  • 8 - Negation and related phenomena

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 169-203

  • Summary

    A common and productive method of expressing sentence negation in Korean is to introduce the negative marker an or a negative auxiliary anh-, which are called ‘short form’ (SFN) and ‘long form’ (LFN) negation, respectively. In this chapter, we show that treating the SFN as a prefix is more viable for various related phenomena such as ordering restrictions among adverbs, lexical blocking cases, plural copying, reduplication process, and the like. As for the LFN employing a negative auxiliary, this chapter shows that the Complex Predicate analysis can allow us to have a streamlined way of analyzing negation phenomena in Korean.

    Short form and long form negation

    The most typical way of expressing sentence negation in Korean is to introduce the negative marker an or mos in the preverbal position or the negative auxiliary anh- in the postverbal position:

    (1) SFN (Short Form Negation)

    a. Kim-un an ttena-ss-ta

    Kim-TOP not leave-PST-DECL

    ‘Kim did not leave.’

    b. Kim-un mos ttena-ss-ta

    Kim-TOP not leave-PST-DECL

    ‘Kim could not leave.’

    (2) LFN (Long Form Negation)

    Kim-un ttena-ci anh-ass-ta

    Kim-TOP leave-CONN not-PST-DECL

    ‘Kim did not leave.’

    The SFN an or mos in (1) occurs in the preverbal position, while the LFN anh- follows the main verb marked with -ci, as given in (2). The different position of the negator distinguishes the two forms, though they are essentially synonymous.

    Historically, the SFN an + V is developed from the adverbial negator ani and the LFN anh- from the combination of ani and ha- ‘do’. This historical development of phonological shortening in fact accompanies the recategorization of the negator from adverb to prefix for the SFN and to auxiliary for the LFN. The recategorization process can even be observed in childhood language uses (see Han and Park 1994, Hagstrom 2002, Han et al. 2007).

    (3) a. an pesu tha

    not bus ride

    ‘I do not take the bus.’

    b. na mos pap hay

    I not meal do

    ‘I can't cook.’

    Such examples, misplacing the negator in front of the object, can often be observed in early childhood, but are eliminated in adult grammar.

    There is much evidence, which we will consider in detail, that the two new forms are recategorized as prefix and auxiliary verb.

  • 2 - Noun phrases

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 26-46

  • Summary

    In Korean, the ordering of NP constituents is known to be quite flexible, but few attempts have been made to provide their precise syntactic structures. This chapter reviews the distributional properties of Korean prenominal expressions as well as relevant ordering constraints, and then offers a surface-oriented analysis of Korean NP structures. This surface-oriented, construction-based analysis, tightly interacting with lexical information, can not only capture the flexible orderings of prenominal expressions in a simple manner but also generate proper and precise NP structures.

    Basic properties

    It is well known that prenominal expressions in Korean display a high flexibility in their distributional possibilities. For example, the prenominal expression chakha-n ‘good-natured’ can either precede or follow a determiner:

    (1) a. ku chakha-n haksayng

    the good.natured-MOD student

    ‘the good-natured student’

    b. chakha-n ku haksayng

    good.natured-MOD the student

    ‘the good-natured student’

    The prenominal expressions in Korean can be largely classified into two types: deterministic and phrasal. The deterministic prenominals, which we call (lexical) determinants here, roughly correspond to determiners in English (cf. Jackendoff 1977, Huddleston and Pullum 2002 for English).

    The lexical determinants have been traditionally classified into five types in accordance with their semantic contribution (see Nam and Ko 1993, Chang 1996, Sohn 1999).

    (2) a. characteristic: say ‘new’, hen ‘old’, yeys ‘old’, ttan ‘other’, …

    b. indicative: i ‘this’, ku ‘the’, ce ‘that’, …

    c. interrogative: enu ‘which’, mwusun ‘what’, etten ‘which’, …

    d. quantificational: motun ‘all’, myech ‘some’, yele ‘several’, …

    e. numeral: han ‘one’, twu ‘two’, sey ‘three’, …

    These lexical determinants combine with a nominal head, forming an NP structure.

    Different from these lexical determinants, there are also phrasal prenominal expressions such as adjectival elements, relative clauses, and genitive-marked possessive NPs:

    (3) a. Adjectival:

    acwu [khu-ko] [pissa-n] [ppalka-n] chayk

    very big-and expensive-MOD red-MOD book

    ‘a very big and expensive red book’

    b. Relative:

    [Mia-ka ilk-un] ku chayk

    Mia-NOM read-MOD the book

    ‘the book that Mia read’

    c. Genitive:

    [Mia-uy] ku chayk

    Mia-GEN the book

    ‘Mia's book’

    These prenominal expressions are marked with the MOD suffix (-(u)n in (3a) and (3b)) or the GEN marker (uy in (3c)) and then combine with the following nominal head.

  • 13 - Relative clause constructions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 283-317

  • Summary

    Relative clause constructions in Korean have been notorious for their complexity in terms of syntax and semantics. This chapter develops a construction-based HPSG analysis of Korean relatives. The analysis is “headdriven” and “constraint-based” in the sense that the head of a relative clause and the declarative constraints on well-defined constructions play a crucial role in the formation of relative clauses. The analysis enables us to eliminate invisible elements (i.e., trace or empty operator) in licensing Korean relative clauses and further allows us to express cross-cutting generalizations among different types of relatives.

    Some key properties

    The Korean relative clause constructions are often characterized by flexible application of the relative morphology, lack of relative pronouns and omission of arguments (see, among others, Na and Huck 1993, Kim 1998b, 2004, Han and Kim 2001, Chae 2012). These properties contribute to the overall structural complexity of the relative clause constructions in Korean. Consider the following two typical relative clause constructions:

    (1) a. [mek-nun] mwul

    eat-MOD water

    ‘water that one drinks’

    b. [Kim-i mek-un] ceskalak

    Kim-NOM eat-MOD chopsticks

    These examples illustrate that the interpretation of relative clauses is in part dependent upon world knowledge or contexts. The natural reading of (1a) is the one in which the direct object is relativized with the subject being unexpressed. It would be difficult to interpret this as a subject relativization since “water” cannot eat/drink anything. The relative clause (1b) could be ambiguous: the most natural reading is “the chopsticks that Kim ate with.” In an imaginary context (where chopsticks are made from edible materials), one may have a reading like “the chopsticks that Kim ate,” where the direct object is relativized.

    Further complications relating to Korean relative clause constructions also could be found from examples like (2) (Na and Huck 1993, Han and Kim 2001):

    (2) a. [NP[S[_i_ j tha-ko tani-nun] cha j -ka] mesci-n]

    ride-CONN drive-Mod car-Nom stylish-Mod

    sinsai

    gentleman

    (lit.) ‘gentlemani who the car j that [hei] is driving is stylish’

    b. [NP [S [_i_ j kac-ko iss-nun] khemphyuthej-ka]

    MAC-i-n] kyoswui

    MAC-COP-MOD professor

    (lit.) ‘professori who the computer j that [hei] has is a MAC’

    In (2), the relative heads sinsa ‘gentleman’ and kyoswu ‘professor’ have been relativized out of the more deeply embedded relative clause, violating the CNPC (Complex Noun Phrase Constraint).

  • 7 - Serial verb constructions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 144-168

  • Summary

    The so-called serial verb construction (SVC) is a complex predicate structure consisting of two or more main verbs but denotes one single complex event with a series of connected subevents. This chapter discusses the grammatical properties of Korean SVCs and provides a lexicalist, construction-based HPSG analysis. In particular, we identify two typical types of SVCs in Korean as well as pseudo-SVCs, each of which has its own idiosyncrasies while sharing general properties with the other, related constructions. We also discuss the monoclausal properties of the SVCs and then offer a construction-based analysis that allows tight interactions between the lexicon and constructional constraints.

    Serial verbs and general properties

    The so-called serial verb constructions (SVC) allow more than one verb to occur without any specific coordination or subordination markings. Korean examples given in (1) are taken to belong to such constructions (Lee, S. 1993, Chung 1993, Chung and Kim 2008, among others):

    (1) a. Mia-ka hakkyo-ey kel-e ka-ss-ta

    Mia-NOM school-to walk-CONN go-PST-DECL

    ‘Mia went to school on foot.’

    b. Mia-ka cwul-ul cap-a tangky-ess-ta

    Mia-NOM rope-ACC hold-CONN pull-PST-DECL

    ‘Mia held a rope and pulled it.’

    The serial verbs here denote actions that are closely connected to each other and can be considered to be part of the same complex event. For instance, (1a) represents an event of walking and its manner while (1b) denotes a complex event of holding and pulling a rope simultaneously.

    The generation of such Korean SVCs, in which the first main verb is marked with the meaningless connective linker -a/e, is quite productive, as attested by the naturally occurring examples with the verb mek-ta ‘eat-DECL’:

    (2) a. nanwu-e mek-ta ‘divide and eat’

    divide-CONN eat-DECL

    b. kkulh-ye mek-ta ‘boil and eat’

    boil-CONN eat-DECL

    c. mantul-e mek-ta ‘make and eat’

    make-CONN eat-DECL

    d. cap-a mek-ta ‘catch and eat’

    catch-CONN eat-DECL

    e. cip-e mek-ta ‘pick up and eat’

    pick.up-CONN eat-DECL

    f. ssip-e mek-ta ‘chew and eat’

    chew-CONN eat-DECL

    g. kwu-we mek-ta ‘roast and eat’

    roast-CONN eat-DECL

    h. ppal-a mek-ta ‘suck and eat’

    suck-CONN eat-DECL

    The examples here, extracted from the Sejong POS-tagged Corpus, show us that the activity verb mek- ‘eat’ can combine with various types of verbs and form the SVCs. Such productivity implies that the SVC is a grammatical process that the language employs.

  • 6 - Verbal nouns and light verb constructions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 121-143

  • Summary

    One of the most widely used constructions in Korean is the so-called light verb construction (LVC) involving a verbal noun (VN) together with the light verb ha- ‘do’. This chapter discusses the light verb's grammatical functions as well as the VN's mixed properties, both of which have provided a challenge to syntactic analyses with a strict version of X-bar theory. It then shows that the mechanism of multiple classification of category types with systematic inheritance can provide an effective way of capturing these mixed properties. The chapter also restates the well-known argument composition properties of the LVC and reinforces them with a construction-based HPSG analysis.

    Basic properties

    In Chapter 4, we have seen that the verb ha- is used as an auxiliary verb:

    (1) a. motwu-ka i chayk-ul ilk-eya ha-yess-ta

    everyone-NOM this book-ACC read-CONN do-PST-DECL

    ‘Everyone had to read this book.’

    b. motwu-ka i chayk-ul ilk-kon ha-yess-ta

    everyone-NOM this book-ACC read-CONN do-PST-DECL

    ‘Everyone used to read this book.’

    The auxiliary verb ha- here induces the meanings of obligation and habitual behavior, respectively. The verb ha- can also be used as a main verb:

    (2) a. Mimi-nun cokumaha-n saep-ul ha-yess-ta

    Mimi-TOP small-MOD business-ACC do-PST-DECL

    ‘Mimi ran a small business.’

    b. Mimi-nun yeyppun kwikeli-lul ha-yess-ta

    Mimi-TOP pretty earing-ACC do-PST-DECL

    ‘Mimi was wearing a pretty earing.’

    Sentence (2a) describes the action of running a business, while (2b) denotes a state of wearing an earing. The use of the verb ha- here is linked to the core meaning of ‘running’ and ‘wearing’ in each case.

    The verb ha- can also be used with an activity-denoting noun (so-called verbal noun):

    (3) a. Mimi-nun DNA kwuco-lul yenkwu ha-yess-ta

    Mimi-TOP DNA structure-ACC research do-PST-DECL

    ‘Mimi studied the DNA structure.’

    b. ku yenghwa-lul hankwuk-eyse chwalyeng ha-yess-ta

    the movie-ACC Korea-LOC shoot do-PST-DECL

    ‘They made the movie at Korea.’

    c. ku ilum-un namwu-lopwuthe yulay ha-yess-ta

    the name-TOP tree-SRC origin do-PST-DECL

    ‘The name was originated from the tree.’

    Unlike the auxiliary and main verb use, the verb ha- in these examples is optional and its absence does not change the sentential meaning. In this sense, the verb ha-, called a light verb, is semantically vacuous, and it is the preceding VN (verbal noun) that denotes the action in question (see, among others, Chae 1996, Kim et al. 2007, Park 2013a).

  • 3 - Case system

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 47-83

  • Summary

    In a free-word-order language like Korean, case markings play a central role in licensing arguments, signaling grammatical functions, and even marking information structure. The case system in Korean displays further intriguing patterns concerning phenomena such as free and forced case alternations, case assignment on non-nominal (adverbial as well as verbal) elements, and so forth. Such a complex array of case patterns has cast theoretical as well as computational challenges. Departing from the previous configurationbased or lexicalist approaches, this chapter tries to provide a treatment of case assignment patterns in Korean within a construction-based HPSG perspective.

    Two basic issues

    Nominal expressions in Korean can carry case markers. Depending on the relationship that these case markers express, cases can be generally divided into two groups: grammatical case markers and semantic case markers. The examples in (1) are some canonical examples:

    (1) a. haksayng-tul-i chayk-ul ilk-ess-ta

    student-PL-NOM book-ACC read-PST-DECL

    ‘The students read a book.’

    b. totwuk-tul-i pomwul-ul sangca-ey swumky-ess-ta

    thief-PL-NOM treasure-ACC box-LOC hide-PST-DECL

    ‘The thieves hid a treasure in the box.’

    The nominative (NOM) and accusative (ACC) grammatical cases here indicate syntactic functions such as subject and object. The grammatical cases NOM and ACC are assigned by a verbal expression like manna- ‘meet’ in (2a), whereas the genitive (GEN) case is licensed by a nominal expression, as seen from (2b):

    (2) a. Mimi-ka/*uy chinkwu-lul/*-uy manna-ss-ta

    Mimi-NOM/*GEN friend-ACC/*GEN meet-PST-DECL

    ‘Mimi met a friend.’

    b. Mimi-uy/*i chinkwu

    Mimi-GEN/*NOM friend

    ‘Mimi's friend’

    Meanwhile, a semantic case like locative (LOC) in (1b) expresses the semantic function of the NP sangca ‘box’. Other semantic cases, closely related to the semantic role of the nominal in question, include goal (GOAL), instrument (INST), source (SRC), comitative (COMIT), and the like, as illustrated in the following:

    (3) a. haksayng-tul-i pesu-lo hakkyo-ey ka-ss-ta

    student-PL-NOM bus-INST school-GOAL go-PST-DECL

    ‘Students went to the school by bus.’

    b. haksayng-tul-i chinkwu-tul-kwa nol-ass-ta

    student-PL-NOM friend-PL-COMIT play-PST-DECL

    ‘Students played with friends.’

    Given these two types of case markings, we could then classify case values as in the following hierarchy:

    The hierarchy tells us that case markings in Korean are first classified into two subgroups: gcase (grammatical or structural case) and scase (semantic case).

  • 14 - Honorification

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 318-332

  • Summary

    Honorific agreement is one of the main properties in languages like Korean, playing a pivotal role in appropriate communication. This makes the deep processing of honorific information crucial in various computational applications such as spoken language translation and generation. We argue that, contrary to the previous literature, an adequate analysis of Korean honorification involves a system that has access not only to morphosyntax but to semantics and pragmatics as well. Along these lines, this chapter offers a constraint-based HPSG analysis of Korean honorification.

    Basic properties of honorific agreement

    Subject agreement

    Honorification, one of the main features of spoken language in Korean, plays a key role in proper and successful verbal communication (Chang 1996, Sohn 1999, Kim and Sells 2008b, Kim 2011). The Korean honorific system basically requires that when the subject is in the honorific form (usually with the marker -nim), the predicate also be inflected with the honorific form -(u)si, as in (1a):

    (1) a. sensayng-nim-i wus-usi-ess-e

    teacher-HON-NOM smile-HON-PST-DECL

    ‘The teacher smiled.’

    b. #sensayng-nim-i wus-ess-e

    teacher-HON-NOM smile-PST-DECL

    ‘The teacher smiled.’

    Examples like (1a) are most natural in the sense that the subject agrees with the matrix verb in terms of honorification. However, those like (1b), where there is disagreement between the two, may be used when the speaker does not honor the referent of the subject (marked by #).

    Considering that agreement disconcord can be used in a limited context, one may assume honorific agreement in Korean is purely pragmatic. This pragmatic direction has led literature to assume that using the -nim and -si form of verbs is a matter of gradience and appropriateness rather than grammaticality (see, among others, Pollard and Sag 1994, Chang 1996, Lee and Ramsey 2000, Aarts 2007). However, one key property that has often been ignored is that honorific agreement must be observed when the subject is non-human, as illustrated by the examples in (2) (Sohn 1999, Kim 2011):

    (2) a. cha-ka o-(*si)-ess-e

    cha-NOM come-HON-PST-DECL

    ‘The car came.’

    b. kwukhoy-ka ku pepan-ul simuy-ha-(*si)-ess-e

    congress-NOM the bill-ACC review-do-HON-PST-DECL

    ‘The congress reviewed the bill.’

    If we rely only on pragmatic information, we would have difficulties understanding why, unlike the disagreement data in (1b), disagreement cases like (2) are hardly found in real language usages.

  • 12 - Topic and focus constructions

  • Jong-Bok Kim, Kyung Hee University, Seoul
  • Book:
    The Syntactic Structures of Korean
    Published online:
    05 July 2016
    Print publication:
    28 July 2016, pp 256-282

  • Summary

    Korean is often called a “topic-prominent language” in the sense that it organizes its syntax to emphasize topic–comment structures. Topic expresses what the given sentence is about while comment is what is being said about the topic. Part of the comment structure includes “focus,” representing new information. This chapter identifies several different types of topic and focus constructions and discusses their grammatical properties. It then offers a construction-based analysis that can in a simple manner state the shared properties among a family of related constructions.

    Topic constructions

    Basic properties

    Topic constructions, representing what the sentence is about, are typically marked by the sentence-initial expression with the marker -(n)un. Consider the following pair:

    a. ai-tul-i pomwul-ul chac-ass-ta

    child-PL-NOM treasure-ACC find-PST-DECL

    ‘The children found the treasure.’

    b. ku pomwul-un [ai-tul-i _chac-ass-ta]

    the treasure-TOP child-PL-NOM find-PST-DECL

    ‘As for the treasure, the children found it.’

    The sentence (1a) has canonical word order, while (1b) topicalizes the object from the underlined position. The topic expression prompts a definite interpretation, and the referent of the topic is thus familiar in discourse.

    Note that not only an argument (subject or object) but also an adverbial expression can be topicalized:

    (2) a. onul-un nalssi-ka cham coh-ta

    today-TOP weather-NOM very good-DECL

    ‘As for today, the weather is really good.’

    b. i tosi-nun ankay-ka manhi kki-n-ta

    this city-TOP fog-NOM much rise-PRES-DECL

    ‘As for this city, the fog rises a lot.’

    In these examples, the adverbial expressions onul ‘today’ and i tosi ‘this city’ function as the topic in the sentences, describing what is being talked about.

    In addition, there is another type in which the topic of a given sentence is a nominal expression not functioning as an argument of the sentence in question.

    (3) a. kkoch-un cangmi-ka alumtap-ta

    flowers-TOP rose-NOM pretty-DECL

    ‘As for flowers, roses are pretty.’

    b. sayngsen-un kotunge-ka masiss-ta

    fish-TOP mackerel-NOM tasty-DECL

    ‘As for fish, mackerel is tasty.’

    In these examples, kkoch-un ‘flowers-TOP’ and sayngsen-un ‘fish-TOP’ are neither an argument of the matrix verb nor an adjunct expression themselves. The sentences headed by pure intransitive verbs here are fully saturated ones even without these topicalized expressions. The grammatical functions of the topicmarked nominals are thus rather “adverbial,” giving an interpretation like “as for flowers” or “as for fish.”

Page 29 of 114