University of Southern California Dissertations and Theses

The speaker's choice of using object-marking in informal spoken Korean

(USC Thesis Other)

The speaker's choice of using object-marking in informal spoken Korean

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

i

THE SPEAKER’S CHOICE OF USING OBJECT-MARKING IN
INFORMAL SPOKEN KOREAN

by
Won Kyung Na
August 2016

A Dissertation Presented to the
FACULTY OF THE USC GRADUA TE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
Doctor of Philosophy
(EAST ASIAN LANGUAGES AND CULTURES)

ii

Acknowledgements
First and foremost, I would like to express my deepest gratitude to my advisor Professor Namkil
Kim for his tremendous help and thoughtful guidance throughout my journey as a graduate
student. I would like to thank my dissertation committee members Professor Audrey Li and
Professor Andrew Simpson for their insightful comments and encouragement, which allowed me
to widen my research from various perspectives. I am also thankful to my qualifying exam
committee members Professor Elsi Kaiser and Processor Roumyana Pancheva for their thoughtful
comments, which allowed me to generate and further develop ideas for my dissertation.
During data collection and writing, Dr. Hee Ju and Dr. Hyunjung Ahn provided needed
encouragement and insights. Their help and advice on both research as well as on my career have
been invaluable. I would also like to recognize my colleagues, Heeyoung Ahn and Chad Walker,
for generously sharing their time and ideas. Especially, I would like to express my appreciation to
Chad Walker for proofreading my dissertation. I would also like to thank Dr. Helen Chung who
generously helped me balance teaching and studying at USC when I tried to juggle teaching and
completing my dissertation.
My time at USC was made enjoyable in large part due to the good friends and groups that became
a part of my life. Special thanks to my best friend Fei Huang, my mentor Dr. Seonkyung Jeon,
and the Korean Bible Study members for their endless support and encouragement. I would also
like to thank my family. Words cannot express how grateful I am to my mother, father, and
younger brother for their unconditional love, support, encouragement, and belief in my abilities.
Lastly, I dedicate this dissertation to my god, my good Father who has carefully guided me
throughout this journey with his unfailing love and patience.

iii

Abstract
This paper examines the speaker’s choice of using the Korean object particle -(l)ul in spontaneous
informal conversation and clarifies that the choice of (l)ul-marking is a device to facilitate the
listener’s selective focusing on either hard-to-process or salient information for successful
communication. (l)ul-marking is frequently used to facilitate the listener’s selective focusing on
difficult-to-process information which requires the listener’s extra attention (i.e., the object NP
with a modifier, the object NP not immediately adjacent to its verb, and the object NP in
sentences with non-canonical word order).
In addition, (l)ul-marking is consistently used in discourse contexts where the speaker disagrees
with what has been said in the prior talk. When there is an ongoing disagreement between the
interlocutors, the speaker needs to direct the listener’s attention to the information that is contrary
to the prior utterance or action of co-present or non-present parties. Disagreement can vary in
range from weak disagreement, such as expressing surprise or wondering about the prior
utterance or action that is contrary to the speaker’s belief or expectation, to serious disagreement
such as showing a negative or sarcastic attitude (e.g., dispreference, disbelief, objection, mockery,
etc.) or asking for the other speaker’s clarification, self-correction, or backdown of the prior talk.
Thus, what the particle -(l)ul does in conversation is more than simply indicating the grammatical
role of the NP in a sentence; specifically, -(l)ul can (i) emphasize an assertion challenging the
prior utterance or action of the co-present or non-present parties in rhetorical questions, self-
addressed questions, and restatements of the speaker’s prior utterance; (ii) initiate repair on the
information involving the referent that needs to be clarified; (iii) distinguish contrastive referents
in discourse contexts of disagreement; and (iv) indicate a discourse topic and project further
descriptions or explanations.

iv

Table of Contents
Acknowledgements ii
Abstract iii
Table of Contents iv
The Yale Romanization of Korean vi
Abbreviations vii
Transcription Conventions viii

Chapter 1: Introduction 1
1.1 Problem 1
1.2 Purposes 3
1.3 Organization 6

Chapter 2: Previous Approaches to Object-Marking 8
2.1 Linguistic Factors Affecting Object-Marking 9
2.1.1 The Size of the Object NP 9
2.1.2 Adjacency Condition and Word Order 10
2.1.3 The Form of the Object NP 13

2.2 Pragmatic Factors Affecting Object-Marking 20
2.2.1 Level of Formality in Speech 20
2.2.2 Shared Knowledge 22
2.2.3 Informational Predictability 25
2.2.4 Informational Salience 28
2.2.4.1 Indicating the Discourse Topic 28
2.2.4.2 Emphasizing an Action 30
2.2.4.3 Initiating Repair 31
2.2.4.4 Distinguishing Contrastive Referents 32
2.3 Summary 36

Chapter 3: Methodology 38
3.1 Discourse Analysis 38
3.1.1 Regularities in Data 38
3.1.2 Communicative Purposes 39

3.2 Data 42
3.2.1 Focus of Analysis 43
3.2.2 Factors Considered 45

Chapter 4: Analysis 46
4.1 Emphasizing an Assertion Challenging the Prior Utterance or Action 49
4.1.1 Challenging the Prior Utterance or Action in Rhetorical Questions 49
4.1.2 Expressing a Feeling of Wondering in Self-Addressed Questions 63
4.1.3 Challenging the Prior Utterance or Action in Restatements 66

v

4.2 Initiating Repair on the Prior Utterance or Action 71
4.2.1 Showing Surprise or Disbelief Using a Partial Repeat of the Prior 72
Utterance with an Adverb
4.2.2 Showing Surprise or Disbelief Using a Candidate Understanding 74
of the Prior Utterance with -nun mali-/iyaki- ‘You Mean’
4.2.3 Showing Disagreement Using a Partial Repeat of the Prior Utterance 79
with the Emphatic Sentence Ending -ko
4.2.4 Showing Disagreement Using the (l)ul-Marked Question 83
Word mewl ‘What’ Plus a Partial Repeat of the Prior Utterance

4.3 Distinguishing Contrastive Referents 86
4.4 Indicating the Discourse Topic 92
4.5 Summary 96

Chapter 5: Discussion 99
5.1 Substitution and Stacking 108
5.2 Double Object Construction 116
5.2.1 Ditransitive Verbs 116
5.2.2 Adverbials 126
5.2.3 Possessor-Possessee 133
5.3 Summary 136

Chapter 6: Conclusion and Implications 138
6.1 Extended Discourse Context of Disagreement 142
6.2 Interaction between the Form, Intonation, and Meaning 145
6.3 Korean Learners’ Use of (l)ul-Marking 147

References 149

vi

The Yale Romanization of Korean
In this dissertation, the Yale Romanization System has been adopted to transcribe Korean spoken
data. The following table is based on Kim (2008).

vii

Abbreviations
An asterisk [ ∗] indicates that the sentence is clearly ungrammatical or inappropriate due to
grammatical constraints or discourse-pragmatic constraints. The following abbreviations are used
in the dissertation.
ADV: adverbializer
CL: classifier
CONJ: conjunctive
DAT: dative particle
DIR: directional particle
EMP: emphatic sentence ending
EVI: evidential marker
GEN: genitive marker
HON: honorific suffix
HOR: hortative particle
INST: instrumental particle
LOC: locative particle
NEG: negative particle
NOML: nominalizer
OBJ: object particle
PL: plural marker
POL: polite suffix
PROS: progressive suffix
PST: past tense suffix
PRES: present tense suffix
Q: question marker
QT: quotative marker
RL: relativizer
SE: sentence ending
SUB: subject particle
TOP: topic particle

viii

Transcription Conventions
The followings are the key features of the transcription conventions adopted in this dissertation.

[ onset of overlapping talk
] end of overlapping talk
= latching of two phrases or sentences
::: lengthened sound
. falling or sentence-final intonation
? rising intonation
, continuing intonation
! intonation of surprise or forcefulness
! more than one line of the talk has been removed from the transcript
(( )) transcriber’s descriptions of talk or action
( ) inaudible or indecipherable talk

1

1. Introduction
1.1 Problem
This dissertation attempts to explain an elusive phenomenon that has been studied for some time
now involving the variations in a speaker’s use of the Korean object particle -(l)ul in informal
conversation. It is generally known that in written Korean, postpositional particles are obligatorily
attached to noun phrases (NPs) to indicate their grammatical roles such as subject or object.
(1) Mary-ka pap-ul mek-ess-ta
Meyli-SUB meal-OBJ eat-PST-SE
‘Mary has eaten a meal.’
For example, the particle -ka indicates that its marked NP ‘Mary’ is the subject, an agent
controlling the action of the verb. The particle -ul marks the the direct object pap ‘meal,’ which
receives the action of the verb mek- ‘to eat.’ However, in spoken Korean, these particles are not
always used. In particular, the omission of the object particle -(l)ul is often observed in informal
conversation.
(2) Speaker A asks B if he has eaten.
A: mwe-ø hay? pap-ø mek-ess-e?
what-(OBJ) do meal-(OBJ) eat-PST-SE
‘What are you doing? Have you eaten a meal?’

B: ani, acik an mek-ess-e
no yet NEG eat-PST-SE
‘No, I haven’t eaten yet.’
In (2), the two objects mwe ‘what’ and pap ‘meal’ are not overtly (l)ul-marked. What makes it
more complicated, however, is that informal conversation is not completely object-particle free.
The use of particle -(l)ul is preferred in certain discourse contexts. See (3)~(7).
(3) nwuka keki-l ka
who there-OBJ go
‘Who would go there?’
[Implied: Nobody would go there.]

(4) mayil cenyek-ul sa-mek-nun-tako-yo?
everyday dinner-OBJ buy-eat-PRES-EMP-POL
‘Are you saying she eats dinner out every day?’

2

[Implied: I can’t believe she eats dinner out every day.]

(5) kuliko i ipwul-ul ne-l cwu-lkey
and this blanket-OBJ you-OBJ give-will
‘And I will give you this blanket.’
[Implied: I insist that I will give you this blanket.]

(6) way myelchi-lul cwe-yo kimchi-lul tal-la ike-ya
why anchovy-OBJ give-POL kimchi-OBJ give-QT this-to be
‘Why are you giving me anchovies? What I want from you is to give me Kimchi!’
[Implied: You should give me Kimchi not anchovies!]

(7) hwacangphwum-ul sa-ss-nuntey eyu amoley toykey nappu-tela. amoley
cosmetics-OBJ buy-PST-CONJ gee Amore very bad-EVI Amore
‘I bought some Amore cosmetics and, boy, they were very bad.’
Previous studies on the Korean object particle -(l)ul have suggested that NPs do not have to be
(l)ul-marked if they are easily identifiable as the object of a sentence or if they are not
emphatically focused (Sohn 1999; Lee & Ramsey 2000). Moreover, “sharedness between
communicators” is also considered to be a determining factor in the occurrence or non-occurrence
of -(l)ul: when NPs are already shared by the speaker and hearer, their grammatical relations do
not necessarily have to be specified by -(l)ul (Lee and Thompson 1987).
More recent studies show that even though the omission of -(l)ul is frequently observed in
informal spoken Korean, object-marking appears consistently in certain discourse contexts: it is a
device to facilitate the listener’s selective focusing on either difficult-to-process (T-H Kim 2008)
or salient information which requires the listener’s attention (Fujii and Ono 2000; Yang 2007).
These previous studies have shed light on the use of the object particle -(l)ul in spoken Korean:
their findings show that the speaker is sensitive to the listener’s mind and signals his or her focus
of attention on hard-to-process or salient information. However, there are no clear explanations
for why speakers tend to overtly produce -(l)ul for more salient information in certain contexts,
and thus we still do not have a full understanding of the mechanism that controls the speaker’s
use of the particle -(l)ul.

3

1.2 Purposes
The purpose of this dissertation is to find the answers to the following questions by analyzing
individual segments of informal conversations in which the use of the Korean object particle -
(l)ul is (or is not) preferred.
(i) What are the significant linguistic or pragmatic factors affecting the speaker’s choice of
using zero- or (l)ul-marking in informal spoken Korean?

(ii) What are the pragmatic functions of -(l)ul in informal spoken Korean?

(iii) How does (l)ul-marking contribute to the achievement of the speaker’s communicative
purposes?

(iv) What are the linguistic forms or patterns that frequently co-occur with -(l)ul?
The first question is asked to investigate the validity of the factors noted in previous studies that
are claimed to determine the speaker’s choice of using zero- or object-marking and to find a set of
linguistically and pragmatically meaningful factors that significantly affect the speaker’s choice
of using -(l)ul in informal spoken Korean. My focus is on first investigating such factors in the
Korean language and then seeking some universal qualities among them across languages.
The second question deals with the pragmatic functions of the particle -(l)ul. The object-
marking function of -(l)ul alone does not explain every role that it plays in natural spoken
discourse. Researchers have claimed that (l)ul-marking is not simply optional in conversation and
is not solely used to indicate the grammatical role of the NP. Rather, (l)ul-marking is a device
available to the speaker in conversation to provide an extra-linguistic signal for either hard-to-
process or salient information, which directs the hearer to pay attention or focus on the marked
elements, and hence facilitates smooth information processing and successful communication.
Thus, the speaker’s choice of using zero- or (l)ul-marking can be affected by linguistic or
pragmatic factors in conversation. To date, the following pragmatic functions of -(l)ul have been
discovered: projecting further descriptions or explanations about a discourse topic which
continues to be talked about in a stretch of discourse, emphasizing an action performed by the

4

actor, distinguishing two or more contrastive referents, and clarifying a referent which is not
clearly understood by participants (Fujii and Ono 2000; Yang 2007). This indicates that the
particle -(l)ul is not optional but is used in certain discourse contexts. The speaker’s choice of
using -(l)ul is also a device available in conversation and not in written language since it relates to
the interactional nature of conversation. The speaker is sensitive to the listener’s mind, and
object-marking helps the listener’s selective focus on information for successful communication
(Fujii and Ono 2000). Therefore, this dissertation will investigate the validity of the pragmatic
functions discussed in previous studies while discovering and describing additional pragmatic
functions of the Korean object particle -(l)ul in particular discourse contexts.
The findings of previous studies have contributed to the discovery of some of the pragmatic
functions of -(l)ul in actual spoken discourse. However, there has yet to be a discussion about the
role of (l)ul-marking in the interpretation of the speaker’s intended meaning. People produce
language to do certain things: to express ideas or beliefs, to explain something, to get other people
to do certain things, to think in a certain way, and so on (Widdowson 2007). The listener has to
make inferences about what is being said by the speaker in a particular context to recognize the
invisible meaning implied in the speaker’s utterances, interpret the speaker’s intention, and act or
reply accordingly. Thus, the third research question above is asked to identify the role of using
the particle -(l)ul in the interpretation of the speaker’s intended meaning and the achievement of
the speaker’s communicative purposes.
The final issue deals with the study of the interaction between the linguistic form and its
pragmatic functions. Discovering regularities in the speaker’s use of -(l)ul and describing
recurring linguistic forms or patterns frequently observed in each pragmatic function of -(l)ul in
actual spoken data would contribute to obtaining a more comprehensive understanding of the
mechanism that controls the speaker’s use of -(l)ul in informal conversation. In addition, the
findings of this dissertation can be applied to the teaching of Korean. Because the use of the
object particle -(l)ul is not simply optional in spoken Korean, describing its pragmatic functions

5

and co-occurring particular linguistic forms or patterns is important work in the development of
more effective teaching strategies for Korean learners.

6

1.3 Organization
This dissertation is organized as follows: Chapter Two is a literature review of the past attempts
to analyze the use of zero- or object-marking in Korean and Japanese. The past analyses of the
phenomenon have originated from various linguistic traditions such as generative grammar,
sociolinguistics, psycholinguistics, and discourse pragmatics. Although a number of significant
discoveries and proposals have been made within the prior literature, there has yet to be an
attempt to integrate the insights from all these different fields into one concerted research effort.
Thus, the purpose of this chapter is to review these previous inquires from different fields of
linguistics and integrate them into a comprehensive set of possible factors, which are to be
investigated in the later chapters.
Chapter Three, Methodology, provides a brief description of the primary research method
used in this dissertation, Discourse Analysis. The spoken data used for the analysis, including the
selection criteria of the object NP and the linguistic and discourse-pragmatic factors examined in
the analysis, will also be introduced.
Chapter Four will discuss the following issues: a) discovering the pragmatic functions of -(l)ul
in informal spoken Korean; b) describing the linguistic forms or patterns frequently observed in
each discourse context where object-marking is preferred. In addition, considering the fact the
object particle -(l)ul is often omitted in informal spoken Korean, when -(l)ul is overtly produced
by the speaker it must help the speaker express the intended meaning to be conveyed through his
or her utterances. Thus, another goal of this chapter is to examine how object-marking helps the
speaker achieve certain communicative purposes and how object-marking facilitates the listener’s
interpretation of the speaker’s intended meaning.
Chapter Five discusses the two closely related grammatical phenomena of “Substitution,”
which refers to the speaker’s choice of using -(l)ul in the place of other particles, and the so-
called “Double Object Construction (DOC),” which refers to the use of -(l)ul for two arguments

7

in a sentence. In particular, DOC has received considerable attention in both the Korean and
Japanese literature (O’Grady 1991; Hong 1991; Han 1991; M-J Kim 2001; Schütze 2001). Most
of these previous studies attempt to demonstrate the syntactic and semantic characteristics of
DOC using case theory. In addition, the data used in the previous research mainly consists of
decontextualized sentences based on native speaker intuition. Thus, the goal of this chapter is to
analyze actual spoken data and discover certain discourse contexts where the speaker tends to
replace other particles with -(l)ul and where the speaker chooses to use (l)ul-marking for more
than one argument in a sentence.
Chapter Six discusses the conclusion and implications of this dissertation. In particular, the
interaction of the linguistic form and intonation in the realization and interpretation of the
speaker’s intended meaning will be discussed as a future research topic.

8

2. Previous Approaches to Object-Marking
In this chapter, previous studies on zero- or object-marking in Japanese and Korean from various
linguistics fields including syntax, sociolinguistics, psycholinguistics, and discourse pragmatics
will be reviewed. Like Korean, Japanese is also a SOV (Subject-Object-Verb) language in which
grammatical relations such as subject and object are marked by post-nominal particles.
(8) Taro-ga gohan-o tabeta
Taro-SUB meal-OBJ ate
‘Taro had a meal.’

(9) A: Taro-ga ima nani-ø shi-teiru?
Taro-SUB now what-(OBJ) do-PROG
‘What is Taro doing now?’

B: gohan-ø tabe-teiru
Meal-(OBJ) eat-PROG
‘(Taro is) having a meal (now).’
The subject NP, Taro in (8) is marked with the particle -ga and the object NP, gohan ‘meal,’ is
marked with -o. The omission of the Japanese object particle -o also frequently occurs in spoken
language, as in (9) where the question word nani ‘what’ and the object NP gohan ‘meal’ are left
unmarked.
Observing the linguistic and pragmatic factors related to the use of unmarked bare forms or o-
marked forms in Japanese will help us understand the use of the Korean object particle -(l)ul in
informal conversation. In addition, the linguistic and pragmatic factors claimed to affect this
phenomenon will be discussed and integrated into a comprehensive set of possible factors.

9

2.1 Linguistic Factors Affecting Object Marking
2.1.1 The Size of the Object NP
It has been observed that the larger the object NP (e.g., NP with a modifier), the more object-
marking is used (Watanabe 1986; Kim 2008). In the analysis of T-H Kim (2008), the number of
object modifiers is counted as a means of measuring how large the object NPs is.
(10) Object NP without a modifier
waynyamyen nay-ka yosay mak hamyense kunyang
because 1SG-NOM these days hard doing just

kyeysok ku mwenya liphothu-ø ssumyense
continuously that what report-( ) writing
‘Because, these days, while I was continuously writing a report…’
(11) Object NP with a modifier
nay-ka [cikum pang-ey iss-nun] ke-l ta
1SG-NOM now room-LOC be-REL thing-ACC all

chiwoyaha-ki-ttaymwuney, an ka-kwu
clean.up-NMLZ-because NEG go-and
‘Since I have to clean up things that are now in my room, (I) won’t go and…’
(T-H Kim, 2008, p.219)
While the object NP liphothu ‘report’ in (10) is not modified by any preceding modifiers, the
object NP ke ‘thing’ in (11) is modified by the three words in brackets, cikum pangey issnun ‘that
are now in my room.’ Here, the use of -(l)ul seems to be more common for longer object NPs
than it is for shorter object NPs. This tendency shows the relationship between word length and
processing difficulty. Longer object NPs tend to carry more information than single object NPs.
For the hearer, longer NPs with modifying words are not as easily recognizable as objects with
respect to their predicates due to their extended length. Thus, the speaker is more likely to mark
such longer NPs with -(l)ul to provide an additional linguistic signal for hard-to-process
information, which may help the hearer decode the sentence by indicating how the utterance is to
be segmented.

10

2.1.2 Adjacency Condition and Word Order
The object NP is often left unmarked when it is immediately followed by the predicate of a
sentence (Tsutsui 1984; Watanabe 1986; Matsuda 1995; Fujii and Ono 2000; T-H Kim 2008).
According to Tsutsui (1984), “unless the speech is very formal, the ellipsis of the o of an NP-o in
a sentence is natural if the NP-o is immediately followed by the predicate of the sentence” (p.132).
(12) a. eiga o/ø mi ni ikimashoo.
movie see to let’s go
‘Let’s go to see a movie.’

b. eiga o/ ??ø hisashiburini mi ni ikimashoo.
after a long while
‘Let’s go to see a movie. We haven’t seen one for a long time.’

c. hisashiburini eiga o/ ø mi ni ikimashoo.
after a long while
‘Let’s go to see a movie. We haven’t seen one for a long time.’
(Tsutsui, 1984, p.133)
If there is an element which intervenes between the o-marked NP and its predicate, such as
hisashiburini ‘after a long while’ in (12b), the omission of o is unnatural. However, if this
intervening element is placed before the o-marked NP as in (12c), the omission of o is preferred.
It has also been claimed that the object NP is more frequently unmarked when it appears
immediately before its predicate. In Watanabe (1986), the referential complexity of the NP is
measured by counting the number of other NPs separating it from the main verb within the same
clause.
(13) Percentage of interfering words between NP and verb

Naked NP Noun Case
[Noun Verb] 93.7% (15) 63.6% (28)
[Noun X Verb] 6.3% (1) 36.4% (16)
Total 100% (16) 100% (44)

(Watanabe, 1986, p.345)

Table (13) shows that when the object NP is adjacent to the verb, it is more likely to be left
unmarked. However, if there are any intervening words between the object NP and its predicate,
the occurrence of bare NPs significantly drops to 6.3%.

11

Similarly, Matsuda (1995) says that any intervening elements between the object NP and its
predicate have a significant effect on the likelihood of the omission of o.
(14) [+adjacent]
hora, koomuin siken-ø ukeru-kara-sa
see government employee’s exam-(ACC) take because-FP
‘See, because I’m going to take the government employee’s exam.’

(15) [-adjacent]
sore-o koo, ma, tumande, wain-o nomi-nagara…
that-ACC like this well eat wine-ACC drink while
‘While eating it like this, drinking wine…’
(Matsuda, 1995, p.184-185)
Because the object NP koomuin siken ‘government employee’s exam’ in (14) is immediately
followed by the verb ukeru ‘to take a (test),’ it is more likely to be left unmarked. When the two
intervening words koo ‘like this’ and ma ‘well’ are placed between the object NP sore ‘that’ and
its predicate tumamu ‘to eat’ in (15), zero-marking is less likely to occur. Fujii and Ono’s (2000)
analysis also shows the same tendency. When there are some intervening elements between the
object NP and its verb, object-marking tends to be preferred.
T-H Kim (2008) claims that the speaker’s object-particle use is closely related to the distance
between the object NP and its predicate. The occurrence of (l)ul-marking increases as the number
of intervening words between the object NP and its verb increases.
(16) na-ø nyuyok-ø ka-l ttay, pesu-ø tha-ko ka-ss-e.
1SG-( ) New York-( ) go time bus-( ) take-and go-PST-sem
‘When I went to New York, I took a bus.’

(17) ku enni-nun chotunghakkyo ttay mikwuk-ey wa-kackwu
that sister-TOP elementary school time USA-LOC come-and
‘She came to the United States when she was in elementary school,

ku cwupyen-ul ta cal anunkeya.
that surrounding area-ACC all well know
so she was very familiar with all the surrounding areas.’
(T-H Kim, 2008, p.201)
For example, there is no intervening word between the object NP and its predicate in (16). The
object pesu ‘bus’ is immediately followed by its predicate, tha- ‘to ride’ and hence the NP is left-
unmarked. On the other hand, if there are any intervening words between the object NP and its

12

predicate, the object NP is preferred to be (l)ul-marked. For example, the two words ta
‘everything’ and cal ‘well’ in (17) come between the object NP ku cwupyen ‘the surrounding area’
and its predicate al- ‘to know’. Thus, the object NP is preferred to be (l)ul-marked. This increase
in preference for (l)ul-marking, however, does not seem to always be gradual. Even though the
object NP ku cwupyen ‘the surrounding area’ in (17) is not adjacent to its predicate al- ‘to know’
because of the two intervening words, it does not necessarily need to be (l)ul-marked to be
identified as the object of the sentence.
Why, then, is this adjacency condition claimed to have a significant effect on the speaker’s
choice in object-marking? Matsuda (1995) points out that when the object NP in a transitive
sentence is strictly adjacent to its verb, it would be relatively easy for the hearer to identify it as
the object of the sentence even in the absence of the object particle. But what if the object NP
does not follow this default Object + Verb type? If the object NP were not immediately adjacent
to the verb, the hearer’s decoding load would significantly increase, and hence object-marking
would help the hearer to decode the sentence. T-H Kim (2008) also claims that the object NP that
occurs closer to its predicate is more easily recognized as the object of the sentence, and hence a
low processing load is assumed for this type of object. Consequently, the object NP is often left
unmarked. On the other hand, when the object NP is more distant from its predicate, it is harder
for the hearer to readily recognize it as the object of a sentence, and therefore object-marking is
more likely to be used for such types of objects to provide an additional linguistic signal for hard-
to-process information.
In languages with SOV word order like Japanese and Korean, the object of a transitive
sentence is normally expected to be immediately adjacent to the verb. When there is no
intervening word between the object NP and its predicate, therefore, the NP is likely to remain
unmarked. However, if any intervening elements appear after the object NP, it is rather
unpredictable for the hearer and hence harder to process, and this may trigger the speaker to use
object-marking.

13

What about object NPs which appear right after the verb then? Because both Japanese and
Korean are verb-final languages with canonical SOV word order, post-predicate object NPs are
unpredictable and harder to process, and thus object-marking is more likely to be used.
(18) totsuzen yoba-rete namae-o
suddenly call-PASS name-DO
‘Suddenly, (he) was called, (his) name’
(Fujii and Ono, 2000, p.14)
Fujii and Ono (2000) show that direct objects that appear after the predicate are more likely to be
o-marked. For example, the direct object namae ‘name’ in (18) appears after the verb yobareru
‘be called’ and is overtly o-marked. Previous studies have found that if there is any intervening
element between the object NP and its predicate, or if the object NP appears after the predicate,
then this may trigger object-marking: the speaker provides an additional linguistic signal for the
object NPs with a non-canonical position in the sentence, and hence facilitate the hearer’s
selective focus on hard-to-process information. However, there is still no clear explanation of the
reason behind the speaker’s choice to use non-canonical SVO word order with object-marking.
There must be something the speaker intends to express or imply by using non-canonical SVO
word order and object-marking that would contribute to the achievement of the speaker’s
communicative purpose.

2.1.3 The Form of the Object NP
The form of the object NP has been claimed to be closely related to the speaker’s choice of using
object-marking (Matsuda 1995; Fujii and Ono 2000; Fry 2003; Kim 2008; Kurumada and Jaeger
2013). Matsuda (1995) examined the variation between the presence and the absence of the object
marker o in Tokyo Japanese and found out that the variation can be accounted for by the form of
the object NP.
(19) Zero-Marking Hierarchy in Tokyo Japanese
wh-pronoun > NP > non-wh-pronoun > clausal NP

14

(20) Wh-Pronoun
nani -ø yaroo-kanna
what-ACC do-FP
‘I wonder what I will do.’

(21) NP
Sorede minna hata-ø hutte-sa…
and everyone flag-ACC wave FP
‘And everyone was waving flags…’

(22) Deictic Pronoun
Tyotoo, kore-o akete
hey this-OBJ open
‘Hey, can you open this?’

(23) Anaphoric Pronoun
Gosunkugi-no konna yatu-ga aru-desyo,
long nail-GEN like this stuff-SUB exist FP
‘You know a long nail like this, right?

are-o densya-e koo oitoite-sa
that-OBJ railway on like this leave FP
We left one on a railway like this.’

(24) Clausal NP
[Sono hen -ni haete-ru kusa]-o tunde…
around there at growing grass-OBJ pick
‘Picking the grass growing around there…
(Matsuda, 1995, p.161-162)
Wh-pronouns (e.g., interrogative words such as nani ‘what’ and dare ‘who’) are most likely to
be zero-marked, followed by lexical NPs, non-wh-pronouns (e.g., deictic pronouns and anaphoric
pronouns), and clausal NPs. Matsuda points out the difference between wh-pronouns and other
kinds of pronouns as follows: Unlike lexical NPs, which carry substantial meanings in themselves,
the primary function of pronouns is to stand for some entity already mentioned or established in
the discourse context, such as the anaphoric pronoun are ‘that’ in (23), or to refer to some entity
in the proximity of the speaker and the hearer, such as a deictic pronoun kore ‘this’ in (22).
Matsuda claims that if the speaker presents these pronouns without o, then the hearer needs to
process them with less contextual information. Thus, it is relatively less likely for non-wh-
pronouns to appear without o.

15

Matsuda’s Zero-Marking Hierarchy in Tokyo Japanese, however, is not necessarily true in all
contexts because if the referent of object NPs is given and identifiable in the immediate or
preceding context, then zero-marking can be preferred. In particular, deictic and anaphoric
pronouns refer to a concrete object or an abstract idea which appears in the immediate or
preceding context. Due to this shared-knowledge, available both to the speaker and the hearer,
their referents can be easily identifiable and hence zero-marking can be preferred. Therefore, the
form of the object NP cannot fully explain the speaker’s choice of zero- or object-marking.
Fujii and Ono (2000) also found that unmarked object NPs are predominantly non-referential,
while o-marked ones are predominantly referential, that is, they refer to specific referents in the
discourse context. They also claim that all occurrences of unmarked objects can be classified into
one of the two following categories: (a) object NPs whose particular identification of referent is
of no interest or is irrelevant, and (b) object NPs whose referent is given and identifiable in
linguistic or extralinguistic contexts. Interrogative and indefinite pronouns, lexicalized
expressions such as verbal nouns and idiomatic expressions, and object NPs that express type or
category, fall into the first category, while demonstratives fall into category (b). Observe the
following examples of category (a).
(25) Indefinite pronoun
tada atte nanka hanashite-ru toka yuu gurai no mon?
just meet something talk-STATE like say about of NOM
‘(You guys) are just getting together and talking about something?’

(26) Interrogative pronoun
ima-goro nani shite-k-ka wakan-nai zo
now-about what do-STAT-QUES understand-not FP
‘(We) don’t know what (she) is doing right now.’
(Fujii and Ono, 2000, p.12-13)

(27) Lexicalized expressions: verbal noun
renkyuu-ake omowazu kaisha de jiman shi- chatta
holidays-after inadvertently company at brag do-ended up
‘I ended up bragging (about it) at work after the holidays’

(28) Lexicalized expressions: idiomatic expression
uchi no guruupu nanka demo are imadani
our of group like even that even now

16

me no iro kaeru wake ja-nai
eye of color change WAKE COP:EMPH-not
‘(We) still get real serious about that, don’t (we)?’
(Fujii and Ono, 2000, p.9-10)

(29) Object NP expressing a type or category
Nanka kanri jimu no maneejaa yatte-ta koro wa
like administrative affairs of manager do:STAT-PAST time TOP
‘Like the time (he) was the manager in the administrative affairs (department)…’
(Fujii and Ono, 2000, p.17)
What do the object NPs in (25)-(29) have in common? They do not have specific referents.
The indefinite pronoun nanka ‘something’ in (25) and the interrogative pronoun nani ‘what’ in
(26) do not refer to a specific or concrete entity. Similarly, the noun jiman ‘bragging’ of the
Object-Verb compound verb jiman suru ‘to do bragging (to brag)’ in (27) expresses an abstract
concept and thus does not have a specific reference. The idiomatic expression me no iro kaeru in
(28) does not refer to specific entities, but it conveys the single concept of ‘(surprised and) getting
serious.’ Next, the expression kanri jimu no maneejaa ‘the manager in the administrative affairs
(department)’ in (29) is used to refer to a type of manager and thus is not referring to a particular
position identifiable in a given context. When object NPs are non-referential or nonspecific, the
identification of their exact referent does not matter in conversation, and hence they are left
unmarked. These NPs are considered low in informational value and emphasis.
Regarding the second category, if the referent of object NPs is given and identifiable in
linguistic or extralinguistic contexts, zero-marking frequently occurs.
(30) Demonstratives
a. nukuzo kore
pull out this
‘(I) will pull (it) out, this!’

b. soo soo soo soo soo soo dakara sore kangaete-te
so so so so so so therefore it think-STAT
‘Yes, yes, yes, so (I) was thinking about it.’
c. sonna no kite toire iku na
such thing wear restroom go PROH
‘Don’t go to the bathroom wearing such a thing!’
(Fujii and Ono, 2000, p.11)

17

In (30a), kore ‘this’ refers to the tape recorder which is placed close to the speaker. In (30b), the
referent of sore ‘it’ is a surprise visit, which is mentioned in the immediately preceding discourse.
In (30c), a demonstrative with a nominalizer, sonna no ‘such a thing’ refers to the dress the
listener is wearing. Note that in these utterances, the demonstratives refer to a concrete object or
an abstract idea which appears in the immediate context, and thus their referent is easily
identifiable. Therefore, these demonstratives do not need to be o-marked.
Fujii and Ono claim that while non-referential or contextually highly identifiable NPs are
prepared to be left unmarked, the majority of o-marked direct objects have specific referents.
(31) atashi Sachiko no saisho no tanjoobi wa oboete-te
I Sachiko of first of birthday TOP remember:STAT
‘I remembered the first time Sachiko had a birthday (here) and

tokei o okutta n da yo ne
watch DO gave NOM COP FP FP
(we) gave (her) a watch.’
(Fujii and Ono, 2000, p.17)
In (31), the speaker is talking about tokei, the watch that she gave to her friend as a birthday gift.
This object NP has a specific referent and it appears to be o-marked. According to Fujii and Ono,
o-marking prefers referential entities and zero-marking prefers non-referential entities, but this
correlation is not categorical: roughly 30% of the direct objects in their data did not follow this
correlational pattern. This indicates that there may be some other factors affecting the speaker’s
choice of zero- or object-marking. Observe (32):
(32) Interrogative pronoun
a: ecey mwe-ø mek-ess-e?
yesterday what-(OBJ) eat-PST-SE
‘What did you eat yesterday?’

b: ecey totaychey mwe-l mek-ess-e?
yesterday in the world what-OBJ eat-PST-SE
‘What in the world did you eat yesterday?’
[Implied: I have no idea what you ate yesterday.]

c: ecey mwe-l mek-ess-tako?
yesterday what-OBJ eat-PST-EMP
‘You ate what?’
[Implied: I can’t believe you ate it yesterday.]

18

If Fujii and Ono’s claim is valid, the interrogative pronoun mwe ‘what’ in an ordinary
information-seeking question as in (32a) does not refer to a specific referent, and thus it is
supposed to be left unmarked. However, it sounds more natural for the same pronoun mwe ‘what’
to be (l)ul-marked as in (32b). Here, the question is uttered not to simply ask for a new piece of
information, but to show the questioner’s negative attitude towards the other speaker’s previous
utterance or action. Object-marking signals an attention or emphasis to its marked element and
therefore the assertion conveyed through the question can be more strongly expressed in (32b): ‘I
have no idea what you ate yesterday.’
Note that the emphatic adverb totaychey ‘(what, who, why, etc.), which expresses the
speaker’s wondering about an unexpected event or state, often co-occurs with (l)ul-marked
elements. This suggests that the speaker’s choice of zero- or (l)ul-marking is not solely
determined by the form of the object NP. Rather, it is closely related to what the speaker intends
to express through utterances.
(33) Demonstratives
a. ike-ø ne cwu-lkey
this-(OBJ) you give-will
‘I will give you this.’

b. ike-l ne-l cw-un-tako
this-OBJ you-OBJ give-PRES-EMP
‘I said I will give you this.’
[Implied: I insist that I give you this!]

c. ike-l ne-l cw-un-tako?
this-OBJ you-OBJ give-PRES-EMP
‘Are you saying (he) gives you this?’
[Implied: I can’t believe he gives you this.]
Demonstratives, which refer to a specific entity or an abstract idea already mentioned or
established in the preceding context, are often left unmarked because their referents are easily
identifiable in the given context. For example, since the referent of the demonstrative ike ‘this’
(thing)’ in (33a) is an entity closer to the speaker, it is easily recognizable in the immediate
context at the time of utterance. Therefore, it is left unmarked. However, the same demonstrative

19

can also be overtly (l)ul-marked as in (33b), which is uttered in a context where there is an
ongoing disagreement between the participants in conversation: the speaker offers something to
the hearer but the hearer refuses to take it. The speaker challenges the hearer’s previous action by
making another offer. In this challenging environment, (l)ul-marking can be used by the speaker
to signal emphasis or attention to its marked elements so that the speaker’s assertion conveyed
through the utterance can be more strongly expressed: ‘I insist that I give you this.’
Similarly, (33c) can be uttered in a discourse context of disagreement, in which the speaker
shows surprise or disbelief regarding the prior utterance or action of the co-present or non-present
party. This also indicates that (l)ul-marking is used by the speaker to direct the listener’s attention
to its marked elements so that the speaker’s intended meaning through utterances such as
expressing disbelief about what has been said which is contrary to the speaker’s belief or
expectation can be more strongly conveyed: ‘I can’t believe (he) gives you this.’
To summarize, previous studies have found that some linguistic factors such as the size,
form, or referentiality of the object NP, the distance between the object NP and its predicate, and
word order may affect the speaker’s choice of zero- or object-marking in spoken language. Zero-
marking is preferred when the referent of the object NP is easily identifiable or non-referential
because these NPs are low in informational value and emphasis. On the other hand, object
marking is frequently used to facilitate the listener’s selective focusing on difficult-to-process
information which requires the listener’s extra attention.
However, the speaker’s choice of zero- or object-marking cannot be solely explained with
those linguistic factors. As shown above, in certain discourse contexts, object-marking can be
used for the listener’s selective focus on its marked element so that it can help the speaker
achieve a certain communicative purpose by expressing something through utterances more
effectively. In the next section, some pragmatic factors claimed to affect the speaker’s use of
object marking in spoken Korean, such as the formality of speech, sharedness between
interlocutors, and the contrastiveness of object NPs, will be examined.

20

2.2 Pragmatic Factors Affecting Object-Marking
Some discourse-pragmatic studies have examined the occurrence and non-occurrence of the
object particle in spoken language in relation to the discourse context, and object-marking has
been found to appear consistently in certain discourse contexts. The following are considered to
be determining factors in the speaker’s choice of zero- or object-marking: Formality of speech
(Tsutsui 1984; Ko 2002), shared-knowledge (Masunaga 1988; Lee and Thompson 1987), and
contrastively focused object NPs (Shin 1982; Tsutsui 1984; C-M Lee 1995; Sohn 1999; T-H Kim
2008; H-J Lee 2010).

2.2.1 Level of Formality in Speech
Tsutsui (1984) claims that the occurrence or non-occurrence of the Japanese object particle o is
closely related to the level of formality in speech. According to Tsutsui’s Case Particle Ellipsis
Rule, the lower the level of formality in speech is, the more zero-marking naturally occurs.
(34) a. Tanaka-sama wa o-sake o/?ø meshiagarimasu.
Tanaka Mr. (very formal) TOP sake (polite) D.( ). drink (honorific)
‘Mr. Tanaka drinks sake. (Very formal)’

b. Tanaka wa sake o/ ø nomu
Tanaka TOP sake D. ( ). drink (informal)
‘Tanaka drinks sake. (Very informal)’
(Tsutsui, 1984, p.91-92)
Japanese has a set of speech styles of address which shows different levels of politeness and
formality. For example, when the speaker addresses someone superior to himself with respect
using honorific forms, as in (34a), zero-marking is less likely to occur. On the contrary, when the
speaker addresses someone inferior or equal to himself, zero-marking can occur with relative
ease, as in (34b).
Ko (2000) analyzed three different speech styles of Korean spoken discourse including
informal telephone conversations between friends, informal narratives in a newsgroup, and
formal TV news scripts. The frequency of zero-marking was found to be closely related to the

21

degree of formality and the speech styles of spoken discourse: zero-marking appeared the most in
informal telephone conversations and the least in formal news scripts. Similar to Japanese,
Korean has a set of speech styles of address which show different levels of politeness and
formality.
In Korean, styles of address consist of two levels in which the speaker addresses the addressee
either as superior (or senior) to himself, or inferior (or junior) or equal to himself. Based on this
two-level distinction, the speaker decides whether to express respect or deference towards the
addressee in speech. Thus, styles of address can be broadly distinguished into four styles on the
basis of the speaker’s decision with respect to deference and formality: deferential informal style,
deferential formal style, plain informal style, and plain formal style. The following dialogue is an
informal conversation between two friends in plain informal speech style.
(35) Speaker A asks B until what time he was drinking last night

à 1 A: yel si-pwuthe swul-ø masy-ess-umyen encey-kkaci masy-ess-nya?
10 hour-from alcohol-(OBJ) drink-PST-if when-until drink-PST-Q
If you started drinking at 10 pm, until what time were you drinking?

à 2 B: yeltwu si-kkaci-nun swul-ø masi-ko nolay pwulu-ko
twelve hour-until-TOP alcohol-(OBJ) drink-CONJ song sing-CONJ
Until 12am, I was drinking and singing songs and then…

3 yeltwu si-pwuthe twu si-kkaci-nun chwumchwu-le ka-ss-e
twelve o’clock-from two o’clock-until-TOP dance-in order to go-PST-SE
…from 12am to 2am, I went dancing
In (35), the plain informal sentence-ending -e/a and the alternative plain informal ending for the
interrogative -nya are used. Note that the object NP swul ‘alcohol’ in line 1 and 2 is left unmarked.
On the contrary, object-marking is consistently used in the following news script, which shows
the deferential formal speech style.
(36) The news anchor is talking about alcohol consumption in Korea.

à 1 A: swul-ul manhi masi-myen meli-to aphu-ko sokto wullengkeli-ko
alcohol-OBJ a lot drink-if head-also sick-CONJ stomach feel nausea-CONJ
If you drink a lot of alcohol, you will have a headache, feel nausea, and

22

à 2 taum nal kosayng-ul manhi ha-key toy-pnita
next day suffering-OBJ a lot do-become-SE
suffer a lot the next day

à 3 Wuli kwukmin-tul-un swul-ul elmana manhi masi-lkkayo?
Our people-PL-TOP alcohol-OBJ how much drink-do you think
How much alcohol do you think Koreans consume?

4 KFDA-nun onul wuli kwukmin-uy
Korean Food & Drug Administration-TOP today our people-GEN
‘Today, the Korean Food and Drug Administration announced

à 5 umcwu silthay cosa kyelkwa-lul palphyohayss-supnita.
drinking real state survey result-OBJ announced-SE
the results of a survey on alcohol consumption in Korea.’
(MBC Evening News: December 23, 2014)

The deferential formal sentence ending -(su)pnita is used in a formal setting, as in (36), in
which the news anchor is addressing the audience. In this formal setting, zero-marking is less
likely to occur. For example, the object NP swul ‘alcohol’ in lines 1 and 3, kyelkwa ‘result’ in line
5, and the noun kosayng ‘suffering’ of the Object-Verb compound verb kosayng hata ‘to do
suffering (to suffer)’in line 2 are overtly (l)ul-marked. The occurrence or non-occurrence of the
object particle is thus closely related to the degree of formality in different speech styles.

2.2.2 Shared Knowledge
Masunaga (1988) shed some light on the pragmatic side of object-marking by proposing a
‘deemphasizing’ factor. The claim is that attaching a final particle to a sentence has a function of
focusing on the verb, which defocuses the object in turn. Therefore, whenever the pertinent NP is
deemphasized or defocused, the object particle o can be dropped.
(37) Hanako-o / ???-ø yonda.
Hanako-ACC invited
‘I invited Hanako.’

(38) Hanako-o / -ø yonda-zo/yo.
Hanako-ACC invited FP
(Masunaga, 1988, p.147)
Considering the fact that o deletion is unnatural in (37), the zero-marking of the object NP is
more acceptable with the presence of the final particle zo or yo as in (38).

23

As mentioned in Matsuda (1995), Masunaga also claimed that the use of shared information
leads to the de-emphasis of the NP and the omission of o, as in (39).
(39) The speaker knows that Yamada is interested in buying a car.

Yamada-kun, kuruma-ø katta?
car-ACC bought
‘Did you buy a car, Yamada?’
(Matsuda, 1995, p.45)
If an object NP refers to an entity which is assumed to be shared by the speaker and the hearer as
in kuruma ‘car’ in (39), the object NP is de-emphasized and thus is likely to be left unmarked.
Masunaga tried to explain this pragmatic aspect of object-marking using the notion of de-
emphasis. However, this explanation is problematic because it is not clear how the presence of
the final particle leads to the deemphasizing of the verb. The final particles do not simply move
focus to the verb as claimed. Rather, they can be used to express the speaker’s attitude towards
the information presented in utterances. In addition, the concept of shared information in
Masunaga’s analysis is very vague.
Lee and Thompson (1987) have suggested that the occurrence or non-occurrence of the object
particle -(l)ul in informal narratives and dialogues is pragmatically determined by “sharedness
between communicators” and the necessity for specification in particular communicative
situations: If there is a greater the amount of sharedness between the interlocutors, such as shared
experiences, shared contexts, and shared cultural backgrounds, it is less necessary for speakers to
specify the grammatical roles played by arguments, which can lead to the omission of the object
particle. Observe (40) for an example of shared experience.
(40) Shared experience
na wekhumayn tollyecwu-lyeko wa-ss-nuntey…
I walkman return-in order to come-PST-CONJ
‘I came to return your walkman, but…’
(Modified from Lee and Thompson 1987)
The referent of the object NP wekhumayn ‘walkman’ in (40) is assumed to be known to both
the speaker and the hearer through the shared experience the interlocutors have been jointly

24

involved in (e.g., the speaker has borrowed a walkman from the hearer). Therefor, the object NP
wekhumayn ‘walkman’ does not need to be (l)ul-marked because the addressee knows exactly
which walkman the speaker is talking about and hence its referent is easily recoverable through
shared experience. Shared contexts can be constructed from the immediate situations of
utterances, such as their time and place of occurrence, the interlocutors involved, the event being
discussed, and so on.
(41) Shared context
a. oppa-hanthey yo chokholleys tu kae cwu-ko siph-ese wa-ss-e
brother-to this chocolate two CL give-want to-CONJ come-PST-SE
‘I came because I wanted to give you these pieces of chocolate.’

b. i-ke ipe kulem
this-thing put on then
‘Try these (pants) on then.’

c. ceki pwa
over there see
‘Look over there.’
(Modified from Lee and Thompson 1987)
The deictic expressions yo ‘this’ in (41a), ike ‘this (thing)’ in (41b), and ceki ‘over there’ in (41c)
refer to something directly perceptible by the interlocutors in conversation. Similar to shared
experience, their referents are easily recoverable in the immediate context and therefore they are
frequently left unmarked.
Sometimes the situational context is not sufficient to understand what the sentence truly means.
Schematic knowledge shared by people in the same group, culture, or society that refers to
customary, normal, or natural ways of thinking about events is also needed. Because what is
familiar or customary can vary across groups, societies, and cultures, schematic knowledge
represents culturally different ways of using language to do things. Idiomatic expressions that
denote single concepts are good examples of conventional ways of expressing something.
(42) Shared cultural background
nay-ka acu khun mam mek-ko cemsim-e chotay-ha-lyeko
I-NOM very big mind eat-CONN lunch-LOC invitation-do-in order to

25

wa-ss-nunte…
come-PST-CONJ
‘I made the hard decision to come and invite you to lunch, but …’
(Modified from Lee and Thompson 1987)
The bare NP mam ‘mind’ in (42) does not refer to any specific referent. Rather, it is part of the
idiomatic expression khun mam mekta, which conveys a single abstract concept ‘to make a hard
decision.’
Lee and Thompson’s principle of “sharedness between interlocutors” provides an account for
the omission of the object particle -(l)ul: When temporal settings, spatial settings, or cultural
backgrounds are shared by interlocutors and when the referent of an object NP is already shared
by the participants in conversation, the grammatical role of the object NP does not necessarily
have to be specified by (l)ul. However, there is still no clear explanation for the speaker’s choice
to use -(l)ul in informal conversation. This interesting issue has been discussed in recent
psycholinguistics and discourse-pragmatics research. I will consider the work done in those areas
in the the following sections.

2.2.3 Informational Predictability
Psycholinguistics researchers have been examining to what extent language and language use are
organized to facilitate efficient communication (Jaeger 2010). One of the earliest observations
related to efficient language production is the link between word frequency and word form (Zipf
1949). Zipf’s Principle of Least Effort says that it is human nature to want the greatest outcome
for the least amount of work. In Zipf’s observation of the distribution of word frequencies in
English, frequent words generally have shorter linguistic forms because speakers can maximize
efficiency by using shorter forms, which take less effort to produce. For example, in the Brown
Corpus of American English, the word “the” is the most frequently occurring word, by itself
accountimg for nearly 7% of all word occurrences (69,971 out of slightly over 1 million). The
second-place word “of” accounts for 3.5% of all words (36,411 occurrences), and it is followed

26

by “and” (28,852). In fact, only 135 vocabulary items account for half of the Brown Corpus.
In more recent research, however, word length has been found to be more strongly related to
word predictability in context than word frequency. In Piantadosi et al. (2011), the relationship
between word length, frequency, and its information content, how surprising/informative a word
is in its context, has been analyzed. The information conveyed by a word can be quantified by its
surprisal, a measure of how unpredictable a word is in its context. Words that are highly
predictable from context convey no significant information and hence have less surprisal value.
For example, in the phrase to be or not to be, the final be is almost entirely predictable from the
preceding context and therefore has no information or surprisal value. On the other hand, words
that are highly unpredictable from context are more informative and hence have high surprisal
values. For instance, the final word kumquat in the phrase to be or not to kumquat is highly
unpredictable and hence high in information and surprisal value. This indicates that word length
can be better predicted by the word’s information content than by its frequency. Longer words
tend to be less inferable in context and hence high in information and surprisal values, while
shorter words are contextually more predictable and hence less informative.
Kurumada and Jaeger (2013) have examined the omission of the Japanese direct object
particle o and found that speakers tend to omit elements whose information content is
contextually predictable, while providing more linguistic signals to convey less predictable
information. Based on the Animacy Scale, the speaker’s choice of zero- or object-marking is
related to the animacy of the object NP.
(43) Animacy Scale
Human > Animate > Inanimate
(Aissen, 2003, p.437)
Subjects have features high on the Animacy Scale, while objects tend to have features lower
on the scale. Object NPs with more subject-like properties, [+ human], are highly unpredictable to
be interpreted as typical objects in a sentence and hence are more likely to be o-marked. On the
other hand, object NPs with typical object properties, [+inanimate], can be readily interpreted as

27

objects and hence are more likely to be left unmarked. In the following two experiments, the
animacy of the direct object and the plausibility of grammatical function assignment of animate
subject- animate object pairs are manipulated.
(44) Animacy
a. Sensee-ga syooboosya-ø mitayo
teacher-SUB fire-engine-(OBJ) saw
‘The teacher saw a fire-engine.’

b. Sensee-ga seeto-o mitayo
teacher-SUB student-OBJ saw
‘The teacher saw a student.’

(45) High plausibility
a. Isya-ga kanzya-ø byoositu-de tiroositayo
doctor-SUB patient-(OBJ) hospital room-at treated
‘The doctor treated the patient in a hospital room.’

b. Low plausibility
Kanzya-ga Isya-o byoositu-de tiroositayo
patient-SUB doctor-OBJ hospital room-at treated
‘The patient treated the doctor in a hospital room.’
(Modified from Kurumada and Jaeger 2013)
While Japanese speakers tend to omit o for prototypical inanimate object NPs such as
syooboosya ‘fire-engine’ in (44a), they tend to produce o for atypical, animate objects such as
seeto ‘student’ in (44b). If both the subject and object are animate in a sentence, both arguments
can be plausible subjects. For example, isya ‘doctor’ and kanzya ‘patient’ in (45) can be the
subject. In this case, speakers tend to omit o when the grammatical function assignment is more
plausible: the doctor treats the patient as in (45a). Speakers tend to produce o when the
grammatical function assignment is less plausible: the patient treats the doctor as in (45b).
Kurumada and Jaeger’s study has shown the close connection between word length and
information predictability, which can explain the speaker’s choice of zero- or object-marking:
Contextually unpredictable object NPs are more informative, and hence object-marking is a
linguistic device used by the speaker to facilitate the listener’s selective focus on salient
information. However, because the predictability of object NPs in this study referred specifically
to word predictability, other discourse contextual factors were not considered. For example, as

28

shown above, Kurumada and Jaeger have claimed that Japanese speakers prefer to use object-
marking for an atypical animate object NP seeto ‘student’ in (44b) while a typical inanimate
object NP syooboosya ‘fire-engine’ in (44a) is preferred to be left unmarked. However, since the
animate object seeto ‘student’ can also be contextually highly predictable due to its close
semantic connection with the subject sensee ‘teacher’ in a sentence, object-marking is not
necessarily needed. This indicates that the speaker’s choice between zero- or object-marking is
not solely determined by the word predictability of the object NP.

2.2.4 Informational Salience
In discourse-pragmatic studies, object marking has also been claimed to mark salient information
in certain discourse contexts (Fujii and Ono 2000; Yang 2007). Salient information refers to
important information that the listener needs to identify or process (Fujii and Ono 2000). Object-
marking is a linguistic device that helps the listener pay attention to or focus on important
information, which can facilitate successful communication. Several discourse factors are claimed
to affect informational salience: the discourse topic, newsworthiness, repair, and contrast (Fujii
and Ono 2000). A similar study has also been done by Yang (2007) in Korean: The (l)ul-marked
element is more emphatically salient and object-marking has pragmatic functions in conversation,
such as marking a discourse topic which continues to be talked about in a stretch of discourse,
emphasizing an action performed by the actor, clarifying a referent which is not clearly
understood by participants, or distinguishing contrastive referents.

2.2.4.1 Indicating the Discourse Topic
Fujii and Ono (2000) claim that object-marking is used to introduce the object NP that the
speaker continues to talk about in an extended conversation. It is a way of signaling the
importance of the object NP in the discourse context, which helps the hearer pay attention to
further descriptions/explanations about the object NP in the subsequent context (Yang 2007).

29

(46) Marking a discourse topic

1 I: hhh ecey-to saypyek neysi-ey kkay kaciko
yesterday.also at dawn four o’clock.at wake up.CONN

à 2 malha-ess-canh-a kkwum-ul kkwu-ess-nuntey
say.PST.you see.FP dearm.ACC dream.PST.given that

3 kay nao-ko koyangi nao-ko…
dog appear.and cat appear.and

‘(I) told (you that I) woke up 4 o’clock yesterday morning. I had a dream and dogs and
cats appeared (in my dream)…’
(Yang, 2007, p.233)
The object NP kkwum ‘dream’ in (46) is overtly (l)ul-marked because it has been introduced
as a discourse topic the speaker continues to talk about in the following utterances. The findings
of previous studies have thus identified one of the pragmatic functions of object-marking,
introducing a discourse topic and signaling further descriptions/explanations about it in the
subsequent context. However, there is still a lack of understanding about the linguistic form
frequently co-occurring with a (l)ul-marked discourse topic.
Note that -(l)ul co-occurs with the connective -(n)untey, which is used to provide background
information, or to establish background in discourse. As mentioned in Kim (1996), -nuntey
indicates a warranted continuation of the relation between the preceding portion of discourse as
the background/topic and the up-coming portion as the elaboration/comment. As such, the
speaker provides a further description or explanation related to the (l)ul-marked object in the
subsequent context to build shared-knowledge between the speaker and the hearer so that the
hearer can understand what the speaker is talking about without difficulty or any
misunderstanding. Therefore, this dissertation will further examine some linguistic forms which
frequently co-occur with (l)ul-marking to find out how (l)ul-marking helps the speaker achieve
certain communicative purposes.

30

2.2.4.2 Emphasizing an Action
According to Yang (2007), (l)ul-marking has also been claimed to have the pragmatic function of
emphasizing an action of the speaker or the actor, as in (47).
(47) Emphasizing an action

1 H: elkwul sacin an po-ketun kuke?
face picture not see.FP that thing
‘(They) don't’ see the face photo (on the pass).’

2 A: kuntey namca-canh-a:::
but man.you see.FP
‘But, you know, it is (the picture of) a man.’

à 3 H: ani sacin-ul cengmal an po-a=
no picture-OBJ really not see.FP
‘No, (they) really don’t see the picture.’
(Yang, 2007, p.228)
Here, the listener has argued against the speaker’s claim about the action involving the direct
object NP sacin ‘picture’ in line 2, and therefore the speaker needs to emphasize the prior
utterance using object-marking when he restates the claim in line 3.
Yang (2007) has captured the phenomenon of object-marking occurring in a challenging
environment, as in (47); that is, when there is an ongoing argument or disagreement between the
participants in conversation, the speaker can use object-marking to emphasize an action of the
speaker or the third party to challenge the other participant’s prior utterance or action as a
sequentially appropriate next response. An adverb such as cengmal ‘really’ (line 3) also conveys
an emphatic reading. If object-marking can be used by the speaker in challenging environments to
signal emphasis or attention to its marked elements so that the speaker’s assertion can be more
strongly expressed, there must be other linguistic forms or patterns frequently used to challenge a
prior utterance or action of the co-present or non-present party in conversation.

31

2.2.4.3 Initiating Repair
Yang (2007) also claims that (l)ul-marking is consistently used when the speaker needs to signal
the listener’s attention to a referent that needs to be corrected or clarified (Yang 2007). When a
participant in conversation has difficulty agreeing with or accepting what has been said, repair
can be initiated to deal with the problematic source (Robinson & Feldman 2010) and to re-
establish a shared-understanding between the speaker and the listener.
(48) Clarifying a referent

1 G: swuswul-to ha-si-ko kulu-si-ess-canh-a::
surgery.also do.HON.and do so.HON.PST.you see.FP
‘(She) also had surgery’

2 A: wuli emma-ka swuswul-ul-yo?
our mom.SUB surgery.OBJ.FP
‘My mom had surgery?’
(Yang, 2007, p.222)
Speaker A initiates repair on Speaker G’s prior utterance that A’s mom had surgery. Because
Speaker A has knowledge of the repeated item in context, wuli emmaka swuswululyo “My mom
had surgery?” in line 2 is uttered not to confirm what Speaker G has said but to re-establish a
shared-understanding between Speaker A and G. Therefore, her utterance is challenging the
accuracy of the information in the repeated item. Here, the direct object NP swuswul ‘surgery’ is
(l)ul-marked to help the listener focus on the salient NP that needs to be corrected or clarified.
Note that initiating repair is also observed in challenging situations. When the speaker thinks
what the other participant has said might be wrong, inaccurate, or hard to accept, the speaker can
initiate repair to show surprise or disbelief about what has been said or even to challenge the
accuracy of the information. Therefore, this dissertation further examines the role of object-
marking in repair situations where the speaker disagrees with or challenges the prior utterance or
action of co-present or non-present parties and asks for a clarification or correction.

32

2.2.4.4 Distinguishing Contrastive Referents
The notion of contrast has been claimed to be one of the important factors that determines the
speaker’s choice of zero- or object-marking: object-marking is consistently used for contrastively
focused referents (Shin 1982; Tsutsui 1984; T-H Kim 2008; H-J Lee 2010). It is generally known
that there are two kinds of focus: contrastive focus and informational focus. Contrastive focus
involves an explicit choice among the set of contextually given alternatives. With a contrastive
focus, a focused element is contrasted with a set of all the other possible candidates that the
predicate of a sentence potentially holds and is identified as the exhaustive subset of this set (Kiss
1998). Contrastive stress can be placed on a focused element. Informational focus, on the other
hand, simply marks new information in a sentence without exhaustively contrasting its focused
element with a set of contextually given entities. For example, the wh-question and answer pairs
represent informational focus. The answer value substituting for the wh-phrase indicates the
identity of the focus. Compare (49) and (50).
(49) A: Who bought a car?
B: John bought a car.
(50) A: Did Mary buy a car?
B: No, it was John who bought a car.
In (49), the non-contrastive information focus merely presents John as new information without
suggesting that John is the only one. Conversely, the focused element John in the contrastive focus
structure (50) receives emphatic stress, and the sentence implies that it was John (and only John)
who bought a car.
Note that unlike informational focus conveying a piece of new information, contrastive focus
does not necessarily need to mark new information. As mentioned in Chafe (1976), the focus of
contrast can be marked by higher pitch and stronger stress, and it is independent of givenness.
Chafe defines the concept of contrast by identifying the three factors: background knowledge
(something the predicate possesses), a set of possible candidates (a set of possible referents that
the predicate can possess), and an assertion of which candidate is the correct one.

33

(51) John

made the salad.
a. Background: someone made the salad.
b. A set of possible candidates: {Bill, David, Sam, John …}
(52) John made the salad, but Mary made the hamburger.
In (51), the speaker assumes that the hearer knows someone made the salad (background
knowledge) and that the hearer has a set of possible candidates in mind who can be that someone.
Then, the speaker tells the hearer John is the right selection for this role of salad-marker as
opposed to other possible candidates that the addressee might have in mind. In this sense, John is
the focus of contrast. It is also possible to have more than one focus of contrast, as in (52).
It has been shown in previous studies that zero-marking is unnatural when the direct object NP
is contrastively focused. Tsutsui (1984) claims that when emphatic stress is placed on the o-
marked NP, it receives a contrastive focus reading, which conveys the idea of exclusivity, “X and
only X,” as in (53).
(53) Boku wa suteeki o/ *ø tabemashita.
I steak D.O. ate
‘I ate (not others but) steak.’
(Tsutsui, 1984, p.93)
The o-marked NP suteeki ‘steak’ with emphatic stress singles out all other possible candidates,
and hence it is the only entity that meets the truth condition represented in the sentence: It was
nothing but steak that the speaker ate. In this case, zero-marking is unnatural.
Shin (1982) also shows that (l)ul-marking is used to convey a contrastive focus reading.
(54) a. acessi! tampay-ø cwuseyo.
Man cigarette give
‘Please give me cigarettes.’

b. acessi tampay-lul cwuseyo.
Man cigarette-ACC give
‘Please give me cigarettes.’
(Shin, 1982, p.121)
For example, (54a) can be uttered when the speaker asks for a pack of cigarettes to a shop
assistant. On the other hand, (54b) can be uttered when the shop assistant has handed over the

34

wrong item so that the speaker needs to ask for cigarettes again. Here, the (l)ul-marked object NP
tampay ‘cigarettes’ is the only item that the speaker wants to get.
T-H Kim (2008) also points out that (l)ul-marked-elements are contrastively focused with
other referents.
(55) kulayse cikum ceyi-pica-lo-nun an tway eyphu-pica-lul pat-aya toyketun
so now J-visa-with-TOP NEG do F-visa-ACC get-should do.SEM
‘So this J-VISA is not useful. It is an F-VISA that (I) must acquire.’
(T-H Kim, 2008, p.243)
In (55), the participants are talking about a type of visa they should apply for in order to study in
the United States. Here, the (l)ul-marked NP eyphu pica ‘F-visa’ is contrasted with the other type
of visa, ceyi pica ‘J-visa,’ and the F-Visa is the only type of visa that allows the holder to study in
the United States.
Lee (2010) proposes that there are predictability asymmetries among the three subtypes of
focus, informational focus, selective focus, and replacing focus, and that different degrees of
predictability affect the speaker’s choice of producing zero- or object-marking.
(56) Informational focus
A: Jinmi-ka mwue-l sa-ss-e?
Jinmi-SUB what-OBJ buy-PST-SE
‘What did Jinmi buy?’

B: khemphyuthe-ø/ khemphyuthe-lul sa-ss-e.
computer-(OBJ)/ computer-OBJ buy-PST-SE
‘(She) bought a computer.’

(57) Selecting focus
A: Jinmi-ka khemphyuthe-lul sasse, hywutaephon-ul sa-ss-e?
Jinmi-SUB computer-OBJ bought cell phone-OBJ buy-PST-SE
‘Did Jinmi buy a computer or a cell phone?

B: khemphyuthe-ø/ khemphyuthe-lul sa-ss-e.
computer-(OBJ)/ computer-OBJ buy-PST-SE
‘(She) bought a computer.’

(58) Replacing focus
A: Jinmi-ka hywutaephon-ul sa-ss-e?
Jinmi-SUB cell phone-OBJ buy-PST-SE
‘Did Jinmi buy a cell phone?

35

B: aniya, khemphyuthe-ø/khemphyuthe-lul sa-ss-e.
no computer-(OBJ)/ computer- OBJ buy-PST-SE
‘It’s a computer she bought.
(Modified from H-J Lee 2010)
The object NP receives information focus when it carries new information, such as the answer to
wh-questions, as in (56). Selecting focus can be placed on the object NP when it is selected from
the pool of candidates given in prior context, as in (57). When the object NP is contrasted with
the previously stated element, replacing focus can be realized, as in (58).
(59) Prediction of Predictability:
Selecting focus > Informational focus > Replacing focus
(H-J Lee, 2010, p.306)
On the Predictability Scale, selecting focus is most predictable because the object NP is
chosen from the given set of candidates in the preceding context. Informational focus shows an
intermediate level of predictability because the object NP carries new information. Replacing
focus shows the lowest level of predictability because the object NP is in contrast with the
previously mentioned element as an alternative substitute. While zero-marking is frequently used
for selecting focus with high predictability and hence low informational value, object-marking is
used for replacing focus with low predictability and hence high informational value.
The previous studies indicate that -(l)ul tends to be used to distinguish contrastive referents.
This dissertation further examines the pragmatic function of -(l)ul that distinguishes contrastive
referents in challenging environments: when the direct object NP in the prior utterance is
dispreferred, disagreed, or even rejected, an alternative substitute can be provided with object
marking. Object-marking directs the hearer’s attention or focus on this salient information, which
helps the hearer’s smooth processing of information and successful interpretation of the speaker’s
intended meaning.

36

2.3 Summary
In summary, previous research has found that both linguistic factors (e.g., the phonological size
and the form of the object NP, the adjacency condition, and word order) and pragmatic factors
(e.g., level of formality in speech, shared-knowledge between interlocutors, information
predictability and salience) may affect the speaker’s choice of producing zero- or object-marking.
Object-marking is a device used by the speaker to help the listener pay attention or focus on
either hard-to-process or salient information for successful communication. Therefore, object-
marking does not solely indicate the grammatical relation of the NP but also has some pragmatic
functions such as indicating the discourse topic, emphasizing an action, clarifying referents that
need to be confirmed or corrected, and distinguishing contrastive referents.
As shown above, the previous studies to date have shed light on the pragmatic uses of object-
marking as a device available for the speaker to facilitate the listener’s selective focus on hard-to-
process or important information for successful communication. However, there has been no
discussion about a more comprehensive understanding of the mechanism that controls the
speaker’s use of object-marking. Many of the pragmatic functions discussed in the previous
studies have something in common; that is, object-marking seems to be consistently used in
discourse contexts in which the speaker disagrees with what has been said in the preceding
context. Disagreement can vary in range from weak disagreement, such as expressing surprise or
wondering about the prior utterance or action, which goes contrary to the speaker’s belief or
expectation, to serious disagreement, such as showing a negative or sarcastic attitude (e.g.,
dispreference, disbelief, objection, mockery, etc.) or asking for a clarification or correction about
what has been said. In such challenging contexts, object-marking is a device used by the speaker
to direct the listener’s attention to salient information and hence the speaker’s intended meaning
so that the utterance can be successfully delivered to the listener. See (60).

37

(60) a. cha-ø takk-ass-e?
car wash-PST-SE
‘Did (you) wash the car?’

b. cha-lul takk-ass-e?
car-OBJ wash-PST-SE
‘Did (you) wash the car?’
[Implied: It’s surprising that you have washed the car. It’s so unlike you.]
[Implied: I don’t believe what you said. I know you didn’t wash the car.]
While (60a) can be uttered when the speaker is asking a yes-no question to the listener, (60b)
can be uttered not to ask for confirmation from the listener but rather to show the speaker’s
disagreement with the other participant’s prior action of washing the car. Thus, the speaker may
express surprise or disbelief about the other participant’s action, which goes contrary to the
speaker’s expectation or belief. In this challenging context, (l)ul-marking can be used by the
speaker to direct the listener’s focus or attention to the salient NP cha ‘car’ so that the speaker’s
intended meaning conveyed through the question can be more strongly expressed. Therefore,
(l)ul-marking helps the listener interpret the speaker’s intention, and act or reply accordingly and
hence the speaker’s communicative purposes can be successfully achieved.
Therefore, this dissertation will identify the pragmatic functions of (l)ul-marking in spoken
Korean and examine the role of (l)ul-marking in the successful achievement of the speaker’s
communicative purposes in conversation. The recurring linguistic forms or patterns related to
each pragmatic function will also be examined. In the next section, the method and data used in
the analysis will be introduced.

38

3. Methodology
In this dissertation, individual segments of actual spoken data are analyzed using Discourse
Analysis to identify certain linguistic environments where object-marking is consistently used. In
the following subsections, the methods of analysis will be briefly explained. The data used for
analysis will also be introduced.

3.1 Discourse Analysis
Discourse Analysis focuses on knowledge about language beyond the sentence, which is needed
for successful communication. Discourse analysis is an analysis of language in use. According to
Paltridge (2006), “Discourse Analysis considers the relationship between language and the
context in which it is used and is concerned with the description and analysis of both spoken and
written interactions” (p.3). Its primary purpose is to provide a deeper understanding of spoken
and written texts and how they become meaningful to their users.

3.1.1 Regularities in Data
Unlike formal linguists trying to find grammatical rules which are fixed and true 100% of the
time, discourse analysts discuss regularities in data. See the following explanation.
The regularities which which the analyst describes are based on the frequency with which a
particular linguistic feature occurs under certain conditions in his discourse data…The regularity in
discourse is a linguistic feature which occurs in a definable environment with a significant
frequency.
(Brown and Yule, 1983, p.22)
Thus, the discourse analyst will try to describe the linguistic forms which occur in data relative to
the environment in which they occur. Therefore, the discourse analyst attempts to discover
regularities in data and to describe them.

39

3.1.2 Communicative Purposes
People produce language to do certain things: to get a message across, to express ideas and
beliefs, to explain something, to get other people to do certain things or to think in a certain way,
and so on. As soon as we realize that we have produced language for a certain communicative
purpose, our actual use of language is defined as a text (Widdowson 2007).
(61) This room is a little bit hot.
Suppose you hear someone say (61). We may know well enough what the word room denotes,
but what the word denotes is not the same as knowing what it is meant to refer to when it occurs
in the phrase this room. The demonstrative pronoun this signals that what is being referred to is a
particular thing that is directly perceptible by both the speaker and the listener. This door is a
particular door in the room the speaker is in.
In other words, in typical discourses we will establish references by relating the text to the
context in which it is located. In this case, context may be constructed directly from the
immediate situation of an utterance, such as time, place, interlocutors, event, and so on. However,
relating the text to the actual situational context is not sufficient to understand what the sentence
truly means. We also need to relate it to the abstract cultural context of what we know to be
conventional. This context is often called schematic knowledge shared by people in the same
group, culture, or society, and it refers to the customary, normal, and natural ways of thinking
about events that exist among the members. What happens in text interpretation is that the
language triggers the recall of some familiar state of affairs, that is, some schematic knowledge.
Observe the following example.
(62) The service last Sunday left much to be desired.
(Widdowson, 2007, p.29)
Suppose you hear someone say (62). When you look up the word ‘service’ in the dictionary,
you will find that it has several semantic meanings. The question is which of them is
pragmatically appropriate in this context. However, when we see the word Sunday, we are able to

40

interpret the meaning of the word ‘service’ as a church service because in our familiar world such
services are customarily held on Sunday. Because what is familiar or customary can vary across
groups, society, and cultures, schematic knowledge represents culturally different ways of using
language to do certain things. For example, the sentence this room is a little bit hot can have
different interpretations in different cultures. In the United States, this sentence could be uttered
to initiate a conversation with another other person in the room. In Korea, however, the same
sentence could be uttered to indirectly request an action to the other person in the room: Please
open the window or Please turn on the air conditioner.
As shown above, Discourse Analysis attempts to find out what a text producer means by a text
and what the text means to the receiver. When this room is a little bit hot is uttered, for example,
the demonstrative pronoun this locates the utterance in a particular context of place that the
participants in conversation share, which enables the hearer to make an appropriate reference
regarding the door in the room. The person who utters this sentence, however, is not just referring
to the current state of affairs, but is performing a kind of illocutionary act, which refers to the
propositional content of what is being said. For example, this sentence might be a warning (e.g.,
The room temperature is somewhat unusual.), or an indirect request (e.g., Would you mind
opening the windows?).
What is said thus takes on a certain communicative value or illocutionary force. In this way, to
the extent that the sentence makes an appropriate connection with context, it can have a certain
reference and a certain force. Moreover, the speaker is doing something else as well. The speaker
is not just acting, but acting upon the other person, to bring about a certain state of mind or course
of action. In case of This room is a little bit hot, the speaker is performing an illocutionary act of
warning or requesting, which brings about a perlocutionary effect. In the case of warning, the
speaker may make everyone evacuate the room immediately. In the case of requesting, the
speaker may make the hearer open the windows or to turn on the air conditioner.

41

In other words, we can say that people produce texts to do something, that is, to realize their
communicative purposes. They make sense of texts, whether spoken or written, by relating them
to what they know of the world they live in, that is, the schematic knowledge that represents the
customary and conventional ways in which their socio-cultural reality is structured. When people
communicate, they have a shared understanding of how they should co-operate. The ways in
which people do this, however, vary across cultures (Paltridge 2006).

42

3.2 Data
The data used in this dissertation consists of the following: 13 transcripts of two-party
spontaneous informal telephone conversations, 1 transcript of the Korean reality TV show
Roommate, 2 transcripts of the Korean food TV show Tasty Guys, and 1 transcript of the Korean
animal TV show Man Feeding a Dog.
First, the telephone conversations were originally recorded as part of the CALLFRIEND
project. The CALLFRIEND Korean telephone speech corpus was collected by the Linguistic
Data Consortium primarily in support of the Language Identification (LID) project, sponsored by
the U.S. Department of Defense. The original database consists of 100 transcribed telephone
conversations with approximately 190,000 words in Korean. The recorded conversations are
between two native speakers of Korean and last up to 30 minutes. All participants were aware of
the fact that they were being recorded and they were given no guidelines about whom to call and
what to talk about. Most calls are between family members or close friends and all calls
originated either in the United States or Canada. The telephone conversations used in the data
analysis include 8 female-female interactions, 4 male-male interactions, and 1 male-female
interaction. Each conversation lasts for about 25-30 minutes, for a total of approximately 390
minutes of the spoken data.
Second, the reality TV show features unscripted informal conversations among 12 celebrities
living together in a shared house. These celebrities, including 7 male and 5 female speakers, have
a variety of occupations and include singers, comedians, and actors. The speakers’ ages range
from the 20s to 50s: 7 are in their 20s, 3 in their 30s, and 1 each in their 40s and 50s. Two
celebrities are non-native speakers of Korean. One is from Hong Kong and the other from Japan.
Each episode is approximately 50 minutes long.
Next, the Korean food TV show Tasty Guys features spontaneous informal interactions among
four celebrities, comprised of one female and three male speakers. In this show, the speakers visit

43

a restaurant, taste food, make comments about it, and provide useful tips on how to fully enjoy
eating the food. Each episode is approximately 50 minutes long. Lastly, the Korean animal TV
show Man feeding a dog features the celebrity’s daily life with their pets: feeding, bathing, and
playing with the pets and building bonds with them. These three Korean TV shows have been
chosen for the analysis of unscripted spontaneous informal conversations between more than two
participants. All speakers use the Seoul Dialect of the Standard Korean language, and all
conversations are transcribed in Yale Romanization.

3.2.1 Focus of Analysis
There are four kinds of NP forms in object slots: (a) omitted NPs, (b) zero-marked NPs, (c) (l)ul-
marked NPs, and (d) NPs marked with other particles.
(63) Omitted NPs
na ø chem mek-epwa
I first eat-try
‘This is the first time I’ve eaten (padak).’
(64) Zero-marked NPs
hwacangphwum-ø san-nal kathi os-ø sa-le ka-ss-e
cosmetics bought -day together clothes buy-in order to go-PST-SE
‘The day I bought cosmetics, I went shopping for clothes, too.’
(65) (l)ul-marked NPs
ecey sinmwun-ul pw-ass-ta, sinmwun-ul pw-ass-nuntey,
yesterday news paper-OBJ see-PST-SE newspaper-OBJ see-PST-CONJ
‘I read a newspaper yesterday and…’
(66) NPs marked with other particles
leymon-un amwulehcito anhkey mek-ul-swu-iss-eyo
lemon-TOP like nothing eat-can-SE
‘I can eat lemons like it’s nothing.’
(67) NPs marked with other particles
ike ha-n taumey kuliko ka-se kongyen-to hay
this do-after and go-CONJ performance-also do
‘I film this show and afterward I also go and perform.’
In (63), the direct object NP, padak (a type of fried chicken with shredded green onions), whose
referent is easily identifiable in the immediate context, does not overtly appear. While the two
direct objects hwacangphwum ‘cosmetics’ and os ‘clothes’ in (64) are left unmarked, the direct

44

object NP sinmwun ‘newspaper’ in (65) is overtly (l)ul-marked. On the other hand, the direct
object leymon ‘lemon’ in (66) is marked with the topic marker (n)un. Similarly, the object NP
kongyen ‘performance’ in (67) is marked with the emphatic particle to ‘also.’
Previous studies have found that unmarked bare object NPs are more frequent than (l)ul-
marked NPs. In Fujii and Ono’s (2000) data, 70% of direct object NPs are left unmarked while
only 30% of NPs are o-marked. Yang (2007) has found the following distribution in the
occurrence of the four types of direct object NPs in conversational data. Unmarked bare NPs
occur more frequently (40%), which is followed by omitted NPs (28%), (l)ul-marked NPs (24%),
and NPs with other particles (8%).
These results indicate that because zero-marking is predominantly used for direct object NPs,
the speaker actively chooses to use object-marking in certain discourse contexts. In previous
studies, object-marking has been described as a device used by the speaker to provide an extra
linguistic signal for hard-to-process information (e.g., post-predicate NPs) or important
information which requires the listener’s attention (e.g., discourse topics, NPs not clearly
understood by the listener, and contrastive referents) to facilitate the smooth processing of
information for successful communication (Fujii and Ono 2000). Therefore, the object particle
does more than simply indicate the grammatical role played by the NP; it can indicate discourse
topics, initiate repair, distinguish contrastive referents, and emphasize an action performed by the
actor (Yang 2007).
While previous studies have examined the pragmatic functions of object-marking in spoken
discourse, what they have missed is that most of those pragmatic functions of object-marking are
observed in certain discourse contexts when the speaker needs to disagree with the previous
utterances or actions of co-present or non-present parties in conversation and that object-marking
plays an important role in the interpretation of the speaker’s intended meaning and the
achievement of the speaker’s communicative purpose. There has also been no discussion about
discovering and classifying recurring linguistic forms or patterns frequently observed in each

45

pragmatic use of object-marking, which would provide a knowledge base toward gaining a more
comprehensive understanding of the mechanism that controls the speaker’s use of -(l)ul in spoken
Korean, and furthermore serve as a useful anchor point from which further research inquiries can
be made.
Therefore, the focus of this study is to investigate the validity of the pragmatic functions of
object-marking discussed in previous studies by considering the discourse context of
disagreement, to discover and classify the linguistic forms or patterns frequently co-occurring
with each pragmatic use of -(l)ul, and to suggest a more comprehensive mechanism that controls
the speaker’s choice of using (l)ul-marking in informal spoken Korean. I will also look into the
instances in which the Korean object particle (l)ul marks something other than the direct object of
a sentence, as in (68) and (69), and explain this extended use of -(l)ul in the discourse context of
disagreement.
(68) nwuka keki-l ka
who there-OBJ go
‘Who would go there?’
[Implied: Nobody would go there.]

(69) i ipwul-ul ne-l cwu-lkey
this blanket-OBJ you-OBJ give-will
‘I will give you this blanket.’
[Implied: I insist that I will give you this blanket.]

3.2.2 Factors Considered
The following factors are considered to examine their validity with respect to the speaker’s
pragmatic use of object-marking in informal spoken Korean: (a) shared knowledge among the
participants in conversation and (b) linguistic forms or patterns frequently co-occurring with (l)ul-
marked elements.

46

4. Analysis
This Chapter examines the discourse contexts in which object-marking is preferred to answer the
following questions: (a) What are the pragmatic functions of the Korean object particle -(l)ul in
informal spoken Korean? (b) What are the roles of (l)ul-marking in the interpretation of the
speaker’s intended meaning and the successful achievement of the speaker’s communicative
purpose? (c) What are the recurring linguistic forms or patterns in each pragmatic use of -(l)ul? In
the following sections, five pragmatic functions of object-marking and frequently co-occurring
linguistic forms or patterns in informal conversation will be discussed.
First, the object particle -(l)ul is frequently used to emphasize an assertion that challenges the
prior utterance or action of the co-present or non-present parties. When the speaker disagrees with
what has been said, object-marking directs the listener’s attention or focus to the information
including the marked element that is disagreed with or challenged, and hence the speaker’s
intended meaning conveyed through the utterance can be more strongly expressed and
successfully interpreted by the hearer without misunderstanding or difficulty. As a result, object-
marking leads to the successful achievement of the speaker’s communicative purpose.
For example, (l)ul-marked NPs are often observed in rhetorical questions. Rhetorical questions
are interrogative in structure but display a strong assertive force and generally do not expect an
answer from the hearer. They are interpreted by the participants in conversation as an assertion
challenging the prior utterance or action of the co-present or non-present parties rather than as an
ordinary information-seeking question. (l)ul-marking is used by the speaker to help the listener
pay attention to important information and hence the assertion conveyed through the question can
be successfully expressed and interpreted by the listener, which leads to the successful
achievement of the speaker’s communicative purposes, such as complaining, arguing, mocking,
blaming, persuading, etc.

47

Similarly, (l)ul-marking is also frequently observed in self-addressed questions expressing the
speaker’s feeling of wondering. When the speaker’s belief or expectation is in contrast with the
current situation, self-addressed questions can be asked. The speaker is not necessarily looking
for a solution to resolve the conflict between his belief or expectation and the current situation.
Rather, the questioner is expressing a feeling of wondering or puzzlement about the situation s/he
is being involved in that goes contrary to his belief or expectation. In this discourse context, (l)ul-
marking can direct the listener’s attention or focus to salient information and hence emphasize the
speaker’s feeling implied in the self-addressed question.
In addition, (l)ul-marking is also used to emphasize the speaker’s assertion in restatements.
When the interlocutor shows uncertainty, disagreement, or rejection of what the speaker has said,
the speaker may restate the previous utterance with -(l)ul. In this case, (l)ul-marking helps the
listener to focus on salient information and hence the speaker’s assertion to counter-challenge the
interlocutor’s uncertainty, disagreement, or rejection can be more strongly conveyed through the
restatement.
Second, the speaker prefers to use -(l)ul when the speaker thinks the prior utterance of co-
present or non-present parties might be wrong, inaccurate, or hard to accept. In this case, repair
can be initiated by the speaker to ask for clarification or correction of the information. (l)ul-
marking helps the hearer to pay attention to salient information which needs to be clarified or
corrected and hence facilitates the interlocutors to re-establish a shared understanding for
successful communication. In this repair situation, partial or full-repeats of the prior utterance
with various linguistic forms and patterns are frequently observed.
Next, (l)ul-marking is used to distinguish two or more contrastive referents, especially in
discourse contexts of disagreement. When the referent of an object NP is dispreferred, challenged,
or rejected by another referent in the conversation, (l)ul can be used to mark those two referents.
This helps the hearer pay attention to this salient information, which leads to the successful
achievement of the speaker’s communicative purpose.

48

Lastly, (l)ul-marking consistently appears when the information introduced in the object NP
continues as the topic of conversation. If the speaker wants to talk about something that the hearer
might not know very well, it would be hard for the hearer to actively participate in the
conversation due to the lack of shared knowledge. Thus, object-marking is a device used by the
speaker to provide a signal that further descriptions/explanations about the object NP are coming
up in subsequent contexts.

49

4.1 Emphasizing an Assertion Challenging the Prior Utterance or Action
The particle -(l)ul is frequently used in challenging environments where the speaker disagrees
with the prior utterance or action of co-present or non-present parties. Here, (l)ul-marking directs
the listener’s attention to salient referents so that the speaker’s assertion conveyed through the
utterances can be emphasized and delivered to the hearer without misunderstanding. Therefore,
object-marking helps the speaker successfully achieve certain communicative purposes.

4.1.1 Challenging the Prior Utterance or Action in Rhetorical Questions
The particle -(l)ul seems to consistently appear in rhetorical questions with wh-question words
such as nwukwu/nwuka ‘who,’ encey ‘when,’ mwues/mwe ‘what,’ and ettehkey ‘how,’ way ‘why,’
etc.
(70) nwu-ka keki-l ka
who-SUB there-OBJ go
‘Who would go there?’
[Implied: Nobody would go there.]

(71) kulen chwuchense-lul kaci-co eti-l ka
that kind of reference letter-OBJ have-and where-OBJ go
‘Where can he go with that reference letter?’
[Implied: He can’t go any there.]

(72) way phiano-lul sa ihay-ka an ka na-n
why piano-OBJ buy understanding-SUB NEG I-TOP
‘Why did he buy the piano? I don’t get it.’
[Implied: There is no reason why he should’ve bought the piano.]

(73) nay-ka mwe-l ettehkey hay-ss-tako-kulay totayche
I-SUB what-OBJ how do-PST-QT what on earth
‘What on earth have I done wrong?’
[Implied: I didn’t do anything wrong.]

(74) mikwuk sal-myense amoley-lul way sse
USA live-though Amore-OBJ why use
‘You’re living in America but why do you use Amore?’
[Implied: There is no reason why you should use Amore.]
It is generally known that wh-questions are commonly used to ask for a piece of new information.
Sometimes, however, they do more than simply seek answers from the hearer. The wh-questions

50

in (70)-(74) can be categorized as rhetorical questions. According to Quirk et al. (1985), “the
rhetorical question is interrogative in structure, but has the force of a strong assertion. It generally
does not expect an answer from the hearer” (p.825). They are interpreted by the participants in a
conversation as an assertion challenging the prior utterance or action of the non-present party as
in (70)-(72) or the co-present party as in (73)-(74) rather than as an ordinary information-seeking
question.
Rhetorical questions can accomplish the speaker’s communicative purposes such as
complaining, as in (72), arguing, as in (73), and mocking, as in (74). Most rhetorical questions are
not designed to expect answers. In (72), for example, a rhetorical question is asked by the speaker
and is immediately followed by another utterance. This means that the speaker does not allow for
an answer to be given.
As shown above, (l)ul-marking is consistently used in rhetorical wh-questions, but why? The
previous studies on rhetorical questions should be examined to answer this question. Some
linguists such as Bolinger (1957) and Quirk et al. (1985) have noted that rhetorical questions are
used to make assertions. As mentioned in Koshik (2005), Bolinger (1957) says certain wh-
questions are implying strong assertions.
(75) Look, it’s John who just came in. Who else?
[‘Nobody came in but John.’]

(76) Who believes such nonsense?
[‘Nobody believes in such nonsense.’]
(Koshik, 2005, p.40)
They are either uttered in a context which cancels possible answers, as in (75), or they contain
expressions indicating undesirability, as in (76). The answer expected is some synonym of ‘zero’:
nobody, nothing, nowhere, none, no reason, etc.
Quirk et al. (1985) examined the syntactic and semantic features of rhetorical questions.
(77) Who knows/cares?
[‘Nobody knows/cares or I don’t know/care.’]

51

(78) What difference does it make?
[‘It makes no difference.’]

(79) How should I know?
[‘There is no reason why I should know.’]

(80) who doesn’t know?
[‘Everybody knows.’]

(81) How couldn’t you remember?
[‘You certainly should have remembered.’]
(Quirk et al., 1985, p.826)
In rhetorical wh-questions, a positive question is equivalent to a strong negative assertion, as in
(77)-(79), while a less common negative question is equivalent to a strong positive assertion, as in
(80)-(81).
These previous studies provide brief syntactic and semantic descriptions of rhetorical
questions. However, the data to date are mainly based on intuition judgments of how these
questions are being used. They fail to see to how, in social interactional contexts, these questions
are heard as making assertions rather than seeking information and understanding the particular
social actions the questioner is trying to accomplish by using them. Thus, the conversation analyst
Koshik (2005) investigated how certain wh-questions are interpreted as a challenge conveying
reversed polarity assertions rather than information-seeking questions.
According to Koshik (2005), wh-questions are interpreted as a challenge primarily because of
their sequential position. When wh-questions are uttered in an already-established environment of
disagreement, accusation, complaints, and the like, challenging is a sequentially appropriate next
response. Wh-questions can also be heard as a challenge because they convey a strong epistemic
stance of the questioner. This means the questioner has greater claim to knowledge and knows the
answer to the question with certainty. This is possible either because the question concerns an
event that the questioner himself has experienced, or because the questioner has the role of expert
in this domain.
(82) Ok how’s it background [‘It’s not background.’]
(83) When have I [‘I have never.’]

52

(Koshik 2005, p.31; p.39)
For example, the wh-question in (82) is uttered by the teacher to challenge the student’s claim,
implying “It’s not background.” Similarly, (83) is used to challenge the accusation in a prior turn.
Here, “When have I” conveys a strong assertion, “I have never.” Koshik (2005) also points out
that when wh-questions are used as a challenge conveying an assertion, the speaker does not
expect an answer from the hearer. When responses are given, however, they are designed to either
align or disalign with the assertion conveyed through the question.
Returning now to the previously asked question, why is the object particle -(l)ul frequently
used in rhetorical questions? When there is an ongoing disagreement between the participants in
conversation, a rhetorical question can be asked to challenge the prior utterance or action of non-
present or co-present parties as a sequentially appropriate next response. In this challenging
environment, object-marking can be used by the speaker to signal the salient elements to the
listener, and hence the speaker’s assertion conveyed through the rhetorical question can be more
strongly expressed and clearly interpreted by the listener without misunderstanding. See Excerpt
1, a telephone conversation between two female speakers, referred to by their alias names, Jane
and Susan. Here, wh-questions are used to challenge a reported action of the non-present party.
Excerpt 1: Jane (J) and Susan (S)
Jane is talking about her colleague Dayoung. Jane helped her with house hunting and found a
one-bedroom apartment for her. Dayoung decided to move into the apartment and asked Jane
to help her find a roommate because the rent was too high.

1 J: kuntey yeki-eyse hawusu weynmanhan ke kwuha-myen[un,
but here-LOC house ordinary thing look for-when-TOP
But when you’re looking for a house for rent

2 S: [ung
Yes

3 J: sepuleys kathun ke ha-myen, ipayk pwul-i-myen ha-canha-[yo
sublet like thing do-if two hundred dollar-to be-if do-SE-PL
If you could find a sublet room, isn’t it $200 or something?

4 S: [ung
Right

53

5 J: kekitaka, yekiceki il pwul osip yuksip-ssik-i-ke[tun
on top of that, here and there one dollar fifty sixty-each-to be-SE
On top of that, it costs $1.50 or $1.60

6 S: [ung
Yes

7 J: ceki pesu tha-ko ilenun ke,
well bus ride-and that kind of thing
Well, taking a bus and things like that

8 [kulemyen han tal-ey choysohan yuksip pwul te tul-canha
then one month-per at least sixty dollar more cost-SE
Then, you spend at least extra $60 per month

9 S: [ung ung
Yeah, right

10 ayu, cwiyak mekess-ni nwuka keki ka-ni?
my goodness rat poison ate-SE who there go-SE
Oh, my goodness! Have you eaten rat poison? Who would go there?

à 11 J: nwuka keki-l ka
Who there-OBJ go
Who would go there?

12 keytaka wenpeytulwum-i-ntey twul-i ettehkey kathi salayo
besides one-bedroom-to be-but two-SUB how together live
Besides, it’s a one-bedroom apartment, so how could two people live there together?

13 am[man kathi hankwuk salam-i-lako kulayto
no matter how together Korea person-to be-even though
Even though they are both Korean

14 S: [ung:::? mal-to an toynta
huh word-EMP not become
Whaaat? That doesn't make sense!
Jane is telling Susan why she thinks Dayoung shouldn’t move into the apartment: She could
find a room for sublease for $200 (lines 1 and 3). She also needs to spend at least extra $60 per
month for transportation (lines 3, 5, 7, and 8). Susan agrees with what Jane has said (lines 2, 4, 6,
and 9) and utters ayu cwiyak mekessni “Oh my goodness, have you eaten rat poison?” in line 10.
This shows that Susan has a negative attitude towards Dayoung’s decision. Just as you would
never eat rat poison unless you were crazy, nobody would move into that apartment unless they
were crazy. The utterance is immediately followed by Susan’s wh-question nwuka keki kani

54

“Who would go there?” which is interpreted as a challenge conveying a negative assertion,
“Nobody would go there,” rather than asking for new information. This is because the question
has been uttered in an already accomplished environment of disagreement in the previous
utterances.
Note that in line 11, Jane agrees with the negative assertion conveyed through Susan’s
question in the previous turn and responds to it by restating the same question with the object
particle (l)ul, nwuka kekil ka “Who would ever go there?” Jane’s question is also heard as an
assertion challenging Dayoung’s decision to move into the apartment. Here, -(l)ul is used by the
speaker to signal the listener’s attention to the salient NP so that the speaker’s assertion conveyed
through the rhetorical question becomes more strongly expressed: “Nobody would ever go there.”
Note that -(l)ul marks something other than the direct object of the sentence, the destination keki
‘there’ of the movement verb ka ‘to go.’
Jane’s question does not even expect an answer from Susan because another wh-question is
immediately uttered in the next turn, keytaka wenpeytulwumintey twuli ettehkey kathi salayo
“Besides, it’s a one-bedroom apartment, so how could two people live together there?” This
refers back to Dayoung’s favor asked to Jane to help her find a roommate and it also conveys a
negative assertion challenging Dayoung’s request to find a roommate: “It’s impossible for two
people to live together in a one-bedroom apartment.” Susan’s response in line 14, malto antoynta
“That doesn’t make sense” also confirms that Jane’s question in line 12 is heard as a challenge
making an assertion rather than seeking information because it aligns with the assertion conveyed
through Jane’s question in the prior turn, “It’s impossible for two people to live together in a one-
bedroom apartment.”
See Excerpt 2, a telephone conversation between a brother and his younger sister, Eric and
Katy.

55

Excerpt 2: Eric (E) and Katy (K)
Eric told Katy about Junseok, a high school student who studied abroad in Australia for two
years. Now Junseok wants to transfer to a new school in the United States and Eric is helping
him find a new school.

1 E: kuliko tto ay-ka cekung-ul mos ha-ni[kka
and also kid-SUB adaptation-OBJ NEG do-CONJ
And because he also had trouble adjusting to his new school

2 K: [ung
Yes

3 E: chwuchense-lang ilen ingkullisi [thukhy-ey tayhayse
reference letter-and this English special conditions-about
Regarding a reference letter and English proficiency

4 K: [ung
Yes

5 E: koyngcanghi pwucengcekulo sse-nwa-ss-e
very negatively write-leave-PST-SE
What they have written is very negative

6 K: a:::
Ohhh

7 E: yenge-ka ceyil mwunceyta mwe ilen sikulo
English-SUB the most problem well this kind of way
His biggest problem is his English proficiency, something like that

8 mak sse-nwa-ss-tela-ko
harshly write-leave-PST-EVI-EMP
They wrote a harsh letter about him

à 9 K: kulen chwuchense-lul kaci-ko eti-l ka-na
that kind of letter-OBJ have-and where-OBJ go-Q
Where could he go with that reference letter?

10 E: kulenikka cikum mwe cohun hakkyo-nun cikum
so now well good school-TOP now
So well now as for good schools

11 kkwum-to mos kkwu-ko
dream-even NEG dream-CONJ
He can’t even dream of transferring to a good school.
Since Junseok had a hard time adjusting to his school life in Australia (line 1), his reference letter
is full of bad comments (lines 3 and 5). In particular, his low English proficiency turns out to be
his biggest problem (lines 7 and 8). After hearing about Junseok’s situation, Katy utters a wh-

56

question in line 9, kulen chwuchenselul kaciko etil kana “Where could he go with that reference
letter?” Her question is not heard as an ordinary question requesting new information, but as a
negative assertion challenging Junseok’s wish to transfer to a new school in the United States.
Note that (l)ul-marking is used for both the direct object chwuchense ‘a reference letter’ and the
non-direct object eti ‘where,’ which emphasizes the negative assertion conveyed through the
question, “He could never go anywhere with such a bad reference letter.” Eric’s response to
Katy’s question in lines 10 and 11 also aligns with the assertion conveyed through the question,
“So now, he can’t even dream of transferring to a good school.”
A rhetorical question, which is interpreted as a challenge rather than an ordinary-information
seeking question, can be uttered by the speaker to achieve a certain communicative purpose, such
as complaining. See Excerpt 3, a telephone conversation between two male speakers, Calvin and
Frank.
Excerpt 3: Calvin (C) and Frank (F)
Calvin is a friend of Frank’s older brother and he tells Frank that his brother has bought a
piano recently.

1 C: kulayse cengsin-i hana-to epse
so mind-SUB one-even not to exist
So it’s been very busy

2 cim yeki ceki phal-ko
stuff here and there sell-and
I’ve been selling my stuff here and there

3 hyeng phiano san ke al-ci?
older brother piano bought know-SE
You know your older brother bought a piano, right?

4 F: tul-ess-eyo
hear-PST-SE
I heard (it)

5 C: ung, kulay-ss-tako
yes it is so-PST-EMP
Yeah, he did (it)

à 6 F: way phiano-lul sa ((laugh)) ihay-ka an ka na-n
why piano-OBJ buy understand-SUB NEG I-TOP
Why did he buy the piano? I don’t even understand

57

7 C: a kuntey hyeng-un, hyeng-i phiano cikum
oh but older brother-TOP older brother-SUB piano now
But now your brother

8 yosay toykey ceki toykey cohaha-ko isse
these days very well very like-PROG
Well, he really enjoys playing the piano these days
Calvin says he has been very busy lately (lines 1 and 2). In the next turn, he asks Frank a tag
question, hyeng phiano san ke alci “You know your brother has bought a piano, right?” to
confirm that Frank knows his older brother has bought a piano. Frank responds to the question by
saying tulesseyo “I heard it.” In line 6, Frank utters a wh-question way phianolul sa “Why did he
buy the piano?” and the question is followed by slight laughter and his other utterance, ihayka an
ka nan “I don’t even understand.” Frank’s wh-question can be interpreted as a harmless
complaint about his brother’s impulse purchase because the question is not designed to expect an
answer from Calvin about the reason why his older brother bought the piano. Here, the direct
object phiano ‘piano’ is (l)ul-marked, which emphasizes the negative assertion conveyed through
the wh-question, “There is no reason why he should have bought the piano.” Therefore, (l)ul-
marking helps the speaker successfully achieve his communicative purpose, complaining about a
prior action of a non-present party.
When a wh-question is heard as an assertion challenging the prior utterance or action, it can
also accomplish the speaker’s particular communicative purpose such as arguing or mocking. See
Excerpt 4, a telephone conversation between Mom who is living in Korea and Daughter, an
international student studying abroad in the United States.
Excerpt 4: Mom (M) and Daughter (D)
Daughter told Mom that she wanted to get married and move somewhere else. After hearing
what her daughter said, Mom gives her some advice.

1 M: kulssey:: emma mal-ul ttokpalo tule
well mom word-OBJ clearly listen
Well, listen to what I’m saying

2 ni-ka hankwuk-eyse hankwuk salam-hako kyelhonha-ko
you-SUB Korea-LOC Korea person-and marry-CONJ

58

If you want to marry someone in Korea and

3 mitum an-eyse chakhan salam-ul manna-ko siph-umyen
faith inside-LOC nice person-OBJ meet-want to-if
If you want to meet a nice guy in the faith

4 ney-ka ney chalim-ina ney mosup-ul
you-SUB your outfit-or your appearance-OBJ
The way you dress or the way you look

5 ney-ka pakkwe-ya tway
you-SUB change-have to
You have to chance it

à 6 D: nay-ka mwe-l ettehkey hay-ss-tako-kulay totaychey?
I-SUB what-OBJ how do-PST-EMP what on earth
What on earth have I done wrong?

7 M: pwala kulay ne-nun ney mosup-i ettehkey sayngki-n ci
look well you-TOP your look-SUB how look-NOML
Look, well you may not know how you look

8 molu-ciman talun salam-i po-l ttay-nun
not to know-CONJ other people-SUB see-when-TOP
But when other people see you,

9 ney-ka thukihan hwacang-ul ha-ko
you-SUB unusual makeup-OBJ do-CONJ
You wear unusual makeup and

10 D: ehyu cham!
Oh my goodness!

11 M: ilehkey os ip-nun-kes-to com chenye-ka
like this clothes wear-RL-thing-also a little bit lady-SUB
When it comes to how you dress, as a lady

12 yamcenhakey ip-eya toy-nuntey,
neatly wear-have to-CONJ
You should dress neatly, but

13 kunyang thak tallapwuth-nun paci-eytaka moca-eytaka
just tight(ly) cling to-RL pants-and hats-and
You wear skin-tight pants and hats
Mom tells Daughter to change the way she dresses and the way she looks if she wants to meet
a nice guy and marry him (lines 1~5). After hearing Mom’s bitter advice, Daughter utters a wh-
question, nayka mewl ettehkey haysstako kulay totaychey “What on earth have I done wrong?” in

59

line 6. This can be heard as a challenge arguing against Mom’s prior utterances: Daughter thinks
there is nothing wrong with her looks and fashion choices. Here, object-marking is used for the
question word mwel ‘what,’ which emphasize a negative assertion conveyed through the question:
“I have never done anything wrong ever.” The assertion conveyed through her rhetorical wh-
question challenges the grounds for Mom’s prior claim, suggesting that there is no basis for
Mom’s claim.
In the next turn, however, Mom counterchallenges Daughter’s argument by providing reasons
for her claim: The way other people see her daughter and the way she sees herself could be
different (lines 7 and 8). People may think she wears unusual makeup (line 9), does not dress
neatly (lines 11 and 12), and wears tight pants and hats (line 13).
Interestingly, when Mom tells Daughter what to do or what not to do if she wants to meet a
nice guy and marry him (lines 1, 3, 4, and 9), object-making is consistently used for the object
NPs such as mal ‘word’ in line 1, salam ‘person’ in line 3, mosup ‘appearance,’ in line 4, and
hwacang ‘makeup’ in line 9. When the speaker tries to make the hearer agree with him or
persuade the hearer to do something, (l)ul-marking can be used to help the the hearer pay
attention to salient object NPs and hence the speaker’s assertion can be more strongly made.
See another example in Excerpt 5, which is taken from the 16th episode of the Korean food
TV show, Tasty Guys, in which where the two speakers are arguing against each other over a
bowl of dongchimi, radish water kimchi.
Excerpt 5: Jake (J) and Michael (M)
Jake requests a bowl of dongchimi and drinks the dongchimi juice from the bowl. Michael
tells Jake that he wants to take a sip, but Jake refuses to share it.

1 J: hana-ssik sikhi-nun ke-ya hana-ssik
one-each order-RL thing-to be one-each
We order one for each person

à 2 M: ya! ike-l ettehkey hana-ssik sikhye mianhakey
hey this-OBJ how one-each order sorry
Hey! How can I order this one for each person? I feel sorry

60

3 J: ani na-nun sikhye-ss-e mianha-myen mos sikhi-nun ke-ye
oh I-TOP ordere-PST-SE sorry-if NEG order-RL thing-to be
Oh, I have already ordered. If you feel sorry, you can’t order it

à 4 M: ani ike taycep-ey tam-un ke-l ettehkey yele kay tal-la kulay
oh I-TOP bowl-LOC put -RL thing-OBJ how several CL give-QT
Man, it’s served in a big bowl so how can I ask for another bowl?

5 J: alasse
All right

((Jake puts his thumb in the bowl of dongchimi and passes the bowl to Min))

à 6 M: sonkalak-ul way phwuk tamka
finger-OBJ why completely put in
Why are you putting your finger in the bowl?

7 J: capswu-l ke-myen capswu-si-[ko
eat-will-if eat-HON-CONJ
Eat if you want

8 M: [sonkalak-ul yek-kaci tamkwu-ney sonkalak-ul!
finger-OBJ here-up to put in-SE finger-OBJ
You put your finger in the bowl up to here! Your finger!
Jake refuses to share the dongchimi with Michael and tells him to ask for one more bowl (line 1).
In the next turn, Michael utters a wh-question, ikel ettehkey hana-ssik sikhye “How can I order
this one for each person?”, which can be heard as a challenge arguing against Jake’s previous
utterance. Here, the object NP ike ‘this,’ which refers to a bowl of dongchimi, is (l)ul-marked to
direct the listener’s attention to the salient object NP so that the speaker’s assertion conveyed
through the question is emphasized: “I can’t ask for another bowl of dongchimi.” Jake says he has
already ordered one for him and if Michael feels sorry for asking for another bowl of dongchimi,
he shouldn’t eat it (line 3).
After hearing Jake’s counterargument, Michael utters another wh-question, taycepey tamun
kel ettehkey yele kay talla kulay “How can I ask for another bowl of dongchimi which is served in
a big bowl?” (line 4). This can also be interpreted as a challenge arguing against Jake’s prior
utterance. Here, (l)ul-marking is used for the object NP taycepey tamun ke ‘the thing served in a
bowl’ and it emphasizes the assertion conveyed through the rhetorical question: “I could never

61

ask for another bowl of dongchimi because it’s served in such a big bowl.” Jake finally passes a
bowl of dongchimi to Michael (line 5). When passing a bowl, however, Jake playfully puts his
thumb into the bowl not to let Michael take a sip. Jake’s playful trick makes Michael laugh and
utter another wh-question sonkalakul way phwuk tamka “Why are you putting your finger in the
bowl?” with the (l)ul-marked NP sonkalak ‘finger’ in line 6, which implies a strong assertion
challenging Jake’s prior action: “You should not put your finger in the bowl!”
Note that when Michael restates Jake’s prior action sonkalakul yek-kaci tamkwuney
sonkalakul “You put your finger in the bowl up to here! Your finger!” in line 8, the previously-
mentioned-and-thus-easily-identifiable object NP sonkalak ‘finger’ is (l)ul-marked. Here, (l)ul-
marking directs the listener’s attention to this important object NP and hence Michael’s assertion
challenging Jake’s prior action of putting his finger into the bowl can be emphasized: “You
should never put your finger in the bowl. Your finger!”
Mocking, another communicative purpose challenging the prior utterance or action, can also
be achieved through rhetorical wh-questions with (l)ul-marking. See Excerpt 6, a telephone
conversation between two female friends, Lucy and Gina.
Excerpt 6: Lucy (L) and Gina (G)
Lucy is talking about Amore, the Korean cosmetics brand.

01 L: kuleko nase hwacangphwum-ul sa-ss-nuntey,
and then cosmetics-OBJ buy-PST-CONJ
And then I bought some cosmetics, but

02 eyu amoley toykey nappu-tela a[moley
ugh Amore very bad-EVI Amore
Ugh, Amore products are not good! Amore

03 G: [e
okay

04 eti? ((laugh)) ya ne wuskin[ta!
where hey you funny
Where? You’re so funny!

05 L: [way?
Why?

62

à 06 G: ne mikwuk sal-myense amoley-lul way sse
you USA live-thought Amore-OBJ why use
You’re living in the United States. Why do you use Amore though?

07 L: amoley-ka ceyil coh-tela
Amore-SUB best good-EVI
Amore is the best

08 langkhom-i com coh-untey talun ke
Lancome-SUB a little bit good-CONJ different thing
Lancome is good and other things like

09 phawute kulen ke langkhom ssu-nuntey,
Powder that kind of thing Lancome use-CONJ
For powder I use Lancome
Lucy tells Gina that she bought some cosmetics from Amore, the Korean beauty brand, but she
didn’t like it at all (lines 1 and 2). Gina says Lucy is very funny (line 4) and utters a wh-question
ne mikwuk salmyense amoley-lul way sse “You’re living in the United States. Why do you use
Amore though?” (line 6) This can be interpreted as a negative assertion mocking Lucy’s prior
action of buying Korean cosmetics. Gina thinks it does not make sense for people living in the
United States to use Korean cosmetics.
Even though Gina’s question seems to ask for the reasons for Lucy’s action, what it truly
implies is a negative assertion. There is no reason for Lucy to use Amore, the Korean cosmetics
brand. Here, (l)ul-marking is used for the object NP Amore, which directs the listener’s attention
to the salient NP and hence emphasizes the negative assertion conveyed through the question:
“There is no adequate reason why you should use Amore.”
To summarize, (l)ul-marking is frequently used in rhetorical questions when the speaker
disagrees with or challenges the prior utterance or action of co-present or non-present parties. In
this challenging environment, (l)ul-marking directs the listener’s attention to salient elements so
that the speaker’s assertion conveyed through rhetorical questions can be emphasized and
successfully delivered to the hearer. As a result, (l)ul-marking helps the speaker successfully
achieve particular communicative purposes such as complaining, arguing, and mocking, etc.
through rhetorical questions.

63

4.1.2 Expressing a Feeling of Wondering in Self-Addressed Questions
In self-addressed questions, (l)ul-marking is also frequently used. When there is a mismatch
between the speaker’s belief or expectation and the situation in which the speaker is currently
involved, a self-addressed question can be asked to express the speaker’s wondering or feeling of
being puzzled about the situation that is contrary to the speaker’s belief or expectation.
See Excerpt 7, a conversation between the three female friends Bae, Sandy, and Yoon in the
30
th
episode of the reality TV show Roommate.
Excerpt 7: Bae (B), Sandy (S), and Yoon (Y)
Sandy and Yoon came over to Bae’s shared house. They ask Bae if she likes living with young
housemates. Bae says it is very fun because she is starting a new lifestyle. Sandy asks Bae
what she likes the most about living with her housemates.

1 B: aytul-i kki-ka nemwu nemchi-nikka kulen ke-l po-ko iss-umyen
people-SUB talent-SUB very overflow-CONJ such thing-OBJ see-PROG-when
Because my housemates are very talented when I see their talent

à 2 nay-ka ettehkey paywu-lul hay-ss-ci
I-SUB how actress-OBJ do-PST-Q
I wonder how I could ever become an actress

3 ilen sayngkak-i tule
this kind of thought-SUB have
That’s how I feel

4 S: kunikka, yeki-n taypon-i eps-canha
that’s right here-TOP script-SUB not to exist-SE
Exactly! there’s no script for this show

5 B: kunikka, kukey nemwu nollawe
that’s right that-SUB very surprising
Right. That’s really amazing

à 6 yayneytul-un cengmal i kki-lul ta ettehkey ha-na
these guys-TOP really this talent-OBJ all how do-Q
I wonder how they deal with their talent

7 ilen sayngkak-i tul-ko manhi paywe tto
this kind of thought-SUB have-CONJ a lot learn too
That’s how I feel and I’m learning a lot too
Bae says that all her housemates are very talented (line 1). Compared to them, Bae thinks she is
not talented enough to be an actress. Her current life as an actress is an unexpected situation

64

which is contrary to her belief. When Bae’s belief about herself conflicts with the situation she is
involved in (e.g., her lack of talent versus her current life as an actress), a self-addressed question
is asked, nayka ettehkey paywulul hayssci “I wonder if I could ever become an actress” (line 2).
Her question is not interpreted as an ordinary information-seeking question. She is not asking
herself how she became an actress. Rather, it is heard as an assertion expressing wondering about
her current situation which goes contrary to her actual belief: “I cannot believe I am an actress
now. I wonder how I was ever able to become an actress because I believe I have no talent.” Here,
(l)ul-marking is used to direct the listener’s attention to this salient NP paywu ‘actress’ which is
involved in the information contrary to the speaker’s belief so that the speaker’s feeling conveyed
through the self-addressed question can be more strongly expressed.
Similarly, in her other self-addressed question yayneytulun cengmal I kkilul ta ettehkey hana
“I wonder how they deal with their talent” in line 6, Bae does not intend to learn about the way
her housemates deal with their endless energy and extraordinary talent. Rather, she is expressing
her surprise or wondering about her housemates who have such great talent. Compared to them,
Bae thinks she has no talent. (l)ul-marking is used for the object NP kki ‘talent’, which
emphasizes the implied meaning hidden behind the self-addressed question: “I can’t even imagine
the life of those talented people because I have no such talent.” See the continued conversation.
Excerpt 7 Extended: Bae (B), Sandy (S), and Yoon (Y)

8 Y: enni-nun nwukwu-lang ceyil cal maca hohup-i yeki-se?
older sister-TOP who-with the most well fit harmony-SUB here-LOC
Who do you find it easy to get along with?

9 B: na-nun kulssey:: elun-tul-i phyenhay. yeki-n ta ay-tul-i-canha
I-TOP well elder-PL-SUB comfortable here-TOP all kid-PL-to be-SE
For me, well I feel more comfortable being with people who are older than I am. Here,
everyone’s younger than I am.

10 S: eme kule-keyss-ta
oh like that-will-SE
Oh, I bet you are.

65

11 B: kunikka cheumey-nun kitay-l tey-ka eps-nun ke-ya, maum-ulo
so at first-TOP rely on-RL place-SUB not to exist-RL-thing-to be heart-INST
So at first I felt like there was no one to rely on.

12 S: chemey himtul-ess-keyss-ta
at first struggle-PST-will-SE
When you first joined the show, you must have had a hard time

à 13 B: kulayse ike-l ettehkey ha-ci kulay-ss-nuntey,
so this-OBJ how do-Q do so-PST-CONJ
So what am I supposed to do with this? I was like that, but

14 kuntey paywu aytul-hako cal maca
but actor people-with well fit
but I work well with other actors.
Yoon asks Bae about the person she gets along well in the shared house where Bea is staying
(line 8). Bae replies that she feels more comfortable being with people who are older than she and
that she is the oldest among her housemates (line 9). This makes her feel like there was no one to
rely on when she first joined the show (line 11). In her self-addressed question ikel ettehkey haci
“What am I supposed to do with this?” in line 12, Bae even admits that she didn’t know what to
do when she first joined the show. Here, (l)ul-marking is used for the object NP ike ‘this’
referring back to Bae’s prior utterance of having no one around, which helps emphasize her
feelings conveyed through the question: “I literally had no idea what to do.”
To summarize, (l)ul-marking is frequently used in self-addressed questions to emphasize the
speaker’s feeling of wondering or being puzzled about the situation which goes contrary to the
speaker’s belief or expectation.

66

4.1.3 Challenging the Prior Utterance or Action in Restatements
It has also been found that (l)ul-marking is consistently used in the restatement of the prior
utterance. In particular, when the hearer expresses rejection, disagreement, or uncertainty about
the speaker’s prior utterance, the speaker may need to restate what has been said to convince or
persuade the hearer. When there is an ongoing disagreement between the participants, object-
marking can be used to help the listener selectively focus on salient elements so that the speaker’s
assertion conveyed through restatements can be more strongly expressed. See another
conversation between Bae (and) and Yoon (Y) from the 30th episode of the reality TV show
Roommate in Excerpt 8.
Excerpt 8: Bae (B) and Yoon (Y)
Yoon is sleeping over at her friend Bae’s shared house. Bae is worried that Yoon might feel
cold when she is sleeping. Bae wants to give Yoon a pair of pants and an extra blanket, but
Yoon says she will be fine.

1 B: ike an-ey ipwul eps-ni?
this inside-LOC blanket not to exist-Q
No blanket inside this?

2 na-n ike an-ey ipwul iss-nuntey
I-TOP this inside-LOC blanket exist-CONJ
I have a blanket inside this

3 ne an chwuwu-lkka? ike cwu-lkka?
you NEG cold-do you think this give-Shall
Don’t you think you might feel cold? Shall I give you this?

4 ike cwu-lkey. ike ppal-ase nehe-noh-un ke-ya
this give-will this wash-and then put in-leave-RL thing-to be
I’ll give you this. I just washed this and put it away (in the closet)

5 Y: i wi-eyta kkyeip-ko ca-lkey
this top-LOC wear-CONJ sleep-will
I’ll just put on more clothes and go to sleep

à 6 B: kuliko i ipwul-ul ne-l cwu-lkey
and this blanket-OBJ you-OBJ give-will
And I’ll give you this blanket

7 Y: aniya ike-man iss-umyen tway
no this-only have-if become
As long as I have this, I should be fine

67

Bae checks Yoon’s bed sheets and comforter (lines 1~2) and tells Yoon that she can give her a
pair of pants (lines 3~4). Yoon says she will just put on more clothes and go to sleep (line 5).
When Bea offers her a blanket in line 6, i ipwulul nel cwu-lkey “I will give you this blanket,” both
the direct object ipwul ‘blanket’ and the recipient ne ‘you’ are (l)ul-marked. Note that the
recipient of the verb cwu ‘to give’ is supposed to be either marked with the particle hanthey ‘to’
or left unmarked in informal conversation. Since Bae’s previous offer has been turned down by
Yoon, object-marking is used to emphasize her assertion counterchallenging Yoon’s previous
action of refusing her offer: “I will give you this blanket. I insist.”
See Excerpt 9, a conversation between a female and male friend, Choi and Ju, which is taken
from the fourth episode of the Korean animal TV show, Man Feeding a Dog.
Excerpt 9: Choi (C) and Ju (J)
Choi is coming over to her friend Ju’s house. In their previous phone conversation, Choi
asked Ju to cook rice so that they could have lunch together at his house. However, Ju
couldn’t do it because he was busy bathing his three dogs.

1 C: pap mek-ess-e?
meal eat-PST-SE
Have you eaten lunch?

2 J: pap?=
meal
Lunch?

3 C: =ung.
Yes

4 J: pap-ul an mek-ess-ci::
meal-OBJ NEG eat-PST-SE
I haven’t eaten it yet

5 C: wuli ppalli pap mek-ca
we quickly meal eat-SE
Let’s eat!

6 J: pap eps-nuntey, amwukesto
rice not to exist-CONJ nothing-even
There is nothing to eat. Nothing

7 C: pap-ul haynohu-lako kulayss-canha nay-ka
rice-OBJ do-IND.Q (I) did it-SE I-SUB
Didn’t I ask you to cook rice?

68

8 J: a kulehci
oh that’s right
Oh, right

((The scene flashes back to Ju’s phone conversation with Choi before she arrived.
Ju called Choi and invited her to his house for lunch.))

9 J: onul ceki cip-ey o-sy-ese siksana hasi-ci
today well house-to come-HON-and meal do-SE
Why don’t you come over to my house today and have a meal?

10 C: pap com hay noh-useyyo
meal a little bit do leave-SE
Please cook rice

11 J: pap-ul hay noh-ulako?

meal-OBJ do leave-EMP
Are you saying you want me to cook rice?

12 C: pap ha-l cwul al-ci?

meal do-NOML know-SE
You know how to cook rice, right?

13 J: cikum palo toy-nun pap-i iss-ki-n iss-nuntey,

now right away become-RL rice-SUB exist-NOML-TOP exist-CONJ
I do have some instant rice

14 C: ai kulen pap silh-ko, papsoth pap hay-cwu-seyyo

no that kind of rice dislike-and rice cooker rice do-give-SE
No, I don’t like instant rice. Please cook rice with a rice cooker.

15 J: kulay alasseyo. kulem pap hay nohu-l theynikka
Okay all right then rice do leave-CONJ
Okay, then I’ll cook the rice.

((The phone call scene ends and the prior conversation between Choi and Ju is continued.))

16 C: hay-ss-e?
do-PST-SE
Did you cook?

17 J: mokyok sikhi-nulako sey mali sikhi-lyemyen [elmana::
bath make-CONJ three CT make-if how much
Because I was giving them a bath… when you bathe three dogs

à 18 C: [kulem pap-ul ppalli hay
then rice-OBJ fast do
Cook rice right now then

19 J: sikan-i kelli-nuntey::
time-SUB take-CONJ

69

It takes a long time

20 C: papsoth etisse, papsoth?
rice cooker where-exist rice cooker
Where’s the rice cooker? The rice cooker

21 J: pathsoth? keki
rice cooker there
The rice cooker? There

à 22 C: samchon-i pap-ul hay
uncle-SUB rice-OBJ do
You cook rice
Choi asks Ju if he has eaten lunch (line 1) and Ju says that he has not eaten yet (line 4). Choi
suggests to Ju that they have lunch together (line 5) but Ju says there is nothing to eat in his house
(line 6). Choi seems to be unhappy to hear Ju’s answer and she reminds him that she had asked
him to cook rice (line 7). Ju finally recalls her request (line 8) and the scene flashes back to Ju’s
phone conversation with Choi before she arrived. Ju called Choi and invited her to his house for
lunch (line 9). Choi asked Ju to cook rice (line 10). Ju said he had some instant rice (line 13), but
Choi wanted him to cook rice with the rice cooker (line 14). Choi finally promised her to cook
rice (line 15).
Getting back to the previous conversation, Choi double-checks if Ju has cooked the rice or not
(line 16). Ju starts to make excuses by saying he was busy bathing his three dogs (line 17). In the
next turn, Choi does not even let him finish his words and asks Ju to cook rice. Even though the
object NP pap ‘rice’ has already been mentioned several times in the previous utterances and
hence its referent is easily identifiable by the participants in the conversation, the NP is still (l)ul-
marked. I claim that this is because the speaker needs to direct the listener’s attention to this
salient object NP so that what the speaker wants from the listener can be more strongly expressed:
“I want you to cook rice right now.”
After telling him what to do, Choi asks Ju where the rice cooker is (line 20). Since Ju rarely
cooks at home, he really doesn’t want to cook. He doesn’t even to into the kitchen and simply
points to the kitchen cabinet where Choi can find the rice cooker (line 21). After witnessing his

70

unenthusiastic attitude, Choi asks him to cook rice again (line 22). This ongoing disagreement
between the two interlocutors makes Choi restate her prior utterance with (l)ul-marking in line 22,
which emphasizes her assertion challenging Ju’s reluctance to cook rice: “You should cook rice
immediately.”
To summarize, (l)ul-marking has a pragmatic function of emphasizing an assertion
challenging the prior utterance or action, and it is frequently used in restatements. When the
speaker’s prior utterance is disagreed with or rejected by the hearer, the speaker may need to
restate his prior utterance to convince or persuade the hearer. Object-marking is used to direct the
listener’s attention to salient elements and hence the speaker’s assertion conveyed through
restatements can be emphasized and successfully delivered to the hearer.

71

4.2 Initiating Repair on the Prior Utterance or Action
Robinson and Feldman (2010) note that when a participant in a conversation has difficulty
hearing, understanding, agreeing with, or accepting what has been said, repair can be initiated to
deal with such trouble. When someone initiates repair on prior utterances, the repair initiation
targets the source of trouble more or less specifically. According to Koshik (2005), the practices
for targeting the source of trouble are as follows: repair initiators such as “huh?” or “what?”;
phrases such as “what do you mean”; wh-question words such as “who,” “where,” “when,” used
alone or together with a partial repeat of the turn that is the source of the trouble (e.g., “met
whom?”); the phrase “you mean” plus a candidate understanding of the prior talk; and a partial or
full repetition of the prior talk with upward intonation (p.112).
When the speaker does not agree with or accept what has been said, repair can be initiated to
ask for clarification or correction about the information mentioned in the prior utterance. The
speaker may show uncertainty, surprise, disbelief, or disagreement, or challenge the accuracy of
the prior utterance that goes contrary to the speaker’s belief or expectation. Similar practices
targeting the source of trouble have also been observed in spoken Korean: a partial repeat of the
previous utterance together with adverbs expressing the speaker’s surprise or disbelief; a partial
repeat of the prior utterance of the other speaker or non-present parties with an adverb expressing
the speaker’s surprise or doubt; a candidate understanding of the turn that is the source of trouble
plus the phrase -nun mali-/iyaki- ‘you mean,’ a partial repeat of the prior talk together with the
emphatic ending -ko; and a partial or full repetition of the prior talk with upward intonation. In
this challenging environment, object-marking directs the hearer’s attention to the information
involving the salient referent that needs to be clarified or corrected and hence helps the re-
establishment of shared understanding between the speaker and the hearer.

72

4.2.1 Showing Surprise or Disbelief Using a Partial Repeat of the Prior Utterance
with an Adverb
Adverbs expressing the speaker’s surprise or doubt about what someone has said often co-occur
with (l)ul-marking in repair situations. (l)ul-marking directs the listener’s attention to the
information involving the marked element which goes contrary to the speaker’s belief or
expectation and hence which needs to be clarified or corrected. See another conversation between
two female speakers Bae and Gukju and one male speaker Andrew taken from the 30
th
episode of
Roommate.
Excerpt 10: Bae (B), Gukju (G), and Andrew (A)
Gukju made fish cake soup for her housemates. While looking for a cooking pot in the kitchen,
Gukju accidently threw away the corn silk tea Bae had made.

1 B: ya kwukcwu-ya
hey Gukju-HOR
Hey, Gujku

2 G: ney?
Me?

3 B: ce okswuswuswuyemcha com masye pwuski nayli-key
that corn silk tea a little bit drink swelling go down-ADV
Drink corn silk tea. It will reduce swelling.

((Gujku doesn’t understand what Bae is talking about and she is confused.))

4 B: emememe ay-tul-i pely-ess-na pota
oh my goodness kid-PL-SUB throw away-PST-seem
Oh my goodness! Somebody must have thrown it away

((Gujku points at the pot.))

5 G: yeki iss-ten ke-yo?
here to exist-used to thing-POL
You mean something over here?

6 B: ni-ka pely-ess-ci? kukey okswuswuswuyemcha-yess-nuntey
you-SUB throw way-PST-SE that corn silk tea-to be-PST-CONJ
You threw it away, right? It was corn silk tea

7 G: mwe kkulhi-taka emwuk kkulh-in naympi-ka i naympi-yess-ketun-yo
something boil-CONJ fish cake boil-RL pot-SUB this pot-to be-PST-CONJ-POL
I was cooking something and this is the pot I used to cook fish cake soup

73

à 8 A: cincca okswuswuswuyem-ul kkulhy-ess-e?
really corn silk-OBJ boil-PST-SE
Did you really boil corn silk?

9 B: kulay
yeah

10 A: emena!
Oh, wow!

11 B: kukey pwuski ceykeey coh-tay
that swelling elimination good-QT
People say it’s good for reducing swelling
Bae calls Gukju (line 1) and tells her to drink corn silk tea, which is good for reducing
swelling (line 3). The object NP okswuswuswuyemcha ‘corn silk tea’ in line 3 is left unmarked
because Bae thinks the listener has shared knowledge about this object NP, that is, that Bae has
made the tea for her housemates, and hence its referent is easily identifiable by the listener.
However, Gukju looks confused because she doesn’t know Bae has made the tea and therefore
she can’t understand why Bae is talking about the corn silk tea so suddently out of the blue.
Seeing Gukju’s reaction, Bae assumes that somebody must have thrown away the tea she had
made (line 4). Gukju points at the pot and asks Bea whether the tea was in the pot (line 5). Bea
realizes that Gukju has thrown away the tea and tells her that what she has thrown away is corn
silk tea (line 6). The object NP okswuswuswuyemcha ‘corn silk tea’ in line 4 and line 6 is easily
dropped because its referent has been already introduced into the discourse context and therefore
the speaker assumes that the listener knows what is being talked about based on their shared
knowledge.
Gukju tells Bae that she accidently threw away the tea in the pot while she was cooking fish
cake soup (line 7). After hearing the story about the corn silk tea, Andrew utters a question in line
8 with the adverb cincca ‘really,’ which expresses the speaker’s surprise or doubt about what has
been said in the prior talk. Here, the object NP okswuswuswuyemcha ‘corn silk tea’ is overtly
(l)ul-marked even though its referent is assumed to be shared by the interlocutors. It is because
Andrew’s question is not uttered just to confirm what has been said in the prior utterances but to

74

show his surprise or disbelief about Bae’s prior action associated with this object NP, making tea
by boiling corn silk. Object-marking directs the listener’s attention to the information involving
the NP okswuswuswuyemcha ‘corn silk tea’ which needs to be clarified. Bae replies to Andrew’s
question that she did make the tea by boiling corn silk (line 9). Andrew’s reaction to her answer in
the next turn, emena “Oh, wow!”, also aligns with the feeling of surprise conveyed through his
prior utterance. Bae also provides a further explanation about the relieving effect of corn silk tea
on swelling (line 11).

4.2.2 Showing Surprise or Disbelief Using a Candidate Understanding of the Prior
Utterance with -nun mali-/iyaki- ‘You Mean’
When the speaker has difficulty understanding what has been said, a candidate understanding of
the trouble source turn plus the phrase -nun malita/iyakita ‘you mean’ can be used to ask for
clarification. This lack of understanding does not only show that the speaker has little knowledge
of the repeated item in context but also indicates that the speaker hardly believes what has been
said because the information in the repeated item is contrary to the speaker’s belief or expectation.
This requires the hearer’s additional explanation or clarification in the next turn. In this
challenging environment, (l)ul-marking directs the listener’s attention to the information
involving the salient referent that needs to be clarified by the listener. Thus, (l)ul-marking helps
reestablish shared understanding between the interlocutors for successful communication. See the
continued conversation of Excerpt 3, a telephone conversation between the two male speakers,
Calvin (C) and Frank (F).
Excerpt 3 Extended: Calvin (C) and Frank (F)
Frank has not been able to get a job, so his older is sending him some money for financial
support.

9 F: ya:: hyeng-i yelsim-i-[ney!
wow older brother-SUB hard-working-to-be-SE
Wow, he’s really into playing the piano!

75

10 C: [ung yelsim-i-tela-ko
yeah hard-working-to be-EVI-EMP
Yeah, I saw him practicing the piano

11 F:

il-to yelsim-i-ko, unhyeki-nun ilcali-lul mos cap-a kacko
work-also hard-working-to be-and Eunhyeok-TOP job-OBJ NEG get-CONJ
He’s also working hard. Well, I haven’t been able to get a job so

12 kwulm-ko anc-ass-nun[tey
starve-and sit-PST-CONJ
I’m starving here

13 C: [kulay?
Really?

14 F: ani hyeng-i ponay-cwu-nun ton sseyo
no older brother-SUB send-give-RL money use
No, I’m using money that my brother sent me

15 C: hyeng-i ton-to ponay-cwe?
older brother-SUB money-also send-give
Does your brother send you money, too?

16 F: apeci-hanthey emeni-hanthey-nun kuke pat-a-ssuki nemwu kule-n ke kath-ase
father-from mother-from-TOP that receive-and-use very like that-seem-CONJ
Because I think it’s too much to get money from my father or mother

17 C: ung
yeah

18 F: ilcali capu-l ttay-kkaci-man
job get-when-until-only
until I get a job

à 19 C: ung hyengi tonu-l ponay-cwu-n-ta-n mal-i[ya?
yes older brother-SUB money-OBJ send-give-PRES-SE-RL word-to be
Yeah, you mean your brother sends you money?

20 F: [yey
Yes

21 C: molla-ss-[ney
not to know-PST-SE
I didn’t know that

22 F: [mek-ko sa-l mankhum-yo
eat-and live-enough to-PL
Enough to eat and live

23 C: ung
Yeah

76

24 F: pap-un [mek-eya toy-l kes ani- ((laugh))
meal-TOP eat-have to-RL-thing not to be
Don’t you think I need to eat

25 C: [ya:: hyeng-i tto kulen tey-ka iss-ess-e?
wow older brother-SUB again like that aspect-SUB exist-PST-SE
Wow, is your brother that kind of guy?
Frank and Calvin talks about Frank’s older brother who is really into playing the piano
recently (lines 9~10). Frank tells Calvin that he is experiencing financial hardship because he has
not been able to get a job (lines 11~12). Calvin is surprised at hearing the news (line 13). Frank
tells Calvin that his financial situation is not too bad because his older brother has been sending
him some money (line 14). In the next turn, Calvin asks Frank a question with a partial repeat of
the prior utterance, hyengi tonto ponay cwe “Does your brother send you money, too?” Frank
assumes that Calvin does not know his older brother is helping him financially and explains more
about the situation: Frank’s brother will continue to help him until he gets a job because Frank
doesn't want to ask for money from his dad and mom (lines 16 and 18).
After hearing Frank’s further explanation about the situation, Calvin asks the same question
again with the phrase -nun mali- ‘you mean’ hyengi tonul ponay cwuntan maliya? “You mean
your brother sends you money?” in line 19. The question is not uttered to confirm what has been
said in the prior utterance. Rather, it is uttered to display the speaker’s lack of understanding
about the information in the prior utterance which goes contrary to the speaker’s belief and
expectation. Calvin hardly believes what Frank has said about his brother, which goes contrary to
Calvin’s expectation: “It is hard to believe that your brother has been sending you money. That’s
not like him, that I know of.” Here, (l)ul-marking is used to direct the listener’s attention to the
salient object NP ton ‘money’ which needs to be clarified and hence helps the interlocutors
reestablish a shared understanding about the event involving the marked NP, Frank getting
financial support from his older brother.
Calvin’s disbelief about the generous act of giving practiced by Frank’s older brother is

77

conveyed through the candidate understanding of the prior utterance plus the phrase -nun mali-
‘you mean,’ which leads to Frank’s further explanation and justification about getting financial
support from his older brother (lines 22 and 24). Frank says the money is just enough to pay his
living expenses. Calvin’s utterance hyengi tto kulen teyka issesse “Is your brother that kind of
person?” in line 25 also aligns with the disbelief conveyed through his prior question in 19.
See Excerpt 11, a telephone conversation between two male friends where (l)ul-marking
appears in a repair situation where the speaker is expressing surprise or disbelief about what has
been said and is asking for clarification.
Excerpt 11: Aron (A) and Tyler (T)
Aron is talking about his Chinese officemate who spends a lot of money on phone bills.

1 A: elma ceney-nun cenhwapi mak kyeysanha-ko iss-tela-ko kun[tey,
while ago-TOP phone bill hard calculate-PROG-EVI-EMP and
I saw him calculating his phone bills lately and

2 T: [ung
yes

3 A: kyayney-ka mwe kyayney-ka tisukhawunthu-lul an hay cwe-ss-tay
they-SUB well they-SUB discount-OBJ NEG do-give-PST-QT
He said they didn’t give him a discount

4 tisukhawunthu-lul hay cwe-ya toy-nun[tey
discount-OBJ do-give-have to-CONJ
They were supposed to give him a discount

5 T: [ung
Yes

6 A: kuntey kuke-lul hay-kaciko mwe kyeysan hay-kaciko ponay-lyeko-kulen-tay
and that-OBJ do-CONJ well calculate-CONJ send-about to-do so-QT
And he is trying to get a discount, get the discount rate reflected on his bills and sent out

7 T: ung
Okay

8 A: ttak po-nikka
just see-CONJ
I had a look at his phone bills

9 T: ung
Okay

78

10 A: ku han thonghwa-ey payk phalsip payk yuksipil pwun-i-n-ka
well one call-per hundred eighty hundred sixty one minute-to be-PRES-Q
I saw some phone calls one hundred eighty, one hundred sixty one or

11 payk phalsip pwun-ccali-to iss-tela
hundred eighty minute-per-also to exist-EVI
one hundred eighty minutes long

12 T: ((laugh)) kulem mwe myech sikan-ul ha-n keya?
then huh how many hour-OBJ do-RL thing-to be
Then, boy, how many hours did he talk on the phone?

à 13 sey sikan-ul hay-ss hay-ss-ta-n yayki-ya?
three hour-OBJ do-PST do-PST-SE-RL story-to be
You mean he talked on the phone for three hours?
14 A: mwe ani kukey tisukhawunthu-ka an tway iss-unikka
well oh that discount-SUB NEG become-CONJ
Oh well it’s because the discount was not applied

15 T: a::: kulayto kulehci
ahhh though that’s it
I see but that’s still too much
Aron is talking about his Chinese officemate who is overusing the phone. Aron saw his
officemate calculating his overcharged phone bills because he wasn’t able to get a discount (lines
1, 3, 4, and 6). Aron even had a look at the phone bills and saw some phone calls that lasted up to
180 minutes long (lines 8, 10, and 11). After asking Aron for how long his officemate has talked
on the phone (line 12), Tylor utters another question with his candidate understanding about the
event plus the phrase -nun yayki- ‘you mean,’ seysikanul hayss haysstan yaykiya “You mean he
talked on the phone for three hours?” (line 13).
This question does not simply confirm what has been said in the previous utterance. Rather, it
is uttered to display the speaker’s lack of understanding and disbelief about the non-present
party’s action in the prior talk, which goes contrary to the speaker’s expectation, and to ask for
clarification about it: Tylor hardly believes what Aron has said about his officemate, who has
spent enormous amounts of time and money on the phone. Here, (l)ul-marking is used to direct
the listener’s attention to the salient NP seysikan ‘three hours,’ whose referent needs clarification
so that the interlocutors can rebuild a shared understanding about the event involving the marked

79

NP for successful communication. Thus, Tylor’s question is interpreted as follows: “It’s hard to
believe that people could spend such a long time talking on the phone.” The question also
initiates Aron’s further explanation about extremely long phone calls in the next turn by saying
that it is because a promised discount has not been reflected in the phone bills (line 14).

4.2.3 Showing Disagreement Using a Partial Repeat of the Prior Utterance with the
Emphatic Sentence Ending -ko
When the speaker thinks that what the other speaker has said might be inappropriate, wrong, or
hard to accept, pre-disagreements, or indications that a disagreement is on the way, can be used to
initiate self-correction or other types of backdown by targeting a trouble source in the prior
utterance (Schegloff 1995). Pre-disagreements can be used to avoid a dispreferred action, or a
disagreement, by providing the other party with an opportunity to self-correct or modify the
utterance in some way, thereby making disagreement unnecessary. The preferred response to a
pre-disagreement is therefore a self-correction, or a backdown, which prevents a dispreferred
action of other-correction or disagreement (Koshik 2005).
A partial repeat of the other speaker’s previous utterance with the emphatic sentence ending
-ko ‘Are you saying~’ can be used as pre-disagreements in informal spoken Korean. The
emphatic sentence ending -ko is derived from the hearsay/quotative construction, -ko ha-. The
indirect quotation marker -ko is followed by a saying verb ha and sentence concluding endings in
indirect quotations. The four sentence concluding endings are possible in the plain formal style of
address based on the four sentential types: declarative -tako, interrogative -nyako, imperative
-lako, and propositive -cako. When the saying verb ha is omitted in this quotative construction,
the complementizer -ko, combined with one of the four sentence concluding endings, -tako,
-nyako, -lako, and -cako, can be used as the emphatic sentence ending which emphasizes the
speaker’s idea and opinion.
It has been found that (l)ul-marking frequently co-occurs with a partial repeat of the other

80

speaker’s prior utterance plus the emphatic sentence ending -ko in repair situations. In this
challenging environment, the speaker shows a pre-disagreement about the other speaker’s prior
utterance which is contrary to the speaker’s belief or expectation to initiate the other speaker’s
self-correction or backdown. (l)ul-marking helps the listener focus on the information involving
the salient referent that needs to be modified or corrected so that the interlocutors can rebuild
shared understanding for successful communication. Revisit Excerpt 9, a conversation between a
female and male friend, Choi and Ju, which is taken from the fourth episode of the Korean animal
TV show, Man Feeding a Dog.
Excerpt 9: Choi (C) and Ju (J)
Ju called Choi and invited her to his house for lunch.

9 J: onul ceki cip-ey o-sy-ese siksana ha-si-ci
today well house-to come-HON-and meal do-HON-SE
Why don’t you come over to my house today and have a meal?

10 C: pap com hay noh-useyyo
meal a little bit do leave-SE
Please cook rice

à 11 J: pap-ul hay noh-ulako?

meal-OBJ do leave-EMP
Are you saying you want me to cook rice?

12 C: pap ha-l cwul al-ci?

meal do-NOML know-SE
You know how to cook rice, right?

13 J: cikum palo toy-nun pap-i iss-ki-n iss-nuntey

now right away become-RC rice-SUB exist-NOML-TOP exist-CONJ
I do have some instant rice

14 C: ai kulen pap silh-ko papsoth pap hay-cwu-seyyo

no that kind of rice dislike-CONJ rice cooker rice do-give-SE
No, I don’t like instant rice. Please cook rice with the rice cooker.

15 J: kulay alasseyo kulem pap hay nohu-l theynikka

Okay all right then rice do leave-CONJ
This specific part of the conversation takes place when Ju calls Choi to invite her to lunch at
his house (line 9). Choi accepts the invitation but she tells him to cook rice for lunch (line 10).
Since Ju rarely cooks at home, Choi’s request to prepare rice is contrary to Ju’s expectation. Ju’s

81

unwillingness to cook rice is expressed in his pre-disagreement using a partial repeat of Choi’s
previous utterance and the emphatic sentence ending -ko ‘Are you saying~’ (line 11). The
question is not uttered to confirm what Choi has said. Rather, the question conveys the speaker’s
disagreement without overtly saying the speaker does not agree with the other speaker’s previous
utterance by providing the other speaker with an opportunity to correct or modify what has been
said in the previous utterance. For example, Choi might volunteer to cook rice for him or even
ask Ju to eat out. Here, (l)ul-marking is used indicate that the information involving the object NP
pap ‘rice’ needs to be corrected or modified in some way: “Are you sure you want me to cook
rice?”
However, Choi does not back down from her request and says that Ju must know how to cook
rice (line 12). In the next turn, since his pre-disagreement fails to initiate the other speaker’s self-
correction or backdown, Ju overtly shows his disagreement by saying he does have some instant
rice (line 13). Again, Choi does not back down and even tells him how she likes her rice to be
cooked (line 14). Ju finally gives in and promises her that he will cook rice (line 15). See another
Example, Excerpt 12, a phone conversation between two female friends.
Excerpt 12: Iris (I) and Whitney (W)
Iris and Whitney are talking about how Whitney met her fiancé.

1 I: a kulemyen ku salam incey wa-kaciko icey sakwi-ess-na poci?
oh then that person now come-CONJ now date-PST-seem to
Oh then the guy just came to the USA and you have been dating him

2 icey mikwuk-ey o-n taumey?
now USA-to come-after
After he came to the USA

3 W: o-ki cen-pwuthe cipsanim han pwun-i sokayhay cwu-lye[ko

come-before-from deacon one HON-SUB introduce give-in order to
The deacon at church wanted me to meet this guy even before I came (to the USA)

4 I: [e
Okay

5 W: kule-ko iss-taka

do so-PROG-CONJ
It was like that and

82

6 I: a kulay icey o-ki cen-pwuthe sokay hay cwe-ss-taka?

oh right now come-before-from introduce do give-PST-CONJ
Oh, then did you meet him before you came (to the USA)?

7 W: wa kacko incey sen pwa-ss-e
come-CONJ now blind date see-PST-SE
I just saw him on a blind date after I came here

à 8 I: wa-se icey sen-ul pwa-ss-[tako?

come-and then now blind date-OBJ see-PST-EMP
Are you saying you came here and met him on a blind date?
9 W: [ung
Yeah
10 I: kyohoy kathun kyohoy tanye-ss-canha
church same church attend-PST-SE
Did you guys attend the same church?

11 W: kuntey kyohoy wa-ss-ess-nuntey nay-ka ku-ttay-nun mos pwa-ss-ess-[ketun
but church come-PST-CONJ I-SUB that-time-TOP NEG see-PST-PST-SE
But when he came to church I didn’t have a chance to see him at the time

12 I: [ung
Okay

13 W: kuletaka nacwungey eti khephisyop-eyse ka-ss-taka,
then later somewhere coffee shop-LOC go-PST-CONJ
Then, I went to a coffee shop somewhere and

14 wuyenhi tto man[na-ss-ketun
by change again meet-PST-SE
I ran into him by chance

15 I: [ung
Okay

16 W: kuntey ku ttay-n ku salam-in cwul molla-ss-ci
but that time-TOP that person-NOML not to know-PST-SE
But at that time, I didn’t know it was him.
Iris is guessing that Whitney has been dating her fiancé since he came to the USA (lines 1 and
2). Whitney replies that the deacon of her church in Korea wanted her to meet him even before
she came to the USA (line 3). Hearing this, Iris tells Whitney that she must have been introduced
to her future fiancé before she came the USA (line 6), but Whitney answers that she just met him
on a blind date after she came to the USA (line 7). Whitney’s reply confuses Iris. She thinks that

83

Whitney must have met him before she came to the USA because she knows that Whitney and
her fiancé used to attend the same church and they were introduced to each other by the deacon of
that church. Thus, Iris’s uncertainty is expressed in her pre-disagreement using a partial repeat of
Whitney’s previous utterance and the emphatic sentence ending -ko ‘Are you saying~’ (line 8).
Her question covertly expresses her disagreement about the information in the other speaker’s
previous utterance that Whitney met her fiancé on a blind after she came to the USA by letting
Whitney self-correct or back down on what has been said. In this repair sentence, -(l)ul-marking
is used to direct the listener’s attention to the salient object NP sen ‘blind date’ so that the
information involving the marked NP can be clarified or corrected by the other speaker.
Iris’s pre-disagreement, however, fails to initiate Whitney’s self-correction or backdown on
her previous utterance. Whitney confirms what she said in the next turn (line 9). As a response,
Iris overtly shows her uncertainty about Whitney’s previous utterance that she met her fiancé on a
blind date in the USA by asking Whitney a confirmation question (line 10): “Don’t you guys
attend the same church?” Whitney provides further explains about the reason why she was
introduced to her fiancé after she came to the USA. Even though Whitney and her fiancé were
attending the same church, she didn’t have a chance to see him at church (line 11). In addition,
she did run into him once at a coffee shop somewhere in Korea (lines 13-14) but she didn’t know
it was him (line 16).

4.2.4 Showing Disagreement Using the (l)ul-Marked Question Word mewl ‘What’
Plus a Partial Repeat of the Prior Utterance
Among the practices for initiating repair on the other speakers’ prior utterance, -(l)ul is
consistently observed in the following pattern: The (l)ul-marked question word mwel plus a
partial repeat of the other speaker’s previous utterance can be used to challenge the
appropriateness or accuracy of the information in the previous turn and to ask for the other
speaker’s self-correction or backdown. This pattern targets the trouble source in the prior turn by

84

being located in the very next turn. See Excerpt 13, a telephone conversation between two male
friends Tony and Paul. In their prior talk, Paul said he didn’t want to talk on the phone any more.
Tony insisted that they keep talking, but, again, Paul said he had nothing to talk about.
Excerpt 13: Tony (T) and Paul (P)
1 T: kulem mwe ha-l keya
then what do-will
What are you going to do, then?

à 2 P: mwe-l mwe hay kyeysok malhay kunyang
what-OBJ what do continuously talk just
What are we going to do? Just keep talking

3 ne kitha paywe-ss-nya acik
you guitar learne-PST-Q yet
Have you learned how to play the guitar?

4 T: kulem ta paywe-ss-ci
of course all learn-PST-SE
Of course, I’ve learned everything

à 5 P: mwe-l tto ta paywe
what-OBJ again all learn
You’ve learned everything?

6 T: kecismal-i-ko kipon-un ta ha-ci
lie-to be-CONJ basic-TOP all do-SE
I’m just kidding but I did master all the basic skills
Tony asks Paul what he is going to do now (line 1). In the very next turn, Paul challenges
Tony’s previous utterance using the (l)ul-marked question word mwel ‘what’ plus a partial repeat
of Tony’s prior turn. Paul’s utterance conveys an assertion challenging the appropriateness of the
question asked by Tony in the previous turn, “Why do you even bother asking me what I’m going
to do now?” and Paul even provides a desirable answer to Tony’s previous question, “Let’s keep
talking.” And then, Paul changes the topic and asks Tony if he has learned how to play the guitar
(line 3). Tony answers Paul’s question in the next turn saying that he has learned everything
about the guitar (line 4).
In the following turn, Paul challenges Tony’s previous utterance using the same pattern, the
(l)ul-marked mwel plus a partial repeat of Tony’s prior turn, “I’ve learned everything,” which

85

conveys a negative assertion challenging Tony’s claim: “It doesn’t make sense that you have
learned everything about the guitar.” As as a response to Paul’s challenge, Tony is backing down
on his prior claim saying that he just mastered all basic skills (line 6). In this repair situation, the
(l)ul-marked question word is used as an indication of disagreement about the other speaker’s
previous utterance.
To summarize, (l)ul-marking is consistently used in contexts where the speaker has difficulty
understanding, agreeing with, or accepting the other speaker’s prior utterance that is contrary to
the speaker’s belief or expectation. In this context of disagreement, repair can be initiated to ask
for the other speaker’s clarification, correction, or backdown of the prior utterance. Because (l)ul-
marking directs the listener’s attention to the information involving salient referents that need to
be clarified or corrected, as a result the interlocutors can re-establish their shared understanding
for successful communication. Thus, (l)ul-marking has a pragmatic function of initiating repair on
the information indicated by the referent which is not clearly understood or is disagreed with by
the speaker.

86

4.3 Distinguishing Contrastive Referents
Previous studies have claimed that the particle -(l)ul is frequently used to distinguish
contrastively-focused referents which convey the idea of exclusivity, “X and only X.” However,
there has been little discussion about certain discourse contexts in which the speaker needs to
distinguish contrastive referents of object NPs and the role of (l)ul-marking in the achievement of
the speaker’s communicative purpose.
I have found that (l)ul-marking has a pragmatic function of distinguishing contrastively
focused referents in discourse contexts of disagreement. When the referent of an object NP is
dispreferred, challenged, or rejected by the other referent in conversation, (l)ul can be used to
mark those two referents. (l)ul-marking directs the hearer’s attention to this salient information,
which not only helps the hearer easily distinguish contrastive referents but also emphasizes the
speaker’s assertion challenging the prior utterance or action involving the marked referents. This
may contribute to the listener’s successful interpretation of the speaker’s intended meaning. See
Excerpt 14, a conversation between three friends Michael, Nicky, and Robert, which is taken
from the 21
st
episode of the Korean food TV show, Tasty Guys.
Excerpt 14: Michael (M), Nicky (N), and Robert (R)
Michael, Nicky, and Robert are eating spicy chicken stew. While Michael and Nicky rarely eat
kimchi, Robert says he must eat Kimchi with the chicken stew because he thinks they have
different flavors of spiciness: intense spiciness of the chicken stew vs. refreshing spiciness of
the kimchi.

1 R: wuli waiphu-pwuthe pelus kochy-eya toy-pnita
our wife-from habit fix-have to-SE
My wife has to fix her habit

2 N: e? waiphu[hanthey?
huh wife-to
Huh? To your wife?

3 M: [e wihemha-pnita i-ke
oh dangerous-SE this-thing
Oh, this’s dangerous.

87

à 4 R: ani lamyen mek-ul ttay maliya kimchi-lul cal an kkenay-cweyo
well ramen eat-when you know kimchi-OBJ well NEG take out-give
Well, you know when I eat ramen, she doesn’t give me kimchi.

à 5 ohilye mak singkewun panchan-ul kkenaycweyo kukey mwe ha-nun cis-iya
instead just bland side dish-OBJ take out-give that what do-RL act-to be
She just gives me a bland side dish instead. What in the world is she doing?

à 6 kimchiccikay kkulhye-ss-nuntey way panchan-ul kimchi-lul an cwu-pnikka
kimchi stew boil-PST-CONJ why side dish-OBJ kimchi-OBJ NEG give-SE
You cooked kimchi stew but why don't you give me kimchi for a side dish?

à 7 way myelchi-lul cweyo
why anchovy-OBJ give
Why do you give me anchovies?

8 N: [a::: ((laugh))
Ahhh

9 M: [myelchi! ((laugh))
Anchovies!

à 10 R: kimchi-lul tal-la ike-ya!
kimchi-OBJ give-QT this-to be
What I’m saying is give me kimchi!
Robert claims that his wife has to fix her annoying habit (line 1). Nicky is surprised by
Michael’s bold comment about his wife (line 2) and Michael expresses his concern about
Robert’s criticism publicly delivered on TV (line 3). Robert says his wife doesn’t let him eat
kimchi when he eats ramen (line 4). Instead, she gives him a not-too-salty side dish and Robert
doesn’t like it (line 5). Here, the two object NPs ‘kimchi’ in line 4 and singkewun panchan ‘a
bland side dish’ in line 5 are overtly (l)ul-marked not only because the two NPs are contrastively
focused, but also because the non-present party’s actions involving the two NPs are disapproved
by the speaker. Robert’s utterance in line 4, “Well, you know when I eat ramen, she doesn’t give
me kimchi,” conveys a strong disagreement about his wife’s prior action involving the object NP
kimchi. She does not let him eat kimchi when he eats ramen. His following utterance in line 5,
“She just gives me a bland side dish instead,” also expresses a strong disagreement about his
wife’s prior action involving the other object NP singkewun panchan ‘a bland side dish.’ She
gives him a bland side dish instead of kimchi.

88

Robert’s question immediately followed by his utterance in line 5, kukey mwe hanun cisiya
“What is the world is she doing?”, also aligns with the speaker’s disagreement conveyed through
his previous utterances: “She shouldn’t do this to me.” In this context of disagreement, -(l)ul can
be used by the speaker to help the listener’s selective focus on those salient referents so that the
speaker’s disagreement about the information indicated by the marked NPs can be strongly
expressed and successfully delivered to the listener.

Similarly, in Robert’s next utterances in lines 6 and 7, the two object NPs ‘kimchi’ and
myelchi ‘anchovies’ are also overtly (l)ul-marked not only because they are contrastively focused,
but also because the two utterances convey the speaker’s assertion challenging the non-present
party’s prior action involving the two object NPs: “You should give me kimchi” in line 6 and
“You shouldn’t give me anchovies” in line 7. Thus, (l)ul-marking helps the listener pay attention
to these salient NPs so that the speaker’s assertion challenging the other speaker’s previous action
can be more strongly expressed.
Robert’s follow-up utterance in line 10 also aligns with the assertion conveyed through his
previous utterances. Here, the previously-mentioned-and-hence-easily-identifiable object NP
‘kimchi’ can be easily left unmarked, but it is still chosen to be (l)ul-marked. Since (l)ul-marking
directs the listener’s attention to this salient NP, the speaker’s assertion challenging the other-
speaker’s previous action involving the other object NP myelchi ‘anchovies’ can be emphasized
and strongly expressed: “What I’m saying is give me kimchi, not anchovies!”
See Excerpt 15, a telephone conversation between two female friends Hanna and Victoria
talking about their other friend Lauren who is suffering financially.
Excerpt 15: Hanna (H) and Victoria (V)
Hanna is talking about her friend Lauren who is having financial difficulties.

1 H: kuntey na kath-umyenun na-lang ku enni-lang cincca pisusha-canha
but I same-if I-with that older sister-with really similar-SE
But if I were her, isn’t her situation very similar to mine?

2 cikum kac-ko iss-nun ton-i han tal-ey ung kulemyen
now possess-PROG-RL money-SUB one month-per well if so

89

If you have this amount money for one month

à 3 cipsey-lul celyakha-tenci ay hakkyo-[lul
rent-OBJ save-or kid school-OBJ
Either you should save money on rent or

4 V: [cokumssik cokumssik moa-kaciko
a little bit a little bit save-CONJ
you should save little by little and then

à 5 H: ay hakkyo-lul an ponay-tenci mwe-l hay-ya toy-can[ha
kid school-OBJ NEG send-or what-OBJ do-have to-SE
you should not send your kid to school, you’ve got to do something right?

6 V: [ung
Right

7 H: na-nun cincca ay hakkyo-to an ponay-ko mak kule-nuntey
I-TOP really kid school-even NEG send-and just like that-CONJ
I don’t even send my kid to school

8 ay hakkyo yuchiwen ceyil pissa-n[tey
kid school preschool the most expensive-CONJ
Preschool tuition is very expensive

9 V: [ung
Right

10 H: ilcwuil-ey twu-pen ka-nun ke opayk pwul-cca[li
one week-per two-time go-RC thing five hundred dollar-worth
It’s $500 for two times a week

11 V: [ung
Okay

12 H: kulen ke han tal-ey opayk pwul-ul ay hakkyo-ey nay-ci
that kind of thing one month-per five hundred dollar-OBJ kid school-to pay-SE
She’s paying $500 for her kid’s preschool tuition

13 cip-ul tto nemwu pissan tey-lul ka-n ke-ya leynthupi-lul
house-OBJ besides very expensive place-OBJ go-RC thing-to be rent-OBJ
She even moved into a very expensive house. I mean the rent

14 V: ahyu::
Oh my goodness

15 H: ku leynthupi-ka chen sapayk pwul-eytaka
that rent-SUB thousand four hundred dollar-on top of that
The rent is $1400 and on top of that

90

16 mwe cenkisey kulen ke ta ha-myen chen chilpayk
well electric bill that kind of thing all do-if thousand seven hundred
Well, it’s $1700 including all the utility fees such as electric bills

17 V: leynthupi-ka chen sapayk pwul-ina tway?
rent-SUB thousand four hundred dollar-as much as become
Her rent is $1400?

18 H: ung kekitaka cenkisey kathun ke-lul ttalo ta nay-myen-un
yes on top of that electric bill like thing-OBJ separately all pay-if-TOP
Yeah. On top of that, if she pays all the utilities fees such as electric bills

19 kasu-hako ta nay-mye-nun chen chilpayk pwul nayn-taynta han tal-ey
gas-and all pay-if-TOP thousand seven hundred dollar pay-QT one month-per
and gas everything, then she pays $1700 per month
Hanna is talking about the financial situation of her friend Lauren, who has to live on her
husband’s low income. Hanna says if she was in her position with the same amount of money
allowed for one month (lines 1 and 2), she would certainly do something to make ends meet by
either saving money on rent or not sending her kid to school (lines 3 and 5). Here, the first action
involving the NP cipsey ‘rent’ in line 3 and the second action including the NP hakkyo ‘school’ in
line 5 appear in parallel pairs with the conjunctive -tenci, which indicates a choice between the
two alternatives: “You should save money on rent. If that is not possible, you should not send
your kid to school.”
Here, (l)ul-marking not only directs the listener’s attention to the two contrastively focused
referents cipsey ‘rent’ and hakkyo ‘school,’ but also emphasizes the speaker’s negative assertion
challenging the non-present party’s prior action involving the two object NPs: “Lauren should’ve
saved some money on rent or she shouldn’t have sent her kid to preschool, but she didn’t do
anything.” Hanna’s follow-up utterances also align with the negative assertion conveyed through
her previous utterances. Lauren is paying $500 for her kid’s preschool tuition (line 12). She is
living in a very expensive house (line 13) and the rent is $1400 (line 15). On top of the expensive
rent, she is paying nearly $1700 total for her rent and all the utilities fees, such as gas and electric
bills (lines 15, 16, 18, and 19).

91

To summarize, -(l)ul has a pragmatic function of distinguishing contrastive referents in
discourse contexts of disagreement. When the referent of an object NP is dispreferred, challenged,
or rejected by the other referent in conversation, (l)ul can be used to mark those two contrastive
referents. Here, (l)ul-marking not only facilitates the listener’s selective focus on salient referents,
but also emphasizes the speaker’s assertion challenging the prior utterance or action involving the
marked referents, which may contribute to the listener’s successful interpretation of the speaker’s
intended meaning.

92

4.4 Indicating the Discourse Topic
The particle -(l)ul is also often used when the information introduced in the object continues to be
talked about as a topic of conversation. This function has been observed in previous studies (Fujii
and Ono 2000; Yang 2007): Object-marking helps the listener focus on the salient element
continuously being talked about in the discourse context. However, what is missing in such
observations in previous studies is that object-marking is used for the object NP not only because
it is continuously being talked about in subsequent contexts but also because there is little shared
knowledge between the speaker and the hearer so that the information related to the object NP is
unknown or unpredictable to the hearer and hence highly informative.
If the speaker wants to talk about something that the hearer might not know well, it would be
hard for the hearer to actively participate in the conversation due to the lack of shared knowledge.
Thus, (l)ul-marking is a device used by the speaker to signal that further descriptions/explanations
about the object NP are coming up in the subsequent context so that the listener can identify and
process important information related to the (l)ul-marked element. See another telephone
conversation between two female friends, Lucy and Gina, talking about U.S. department store
return policies in Excerpt 16.
Excerpt 16: Lucy and Gina
Lucy is talking about how to take advantage of return policies at U.S. department stores.

1 L: kulenikka wuli ettehkey ha-nya-myen-un phathi kathun ke iss-canha.
so we how do-SE-if-TOP party like thing exist-SE
So what we do is, when you have a party or something

à 2 kulem ka-se tuleysu-lul thak sa-nun ke-ya.
then go-and then dress-OBJ just buy-RC thing-to be
Then you go (to the department store) and just get a dress

3 G: ung
Okay

4 L: kulem ip-ko taumey ka-se
then wear-CONJ next time go-and then
Then, wear (the dress) and go to (the department store) next time and then

93

5 tuleysu-lang kwutwu-lang payk-ilang ta sa
dress-and shoes-and bag-and all buy
buy the dress, a pair of shoes and a bag too

6 ip-e po-ko nase ku nal phathi-ey ka-se ta kwutwu-to sin-ko
wear-try-CONJ that day party-to go-CONJ all shoes-also wear-and
Try the dress on and then you go to a party on that day wearing the shoes you bought

7 mak ta ha-canha suthakhing kathun ke-nun an toy-ciman
just all do-SE stockings like thing-TOP NEG possible-but
You can try everything on, right? But not stockings

8 G: ung
Okay

9 L: ttan ke ta ha-n taumey ku taum nal
different thing all do-after that that next day
You use every item you bought and the next day

10 tolo ta kassta cwe ton-ulo ta pat-a-wa
back all bring money-with all receive-and then-come
You bring everything back (to the department store) and take your refund money
Lucy tells Gina that some people take advantage of lenient refund policies at department
stores in the United States. If you buy something at the department store, you can easily return it
and get your money refunded. In case you want to go to a party with a new dress, but you cannot
afford it (line 1), just get a new dress (line 2) and other items that can go along with the dress
such as shoes and bags (line 5) at the department store, wear everything, go to the party (line 6),
and then return everything the next day to get your money refunded (lines 9 and 10).
The object NP tuleysu ‘dress’ is first introduced by Lucy in line 2, and she continues to talk
about it in her strings of utterances. Here, (l)ul-marking is used not only to direct the listener’s
attention to this discourse topic, but also to indicate that further descriptions/explanations about
the object NP will be provided in the subsequent context to build shared knowledge between the
speaker and the listener. Because Gina is currently living in Korea, she might not be familiar with
U.S. department store return policies. Lucy assumes that there is a lack of shared knowledge
available to Gina about the information related to the object NP tuleysu ‘dress.’ Thus, -(l)ul is
used for the object NP to indicate that Lucy is going to talk about something the hearer might not

94

know in the subsequent context.
In this case, interestingly, -(l)ul consistently co-occurs with the conjunctive -(n)untey which is
generally known to be used for providing background information or establishing the background
in discourse contexts. As Kim (1996) pointed out, -nuntey indicates a warranted continuation
relation between the preceding portion of the utterance as the background/topic and the up-
coming portion as the elaboration/comment. Thus, the speaker provides further
descriptions/explanations related to the (l)ul-marked object in the subsequent context to build
shared knowledge so that the listener can understand what the speaker is talking about without
difficulty or misunderstanding. See Excerpt 17, another telephone conversation between Lucy
and Gina.
Excerpt 17: Lucy (L) and Gina (G)
Lucy is talking to someone at her church who is wearing a T-shirt made to commemorate the
Saint Louise Great Flood of 1993.

1 L: e ku hay emchengnan hongswu-ka na-se
yeah that year huge flood-SUB occur-CONJ
Yeah, a huge flood occurred that year and

2 ku tosi-ka keuy ta mwul-ey camkyess-ess-tay
that city-SUB almost all water-in sank-used to-IND.Q
I heard almost every part of the city sank under the water

3 G: e
Okay

à 4 L: kulay-kacko ku thisyechu-lul pha-nuntey
like that-CONJ that T-shirt-OBJ sell-CONJ
So they sell that T-shirt and

5 kwusip sam nyento-i-n-ka kwusip i nyento-i-n-ka
ninety three year-to be-SE ninety two year-to be-SE
It was sometime around 1993 or 1992

6 hayethun ku-ttay that isscanha hongswu-ey tayhayse
anyway that-time well you know flood-about
Anyway at that time, well you know, about the flood

7 G: ung
Okay

95

8 L: kulen kinyem thi nwuka keki sal-ass-tako
that kind of commemoration T-shirt somebody there live-PST-EMP
that T-shirt commenmorating the Great Flood in Saint Louise

à 9 ku thi-lul ip-ko wa-ss-tela-ko.
that T-shirt-OBJ wear-CONJ come-PST-EVI-EMP
I saw someone at church, and he was from Saint Louise and wearing that T-shirt

10 han sam nyen cengto iss-ess-tay
approximately three year around to be-PST-QT
He said he lived there for about three years.

11 G: e e
Yeah, I see
!
12 L: kuntey hongswu-ka na-ss-ess-tanun ke-ya.
and flood-SUB occur-PST-PST-SE-RL thing-to be
And he told me about the flood
à 13 kulen thisyechu-lul ip-ko iss-nun keya
that kind of T-shirt-OBJ wear-PROG-RL thing-to be
I saw him wearing that T-shirt

14 kulayse nemwu wusky-e kacko
so very funny-CONJ
So it was very hilarious

The object NP thisyechu ‘T-shirt’ is first introduced into conversation by Lucy in line 4 and it
is continuously talked about in a string of her utterances. There was a huge flood in Saint Louise
in 1993 (line 1). As a result of the flood, almost every part of city sank under water (line 2) and
people started to make T-shirts and sell them to remember the Great Flood (lines 4, 5, 6 and 8).
Lucy saw someone wearing that T-shirt at church (line 9) who lived in Saint Louise for about
three years (line 10). The person also told Lucy about the flood (line 12) so she thought it was
very funny. Since Gina does not know Lucy’s story about the T-shirt, all the information related
to this discourse topic is salient information that Gina needs to identify and process. Thus, the
(l)ul-marked object NP thisyechu ‘T-shirt’ indicates that further explanations/descriptions related
to the marked NP will be coming up in the subsequent context to establish shared knowledge
between the interlocutors so that Gina can clearly understand what Lucy is talking about without
difficulty.

96

4.5 Summary
Previous studies (Fujii and Ono 2000; Yang 2007; Kim 2008) have shed light on the pragmatic
use of -(l)ul as a device available to the speaker to facilitate the listener’s selective focus on either
hard-to-process or salient information for successful communication. This means that the speaker
is sensitive to the listener’s mind and provides more linguistic signals indicating important
information that the listener needs to identify or process. Then, what makes such information
more important in discourse contexts?
What has been missed in the observations of previous studies is a more comprehensive
understanding of the mechanism that controls the speaker’s use of -(l)ul. That is, (l)ul-marking is
consistently used in discourse contexts where the speaker disagrees with what has been said in the
prior talk. When there is an ongoing disagreement between the interlocutors, the speaker needs to
direct the listener’s attention to the information which goes contrary to the prior utterance or
action of co-present or non-present parties. Disagreement can vary in range from weak
disagreement, such as expressing surprise or wondering about the prior utterance or action which
goes contrary to the speaker’s belief or expectation, to serious disagreement, such as showing a
negative or sarcastic attitude (e.g., dispreference, disbelief, objection, mockery, etc.) or asking for
the other speaker’s clarification, self-correction, or backdown on prior talk.
In such discourse contexts of disagreement, object-marking is a device used by the speaker to
direct the listener’s attention to the important information that disagrees with or challenges the
prior utterance or action, or which needs to be clarified or corrected by the listener. Thus, object-
marking helps the listener interpret the speaker’s intention and act or reply accordingly, and as a
result the speaker’s communicative purpose can be successfully achieved.
Therefore, what the particle -(l)ul does in conversation is more than simply indicating the
grammatical role of the NP in a sentence. First, -(l)ul has a pragmatic function of emphasizing an
assertion challenging the prior utterance or action of the co-present or non-present parties in

97

rhetorical questions, self-addressed questions, and restatements of a speaker’s prior utterance,
which helps the speaker successfully achieve certain communicative purposes such as
complaining, arguing, mocking, blaming, and persuading, etc.
Second, (l)ul-marking is used to initiate repair on the other speaker’s prior utterance which
goes contrary to the speaker’s belief or expectation and hence might be wrong, inaccurate, or hard
to believe or accept. In repair situations, the speaker may express surprise or disbelief about what
has been said and ask for clarification using a partial repeat of the other speaker’s prior turn
together with adverbs expressing the speaker’s surprise or doubt, such as cengmal or cincca
‘really,’ or a candidate understanding of the prior utterance plus the phrase -nun mali-/iyaki- ‘you
mean.’ A partial repeat of the trouble source turn with the emphatic sentence ending -ko ‘Are you
saying~’ and the (l)ul-marked question word mwel plus a partial repeat of the other speaker’s
previous utterance can also be used to show the speaker’s disagreement about what has been said
in the previous turn and to ask for the other speaker’s self-correction or backdown of the prior
turn. In this kind of repair situation, (l)ul-marking helps the listener focus on the information
involving salient referents that need to be modified or corrected, and hence interlocutors can re-
build shared understandings for successful communication.
Next, -(l)ul has a pragmatic function of distinguishing two or more contrastive referents
especially in challenging contexts. When the referent of an object NP is dispreferred, challenged,
or rejected by the other referent in conversation, (l)ul can be used to mark the two referents.
Object-marking directs the hearer’s attention to this salient information, which not only helps the
hearer easily distinguish contrastive referents but also emphasizes the speaker’s assertion
challenging the prior utterance or action involving the marked referents. This may contribute to
the listener’s successful interpretation of the speaker’s intended meaning.
Lastly, (l)ul-marking consistently appears when the information introduced in the object NP
continues as topic of conversation. If the speaker wants to talk about something that the hearer
might not know well, it would be hard for the hearer to actively participate in conversation due to

98

the lack of shared knowledge. Then, the speaker needs to build shared knowledge so that the
listener can understand what the speaker is talking about without difficulty or misunderstanding.
In this case, (l)ul-marking is used to indicate that further descriptions or explanations related
about the discourse topic will be coming up in the subsequent context. The conjunctive -nuntey,
which indicates a warranted continuation relation between the preceding portion of the utterance
as the background/topic and the up-coming portion as the elaboration/comment, often co-occurs
with the (l)ul-marked discourse topic.
In the next Chapter, the two closely related grammatical phenomena of “Substitution” and
“Double Object Construction” will be discussed. Considering the fact that both particles and their
marked elements are frequently omitted in spoken Korean due to shared knowledge between the
interlocutors, the speaker’s choice of replacing other particles with the object particle -(l)ul or
using double (l)ul-marking is expected to be observed in certain discourse contexts where the
speaker needs to direct the listener’s selective focus on salient information so that the speaker’s
intended meaning can be successfully expressed and delivered to the listener.

99

5. Discussion
This Chapter discusses the two closely-related grammatical phenomena of “Substitution,” which
refers to the speaker’s choice of using -(l)ul in the place of other particles, and “Double Object-
Marking,” which refers to the use of object-marking for two arguments in a sentence. In Korean
literature, there has been an effort to discover the meaning of the particle -(l)ul which can explain
the phenomenon of using (l)ul in the place of other particles, or using double (l)ul-marking (Hong
1986, 1987; Jung 1988; Lee 1989; Ko 2000).
Hong (1986, 1987) points out the following characteristics of -(l)ul and claims that the particle
-(l)ul functions as a special particle conveying an emphatic reading.
(84) -(l)ul stacked on the other particle
John-un sewul-ey +(nun/to/man/lul) ka-ss-ess-ta.
John-TOP Seoul-to-(TOP/also/only/OBJ) go-PST-PST-SE
‘John has been to Seoul (but not Busan).’
‘John has also been to Seoul.’
‘John has only been to Seoul.’
‘John has been to Seoul.’

(85) -(l)ul added to non-nominal elements
i cha-nun ppalli-lul ka-ci-lul anha.
this car-TOP fast-OBJ go-NOM-OBJ NEG
‘This car doesn’t run fast.’

(86) -(l)ul added to adverbials expressing distance, duration, weight, price, etc.
a. John-un hakkyo-pwuthe yekk-kaci i-khillo-lul ttwi-e-ka-ss-ta.
John-TOP school-from station-to two-kilometer-OBJ run-go-PST-SE
‘John ran two kilometers from school to the station.’

b. John-un kongwen-eyse twu sikan-ul ponay-ss-ta.
John-TOP park-LOC two hour-OBJ spend-PST-SE
‘John spent two hours at the park.’

c. John-un sal-ul sip-khillo-lul ppay-ss-ta.
John-TOP weight-OBJ ten-kilogram-OBJ lose-PST-SE
‘John has lost 10 kilos.’

d. John-un chayk-kaps-ulo sip talle-lul nay-ss-ta.
John-TOP book-price-INST ten dollar-OBJ pay-PST-SE
‘John paid $10 for the book.’
(Modified from Hong, 1987, p.53-54)

100

-(l)ul can be stacked on the existing particle in the same way as other special particles carrying a
certain meaning, such as the topic particle -(n)un conveying a contrastive reading, the emphatic
particle to ‘also,’ and the delimiter particle man ‘only,’ as in (84). -(l)ul can also be added to non-
nominal elements such as the manner adverb ppalli ‘fast’ or the long form negation pattern ci
anh- as in (85) as well as adverbials expressing distance, duration, weight, price, etc., as in (86).
Hong (1987) also claims that the meaning of -(l)ul is related to the notion of ‘wholeness.’
(87) a. John-un ilhepeli-n koyangi-lul chac-ule on/yele tongney-lul
John-TOP lose-RL cat-OBJ find-in order to whole/several town-OBJ

ta tolatanye-ss-ta.
all wonder around-PST-SE
‘John wondered around the whole town/several towns to find his cat.’

b. *John-un ilhepeli-n koyangi-lul chac-ule on/yele tongney-eyse
John-TOP lose-RL cat-OBJ find-in order to whole/several town-LOC

ta tolatanye-ss-ta.
all wonder around-PST-SE
‘John wondered around the whole town/several towns to find his cat.’
As mentioned in Ko (2002), Hong (1987) claims that since (l)ul conveys the meaning ‘wholeness,’
the noun modifiers expressing ‘wholeness’ or ‘plurality’ such as on ‘whole’ or yele ‘several’ can
be combined with the (l)ul-marked noun tongney ‘town,’ as in (87a). On the other hand, since the
action locative particle -eyse ‘at, in, on’ has the meaning of ‘partness,’ which refers to a certain
place where the action of the verb is taking place, the same modifiers cannot be combined with
the LOC-marked noun tongney ‘town,’ as in (87b).
Jung (1988) also attempts to explain the difference between the directional particle -ey
indicating a goal or destination and -(l)ul using the notion of ‘wholeness’ and ‘partness.’
(88) a. John-i kyeytan-ey/-ul olu-n-ta.
John-SUB stairs-to/-OBJ climb-PRS-SE
‘John is climbing the stairs.’

b. John-i acik kyeytan-*ey/-ul olu-n-ta.
John-SUB still stairs-to/-OBJ climb-PRS-SE
‘John is still climbing the stairs.’

101

(89) a. John-un pyek-ey/-ul chilha-n-ta.
John-TOP wall-LOC/-OBJ paint-PRS-SE
‘John paints the wall.’

b. John-un pyek-*ey/-ul ta chilhay-ss-ta
John-TOP wall-LOC/-OBJ all paint-PST-SE
‘John has painted the wall.’
(Modified from Jung 1988)
Jung (1988) claims that while the directional particle -ey used with movement verbs such as oluta
‘to climb’ requires one part of the stairs as its destination, -(l)ul requires the whole of the stairs as
its destination. For example, if the speaker is in the middle of climbing up the stairs to get to a
certain place, but has not yet reached the end point, as in (96b), the whole of the stairs should be
the destination of the speaker’s action of climbing. Therefore, -ey cannot added to the destination
noun kyeytan ‘stairs’ in (88b).
In case of non-movement verbs such as chilhata ‘to paint’ in (89), both -ey indicating ‘partial
involvement’ and -(l)ul expressing ‘whole involvement’ can be used for the location noun pyek
‘wall,’ as in (89a). However, in (89b), only -(l)ul can be added to the location noun pyek ‘wall’
because the adverb ta ‘all’ indicates that the action of painting has been completed in a designated
place.
However, Ko (2000) argues that the notion of ‘wholeness’ cannot explain the meaning of
-(l)ul. See the following examples.
(90) a. pangkum wuli-tul aph-ey/-ul cinaka-n cha-ka
just we-PL front-LOC/-OBJ pass-RL car-SUB

konghang-ey ka-nun cha-ta.
airport-to go-RL car-to be-SE
‘The car which just passed by us is going to the airport.’

b. simsawiwen-i John ccok-ey/-ul cemswulul manhi cwe-ss-ta.
judge-SUB John side-LOC/-OBJ score-OBJ a lot give-PST-SE
‘The judge gave John higher scores.’

c. uysa-ka John-uy oynccok engtengi-ey/lul cwusa-lul noh-ass-ta.
doctor-SUB kid-GEN right side buttocks-LOC/-OBJ shot-OBJ put-PST-SE
‘The doctor gave John a shot in the right side of his buttock.’
(Modified from Ko 2000, p.200)

102

Even though the noun denoting a spatial dimension, such as aph ‘front’ in (90a), and the
dependent noun ccok ‘side’ do not express the meaning of ‘wholeness,’ they can be still (l)ul-
marked. Similarly, the (l)ul-marked noun engtengi ‘buttock’ in (90c) cannot be interpreted such
that the doctor gave John a shot everywhere in his right buttock. Therefore, Ko (2000) claims that
‘wholeness’ cannot be the inherent meaning of -(l)ul and suggests that -(l)ul indicates the
speaker’s choice of a referent among a set of possible candidates which can be the end point of
the action or effect of the predicate. See (91)-(93).
(91) A: John-un yocum nwukwu-lul manna?
John-TOP recently who-OBJ meet
‘Who is John meeting recently?’

B: amwu-na/-*lul manna.
any-ever/-OBJ meet
‘He’s meeting anyone he’s interested in.’

(92) A: John-un mwe-lul cal meke?
John-TOP what-OBJ well eat
‘What does John like to eat?’

B: amwuke-na/*-lul cal meke
anything-ever/-OBJ well eat
‘He eats anything.’

(93) a. ecey hoksi nwukwu-*lul/ -ø manna-n salam epse?
yesterday by chance who-OBJ meet-RL person not to exist
‘Didn’t you meet anyone yesterday by chance?’

b. mwe-*ul/ -ø mek-ul manha-n-kes epse?
what-OBJ eat-worth eating-RL-thing not to exist
‘Isn’t there anything that I can eat?’
The indefinite pronouns amwuna ‘anyone’ in (91) and amwukena ‘anything’ (92) are the
objects receiving the action of the verb manna- ‘to meet’ and mek- ‘to eat,’ respectively. However,
they cannot be (l)ul-marked because the meaning of -(l)ul, which indicates the selective choice of
a referent among other possible candidates, is contradictory to the meaning of the two indefinite
pronouns which do not refer to any person or thing in particular. Similarly, when the question
word nwukwu ‘who’ as in (93a) and mwe ‘what’ as in (93b) are used as indefinite pronouns
‘anyone’ and ‘anything’ respectively, (l)ul cannot be used.

103

Ko (2000) also points out that when there is only one referent available to be interpreted as the
recipient of the action of the verb in discourse contexts, -(l)ul cannot be used because the
meaning of (l)ul, which indicates the speaker’s selective choice of a particular referent among a
set of possible candidates, requires more than one possible referent in discourse contexts.
(94) a. nay-ka cikum ne-??lul /-ø po-le ka-lkey.
I-SUB now you-OBJ see-in order to go-will
‘I’m coming to see you right now.’

b. ike-??lul / -ø ettehkey meke?
this-OBJ how eat
‘How can I eat this?’

c. isang iss-umyen i mwulken-??ul /-ø pakkwu-le
defection exist-if this item-OBJ exchange-in order to

o-l keyey-yo.
come-will-POL
‘If something is wrong with this item, I will come back to exchange it.’
(Ko, 2002, p.206)
For example, in a telephone conversation, as in (94a), the second person pronoun ne ‘you’ can
only refer to the other speaker talking on the phone. Similarly, the demonstrative pronoun ike
‘this (thing)’ in (94b) and the object NP i mwulken ‘this item’ in (94c) can only refer to a specific
entity closer to the speaker in the immediate context.
Ko (2000) also claims that the speaker would choose to use (l)ul over other possible particles
based on the speaker’s judgment. When the speaker thinks that an entity can function as the end
point of the action or effect of the verb, (l)ul-marking is used.
(95) a. John-un hakkyo-ey ka-ss-ta
John-TOP school-to go-PST-SE
‘John went to school.’

b. John-un hakkyo-ey-lul ka-ss-ta
John-TOP school-to-OBJ go-PST-SE

c. John-un hakkyo-lul ka-ss-ta
John-TOP school-OBJ go-PST-SE

d. John-un hakkyo-ø ka-ss-ta
John-TOP school go-PST-SE

104

(96) a. John-i Mary-eykey chayk-ul cwu-ess-ta
John-SUB Mary-to book-OBJ give-PST-SE
‘John gave Mary a book.’

b. John-i Mary-lul chayk-ul cwu-ess-ta
John-SUB Mary-OBJ book-OBJ give-PST-SE
‘John gave Mary a book.’

c. John-i Mary-ø chayk-ul cwu-ess-ta
John-SUB Mary-OBJ book-OBJ give-PST-SE
‘John gave Mary a book.’
For example, in a sentence with the movement verb ka- ‘to go,’ -(l)ul can be stacked onto the
directional particle -ey, as in (95b). -(l)ul may also replace -ey as in (95c) or not even be used, as
in (95d). Ko (2000) claims that the use of object-marking in the above sentences is based on the
speaker’s judgment. Regardless of the presence or absence of the directional particle -ey, if the
speaker thinks that the place noun hakkyo ‘school’ can function as the end point of the action of
the movement verb ka- ‘to go,’ -(l)ul can be used. Similarly, -(l)ul can also be used to mark Mary,
the recipient of the transferred item chayk ‘book,’ as in (96b), because the speaker sees Mary as
the end-point of the action of the verb cwu- ‘to give.’
Ko’s (2000) insights have captured an important point that the use of object-marking is not an
obligatory grammatical phenomenon, but is rather a choice made by the speaker. However, Ko’s
proposal is problematic for a few reasons. First, the data used in the analysis mainly consist of
decontextualized sentences based on a native speaker’s intuition. Thus, there is no consideration
of pragmatic-discourse factors which can affect the speaker’s choice of using zero- or (l)ul-
marking. In partcular, Ko has failed to consider the importance of shared knowledge between the
interlocutors in the speaker’s choice of zero- or (l)ul-marking. The referent of an entity which is
easily identifiable or recoverable in discourse contexts is likely to be left unmarked. For example,
the three object NPs in (94), ne ‘you,’ ike ‘this,’ and i mwulken ‘this item’ are left unmarked not
because of the lack of a set of possible candidates, which is a required condition for the
realization of the ‘selective choice among others’ meaning of -(l)ul, but because their referents

105

are already shared by the interlocutors in the discourse context, and hence they are easily
identifiable as the object of the sentences even in the absence of -(l)ul.
Next, the meaning of -(l)ul proposed by Ko, the speaker’s selective choice of a referent among
other possible candidates, cannot fully explain the various pragmatic uses of -(l)ul in spoken
Korean. The speaker’s use of (l)ul-marking helps the listener’s selective focusing on salient
information so that the speaker’s intended meaning through the utterance can be successfully
expressed and delivered to the listener. For instance, (l)ul-marking can be used in the sentences in
(94) in certain discourse contexts where the speaker disagrees with the prior utterance or action of
non-present or co-present parties.
(97) a. nay-ka cikum ne-lul po-le ka-lkey.
I-SUB now you-OBJ see-in order to go-will
‘I’m coming to see you right now.’
[Implied: I insist that I’m coming to see you right now.]

b. ike-l ettehkey meke
this-OBJ how eat
‘How can I eat this?’
[Implied: I can’t eat this.]

c. isang iss-umyen i mwulken-ul pakkwu-le o-l keyey-yo.
defection exist-if this item-OBJ exchange-in order to come-will-POL
‘If something is wrong with this item, I will definitely come back to exchange it.’
(l)ul-marking can be used in (97a) when the other speaker on the phone doesn’t want to meet
the speaker now, and thus the speaker needs to restate his prior utterance to strongly express his
wish or determination. The demonstrative pronoun ike ‘this’ is easily identifiable in the
immediate context and hence it is often left unmarked, but object-marking can be used in a
rhetorical question conveying the speaker’s assertion challenging the prior utterance or action of
the other speaker, as in (97b). Here, the question is not heard as an ordinary wh-question asking
for a new piece of information. Rather, it is interpreted as an assertion conveying the speaker’s
strong unwillingness to eat. Similarly, object-marking can also be used for the easily-identifiable
object NP i mwulken ‘this item’ to emphasize the speaker’s assertion challenging the prior
utterance of the other speaker. For example, if the owner of the clothing store says that exchange

106

is not allowed for any items in any circumstances, the speaker can argue against it by saying that
exchange should be allowed for defective items.
Lastly, Ko’s proposal cannot explain the difference in meaning when -(l)ul is stacked onto the
directional particle -ey as in (95b) or when -(l)ul is used in the place of -ey as in (95c). In addition,
there is no explanation about the reason why the speaker chooses to use (l)ul-marking for more
than one entity in a sentence.
Therefore, the goal of this chapter is to analyze actual spoken data and discover and describe
the following discourse contexts: where the use of -(l)ul is preferred in place of other particles;
where -(l)ul is stacked onto existing particles; and where double (l)ul-marking is used. These
questions will be answered based on the following hypothesis adopted from the findings of the
previous chapters: (l)ul-marking is a device available to the speaker in conversation to provide an
extra linguistic signal for either hard-to-process or important information that the listener needs to
identify or process.
Thus, substitution, stacking, and double (l)ul-marking are expected to be used by the speaker
to direct the listener’s attention to either hard-to-process or salient information that the listener
has to identify or process. In particular, substitution, stacking, and double (l)ul-marking would be
consistently used in discourse contexts where the speaker disagrees with or challenges the prior
utterance or action of non-present or co-present parties. Disagreement can vary in range from
weak disagreement, such as expressing surprise or wondering about the prior utterance or action
which goes contrary to the speaker’s belief or expectation, to serious disagreement, such as
showing a negative or sarcastic attitude (e.g., dispreference, disbelief, objection, mockery, etc.) or
asking for a clarification or correction of the previous utterance or action. In these challenging
contexts, substitution, stacking, or double (l)ul-marking can be used by the speaker to direct the
listener’s attention to salient information that needs to be identified, processed, clarified, or
corrected, and hence the speaker’s intended meaning challenging the prior utterance or action can
be more strongly expressed and successfully delivered to the listener.

107

I predict that substitution, stacking, and double (l)ul-marking will have the following
pragmatic functions in informal conversation, which will be discussed in the next subsections.
(i) Emphasizing an assertion challenging the prior utterance or action
(e.g., wondering, complaining, arguing, mocking, blaming, regretting, persuading, etc.)

(ii) Initiating repair on the information involving the referent that needs to be clarified,
corrected, or modified

(iii) Distinguishing contrastive referents in discourse contexts of disagreement

(iv) Indicating a discourse topic and projecting further descriptions or explanations
about the discourse topic in the subsequent context

108

5.1 Substitution and Stacking
In the case of movement verbs such as ka- ‘to go,’ o- ‘to come,’ and tani- ‘to attend,’ and teyliko
k-/o-/tani- ‘to take someone/to bring someone/to take someone around,’ the destination noun is
frequently left-unmarked as in (98).
(98) a. nayil eti-ø ka?
tomorrow where go
‘Where are you going tomorrow?’

b. maynnal hakkyo-ø wa-ss-ta ka-ss-ta ha-myense
every day school come-PST-CONJ go-PST-CONJ do-CONJ
‘I’m just coming and going to school every day.’

c. Disney Land-ø nol-le ka-lyeko
Disney Land pay-in order to go-CONJ
‘I’ve decided to go to Disney Land.’

d. ecey nolaypang-ø ka-ss-e
yesterday karaoke go-PST-SE
‘I went to karaoke yesterday.’
e. micangwen- ø teyliko tani-myense meli calla cwu-n kes iss-cyo
beauty shop take around-CONJ hair cut-give-RL thing exist-SE
I have been taking him to the beauty shop for a haircut
(l)ul-marking of the destination noun and -(l)ul stacked on the direction particle -ey are used less
frequently and observed in certain discourse contexts. See the continued telephone conversation
between Eric and Katy in Excerpt 2.
Excerpt 2 Extended: Eric (E) and Katy (K)
Eric told Katy about Junseok, a high school student who studied abroad in Australia for two
years. Junseok wants to transfer to a new school in the United States and Eric is helping him
find a new school. Junseok’s parents want him to transfer to one of the top U.S. high schools
with few Korean students, but Eric thinks it is hard for Junseok to get admission because of
his weak reference letter from his previous school.
!
12 E: kyay chinkwu-tul-to taykay kulen sipwi kwen an-ey
that kid friend-PL-also usually that kind of 10
th
place inside-LOC

13 tu-nun hakkyo
enter-RL school
Most of his friends also go to one of the top 10 high schools

14 K: ung
Okay

109

15 E: cen mikwuk-eyse
whole USA-LOC
Across the U.S

16 K: e
okay

17 E: kulen hakkyo-tul-i-la kulen tey-nun
that kind of school-PL-to be-because that kind of place-TOP
That kind of top schools

18 mwe kwukmwul-to epse
well soup-even not to exist
Well, it would be impossible to get admission

19 K: kulehci aiko::
right Oh
Oh, you’re right

20 E: kulen ke mal-ko cikum kunyang sisikholkholhan tey
that kind of school not-and now just mediocre place

ilehkey alapo-myen-un
like this look for-if-TOP
If I just look for mediocre schools instead of top schools

21 K: ung
Okay

22 E: hankwuk salam-man samsip myeng ((laugh))
Korea person-only thirty CL
They would have at least thirty Korean students

23 K: ip-ey mac-nun ttek-i eps-ney cohun tey-nun elyep-ko
mouth-to fit-RL rice cake-SUB not to exist-SE good place-TOP difficult-CONJ
It is hard to find a school for Junseok. And all those good schools are hard to get into

24 E: cohun tey-nun hankwuk salam manh-aya sene myeng-to an tway
good place-TOP Korea person many-CONJ three-four CL-even NEG become
There are only three to four Korean students in those good schools

25 K: ung
Okay

26 E: mos tuleka-nikka
NEG enter-CONJ
He just can’t get in

à 27 K: kulem kulen ay-nun enu hakkyo-lul ka-na
then that kind of kid-TOP which school-OBJ go-Q
To which school can he transfer, then?

110

28 E: kulayse nayka han cikum ilkop kwuntey cengto poppa-kaciko
so I-SUB around now seven CL about select-CONJ
So now I picked seven schools and

29 alapo-lyeko ha-nun cwung-iya cikum
look for-CONJ-RL middle-to be now
I’m going to do some research about them
Most of Junseok’s friends go to one of the top 10 high schools across the U.S (lines 12, 13,
and 15), but Eric thinks it would be impossible for Junseok to get admission from those top
schools (lines 17 and 18). Eric says if he looks for mediocre schools instead, they would have at
least 30 Korean students (lines 20 and 22). After hearing what Eric has said, Katy says that it is
hard to find a school for Junseok because top schools are hard to get into and mediocre schools
have too many Korean students (line 23). Eric adds that all those good schools only have three to
four Korean students (line 24), but Junseok just can’t get in those top schools (line 26).
After hearing Eric’s concern about Junseok, Katy utters a wh-question, kulem kulen aynun enu
hakkyolul kana “Then, to which school can Junseok transfer?” in line 27. Her question is not
uttered to ask for a new piece of information. Rather, it is asked to express the speaker’s negative
assertion challenging the prior action of Junseok’s parents. They want him to transfer to one of
the top U.S. high schools, but it is almost impossible for him to get admission from those good
schools because his reference letter is full of bad comments about him. However, Junseok’s
parents would never let him transfer to an ordinary school with many Korean students. They
insist that he should transfer to one of the top schools with few Korean students.
Considering the gap between Junseok’s current situation and the high expectations of
Junseok’s parents, Katy thinks that it is impossible to find a school for Junseok. There will be no
school qualified enough to meet all the high standards of Junseok’s parents and generous enough
to accept a student like Junseok who has a very weak reference letter. Therefore, Katy’s wh-
question in line 27 conveys her strong negative assertion: “Junseok has nowhere to go. He is only
looking for top schools with few Korean students but he can’t even meet their qualifications for

111

admission.” Here, (l)ul-marking is used with the salient NP hakkyo ‘school,’ which emphasizes
Katy’s negative assertion conveyed through her rhetorical question.
Moreover, (l)ul-marking can also be used to mark a discourse topic when the information
introduced by the destination noun is continuously being talked about in the subsequent discourse
context. Revisit Excerpt 15, a telephone conversation between two female friends, Hanna and
Victoria.
Excerpt 15: Hanna (H) and Victoria (V)
Hanna is talking about her friend Lauren who is having financial difficulties.

1 H: kuntey na kath-umyenun na-lang ku enni-lang cincca pisusha-canha
but I same-if I-with that older sister-with really similar-SE
But if I were her, isn’t her situation very similar to mine?

2 cikum kac-ko iss-nun ton-i han tal-ey ung kulemyen
now possess-PROG-RL money-SUB one month-per well if so
If you have this amount money for one month

3 cipsey-lul celyakha-tenci ay hakkyo-[lul
rent-OBJ save-or kid school-OBJ
Either you should save money on rent or

4 V: [cokumssik cokumssik moa-kaciko
a little bit a little bit save-CONJ
you should save little by little and then

5 H: ay hakkyo-lul an ponay-tenci mwe-l hay-ya toy-can[ha
kid school-OBJ NEG send-or what-OBJ do-have to-SE
you should not send your kid to school, you’ve got to do something right?

6 V: [ung
Right

7 H: na-nun cincca ay hakkyo-to an ponay-ko mak kule-nuntey
I-TOP really kid school-even NEG send-and just like that-CONJ
I don’t even send my kid to school

8 ay hakkyo yuchiwen ceyil pissa-n[tey
kid school preschool the most expensive-CONJ
Preschool tuition is very expensive

9 V: [ung
Right

10 H: ilcwuil-ey twu-pen ka-nun ke opayk pwul-cca[li
one week-per two-time go-RC thing five hundred dollar-worth

112

It’s $500 for two times a week

11 V: [ung
Okay

12 H: kulen ke han tal-ey opayk pwul-ul ay hakkyo-ey nay-ci
that kind of thing one month-per five hundred dollar-OBJ kid school-to pay-SE
She’s paying $500 for her kid’s preschool tuition

à 13 cip-ul tto nemwu pissan tey-lul ka-n ke-ya leynthupi-lul
house-OBJ besides very expensive place-OBJ go-RC thing-to be rent-OBJ
She even moved into a very expensive house. I mean the rent

14 V: ahyu::
Oh my goodness
15 H: ku leynthupi-ka chen sapayk pwul-eytaka
that rent-SUB thousand four hundred dollar-on top of that
The rent is $1400 and on top of that

16 mwe cenkisey kulen ke ta ha-myen chen chilpayk
well electric bill that kind of thing all do-if thousand seven hundred
Well, it’s $1700 including all the utility fees such as electric bills

17 V: leynthupi-ka chen sapayk pwul-ina tway?
rent-SUB thousand four hundred dollar-as much as become
Her rent is $1400?

18 H: ung kekitaka cenkisey kathun ke-lul ttalo ta nay-myen-un
yes on top of that electric bill like thing-OBJ separately all pay-if-TOP
Yeah. On top of that, if she pays all the utilities fees such as electric bills

19 kasu-hako ta nay-mye-nun chen chilpayk pwul nayn-taynta han tal-ey
gas-and all pay-if-TOP thousand seven hundred dollar pay-QT one month-per
and gas everything, then she pays $1700 per month
Hanna is talking about the situation of her friend Lauren, who has to live on her husband’s low
income. Hanna says that if she were in her position with the same amount of money for one
month (lines 1 and 2), she would certainly do something to make ends meet by either saving
money on rent or not sending her kid to school (lines 3 and 5). Hanna says that she doesn’t even
send her kid to preschool (line 7) because the tuition is very expensive (lines 8 and 10). Hanna
also says Lauren is not only paying $500 for her kid’s preschool tuition (line 12) but also is living
in a very expensive house (line 13).

113

In Hanna’s utterance in line 13, the two destination nouns cip ‘house’ and pissan tey
‘expensive place’ of the movement verb ka- ‘to go’ are (l)ul-marked because the speaker needs to
direct the listener’s attention to this salient information that will be continuously being talked
about in the discourse context. Here, (l)ul-marking is a device used by the speaker to signal that
further descriptions or explanations about the marked element are coming up in the subsequent
context. Thus, in a string of Hanna’s follow-up utterances, she is talking about her friend
Lauren’s expensive house rent (lines 15, 16, 18, and 19).
Note that -(l)ul can also be stacked to the directional particle -ey, which can help the listener
distinguish contrastive referents in challenging discourse contexts. See Excerpt 18, a telephone
conversation between two female friends, Kristal (K) and Silvia (S).
Excerpt 18: Kristal (K) and Silvia (S)
Silvia tells Kristal that she has been taking her son Taekgeun to the beauty shop for a haircut
since he was three, but she has never taken her daughter Sojeong to the beauty shop until
recently. She felt sorry for her daughter so took her to the beauty shop today.

1 S: kulayse po-nikka nay-ka thaykkuni-nun sey sal ttay-pwuthe
so see-CONJ I-SUB Taekgeun-TOP three CL when-from
So I thought about it and as for Taekgeun ever since he was three

2 micangwen- ø teyliko tani-myense meli calla cwu-n kes iss-cyo
beauty shop take around-CONJ hair cut-give-RL thing exist-SE
I have been taking him to the beauty shop for a haircut
!
3 yethay-kkaci socengi melikhalak-to eps-ciman hanpen-to
now-until Sojeong hair-even not to exist-CONJ one time-even
Sojeong doesn’t even have hair but until now

à 4 micangwen-ey-lul an teyli-ko ka-ss-canha-yo

beauty shop-to-OBJ NEG take-and go-PST-SE-POL
I’ve never taken her to the beauty shop.

5 K: wenlay kulen-un key-eyyo
yecaay-tul-un kunyang kilu-ko
just like that-RL thing-to be girl-PL-TOP just grow-CONJ
That’s just how you should do. For girls, you just let them grow their hair

6 mith-ey-man calu-myen toy-nuntey mwe ha-le teyliko ka.
bottom-LOC-only cut-if become-CONJ what do-in order to take
and trim the ends of it so you don't need to take them to the beauty shop

114

7 S: kukes-to tto guilty feeling-i o-tela-kwu.
that-also again guilty feeling-SUB come-EXP-EMP
But that also makes me feel guilty

8 K: aiko chammal
Oh really
!
à 9 S: kulaykaciko onul micangwen-ey-lul teyliko ka-se
so today beauty shop-to-OBJ take-CONJ
So I took her to the beauty shop today and

10 K: ung
Okay

11 S: khwuphon-to epsi ka kaciko kwu pwul-ccali-lul call-ass-ta
coupon-even without go-CONJ nine dollar-worth-OBJ cut-PST-SE
She got a $9 haircut because I forgot to take a coupon
Silvia just realized that she has been taking her son to the beauty shop ever since he was three
years old (lines 1-2) but she has never taken her daughter to the beauty shop until now (lines 3-4).
While micangwen ‘beauty shop,’ the destination of the movement verb teyliko tani- ‘to take
around,’ is zero-marked in line 2, the same noun is marked with the stacked form -eylul in line 4.
Because the two actions involving the destination micangwen ‘beauty shop’ are contrasted with
each other (e.g., Saliva’s son has regularly taken to the beauty shop vs. Silvia’s daughter has
never taken to the beauty shop), -(l)ul can be added to the ey-marked destination noun micangwen
‘beauty shop’ in line 4. Here, (l)ul-marking direct the listener’s attention to this salient NP whose
referent is involved in an action sharply contrasted with the speaker’s prior action involving the
same destination and therefore the listener can easily distinguish the two contrastive actions.
After hearing Silvia’s concern, Kristal tells Silvia that she doesn’t really need to take her
daughter to the beauty shop because she can just let her daughter grow her hair and trim the ends
of it (lines 5 and 6). However, Silvia still feels sorry for not taking her daughter to the beauty
shop (line 7) and she starts to talk about her daughter’s first visit to the beauty shop today (line 9).
In Silvia’s utterance in 9, -(l)ul can be added to the ey-marked destination micangwen ‘beauty
shop’ because the information introduced by this NP is continuously being talked about in the

115

subsequent context. Thus, (l)ul-marking is used to indicate that further descriptions or
explanations related to the discourse topic will be coming up in the subsequent context so that the
listener can easily identify and process the salient information. Once micangwen ‘beauty shop’ is
introduced as a discourse topic in line 9, Silvia tells Kristal a story about her daughter’s first visit
to the beauty shop. Silvia’s daughter accidently got an expensive hair cut because Silvia forgot to
bring a coupon (line 11).
To Summarize, (l)ul-marking can be used in certain discourse contexts where the speaker
directs the listener’s attention to salient information that needs to be identified or processed. Thus,
-(l)ul can be used to mark the destination noun of the movement verbs such as ka- ‘to go,’ o- ‘to
come,’ and tani- ‘to attend,’ and teyliko k-/o-/tani- ‘to take someone/to bring someone/to take
someone around’ and the following pragmatic functions have been performed by -(l)ul regardless
of the absence or presence of the directional particle -ey.
(99) Emphasizing an assertion challenging the prior utterance or action
kulem kulen ay-nun enu hakkyo-lul ka-na
then that kind of kid-TOP which school-OBJ go-Q
‘To which school can he transfer then?’
[Implied: There is no school that he can transfer to.’]

(100) Distinguishing contrastive referents in discourse contexts of disagreement
hanpen-to micangwen-ey-lul

an teyli-ko ka-ss-canha-yo
one time-even beauty shop-to-OBJ NEG take-and go-PST-SE-POL
‘I’ve never taken her to the beauty shop.’

(101) Indicating a discourse topic and projecting further descriptions or explanations
about the discourse topic in the subsequent context
a. cip-ul tto nemwu pissan tey-lul ka-n ke-ya leynthupi-lul
house-OBJ besides very expensive place-OBJ go-RL thing-to be rent-OBJ
‘She even moved into a very expensive house. I mean the rent.’

b. kulaykaciko onul micangwen-ey-lul teyliko ka-se
so today beauty shop-to-OBJ take-CONJ
‘So I took her to the beauty shop today and…’

116

5.2 Double Object Construction
It is generally known that Korean allows two NPs marked with the object particle -(l)ul but
Japanese does not. This so-called “Double Object Construction (DOC)” has received considerable
attention both in the Korean and Japanese literature (O’Grady 1991; Hong 1991; Han 1999; M-J
Kim 2001; Schütze 2001). However, most previous studies have failed to explain the reasons and
intentions behind a speaker’s choice of using double (l)ul-marking because their analyses are
mostly based on decontextualized sentences and case theory. Therefore, the goal of this section is
to discuss four types of DOC and attempt to discover certain discourse contexts where the
speaker chooses to use double (l)ul-marking by analyzing actual spoken data.

5.2.1 Ditranstive Verbs
A ditransitive verb is a verb which takes a subject and two objects. The two objects refer to a
theme (a transferred item) and a goal (a recipient of the transferred item). These objects may also
be called direct and indirect.
(102) John-i Mary-eykey/-ul chayk-ul cwu-ess-ta
John-SUB Mary-DAT/OBJ book-OBJ give-PST-SE
‘John gave Mary a book.’
The goal argument Mary in (102) can be marked with either the dative particle -eykey ‘to’ or the
object particle -(l)ul. Hong (1991) claims that there is a difference in accomplishment semantics
(i.e., the end point of the theme argument) between the DAT-marked and (l)ul-marked goal.
(103) a. John-i Mary-eykey chayk-ul cwu-ess-ciman Mary-nun
John-SUB Mary-DAT book-OBJ give-PST-CONJ Mary-TOP

pat-ci an-ass-ta
receive-NEG-PST-SE
‘John gave a book but she didn’t get it.’

b. *John-i Mary-lul chayk-ul cwu-ess-ciman Mary-nun
John-SUB Mary-DAT book-OBJ give-PST-CONJ Mary-TOP

pat-ci an-ass-ta
receive-NEG-PST-SE
‘John gave Mary a book but she didn’t get it.’

117

For example, Hong (1991) claims that when the goal Mary is marked with the dative particle
eykey as in (103a), it only indicates that the theme chayk ‘book’ has been transferred to the
recipient Mary. On the other hand, the (l)ul-marked recipient Mary as in (103b) receives an
accomplishment reading that the book has been successfully delivered to the recipient Mary so
she got the book. Therefore, (103b) is claimed to be ungrammatical. However, the (l)ul-marked
goal argument does not mean that the theme chayk ‘book’ should necessarily arrive at the
endpoint. Thus, there seems to be no difference in accomplishment semantics between the goal
argument marked with the dative particle eykey and the object particle -(l)ul.
Others say that object-marking can be attached to goal arguments to mark focus (Han 1999;
Schütze 2001). Based on the ‘Two Case Layers’ hypothesis, Han (1999) claims that the semantic
dative particle eykey on goal arguments can be shifted to the pragmatic case -(l)ul for the purpose
of focus structure marking, as in (104).
(104) John-i Mary-lul chayk-ul cwu-ess-ta
John-SUB Mary-OBJ book-OBJ give-PST-SE
‘It was to Mary that John gave a book.’
Similarly, based on the analysis of -(l)ul stacked on the dative marker of a goal argument,
Schütze (2001) claims that the stacked particle -(l)ul should be analyzed not as a case marker but
as a focus marker. For example,
(105) a. John-i nwukwu-eykey chayk-ul cwu-ess-ni?
John-SUB who-DAT book-OBJ give-PST-Q
‘Who did John give a book to?’

b. John-i Mary-eykey-(lul) chayk-ul cwu-ess-ta
John-SUB Mary-DAT-ACC book-OBJ give-PST-SE
‘It was Mary that John gave a book to.’
In the answer constituent of a goal wh-question, “Who did John give the book to?” -(l)ul can be
optionally stacked on the DAT-marked goal Mary, which induces a focus reading. If Schütze’s
claim is valid, the stacked particle -(l)ul should be compatible with the focus marker such as man
‘only’, which enables a stronger exhaustive focus reading, as in (106).

118

(106) John-i Mary-eykey-man-ul chayk-ul cwu-ess-ta
John-SUB Mary-DAT-only-OBJ book-OBJ give-PST-SE
‘It was only to Mary that John gave a book to.’
Schütze (2001) has explained the syntactic and semantic difference of stacked -(l)ul using a focus
analysis, but there are some limitations in Schütze's proposal. The occurrence of case stacking is
very rare in Korean, and speakers appear to show significant differences in their individual use of
language concerning the acceptability of sentences with stacked (l)ul (Yoon 2004).
Han (1999) and Schütze's (2001)’s analysis, however, has captured one of the pragmatic
functions of (l)ul-marking that distinguishes contrastive referents in discourse contexts of
disagreement. When the referent of an object NP is dispreferred, challenged, or rejected by the
other referent in conversation, (l)ul can be used to mark those two referents. (l)ul-marking directs
the hearer’s attention to this salient information, which not only helps the hearer easily
distinguish contrastive referents but also emphasizes the speaker’s assertion challenging the prior
utterance or action involving the marked referents. Unlike most previous research, whose data are
mainly based on decontextualized sentences, in this dissertation individual segments of informal
conversation are analyzed to discover the recurring patterns in the speaker’s use of double (l)ul-
marking. The current analysis of the spoken data resulted in the following findings.
First, the shared knowledge between the interlocutors in the discourse context allows both the
goal and theme argument to be omitted. Revisit the continued conversation of Excerpt 3, a
telephone conversation between the two male speakers, Calvin (C) and Frank (F).
Excerpt 3 Extended: Calvin (C) and Frank (F)
Frank has not been able to get a job, so his older is sending him some money for financial
support.

9 F: ya:: hyeng-i yelsim-i-[ney!
wow older brother-SUB hard-working-to-be-SE
Wow, he’s really into playing the piano!

10 C: [ung yelsim-i-tela-ko
yeah hard-working-to be-EVI-EMP
Yeah, I saw him practicing the piano

119

11 F:

il-to yelsim-i-ko, unhyeki-nun ilcali-lul mos cap-a kacko
work-also hard-working-to be-and Eunhyeok-TOP job-OBJ NEG get-CONJ
He’s also working hard. Well, I haven’t been able to get a job so

12 kwulm-ko anc-ass-nun[tey
starve-and sit-PST-CONJ
I’m starving here

13 C: [kulay?
Really?

14 F: ani hyeng-i ponay-cwu-nun ton sseyo
no older brother-SUB send-give-RL money use
No, I’m using money that my bother sent me

à 15 C: hyeng-i ton-to ponay-cwe?
older brother-SUB money-also send-give
Does your brother send (you) money too?

16 F: apeci-hanthey emeni-hanthey-nun kuke pat-a-ssuki nemwu kule-n ke kath-ase
father-from mother-from-TOP that receive-and-use very like that-seem-CONJ
Because I think it’s too much to get money from my father or mother

17 C: ung
Yeah

18 F: ilcali capu-l ttay-kkaci-man
job get-when-until-only
Until I get a job

à 19 C: ung hyengi tonu-l ponay-cwu-n-ta-n mal-i[ya?
yes older brother-SUB money-OBJ send-give-PRES-SE-RL word-to be
Yeah, you mean your brother sends (you) money?

20 F: [yey
Yes
Frank tells Calvin that his financial situation is not too bad because his older brother has been
sending him some money (line 14). In the next turn, Calvin asks Frank a question with a partial
repeat of the prior utterance, hyengi tonto ponay cwe “Does your brother send (you) money too?”
In Calvin’s question in line 15, the goal argument Frank is omitted because Frank is the only one
available in the immediate context who can be the the recipient of the money that his brother has
been sent and thus its referent is easily identifiable or recoverable even without meaning it. In
Calvin’s other question “You mean your brother sends (you) money?” in line 19, the goal

120

argument Frank is also omitted. Revisit Excerpt 14 for another example of the omission of the
goal argument.
Excerpt 14 Revisited: Michael (M), Nicky (N), and Robert (R)
Michael, Nicky, and Robert are eating spicy chicken stew. While Michael and Nicky rarely eat
kimchi, Robert says he must eat kimchi with the chicken stew because he thinks they have
different flavors of spiciness: intense spiciness of the chicken stew vs. the refreshing spiciness
of kimchi.
!
4 R: ani lamyen mek-ul ttay maliya kimchi-lul cal an kkenay-cweyo
well ramen eat-when you know kimchi-OBJ well NEG take out-give
Well, you know when I eat ramen, she doesn’t give me kimchi.

5 ohilye mak singkewun panchan-ul kkenaycweyo kukey mwe ha-nun cis-iya
instead just bland side dish-OBJ take out-give that what do-RL act-to be
She just gives me a bland side dish instead. What in the world is she doing?

à 6 kimchiccikay kkulhye-ss-nuntey way panchan-ul kimchi-lul an cwu-pnikka
kimchi stew boil-PST-CONJ why side dish-OBJ kimchi-OBJ NEG give-SE
You cooked kimchi stew but why don't you give me kimchi for side dish?
!
Robert says his wife doesn’t let him eat kimchi when he eats ramen (line 4). She gives him a
bland side dish instead of kimchi and Robert does not like it (line 5). In Robert’s next utterance in
line 6, the goal argument of the verb cwu- ‘to give’ is omitted because Robert is the only possible
candidate available in the immediate context who can be the recipient of the side dish kimchi, and
hence the referent of the goal argument is easily recoverable based on the shared knowledge
between the interlocutors in the conversation. The theme argument kimchi in line 5 is (l)ul-
marked, which emphasizes the speaker’s assertion conveyed through his question: “You should
give me kimchi!”
Moreover, the theme argument is also often omitted, as in Excerpt 19, a conversation between
two male friends, Ted and William. Ted wants to give away his video cassette recorder (VCR) to
one of his friends, Taehun or Hyeongdae.
Excerpt 19: Ted (T) and William (W)
Ted wants to give away his VCR to one of the two candidates, Taehun or Hyeongdae.

121

à 1 T: nay-ka ayey kunyang kacta cwu-ci ma-lkka
I-SUB at all just bring-NEG-shall
Shall I just not even bring it?

2 kulayto twu kay ta- twu kay-lul tul-ko ka-l swu-nun-
though two CL all two CL-OBJ carry-and-go-can-TOP
But I can’t take both anyway

3 W: il-un thayhwuni-ka manhi hay-cw-ess-nuntey
work-TOP Taehun-SUB a lot do-give-PST-CONJ
As for work, Taehun helped you a lot

à 4 T: kulekey maliya kulehta-ko hay-to thayhwuni-lul cwu-lyeni
That’s what I’m saying it is so-QT-CONJ Taehun-OBJ give-CONJ
That’s what I’m saying. But when I think that I should give it to Taehun

5 W: kukes-to ku[lehci ((laugh))
that-also it is so
That’s also a problem

6 T: [Hyengtay-ka tto kulehko.
Hyeongdae-SUB again like that
Hyeongdae is still on my mind
Ted is talking about to whom he should give his VCR, Taehun or Hyeongdae. He is having a hard
time making a final decision. He even thinks he should not give it away (line 1). In his utterance
in line 1, the theme argument VCR has been omitted because its referent is assumed to be known
to both the speaker and the listener and thus it is easily recoverable in context. In the next
utterance, Ted changes his mind to give it away because he cannot take two VCRs anyway (line
2). William tells Ted to consider the fact that Taehun has helped him a lot (line 3). Ted agrees
with William’s point but when he thinks of giving his VCR to Taehun (line 4), Hyeongdae is still
on his mind.
Another interesting finding is that when the goal argument is contrasted with another
candidate in a challenging context, it is likely to be (l)ul-marked. For example, the goal argument
Taehun in line 4 is (l)ul-marked not only because it is contrastively focused with the other
candidate Hyeongdae in line 6, but also because the action involving this goal argument Taehun
(i.e., Ted is thinking of giving his VCR to Taehun) is in disagreement with the action involving
the other candidate Hyeongdae in the follow-up utterance (i.e., Ted still thinks his VCR should be

122

given to Hyeongdae). (l)ul-marking directs the listener’s attention to the two contrastive actions
involving Taehun and Hyeongdae so that the listener can easily distinguish and process this
important information.
Lastly, because a goal and a theme argument are often dropped based on the shared
knowledge available to the interlocutors, the speaker may use (l)ul-marking for both arguments to
achieve a certain communicative purpose. See Excerpt 20, another conversation between a male
speaker Ju and two female speakers Choi and Park taken from the fourth episode of the Korean
animal TV show, Man Feeding a Dog.
Excerpt 20: Choi (C), Ju (J), and Pakr (P)
After having lunch, Ju asks Choi how to make handmade snacks for dogs.

1 J: swucey kansik-ul mwel-lo mantule?
handmade snack-OBJ what-INST make
With what do you make handmade snacks for your dogs?

2 C: na-nun yayney tanhopak salma-cwe maynnal
I-TOP this kid sweet pumpkin boil-give every day
I give them steamed sweet pumpkin every day

3 J: haa:: ((sigh))

4 P: cengseng-i-cyo ((laugh))
effort-to be-SE
It’s such devotion, isn’t it?

5 J: na-nun nay kes-to ceytaylo hay mek-ci anh-ko
I-TOP my thing-even right way do eat-NEG-CONJ
I don’t even cook my own food

6 cenpwu pan insuthenthu hay mek-ko
all half instant do eat-CONJ
I always use half-cooked instant food

7 C: samchon-i iltan sey mali-lul ttean-ass-umyen incey [hay cwe-ya-ci
uncle-SUB once three CL-OBJ take-PST-when now do give-have to-SE
When you adopt three dogs, now you should cook and feed them

8 P: [kuchyo hay-ya-cyo
right do-have to-SE
That’s right. You should

9 mal nao-n kimey onul kunyang kansik hanpen
word come out-CONJ today just snacks once

123

Since we’re talking about homemade snacks, why don’t we just try making some today?

10 J: hanpen mantu-le pwa
once make-try
Go ahead

11 C: kathi mantul-eya-ci
together make-have to-SE
We have to make it together

((Choi and Park wash and boil sweet pumpkins, sweet potatoes, and broccoli for dog snacks.
Ju is watching them prepare the ingredients.))

à 12 J: ani ike-l kangaci-lul cwu-lyeko i cohun ke-l ta salma?
no this-OBJ puppy-OBJ give-in order to this good thing-OBJ all boil
Wait, do you boil all these good ingredients just to give them to the dogs?

13 nay-ka mek-keyss-ta ya
I-SUB eat-will-SE oh my
Oh my, I would rather eat it

Ju asks Choi how to make homemade snacks for his dogs (line 1). Choi says she gives her
dogs steamed sweet pumpkins every day (line 2). Ju’s deep sigh in the next turn shows that he is
surprised by Choi’s answer (line 3). Park is also surprised by Choi’s devotion to her dogs (line 4).
Ju says he does not even cook for himself and that he only eats instant food (lines 5-6), which
shows that Ju doesn’t seem to like to make snacks for his dogs because he thinks it’s too much
work. However, Choi thinks that he should start cooking and should be a responsible father for
his three adopted dogs (line 7). Park also agrees with what Choi says (line 8) and suggests to
make handmade snacks for his dogs (line 9). Ju tells Park to cook (line 10) and tries to walk away,
but Choi insists that he cooks together with them (line 11). Ju reluctantly goes into the kitchen
and watches Choi and Park prepare and boil the ingredients for dog snacks, such as sweet
pumpkins, sweet potatoes, and broccoli.
After watching the preparation process for making dog snacks, Ju utters a question, ikel
kangacilul cwulyeko i cohun kel ta salma? “Do you boil all those good ingredients just to give
them to the dogs?” in line 12. His question is not uttered to ask for confirmation from the listener,
but rather to express the speaker’s negative assertion challenging the prior action of the other

124

speakers. Here, (l)ul-marking is used for both the goal and theme argument kangaci ‘puppy’ and
ike ‘this’ of the verb cwu- ‘to give,’ which emphasizes the speaker’s negative assertion conveyed
through his utterance: “It’s too much to boil all those good ingredients just to give them to the
dogs.” Ju’s follow-up utterance in the next turn, nayka mekkeyssta “I would rather eat it,” also
aligns with the negative assertion implied in his previous utterance.
Double (l)ul-marking is also used when the speaker restates his or her prior utterance,
especially when the interlocutor expresses rejection, disagreement, or uncertainty about the
speaker’s prior utterance. Revisit the conversation between Bae (and) and Yoon (Y) in Excerpt 8.
Excerpt 8 Revisited: Bae (B) and Yoon (Y)
Yoon is sleeping over at her friend Bae’s shared house. Bae is worried that Yoon might feel
cold when she is asleep. Bae wants to give Yoon a pair of pants and an extra blanket, but Yoon
says she will be fine.

1 B: ike an-ey ipwul eps-ni?
this inside-LOC blanket not to exist-Q
Isn’t there a blanket inside this?

2 na-n ike an-ey ipwul iss-nuntey
I-TOP this inside-LOC blanket exsit-CONJ
I have a blanket inside this

3 ne an chwuwu-lkka? ike cwu-lkka?
you NEG cold-do you think this give-Shall
Aren’t you going to feel cold? Do you want his?

4 ike cwu-lkey. ike ppal-ase nehe-noh-un ke-ya
this give-will this wash-and then put in-leave-RL thing-to be
I’ll give you this. I just washed it and put it away (in the closet)

5 Y: i wi-eyta kkyeip-ko ca-lkey
this top-LOC wear-CONJ sleep-will
I’ll just put on more clothes and go to sleep

à 6 B: kuliko i ipwul-ul ne-l cwu-lkey
and this blanket-OBJ you-OBJ give-will
And I’ll give you this blanket

7 Y: aniya ike-man iss-umyen tway
no this-only have-if become
As long as I have this, it should be fine

125

Bae wants to give Yoon a pair of pants and an extra blanket because she is worried that Yoon
might feel cold while asleep (line 3). Bae offers Yoon a pair of pants (line 4), but Yoon politely
turns them down by saying that she will just put on more clothes and go to sleep (line 5). When
making her second offer in line 6, both the goal and theme argument ne ‘you’ and ipwul ‘blanket’
are (l)ul-marked. This double (l)ul-marking emphasizes her action of giving an extra blanket to
the hearer which has been turned down by the hearer in the prior utterance. When the hearer
expresses rejection, disagreement, or uncertainty about the speaker’s prior utterance, the speaker
may need to restate what has been said to convince or persuade the hearer. When there is an
ongoing disagreement between the interlocutors, object-marking can be used to help the listener’s
selective focus on salient elements so that the speaker’s assertion conveyed through restatements
can be more strongly expressed.
Thus far, the speaker’s use of (l)ul-marking in ditransitive sentences has been analyzed. The
results show that both goal and theme arguments are often omitted when their referents are easily
identifiable or recoverable in discourse context based on the shared knowledge available to the
interlocutors. Next, the goal argument is likely to be (l)ul-marked when there is a contrastive
referent in the preceding or subsequent context. Lastly, double (l)ul-marking may occur when the
speaker needs to emphasize the speaker’s assertion challenging the prior utterance or action of
non-present or co-present parties.

126

5.2.2 Adverbials
In Korean, certain adverbials can be (l)ul-marked, as illustrated in (107)-(110).
(107) Durational adverbials
na han sikan sasip pwun-ul pesulul tha-n-ta
I one hour forty minute-OBJ bus-OBJ ride-PRES-SE
‘I was riding on a bus for 1 hour and 40 minutes.’

(108) Path length adverbials
John-i onul swuyeng-ul il khillo-lul hay-ss-ta
John-SUB today swimming-OBJ one kilometer-OBJ do-PST-SE
‘John swam 1 kilometer today.’

(109) Multiplicative adverbials
John-i ilcwuil-ey swul-ul yel pen-ul masi-n-ta
John-SUB one wekk-per alcohol-OBJ ten time-OBJ drink-PRES-SE
‘John drinks 10 times a week.’
(110) Quantity
a. swuep-ul sey kay-lul tut-nun-ta
class-OBJ three CL-OBJ listen-PRES-SE
‘(I’m) taking three classes.’

b. John-i na-eykey chayk-ul twu kwen-ul cwu-ess-ta
John-SUB I-to book-OBJ two CL-OBJ give-PST-SE
‘John gave me two books.’
This phenomenon has received considerable attention in the literature (Wechsler and Lee 1996;
Schütze 1996; M-J Kim 2001). Wechsler and Lee (1996) claims that (l)ul-marked adverbials
expressing durations of time, path length, or the multiplication of an action or event are
Situational Delimiters, which “temporally quantify or delimit the situation expressed by the
predicate.” The defining semantic property of Situational Delimiters is that they are “Extensive
Measures” which satisfy the condition of “Additivity.”
(111) Additivity: m(x ⊕ y) = m(x) + m(y), if x and y do not overlap.
(Wechsler and Lee, 1996, p.645)
(111) can be re-phrased as this: the sum of the measured x and y equals the sum of measured x
and measured y. For example, Wechsler and Lee (1996) compare average driving speed with
duration. Suppose an event x of driving at an average speed of 50 miles per hour for one hour and
an event y of driving at an average speed of 40 miles per hour for one hour. Both average speed

127

and duration are measure functions. The sum of the duration of the two events will be two hours
(1 hour + 1 hour = 2 hours). However, the sum of the average speed of the two events will be 45
instead of 90. Thus, duration is an extensive measure which satisfies the condition of additivity,
while average driving speed is not.
Based on this criteria, Wechsler and Lee (1996) have captured why (l)ul-marking can be used
for certain types of adverbials such as duration, path length, and multiplicative adverbials but not
for location and manner adverbials such as quickly, intelligently, or halfheartedly. In case of
location and manner adverbials, the sum of the two events are not additive and hence they are not
extensive measures.
Following Wechsler and Lee (1996), M-J Kim (2001) also claims that (l)ul-marked adverbials
have a delimiting function which sets a limit on duration, path length, or multiplication of the
event denoted by the predicate. M-J Kim also says that not all types of adverbials can be (l)ul-
marked. For example, locative, temporal, manner adverbials and and frequency adverbs can not
be (l)ul-marked.
(112) Locative adverbials
John-un Mary-lul Seoul-ese/-*(l)ul manna-ess-ta.
John-TOP Mary-OBJ Seoul-in/-OBJ meet-PST-SE
‘John met Mary in Seoul.’

(113) Temporal adverbials
John-un Mary-lul se-si-ey/-*(l)ul manna-ess-ta.
John-TOP Mary-OBJ three-o’clock-at meet-PST-SE
‘John met Mary at three o’clock.’

(114) Manner adverbials
John-un ppali-*lul kel-ess-ta.
John-TOP quickly-OBJ walk-PST-SE
‘John walked quickly.’
(115) Frequency adverbials

John-nun ttayttaylo-*lul sanchayk-ul naka-ess-ta.
John-TOP sometimes-OBJ walk-OBJ go out-PST-SE
‘Sometimes John went out for a walk.’
(M-J Kim, 2001, p.8)

128

Wechsler and Lee (1996) and M-J Kim (2001)’s proposal is meaningful because they have
explained semantic restrictions and differences between (l)ul-marked and non-(l)ul marked
adverbials. However, because the data used in their analyses consist of decontextualized
sentences based on native speaker intuition judgments, they have failed to consider the pragmatic-
discourse factors which can significantly affect the speaker’s choice of using (l)ul-marking in
adverbials. Consider the following examples.
(116) Manner adverbs
a. ku aki-nun camsi-to coyonghi-lul an iss-nun-ta.
that kid-TOP for a second-even quietly-OBJ NEG exist-PRES-SE
‘The kid does not stay quiet for a second.’
[Implied: The parents should do something to calm him down.]

b. chopo wuncen i-ese ppalli-lul ka-ci-lul mos ha-n-ta.
novice driving to be-CONJ quickly-OBJ go-NOML-OBJ NEG do-PRES-SE
‘(He) can’t go fast because he is a novice driver.’
[Implied: He should go faster.]

(117) Frequency adverbs
a. John-un yeca chinkwu-wa heyeci-ko nase mayil-ul
John-TOP girl friend-with break up-CONJ every day-OBJ

swu-lul masy-ess-ta.
alcohol-OBJ drink-PST-SE
‘John has been drinking every day ever since he broke up with his girlfriend.’
[Implied: I cannot believe that John has been drinking every day ever since he
broke up with his girlfriend.]

b. eti-eyse-tun enu swunkan-ey-tun hangsang-ul
Where-LOC-or which moment-LOC-or always-OBJ

nolyekha-nun casey-lo sal-aya-ci!
make an effort-RL attitude-INST live-have to-SE
‘Wherever you are and at whatever moment, you should live your life with a
hard-working attitude.’
(l)ul-marking can be used for manner or frequency adverbs in certain discourse contexts:
where the speaker disagrees with or challenges the prior action involving manner adverbs such as
coyonghi ‘quietly’ in (116a) and ppalli ‘quickly’ in (116b); where the speaker expresses surprise
or disbelief about the action or event involving frequency adverbs such as mayil ‘every day’ in
(117a), which goes contrary to the speaker’s belief or expectation; or where the speaker

129

emphasizes his or her assertion challenging the interlocutor’s uncertainty, disagreement, or
rejection of the action involving frequency adverbs such as hangsang ‘always’ in (117b).
Another interesting analysis of (l)ul-marked adverbials has been proposed by Schütze (1996).
As mentioned in M-J Kim (2001), Schütze (1996) claims that -(l)ul is a focus-marker and
therefore that (l)ul-marked adverbials can have a distinct pitch accent as if they are focused.
Schütze’s analysis has captured one of the pragmatic functions of object-marking that
emphasizes the speaker’s assertion challenging the previous utterance or action which goes
contrary to the speaker’s belief or expectation. In this case, (l)ul-marked adverbials are often
accompanied with emphatic stress and -(l)ul can be easily interchangeable with the particle -ina
which expresses the meaning of ‘more than the expected quantity.’
(118) Durational adverbials
na han sikan sasip pwun-ul/-ina pesu-lul tha-n-ta
I one hour forty minute-OBJ bus-OBJ ride-PRES-SE
‘I was riding on a bus for 1 hour and 40 minutes.’
[Implied: It was too long to ride on a bus for 1 hour and 40 minutes.]

(119) Path length adverbials
John-i onul swuyeng-ul il khillo-lul/-ina hay-ss-ta
John-SUB today swimming-OBJ one kilometer-OBJ do-PST-SE
‘John swam 1 kilometer today.’
[Implied: It’s more than I expected that John can swim 1 kilometer.]

(120) Multiplicative adverbials
John-i ilcwuil-ey swul-ul yel pen-ul/-ina masi-n-ta
John-SUB one wekk-per alcohol-OBJ ten time-OBJ drink-PRES-SE
‘John drinks 10 times a week.’
[Implied: It’s too much to drink 10 times a week.]

(121) Quantity
a. swuep-ul sey kay-lul/-na tut-nun-ta
class-OBJ three CL- OBJ listen-PRES-SE
‘(I’m) taking three classes.’
[Implied: It’s too much to take three classes.]

b. John-i na-eykey chayk-ul twu kwen-ul/-na cwu-ess-ta
John-SUB I-to book-OBJ two CL-OBJ/ give-PST-SE
‘John gave me two books.’
[Implied: It’s more than I expected that I got two books from John.]

130

The analysis of actual spoken data has also revealed other pragmatic functions of (l)ul-marked
adverbials. First, (l)ul-marked adverbials often appear in repair situations where the speaker
expresses surprise or disbelief about the prior utterance and asks for clarification about the
information involving the marked adverbial. Revisit Excerpt 11, a telephone conversation
between two friends Aron and Tyler.
Excerpt 11 Revisited: Aron (A) and Tyler (T)
Aron is talking about his Chinese officemate who spends a lot of money on phone bills.
!
8 A: ttak po-nikka
just see-CONJ
I had a look at his phone bills

9 T: ung
Okay

10 A: ku han thonghwa-ey payk phalsip payk yuksipil pwun-i-n-ka
well one call-per hundred eighty hundred sixty one minute-to be-PRES-Q
I saw some phone calls one hundred eighty, one hundred sixty one or

11 payk phalsip pwun-ccali-to iss-tela
hundred eighty minute-per-also to exist-EVI
one hundred eighty minutes long

à 12 T: ((laugh)) kulem mwe myech sikan-ul ha-n keya?
then huh how many hour-OBJ do-RL thing-to be
Then, boy, for how many hours did he talk on the phone?

à 13 sey sikan-ul hay-ss hay-ss-ta-n yayki-ya?
three hour-OBJ do-PST do-PST-SE-RL story-to be
You mean he talked on the phone for three hours?
Aron is talking about his Chinese officemate’s ridiculously overcharged phone bills. He said that
he saw the bills once (line 8) and and the longest call was almost 180 minutes long (lines 10-11).
In line 12, Tylor is very surprised by this unbelievably long phone call and asks a question, “For
how many hours did he talk on the phone?” His question is not uttered to ask for a piece of new
information, but rather to express the speaker’s surprise about the other speaker’s prior action
involving the adverbial expression myech sikan ‘how many hours.’ Thus, this (l)ul-marked
adverbial phrase emphasizes the speaker’s feeling of surprise about the event involving the

131

marked element: “It’s very surprising that he was on the phone for such a long time.” In Aron’s
other question in line 13, the adverbial phrase sey sikan ‘three hours’ is (l)ul-marked, which also
emphasizes the speaker’s surprise or disbelief about the prior action: “It’s unbelievable that he
was talking on the phone for three hours.”
Note that if the noun thonghwa ‘talking on the phone’ of the Object-Verb compound verb
thonghwa-ha- ‘to talk on the phone’ appears in the above sentences, double-(l)ul marking can be
naturally used.
(122) Durational adverbials
a. thonghwa-lul myech sikan-ul ha-n ke-ya?
talking on the phone-OBJ how many hour-OBJ do-RL-thing-to be
‘Then, for how many hours did he on the phone?’
[Implied: It’s surprising that he was on the phone for such a long time.]

b. thonghwa-lul sey-sikan-ul hay-ss-ta-n yayki-ya?
Phone call-OBJ three-hour-OBJ do-PST-SE-RL story-to be
‘Are you saying he was on the phone for three hours?’
[Implied: I can’t believe that he was on the phone for three hours.]
Double (l)ul-marking can also be used naturally for other adverbials when the speaker initiates
repair on the information involving the referent that needs to be clarified, corrected, or modified.
(123) Path length adverbials
John-i onul swuyeng-ul il khillo-lul hay-ss-ta-n maliya?
John-SUB today swimming-OBJ one kilometer-OBJ do-PST-SE-RL you mean
‘You mean John swam 1 kilometer today?’
[Implied: I don’t think John swam 1 kilometer. Maybe just 100 meters.]

(124) Multiplicative adverbials
John-i ilcwuil-ey swul-ul yel pen-ul masi-n-tako?
John-SUB one wekk-per alcohol-OBJ ten time-OBJ drink-PRES-EMP
‘Are you saying John drinks 10 times a week?’
[Implied: I don’t think John drinks 10 times a week. Maybe 5 times.]

(125) Quantity
a. ne swuep-ul sey kay-lul tut-nun ke-ya?
you class-OBJ three CL-OBJ listen-PRES thing-to be
‘Are you taking three classes?’
[Implied: I thought you’re taking only one class. Aren’t you?]

b. John-i ne-hanthey chayk-ul twu kwen-ul cwu-ess-tako?
John-SUB you-to book-OBJ two CL-OBJ give-PST-EMP
‘Are you saying John gave you two books?’
[Implied: I don’t think John gave you two books. It was three.]

132

Second, (l)ul-marked adverbials also indicate that further descriptions or explanations about
the marked elements will be coming up in the subsequent context so that the listener can easily
identify and process the information that he/she may not know. See Excerpt 21, a telephone
conversation between two friends Elaine (E) and Rachel (R).
Excerpt 21: Elaine (E) and Rachel (R)
Elaine is talking about her colleague who is a new a Ph.D. student in her department.

à 1 E: kuntey incey nayka swuep-ul kathi paksa kwaceng ke-lul
by the way now I-SUB class-OBJ together Ph.D. course thing-OBJ
By the way, now I’m taking a Ph.D. course class with (her)

à 2 hana-lul tut-ketun
one-OBJ listen-SE

3 R: ung
Okay

4 E: kuntey kuke-ey tayhayse koyngcanghi suthuleysu-lul pat-tela-ko
but that-about very stress-OBJ receive-EVI-EMP
But that’s giving her a lot of stress.
Elaine is talking about her colleague who has recently started a Ph.D. program in her department.
She is taking a Ph.D. course with this new student now (line 1). In Elaine’s utterance in line 1, the
two adverbial expressions paksa kwaceng ‘Ph.D. course’ and hana ‘one’ are (l)ul-marked to
indicate that the speaker is going to provide further descriptions or explanations about the marked
elements in the subsequent context. Therefore, the two (l)ul-marked adverbials can be considered
additional information related to the discourse topic swuep ‘class’ that the listener needs to
identify and process.

133

5.2.3 Possessor-Possessee
When two NPs have a relationship of inalienable possession including whole-part or kinship
relations, -(l)ul can be used in the place of the genitive particle -uy which marks a possessor (Han
1999).
(126) Possessor-Possessee (inalienable)
a. Mary-ka John-uy tali-lul cha-ss-ta
John-SUB John-GEN leg-OBJ kick-PST-SE
‘Mary kicked John’s leg.’

b. Mary-ka John-ul tali-lul cha-ss-ta
Mary-SUB John-OBJ leg-OBJ kick-PST-SE
‘Mary kicked John in the leg.’

(127) Possessor-Possessee (inalienable)
a. Mary-ka namwu-uy kaci-lul chi-ess-ta.
Mary-SUB tree-GEN branch-OBJ prune-PST-SE
‘Mary pruned the branches of the tree.’

b. Mary-ka namwu-lul kaci-lul chi-ess-ta.
Mary-SUB tree-OBJ branch-OBJ prune-PST-DEC
‘Mary pruned tree branches.’
In (126b), the first (l)ul-marked NP John stands in an inalienable relationship with the second
(l)ul-marked NP tali ‘leg’ as the possessor of the body part. Similarly, the first (l)ul-marked NP
namwu in (127b) stands in an inalienable relationship with the second (l)ul-marked NP kaci
‘branch’ as the possessor of part of a tree. Han (1999) points out that when the two NPs stand in
an alienable relationship, such -uy to (l)ul alternation is not available, as in (128).
(128) Possessor-Possessee (alienable)
Mary-ka tongsayng-uy/*-ul chak-ul peli-ess-ta.
Mary-SUB brother-GEN / OBJ book-ACC throw out-PST-SE
‘Mary threw out her brother’s book.’
Interestingly, Wunderlich (2014) has found that this (l)ul-marked possessor and (l)ul-marked
possessee construction is sometimes accepted in Japanese. See (129).
(129) a. Double o constraint
Mary-ka John-no/*-o asi-o ketta.
Mary-SUB John-GEN/OBJ leg-OBJ kicked
‘Mary kicked John’s leg.’

134

b. The two (l)ul-marked possessor and possessee are separated by adverbials
Mary-ka John-o kinoo undoozyo-de asi-o ketta.
Mary-SUB John-OBJ yesterday playground-LOC leg-OBJ kicked
‘Yesterday, Mary kicked John’s leg at the playground.’

c. The Possessor is scrambled
John-o Mary-ga asi-o ketta.
John-OBJ Mary-SUB leg-OBJ kicked
‘John, Mary kicked (his) leg.’
(Modified from Wunderlich 2014)
It is generally known that Japanese does not allow the use of object-marking for more than one
argument in a sentence, as in (129a). However, double object-marking can be allowed under
certain circumstances: When two o-marked arguments are separated by adverbials or when the
possessor nominal appears in the sentence-initial position due to scrambling, double object-
marking can be allowed. For instance, in (129b), both the possessor John and the possessee asi
‘leg’ can be o-marked because the two arguments are separated by two adverbials kinoo
‘yesterday’ and undoozyo-de ‘at the playground.’ In addition, a sentence with non-canonical word
order, as in (129c), where the possessor John appears in the sentence initial position, both the
possessor John and the possessee asi ‘leg’ can be o-marked.
The findings of Wunderlich’s (2014) work can be applied to Korean. (l)ul-marking is a device
used by the speaker in conversation to provide an extra linguistic signal for either hard-to-process
or salient information, which helps the hearer to pay attention or focus on the marked elements
and hence facilitates smooth information processing and successful communication. In languages
with SOV word order like Japanese and Korean, a subject usually appears in the sentence-initial
position and an object is normally expected to be immediately adjacent to the sentence-final verb
in transitive sentences. If the object NP is not immediately adjacent to the verb, the hearer’s
decoding load would significantly increase and hence (l)ul-marking would help the hearer to
decode the sentence as in (129b).
What if the arguments in a sentence do not follow this default SOV word order? The hearer’s
decoding load would also significantly increase and hence object-marking would help the hearer

135

to decode the sentence. Therefore, a sentence with non-canonical OSV word order as in (129c) is
highly unpredictable, and therefore object-marking is more likely to be used for such types of
objects to provide more linguistic signals for hard-to-process information.

136

5.3 Summary
In this Chapter, the two grammatical phenomena of “Substitution,” the use of (l)ul in the place of
other particles, and the so-called the “Double Object Construction (DOC)” in conversational
Korean have been discussed based on the following hypothesis: Object-marking is a device used
by the speaker to provide an extra linguistic signal for either hard-to-process or important
information that the listener needs to identify or process. Thus, substitution, stacking, and double
object-marking are expected to be used by the speaker to direct the listener’s attention to either
hard-to-process or salient information that the listener has to identify or process.
The analysis of actual spoken data has discovered that substitution, stacking, or double object-
marking is consistently used by speaker in certain discourse contexts to perform the following
pragmatic functions.
(i) Emphasizing an assertion challenging the prior utterance or action
(e.g., wondering, complaining, arguing, mocking, blaming, regretting, persuading, etc.)

(ii) Initiating repair on the information involving the referent that needs to be clarified,
corrected, or modified

(iii) Distinguishing contrastive referents in discourse context of disagreement

(iv) Indicating a discourse topic and projecting further descriptions or explanations
about the discourse topic in the subsequent context
The findings in this chapter are as follows: First, (l)ul can be used to mark the destination
noun to perform the pragmatic functions above regardless of the absence or presence of the
directional particle -ey. Second, in sentences with the ditranstive verbs cwu- ‘to give,’ both the
goal and the theme argument are often omitted when their referents are easily identifiable or
recoverable based on the shared knowledge established between the interlocutors. The goal
argument tends to be (l)ul-marked if a contrastive referent appears in the preceding or subsequent
context. Double (l)ul-marking can be used when the speaker needs to emphasize his or her
assertion challenging the prior utterance or action of non-present or co-present parties.

137

Next, (l)ul-marked adverbials are often used to emphasize the speaker’s assertion challenging
the previous utterance or action which goes contrary to the speaker’s belief or expectation. They
also appear in repair situations in which the speaker expresses surprise or disbelief about the prior
utterance and asks for clarification about the information involving the marked adverbials. In
addition, two (or more) (l)ul-marked adverbials can be considered as additional information
related to the discourse topic that the listener needs to identify and process.
Lastly, when two NPs have a relationship of inalienable possessor and possessee, double
(l)ul-marking can be used when the two arguments are separated by adverbials or when the
possessor NP appears in the sentence-initial position due to scrambling. If the object NP is not
immediately adjacent to its predicate, the hearer’s decoding load would significantly increase and
hence (l)ul-marking would help the hearer to decode the sentence. In sentences with non-
canonical word order, the hearer’s decoding load would also significantly increase and hence
(l)ul-marking is likely to be used to help the hearer decode the sentence.

138

6. Conclusion and Implications
In this dissertation, the speaker’s choice in the use of zero- or (l)ul-marking in informal spoken
Korean has been analyzed. Both linguistic factors (i.e., the phonological size and the form of the
object NP, adjacency conditions, and word order) and pragmatic factors (i.e., level of formality in
speech, shared-knowledge between interlocutors, information predictability, and information
salience) have been found to affect the speaker’s choice of (l)ul-marking. (l)ul-marking is
frequently used to facilitate the listener’s selective focus on difficult-to-process information
which requires the listener’s extra attention (i.e., the object NP with a modifier, the object NP not
immediately adjacent to its verb, and the object NP in a sentence with non-canonical word order).
However, the speaker’s choice of using -(l)ul cannot be solely explained with those linguistic
factors. In particular, shared-knowledge between interlocutors plays an important role in the
speaker’s choice. When temporal, spatial settings, or cultural backgrounds are shared by
interlocutors and when the referent of an object NP is already shared by the participants in
conversation, the grammatical relation of the object NP does not necessarily have to be specified
by (l)ul (Lee and Thompson 1987).
Previous studies have found that object-marking is a device used by the speaker to help the
listener pay attention or focus on salient information for successful communication (Fujii and
Ono 2000; Yang 2007). Therefore, object-marking does not solely indicate the grammatical
relation of the NP, but it also performs the following pragmatic functions: indicating the
discourse topic, emphasizing an action, clarifying referents that need to be confirmed or corrected,
and distinguishing contrastive referents.
Previous studies have shed some light on the pragmatic use of -(l)ul as a device available to
the speaker to facilitate the listener’s selective focus on either hard-to-process or salient
information for successful communication. However, there has been no discussion about a more
comprehensive understanding of the mechanism that controls the speaker’s use of (l)ul-marking.

139

What is missing in the observations made by previous studies is that (l)ul-marking seems to be
consistently used in discourse contexts where the speaker disagrees with what has been said in the
preceding context. Disagreement can vary in range from weak disagreement, such as expressing
surprise or wondering about the prior utterance or action which goes contrary to the speaker’s
belief or expectation, to serious disagreement, such as showing a negative or sarcastic attitude
(e.g., dispreference, disbelief, objection, mockery, etc.) or asking for a clarification or correction
about what has been said. In such challenging contexts, (l)ul-marking is a device used by the
speaker to direct the listener’s attention to salient information and hence the speaker’s intended
meaning through the utterance can be successfully delivered to the listener.
Therefore, what (l)ul-marking does in conversation is more than indicating the grammatical
role of the NP in a sentence. This dissertation also provides detailed descriptions and explanations
about the linguistic forms or patterns that frequently co-occur with each pragmatic use of -(l)ul.
First, (l)ul-marking emphasizes an assertion challenging the prior utterance or action of co-
present or non-present parties in rhetorical questions, self-addressed questions, and restatements
of the speaker’s prior utterance, which helps the speaker successfully achieve communicative
purposes such as complaining, arguing, mocking, blaming, and persuading, etc.
Second, (l)ul-marking is used to initiate repair on the other speaker’s prior utterance which
goes contrary to the speaker’s belief or expectation. In repair situations, the speaker may express
surprise or disbelief about what has been said and ask for a clarification using a partial repeat of
the other speaker’s prior turn together with the adverbs cengmal or cincca ‘really’ or a candidate
understanding of the prior utterance plus the phrase -nun mali-/iyaki- ‘you mean.’ A partial repeat
of the trouble source turn with the emphatic sentence ending -ko ‘Are you saying~’ and the (l)ul-
marked question word mwel plus a partial repeat of the other speaker’s previous utterance can
also be used to challenge the appropriateness or accuracy of the information in the previous turn
and to ask for the other speaker’s self-correction or backdown. In repair situations, (l)ul-marking
facilitates the listener’s selective focusing on the information involving the referents that need to

140

be modified or corrected. As a result, interlocutors can re-build shared understanding for
successful communication.
Next, (l)ul-marking has a pragmatic function of distinguishing two or more contrastive
referents, especially in challenging contexts. When the referent of an object NP is dispreferred,
challenged, or rejected by another referent in a discourse context, (l)ul can be used to mark those
two contrastive referents. This not only helps the hearer easily distinguish contrastive referents
but also emphasizes the speaker’s assertion challenging the prior utterance or action involving the
marked referents, which may contribute to the listener’s successful interpretation of the speaker’s
intended meaning.
Lastly, (l)ul-marking consistently appears when the information introduced in the object NP is
continuously being talked about in conversation. If the speaker wants to talk about something that
the hearer might not know, it would be hard for the hearer to actively participate in conversation
due to the lack of shared knowledge. Then, the speaker needs to build shared knowledge so that
the listener can understand what the speaker is talking about without difficulty or
misunderstanding. In this case, (l)ul-marking is used to indicate that further descriptions or
explanations about the discourse topic will be provided in the subsequent context. The
conjunctive -nuntey, which indicates a warranted continuation relation between the preceding
portion of the utterance as the background/topic and the up-coming portion as the
elaboration/comment, often co-occurs with the (l)ul-marked discourse topic.
The findings about the pragmatic functions of (l)ul and co-occurring linguistic forms or
patterns for each pragmatic use are used to explain the two controversial grammatical phenomena
of
“Substitution,” which refers to the speaker’s choice of using -(l)ul in the place of other
particles, and the so-called “Double Object Construction (DOC).” It has been found that
substitution, stacking, and double (l)ul-marking are used by the speaker to direct the listener’s
attention to either hard-to-process or salient information that the listener has to identify or process.

141

The analysis of actual spoken data has discovered that substitution, stacking, or double (l)ul-
marking is consistently used by speakers in certain discourse contexts to perform the following
pragmatic functions.
(i) Emphasizing an assertion challenging the prior utterance or action
(e.g., wondering, complaining, arguing, mocking, blaming, regretting, persuading, etc.)

(ii) Initiating repair on the information involving the referent that needs to be clarified,
corrected, or modified

(iii) Distinguishing contrastive referents in discourse contexts of disagreement

(iv) Indicating a discourse topic and projecting further descriptions or explanations
about the discourse topic in the subsequent context

142

6.1 Extended Discourse Context of Disagreement
This dissertation provides detailed descriptions and explanation about the linguistic forms or
patterns that frequently co-occur with each pragmatic use of object-marking in discourse contexts
of disagreement, which can be extended to other linguistic forms or patterns expressing the
speaker’s disagreement about the prior utterance or action. The sentence ending -ney is used to
express the speaker’s opinion about what he or she perceives. More specifically, the speaker’s
opinion is contradictory or contrary to what he or she had thought or anticipated. For instance,
(130) John-i pap-ul mek-ney!
John-SUB meal-OBJ eat-SE
‘Oh, John is eating!’
[Implied: I can’t believe he’s eating!]
(130) can be uttered in a certain context where there is a conflict between the speaker’s belief or
expectation and the current situation. For example, the speaker knows John has not been able to
eat much because he is sick, so the event of John’s eating is not what the speaker had anticipated.
In this case, -(l)ul is likely to be used for the object NP pap ‘meal’ to direct the listener’s attention
to this salient information which is contrary to the speaker’s belief or expectation. Thus, (l)ul-
marking helps the speaker to more strongly express the intended meaning conveyed through the
utterance.
Similarly, (l)ul-marking is likely to co-occur with the adverb ta showing the speaker’s
attitude towards an unexpected situation which goes contrary to the speaker’s belief or
expectation. The original meaning of the adverb ta is ‘completely’ or ‘all,’ but it can also indicate
that things go in an unexpected direction. Thus, the adverb ta expresses the speaker’s surprise,
wonder, or sarcastic attitude towards an unexpected event which goes contrary to the speaker’s
belief or expectation. See (131).
(131) weyniliya ne-ka chengso-lul/ ??-ø ta ha-ko?
what a surprise you-SUB cleaning-OBJ/ (OBJ) all do-SE
‘What a surprise! What makes you do cleaning?’
[Implied: I can’t believe you’re cleaning up your room!]

143

The expression weynil and the adverb ta indicate that the event described in the sentence is
unusual and unexpected for the speaker. The speaker knows that the hearer rarely cleans up his
room. Thus, (131) can be uttered when the speaker challenges the other speaker’s prior action of
cleaning his room by showing surprise or even a sarcastic attitude. In this challenging context, the
object NP chengso ‘cleaning’ in ha-compound ‘to do Noun’ is (l)ul-marked to direct the listener’s
attention to this salient NP. Thus, the speaker’s assertion challenging the other speaker’s prior
action is emphasized and hence successfully delivered to the listener: “I can’t believe you have
cleaned your room!” In this particular context, interestingly, (l)ul-marking seems to be not just
preferred but obligatory. If the object NP chengso ‘cleaning’ is left unmarked, the sentence
sounds very unnatural.
The speaker’s choice of using (l)ul-marking can also be extended to broader discourse
contexts where the speaker disagrees with or challenges the prior action of the other speaker or
sometimes that of the speaker himself. See (132).
(132) mwe ilen ke-l/ ??-ø ta
oh like this thing-OBJ all
‘Oh, why do you bother bringing such things?’
[Implied: You shouldn’t have!]
When receiving a gift, it is customary to politely refuse a gift at first by saying mwe ilen kel ta
“Why do you bother bringing such things?” as in (132). In this utterance, the speaker is
challenging the other speaker’s action of presenting a gift. The adverb ta ‘all’ indicates that the
event described in the sentence is unusual and unexpected for the speaker. In addition, the object
NP ilen ke ‘such things’ is (l)ul-marked to direct the listener’s attention to the NP which goes
contrary to the speaker’s expectation. As a result, the speaker’s assertion challenging the other
speaker’s prior action is emphasized and hence successfully delivered to the listener: “You
brought me a gift? You shouldn’t have!” Like the example above, (l)ul-marking seems to be not
just preferred but obligatory in this context. If the object NP ilen ke’ is left unmarked, the
sentence sounds very unnatural.

144

Similarly, when a compliment is given, the recipient would typically respond with a denial.
(133) A: swuko-ha-sy-ess-supnita
effort-do-HON-PST-SE
‘Thank you for your effort.’

B: mwe-l-yo
what-OBJ-POL
‘Not at all.’
For example, when someone expresses gratitude for your hard work as in (133), you would
humbly say mwelyo “Not at all” to show that you are totally undeserving of such praise. The
question word mwe is overtly (l)ul-marked, which can be literally translated, “Be thankful for
what?” This conveys the speaker’s assertion challenging the other speaker’s prior action: “Please
don’t be thankful for what I have done. It was nothing.” Interestingly, (l)ul-marking also seems to
be obligatory in this case.
When the speaker’s intended meaning conveyed through the utterance is to challenge the prior
action of the other speaker, such as blaming, as in (134), or that of the speaker himself, such as
regretting as in (135), (l)ul-marking can be often used.
(134) Blaming
a. keki-l ka-myen ettekhay
there-OBJ go-if do what
‘What am I supposed to do if you go there?’
[Implied: You should not have gone there.]

b. keki-l ka-myen toy-keyss-nya?
there-OBJ go-if become-will-Q
‘Do you think you’re allowed to go there?’
[Implied: You should not have gone there.]

(135) Regretting
emma ma-lul tul-ess-eya hay-ss-e
mom word-OBJ listen to-PST-have to-PST-SE
‘I should’ve listened to my mom.’

145

6.2 Interaction between the Form, Intonation, and Meaning
Further research is required to determine the interaction of the grammatical form and intonation
in the interpretation of the speaker’s intended meaning through utterances. For instance, wh-
words in Korean can function either as a wh-pronoun in a wh-question or as an indefinite pronoun
in a yes/no question as in (136)-(138).
(136) Information-seeking wh-question
eti ka?
where go
‘Where are you going?’

(137) Yes/no question
eti ka?
where go
‘Are you going somewhere?’

(138) Rhetorical question
kaki-n eti-l ka?
go-TOP where-OBJ go
‘Where are you going?’
[Implied: Don’t go!]
The wh-word eti can function either as the wh-pronoun ‘where’ in an ordinary information
seeking wh-question as in (136) or as the indefinite pronoun ‘somewhere’ in a yes/no question as
in (137). The question eti ka seems to be ambiguous between “Where are you going?” and “Are
you going somewhere?” but the two different types of questions can be distinguished by their
sentence-final intonation: The wh-question ends in a falling intonation while the yes/no question
ends in a high rising intonation. Then, how can the speaker distinguish these two types of
questions from rhetorical questions with (l)ul-marked elements as in (138)? Would rhetorical
questions have a different sentence-final intonation from ordinary information-seeking questions
or yes/no questions? The investigation of this issue would provide more comprehensive
mechanism controlling the speaker’s use of object-marking and the realization of the speaker’s
intended meaning.
Next, it would be worth examining the role of intonation in the interpretation of wh-questions

146

either as an ordinary information-seeking question or as a rhetorical question. See Excerpt 22, a
telephone conversation between two female friends, Ann and Betty.
Excerpt 22: Ann (A) and Betty (B)
Ann told Betty about his son Dongbin’s mischievous behavior in her friend’s house. Ann was
invited to her friend’s house and her son Dongbin was caught drinking a bottle of perfume
there.

1 A: kuntey tongpini-ka ku cip hyangswu-lul
but Dongbin-SUB tha thouse perfume-OBJ
But in their house,

2 cokuman ke-lul mek-ess-ta
small thing-OBJ eat-PST-SE
Dongbin drank a small bottle of perfume

à 3 B: mwe-l meke
what-OBJ eat
He drank what?

4 A: hyangswu-lul mek-ess-e hyangswu
perfume-OBJ eat-PST-SE perfume
a bottle of perfume

à 5 B: hyangswu-lul ettehkey meke
perfume-OBJ how eat
How could he drink a bottle of perfume?

6 A: ci-ka kunyang umlyoswu kathi mek-ess-ta-nikka
self-SUB just drink like eat-PST-SE
He drank it like he was drinking soda
Ann tells Betty that her son Dongbin got caught drinking a bottle of perfume at her friend’s house
(line 1-2). Betty’s question in line 3, mwel meke “He drank what?” can be interpreted either as a
question confirming what she has heard, or as a challenge displaying surprise or disbelief
concerning the action described in the previous turn. Similarly, Betty’s another question in line 5
can be ambiguous either as an ordinary information-seeking question, “Please describe to me how
he drank a bottle of perfume” or as an assertion challenging Dongbin’s prior action, “There is no
way he could drink a bottle of perfume.” Furthermore, it would be also worth investigating
whether there are any other factors affecting the interpretation of wh-questions either as
information seeking questions or as rhetorical questions.

147

6.3 Korean Learners’ Use of (l)ul-Marking
The findings of this dissertation can be used to develop more effective and attuned teaching
strategies for English-speaking L2 learners of Korean because they give us a deeper and
expanded understanding the pragmatic functions of the so-called Korean object particle (l)ul. The
role of (l)ul-marking in spoken Korean is more than the indicator of the object in a sentence
which receives the action of a verb. It is a device to facilitate the listener’s selective focusing on
either difficult-to-process or salient information that requires the listener’s attention. It also
performs a number of pragmatic functions, such as emphasizing an assertion challenging the prior
utterance or action; initiating repair on the information involving a referent that needs to be
clarified, corrected, or modified; distinguishing contrastive referents; indicating a discourse topic;
and projecting further descriptions or explanations about the discourse topic in the subsequent
context.
Because English lacks this type of marker, English-speaking learners of Korean tend to have
difficulty grasping its nature and uses. Moreover, since (l)ul has more than one pragmatic
function, describing its main functions and co-occurring linguistic forms or patterns for each
pragmatic use of (l)ul is important work. Therefore, the findings of this dissertation can be
applied to classroom settings todirectly help learners understand that (l)ul-marking is not merely
an optional choice, but is actually preferred in certain discourse contexts, and also allow them to
discover the pragmatic uses of -(l)ul in actual conversation.
Moreover, for future research it would be worth investigating the differences between Korean
native speakers and English-speaking learners of Korean with respect to their use of (l)ul-marking
in informal conversation. Because English-speaking learners of Korean may not have a
comprehensive understanding of the pragmatic functions of (l)ul in spoken discourse, one could
hypothesize that zero-marking would be more likely to be used. Korean native speakers, on the
other hand, would likely show more variations in their use of (l)ul-marking because they would

148

know in what discourse contexts the use of zero- or (l)ul-marking would be preferred.

149

References
Aissen, J. (2003). Differential Object Marking: Iconicity vs. Economy. Natural Language &
Linguistic Theory, 21 (3), 435-483.
Bolinger, D. (1957). Interrogative structures of American English. Tuscaloosa: University of
Alabama Press.
Brown, G. and Yule, G. (1983). Discourse Analysis. Cambridge University Press
Chafe, W. (1976). Givenness, contrastiveness, definiteness, subjects, topics, and point of view. In
Li (Ed.), New York: Academic Press, 27-55.
Fry, J. (2003). Ellipsis and wa-marking in Japanese conversation. New York, N.Y: Routledge.
Fujii, N., & Ono, T. (2000). The occurrence and non-occurrence of the Japanese direct object
marker o in conversation. Studies in Language, 24 (1), 1-39
Givón, T. (1990). Syntax: a functional-typological introduction. Vol.2. Amsterdam, Philadelphia:
John Benjamins.
Jackendoff, R. (1972). Semantic Interpretation in Generative Grammar. Cambridge, Mass. MIT
Press.
Jaeger, T. F. (2010). Redundancy and reduction: Speakers manage syntactic information density.
Cognitive Psychology, 61, 23-62.
Han, J. (1999). Grammatical Coding of Information Structure in Korean: a Role & Reference
Grammar (RRG) Account. Doctoral dissertation, State University of New York at Buffalo.
Hong, K-S. (1991). Argument Selection and Case Marking in Korean. Ph.D. dissertation in
Stanford University.
Jung, H-J. (1988). ‘-ey’lul cwungsimulo pon thossiuy uymi: ‘-ey’wa ‘ko, lul’uy uymi pikyo.
Korean Linguistics, 17. The Society of Korean Linguistics.
Kim, H-Y. (1996). Functions of the verbal affix -nuntey in Korean conversation. Language
Research, 32 (4), 637-663.
Kim, M-J. (2001). Accusative adverbials in Korean: Delimiting phrase and case. Ms., University
of Massachusetts-Amherst.
Kim, T-H. (2008). Subject and Object Marking in Conversational Korean. Doctoral Dissertation.
State University of New York.
Kiss, K. (1998). Identificational Focus versus Information Focus. Language 74 (2), 245-273.
Ko, E-S. (2000). A discourse analysis of the realization of objects in Korean. Japanese/Korean
Linguistics, 9, 195-208. Stanford: CSLI Publications
Koshik, I. (2005). Beyond rhetorical questions: assertive questions in everyday interaction.
Amsterdam: John Benjamins.
Kuno, S. (1973). The structure of the Japanese language. The MIT Press.
Kurumada, C., & Jaeger, T.F. (2013). Communicatively efficient language production and case-
marker omission in Japanese. Proceedings of the 36
th
Annual Meeting of the Cognitive
Science Society.
Hong, J-S. (1986). Kyocha Cangsopoe Kwumwun Yenkwu. Hangul, 191, 53-76.
Hong, J-S. (1987). Hyentay Hankwuke Tongsakwumwunuy Yenkwu. Top Publishing Company.
Lee, C-M. (1995). Definite/ specific and case marking in Korean. Theoretical Issues in Korean
Linguistics. CSLI publications, Stanford, 245-341.
Lee, H-S. and Thompson, S.A. (1987). A Discourse Account of the Korean Accusative Marker.
Santa Barbara Papers in Linguistics, 1, 23-49.
Lee, H-J. (2010). Explaining variation in Korean case ellipsis: Economy versus iconicity. Journal
of East Asian Linguistics, 19 (4), 291-318.
Lee, I-S., & Im, H-B. (1996). Korean syntax. Hakyensa.

150

Lee, I-S., & Ramsey, S. R. (2000). The Korean Language. State University of New York Press.
Masunaga, K. (1988). Case deletion and discourse context. In papers from the Second
International Workshop on Japanese Syntax, edited by W. J. Poser. Stanford: CSLI. 145-154.
Matsuda, K. (1995). Variable zero-marking of o in Tokyo Japanese. Doctoral Dissertation.
University of Pennsylvania.
O’Grady, W. (1991). Categories and case: The sentence structure of Korean. Amsterdam: John
Benjamins.
Paltridge, B. (2006). Discourse Analysis: An Introduction. London: Continuum.
Piantadosi, S. T., Tily, H. and Gibson, E. (2011). Word lengths are optimized for efficient
communication. Proceedings of the National Academy of Sciences, 108
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London: Longman.
Robinson, J. D., & Kevoe-Feldman, H. (2010). Using full repeats to initiate repair on others’
questions. Research on Language and Social Interaction, 43, 232-259.
Schutze, C. T. (2001). On Korean" Case stacking": The varied functions of the particles ka and
lul. Linguistic review, 18 (3), 193-232.
Shin, H-S. (1982). A study on accusative case maker -lul. Korean Journal of Linguistics, 7,
119-139
Sohn, H-M. (1999). The Korean Language. New York: Cambridge University Press.
Tsutsui, M. (1984). Particle ellipsis in Japanese. Doctoral Dissertation. University of Illinois at
Urbana-Champaign.
Watanabe, Y. (1986). Two kinds of ellipsis in Japanese discourse: A Quantitative text study.
Studies in Language, 10 (2), 337-351.
Widdowson, H.G. (2007). Discourse Analysis. Oxford University Press.
Wunderlich, D. (2014). Variations of double nominative in Korean and Japanese. Studies in
Language and Cognition, 339.
Wechsler, S., & Lee, Y-S. (1996). The domain of direct case assignment. Natural Language and
Linguistic Theory, 14, 629-664.
Yang, H-J. (2007). Grammar as interactional resources in spontaneous Korean conversations.
Doctoral Dissertation. University of Colorado.
Yoon, J-H. (2004). Non nominative (major) subjects and case stacking in Korean, Typological
Studies in Langauge 61 (2), 265-314.
Zipf, G. K. (1949). Human behavior and the principle of least effort: An introduction to human
ecology. Addison-Wesley.

Abstract (if available)

Abstract This paper examines the speaker’s choice of using the Korean object particle -(l)ul in spontaneous informal conversation and clarifies that the choice of (l)ul-marking is a device to facilitate the listener’s selective focusing on either hard-to-process or salient information for successful communication. (l)ul-marking is frequently used to facilitate the listener’s selective focusing on difficult-to-process information which requires the listener’s extra attention (i.e., the object NP with a modifier, the object NP not immediately adjacent to its verb, and the object NP in sentences with non-canonical word order). ❧ In addition, (l)ul-marking is consistently used in discourse contexts where the speaker disagrees with what has been said in the prior talk. When there is an ongoing disagreement between the interlocutors, the speaker needs to direct the listener’s attention to the information that is contrary to the prior utterance or action of co-present or non-present parties. Disagreement can vary in range from weak disagreement, such as expressing surprise or wondering about the prior utterance or action that is contrary to the speaker’s belief or expectation, to serious disagreement such as showing a negative or sarcastic attitude (e.g., dispreference, disbelief, objection, mockery, etc.) or asking for the other speaker’s clarification, self-correction, or backdown of the prior talk. ❧ Thus, what the particle -(l)ul does in conversation is more than simply indicating the grammatical role of the NP in a sentence

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The notion of topic-comment constructions and the meaning of the Korean topic marker '-(n)un'

PDF

Knowledge and use of evidentials in L1 and L2 Korean

PDF

Discourse particles in Mandarin Chinese

PDF

Vietnamese sentence final particles

PDF

A corpus-based discourse analysis of Korean grammatical constructions: Focus on the multifold functions and meanings of the pragmatic construction e kaciko

PDF

Vietnamese pronouns in discourse

PDF

The theory of empty noun in Chinese: with special reference to the right node raising construction

PDF

Exploring the effects of Korean subject marking and action verbs’ repetition frequency: how they influence the discourse and the memory representations of entities and events

Asset Metadata

Creator Na, Won Kyung (author)

Core Title The speaker's choice of using object-marking in informal spoken Korean

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program East Asian Languages and Cultures

Publication Date 08/04/2016

Defense Date 06/22/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag challenging the prior utterance or action,Disagreement,discourse analysis,double object construction,OAI-PMH Harvest,object particle

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kim, Namkil (committee chair), Li, Audrey (committee member), Simpson, Andrew (committee member)

Creator Email wonkyoungna@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-297314

Unique identifier UC11281218

Identifier etd-NaWonKyung-4744.pdf (filename),usctheses-c40-297314 (legacy record id)

Legacy Identifier etd-NaWonKyung-4744.pdf

Dmrecord 297314

Document Type Dissertation

Format application/pdf (imt)

Rights Na, Won Kyung

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA