Indexicality and honorific speech level choice in Korean.
Strauss, Susan ; Eun, Jong Oh
Abstract
The use of Korean honorifics is generally dependent upon such
social factors as age, profession, socioeconomic status, and so forth.
Traditional accounts of the two Korean honorific verbal suffixes,
namely, the deferential and the polite forms, explain the use of each on
the basis of relative status: the deferential is the more formal of the
two, used when addressing persons of higher social status; the polite
form is used when addressing persons of equal or higher status, but not
so high as to require the deferential form. However, cursory examination
of naturally produced oral discourse reveals that speakers often
alternate between the two forms within the same stretch of talk and
while addressing the same interlocutor. Using a database of
approximately eight hours of naturally occurring speech from a variety
of discourse genres, this article proposes an alternative analysis of
the two honorific speech levels. Rather than the static, relative status
account put forth by traditional linguistic and sociolinguistic views,
we propose instead that these forms differ in terms of the semantic
feature of +/-BOUNDARY vis a vis speaker and interlocutor and their
respective domains of cognition and/or experience. That is, when
speakers use the deferential form (+BOUNDARY), they index a stance of
EXCLUSION with the interlocutor, such that the interlocutor is
positioned as outside the sphere of the speaker's cognitive and/or
experiential domains," discourse marked with the deferential form
is thus framed as detached, objective, and authoritative. In contrast,
when speakers use the polite form (-BOUNDARY), they index a stance of
INCLUSION. Essentially, then, the deferential form creates bounded
distance between speaker and addressee, while the polite form
establishes and/or reinforces common ground.
1. Introduction (1)
According to Sohn (1994, 1999), Korean is a typical honorific
language, the patterns of which are extraordinarily systematic, if not
perhaps the most systematic of the world's known languages (Sohn
1994: 7, 1999: 17). And while these patterns are indeed systematic, they
are also quite complex and elaborate, deriving from an array of lexical
and morphosyntactic markers--markers such as honorific and humble
pronouns, honorific address terms, titles, vocative suffixes, verbal
infixes, honorific lexical items (i.e. nouns, verbs, and case marking
particles), as well as a set of honorific verbal suffixes. The
linguistic elements comprising the honorific system for Korean are
summarized below in Figure 1.
[FIGURE 1 OMITTED]
The various combinations of these markers serve to express varying
degrees of politeness, deference, and other social attitudes toward
interlocutors, as well as toward certain subject and object referents.
Essentially, the use of these forms and speech levels is largely
dependent upon such factors as age, gender, profession, and status, as
they apply to speaker, addressee, and object/subject referent. Simply
put, when speakers address or refer to others who rank higher in one or
more of these general categories, their speech contains some aspect of
honorific marking.
Examples (1) and (2) are invented examples to illustrate the
potential complexity of the Korean honorific system as encapsulated in
Figure 1. With respect to honorific marking, example (1) contains only
the polite sentence ender -yo, and could be uttered by an adult to a
status equal about a person whose status is either equal to or lower
than that of both speaker and addressee.
(1) Sumi cokum cen-ey cip-ey ka-ss-e-yo
Sumi little before-DAT house-to go-PST-H (POL)
'Sumi went home a little while ago.'
Example (2) is less simple. It contains four instances of honorific
marking, broadly indicating what we will refer to here as "status
differential" (2) between the speaker and the subject referent,
Professor Kim, as well as between the speaker and the addressee.
(2) Kim kyoswu-nim cokum cen-ey tayk-ey
Kim Professor-H (T) little before-DAT house (HLI)-DAT
ka-sy-ess-supnita.
go-H (VI-)PST-H (DEF)
'Professor Kim went home a little while ago.'
The four honorific categories exhibited in (2) are as follows: the
deferential honorific title -nim (Sohn 1994: 289), the honorific lexical
item for 'house,' tayk, the verbal infix -si-, and finally,
the deferential verbal suffix -(s) upnita.
The present article centers on the subset of honorific verbal
suffixes in Figure 1, that is, the polite form (-yo) and the deferential
form (-[s]upnita), with a view to determine with greater precision what
it is that motivates speakers' choice of one form over the other.
This question arises from the simple fact that, even with only a cursory
look at naturally occurring speech in Korean, speakers alternate between
the two forms within the same discourse, in spite of the fact that their
addressees remain constant. In fact, we find that speakers often shift
freely from one predominant form to the other and back again within a
very short spate of talk and while addressing the same interlocutor or
audience. It is our claim that such shifting occurs with a striking
systematicity, and we will attempt to account for this seeming
regularity throughout the remainder of the article.
Based on a close, micro-level analysis of a corpus of authentic
oral data produced by a variety of speakers in a variety of contexts, we
will demonstrate that inherent social factors alone cannot account for
the deployment of these forms in discourse. Further, we will illustrate
that through the use of the two forms, speakers are indexing particular
stances vis a vis their interlocutors by excluding the interlocutor(s)
from their personal spheres of experience and/or cognition or including
them within such spheres. We demonstrate that the pragmatic inferences
of EXCLUSION and INCLUSION emerge from the meaning signal of +BOUNDARY
and -BOUNDARY for the deferential and polite forms, respectively.
2. Speech levels in Korean--verbal suffixes
Sohn (1994, 1999) identifies six speech levels for Korean,
designating them as plain, intimate, familiar, blunt, polite, and
deferential, and indicates that each can cut across all four sentence
types (i.e. declarative, interrogative, propositive, and imperative) in
addition to the three sentential moods (i.e. indicative, retrospective,
and requestive). (3) A variation of Sohn's (1999) taxonomy for
declarative forms appears in Figure 2 below.
The forms in Figure 2 are explicitly ordered on the basis of
ascending degrees of formality, with the plain form expressing the
lowest formality level, and the deferential form, the highest.
Non-honorific speech styles (plain, familiar, intimate, blunt) are used
among intimates and in-group members, or in unequivocally downward
directions of address by the speaker to his or her interlocutor. The two
honorific speech styles, namely, the polite and deferential, are
prototypically used among non-intimate adults of relatively equal rank.
(See Sohn 1994, 1999; Lee and Ramsey 2000; Martin 1964; among others).
Overall, the polite form is used more frequently and in broader
contexts than the deferential form. Lukoff (1982: 122) refers to the
polite speech style as the "honorific ordinary style." Lee and
Ramsey (2000: 259) indicate that the "polite form is broadly used
in virtually any situation where polite language is called for. It is
the all-purpose style used with superiors and inferiors alike." Lee
and Ramsey cite specific interactional contexts where the polite form
would be typically used, such as those between store clerks and
customers, children to their parents, students to their teachers, and
even lost foreigners in Seoul receiving directions. Sohn (1994: 9)
refers to the polite level as the "informal counterpart of the
deferential level" and points out that it is the most commonly used
speech level of the six that he has categorized.
The deferential style, in contrast, appears in more narrow contexts
and more formal situations. It is the prototypical sentence ender used
in news broadcasts, lectures, and official announcements, as well as in
more interactive formal contexts in which inferiors address their
superiors (Sohn 1994, 1999; Lee and Ramsey 2000; Lukoff 1982).
According to some accounts, gender also seems to be a factor
influencing deferential and polite speech styles. That is, in other than
the formal oral discourse contexts as mentioned above, in which only the
deferential style is said to be used, the deferential and polite forms
are discussed as patterning somewhat differently in male speech and
female speech. In one elementary Korean textbook, (6) for example, there
is an added note that females use the informal style when meeting for
the first time, implying that males would opt for the deferential style
in the same context. Sohn (1994: 10) indicates that the deferential
level is "usually used by males" (with the implication that it
is not commonly used by females), and later points out that when males
do use the form, they tend to mix both the deferential and the polite in
the same discourse, while females tend to use only the polite speech
level (Sohn 1999: 271).
The fact that speakers do mix these forms, then, has been
recognized by linguists, as evidenced by Sohn's quote above. Martin
(1964: 411) also states that "a Korean may open up a conversation
with the deferential style, slip into the polite style, and then
occasionally throw in a deferential form." And Lee and Ramsey
(2000: 261) point out that speakers might mix these within the same
discourse while interacting with the same participants, depending upon
"the feel of the situation and the atmosphere that one wishes to
convey." These discussions are vague at best, and do not account
for the motivations behind such mixing, nor do they propose any type of
systematicity which might underlie it.
Descriptive rules for the use of honorific speech levels are based
entirely on sentence-level examples and examples that are either
invented or excised from the larger spates of discourse in which they
may have originally been uttered. Further, they rely heavily on
contrived scenarios to explain apparent inconsistencies, and ultimately
arrive at no unified method of accounting for the absolute richness and
complexity of the phenomenon in question. In fact, we find enough
exceptions to descriptive rules to inspire us to reassess the issue from
a new perspective. First, we find that adult speakers tend to frequently
mix polite and deferential forms in ways that are more systematic than
simply creating a particular "feel" of a situation or a
certain "atmosphere." Second, we find that females do use the
deferential form, also in ways not consistent with previous accounts.
And third, we find instances of the use of the deferential form where
the direction of address is the polar opposite of what is reported in
previous accounts--that is, we find its use by higher status speakers to
addressees of unequivocally lower status. These inconsistencies motivate
us to question whether in fact it is strictly social factors and
concomitant issues of degree of formality which underlie the honorific
speech levels or whether something different may actually be at work. We
will be building on our earlier work motivated by the same
inconsistencies (Eun and Strauss 2004), and hope to provide a unified
analysis of this phenomenon in Korean. (7)
3. Deference, politeness, and Korean speech levels
Honorifics and honorific systems of language are commonly treated
in linguistic and sociolinguistic literature under the broad rubric of
deference (Brown and Gilman 1960; Martin 1964; Brown and Levinson 1987;
Hwang 1990; Duranti 1997; Agha 1994, among others). Treichler et al.
(1984: 65) define deference as "power as a social fact, established
a priori by the differential positions of individuals or groups within
the social structure." In this sense, what underlies deference is a
set of social, interactional, and political stratifications whereby
interactants exhibit an awareness of such power and such differential
through both verbal and nonverbal modes of communication. Deference is
manifested in its near-absolute form through such ritualistic behaviors
as a bow or other types of routinized greetings (Scott 1990: 24) and
more variably through the use of socially appropriate honorific language
(e.g. address terms, honorific pronouns and verb forms, distancing
markers, and so forth). These and other linguistic acts of deference may
be enacted simply because they are automatic and habitual, for the sake
of preserving a sense of conformity and social convention, for the
purpose of achieving some element of personal gain, or out of sincere
respect for the person in the superior position (Scott 1990: 24).
And because deference is rooted within systems of social role and
hierarchy, its underpinnings and enactments are distinct from those of
politeness. That is, inherent in these socio-interactional systems are
sets of conventions and expectations, the interactants' response to
which are essentially limited to sets of repertoires of ritual and
routine, and to closed, finite sets of honorific/deferential markings,
with some but little room for variation. Politeness, on the other hand,
at least according to the essential principles of the theory set forth
by Brown and Levinson (1987), is couched in a universal, abstract
psychosocial concept of "face" or more specifically in its
underlying bifurcation into negative face wants (the desire to be
unimpeded and/or unintruded by others) and positive face wants (the
desire to be approved of and/or accepted by others).
On the basis of Brown and Levinson's model, certain acts are
intrinsically face-threatening to the speaker's or addressee's
positive face, negative face, or to both poles at once. A prototypical
speech act that constitutes a face-threatening act (FTA) to the
hearer's positive face is a criticism or scolding; one that
constitutes a threat to the hearer's negative face is a demand,
order, or request. According to this model, speakers and hearers in
interactional contexts select from a number of strategies which could
affect the production of such an FTA, ranging from simply proceeding
with its execution, amplifying the severity of the FTA, mitigating the
severity of the FTA, or choosing not to produce it at all. Politeness,
then, according to this model and others deriving from it, emerges from
personal relationships between speakers and addressees and their varying
sensitivities and attention to mutual face wants--not from inherent
social roles. Thus, potential strategies, enactments, and
representations of politeness, both linguistic and not, are fluid,
choice-driven, dynamic, and virtually unlimited in number (Brown and
Levinson 1987; Hwang 1990; Agha 1993, 1994).
Matsumoto (1988: 415) presents an example to illustrate the
interplay of face, politeness, and honorifics for the case of Japanese,
though Matsumoto's explanation actually centers on deference, not
politeness or face. By using three variations of the simple declarative
sentence "Today is Saturday" in Japanese, an utterance
essentially devoid of face-threatening potential in its own right,
Matsumoto underscores the importance of social context in languages with
morphosyntactic honorific systems. The sentence triplet appears in
example (3) below.
(3) a. Kyoo wa doyoobi da. [non-honorific]
Today TM Saturday COPULA-PLAIN
b. Kyoo wa doyoobi desu. [honorific]
Today TM Saturday COPULA-POLITE
c. Kyoo wa doyoobi de gozai masu [honorific]
Today TM Saturday COPULA-SUPER POLITE
'Today is Saturday.'
(from Matsumoto 1988: 417)
The (a) version of the sentence contains the use of the copula in
the plain form; the (b) and (c) versions contain the copula used with an
honorific form (i.e. the so-called polite and super-polite,
respectively). Matsumoto's point here is that each version takes
into account the social relationships between speakers and addressees or
writers and readers. That is, the version in (a) would be an appropriate
form for expository writing or for casual conversation between persons
in a close relationship. The (b) version has the broadest use of the
three and could be uttered to a stranger, an acquaintance, or to a
person of a higher social status than the speaker, while the (c) version
would be restricted to a formal interaction. What is crucial in the
choice of one form over the other two, according to Matsumoto, is the
"speaker's assessment of the social context"--the very
factor underlying the concept of deference discussed earlier. In other
words, when the speaker deems that his/her relationship with the
interlocutor calls for honorific marking, then the honorific suffix(es)
should be used. "If [the speaker] were to choose intentionally from
a form that the addressee would not expect, it will cause either
incomprehension or an implication that would lead to embarrassment and
loss of face (Matsumoto 1988: 416)." (8) Again, the
"choice" is not really a choice on the part of the speaker; it
is essentially an obligation, and one that is driven by static,
macrosociological forces such as power or status differential.
We will now complexify the issue of deference and politeness by
viewing the utterance in a way not discussed (nor intended) by
Matsumoto. Superficially, the utterance "Today is Saturday" is
indeed polarity neutral with respect to positive or negative
face-threatening potential. However, what is not mentioned by Matsumoto
is that if the speaker is correcting his/her interlocutor, who may have
just said something like "Today is Sunday," the pragmatic
force of the utterance "Today is Saturday" could actually be a
threat to the interlocutor's positive face, since it could serve as
a correction or repair initiator (Sacks et al. 1974; Schegloff et al.
1977) to a prior statement. By the same token, if the speaker is
reminding the interlocutor of a previously made commitment that the
interlocutor had promised to carry out on Saturday, the utterance could
easily be construed as a threat to the interlocutor's negative
face, since it obligates him/her to make good on the promise or to
otherwise find a way out of it. (9) Thus, a statement in any language
which contains lexicopragmatic systems of marking honorifics and
deference can be deferential and polite in some contexts and deferential
and impolite in others, at least with respect to the benchmarks of
face-related politeness established in the Brown and Levinson model.
In the case of honorific speech levels in Korean, we clearly
recognize that social stratification is an inherent factor when it comes
to selecting an honorific form (i.e. polite or deferential) over
non-honorific ones (plain, intimate, familiar, blunt). We equally
recognize that aspects of deference, from a social point of view, and
aspects of politeness, from an interpersonal one, crosscut the whole of
interaction in Korean society. However, the two honorific forms under
investigation seem to pattern in actual discourse in a far more complex
way than previous literature on deference and politeness has addressed.
We will be adding some new dimensions to these issues as they pertain specifically to Korean honorific speech levels, since the alternation between its two honorific forms is only marginally predictable from the
point of view of deference as previously discussed, and not at all
predictable from the point of view of politeness. What we will be
proposing here instead is that the choice of one so-called honorific
form over the other seems to be squarely related to notions of degree of
relative sharedness of domains of experience and cognition between and
among interlocutors--concepts that have little, if anything, to do with
traditional concepts of deference and its attendant sources in macro
levels of social structure per se, or politeness and its attendant
sources in micro levels of interpersonal relationships and interaction.
4. Data and methodology
In order to more deeply investigate honorific speech level
alternation in Korean and to provide a more precise explanation of such
alternation, we have analyzed approximately eight hours of naturally
occurring oral data. Our database for this project consists of both
spontaneously produced and scripted speech. The spontaneous data include
informal conversations between graduate students, excerpts from an
English language classroom lesson in a South Korean middle school, a
radio talk show, two television talk shows, and an English language
instruction television broadcast. The scripted speech includes a
twenty-minute segment of an MBC news broadcast, a collection of 70
television commercials, four segments of the children's television
program "Ppo Ppo Ppo," and three religious sermons. (10)
Interwoven within this corpus are instances of monologic speech,
conversational dialogue, announcements, and personal narratives, all of
which are significant in terms of illuminating the complex patterns of
speech style alternation within extended discourse.
The data were transcribed according to the conventions of
Conversation Analysis (Atkinson and Heritage 1984) and Romanized using
the Yale system. We coded and counted all instances of the polite form
as it appeared with verbal predicates (11) and all instances of the
deferential form. In addition, we have isolated a number of modals and
other interactional particles which pattern systematically with the
target forms.
Table 1 below presents a simple, quantitative illustration of the
frequency and distribution of the target forms throughout the data.
As we can see from the table, none of the datasets evidences an
exclusive use of one form over the other--testimony to the fact that
discourse genre or social stratification alone is not a reliable
predictor of honorific sentence enders in Korean. (13) However, the news
broadcast (dataset 10) does contain near exclusive use of the
deferential, and the "earthquake data" (dataset 1), near
exclusive use of the polite. In this respect, each of these datasets
patterns in complete accordance with the current literature.
The news broadcast represents a formal, highly stylized genre of
public discourse, and as such, it is not surprising that every finite
verb in the entire dataset is inflected with deferential verbal
morphology. (14) Example (4) below presents three of the 22 segments in
the broadcast:
(4) MBC News, 19 November 2002, 6:00 p.m. edition [excerpts]
Segment 2
Male
Anchor: hakkyo swuep-ey tayhan haksayng-tul-uy pwulman-i
School lesson-DAT about student-PL-GEN discontent-SM
kyengcey hyeplyek kikwu OECD hoywen-kwuk kawuntey
economy cooperation organization OECD member-country
among
kacang noph-un kes-ulo nathana-ss-supnita
most high-ATTR NML-INST appear PST-DEF
'It is reported [DEF] that the level of discontent among
Korean students' is the highest among member countries of the
Organization for Economic Cooperation and Development (OECD)"
Segment 9
Male
Anchor: o-nun ichen-sa-nyen-puthe catongcha soyuca-tul-un
Come-ATTR 2004 year from car owner-PL-TM
tay-mwul pohem-ey pantusi kaip-hay-ya ha-mye
about-things insurance-LOC definitely join-do-have to do-
CONN
conghap pohem-ey kaip-ha-ci anh-un salam-tul-i umcwu-na
composite insurance-LOC join-do-COMM not-ATTR person-PL-SM
drinking-or
mwu-myenhe sako-lul nay-ss-ul ttay chwuka pwutam-ul hay-ya
no-license accident-OM cause-PST-ATTR when extra burden-
OM do-have to
ha-nun caki pwutam-kum ceyto-ka toip-toy-pnita.
Do-ATTR self burden-money system-SM introduction-
become-DEF
Kyucey kayhyek wiwenhoy-nun onul catongcha sako
phihayca-ey tayhan
Regulation reform committee-TM today car accident victim-
LOC about
poho-lul kanghwa-haki wihay-se i-kathun nayyong-uy
catongcha
Protection-OM enhance-do-NML in order to this-like
content-GEN car
sonhay-paysang pocang-pep kayceng-an-ul simuy uykel-hay-ss-supnita.
Damage-compensation guarantee-law revision-proposal-OM
Review pass-do-PST-DEF
kayceng-an-un tay-mwul sako-ka na-ss-ul ttay posang-i
Revision-proposal-TM about-things accident-SM appear-PST-
ATTR when compensation-SM
ceytaylo illwu-e cil swu iss-tolok conghap-pohem-ey kaip-ha-ci
properly achieve-CONN become can exist-so that composite
insurance-LOC join-do-CONN
ahn-un chalyang soyuca-tul-to motwu ilceng-ayk isang-uy
not-ATTR vehicle owner-PL-also all certain-amount over-
GEN
tay-mwul pohem-ey pantusi kaip-ha-tolok hay-ss-supnita.
about-things insurance-LOC definitely join-do-so that do-
PST-DEF
From the coming year 2004, car owners must purchase property damage
insurance for their vehicles and the "Extra Burden System"
will be instituted [DEF], which means that automobile owners who
don't have full coverage must pay an extra charge to cover property
damage incurred while driving drunk or while driving without a license.
The Regulatory Reform Committee reviewed and aspassedd [DEF] the
proposal today, revising the Guarantee for Automobile Damage such that
it would provide better protection for accident victims: Automobile
owners who don't have full coverage expected to pay [DEF] property
damage premium that exceeds the minimum amount (determined by the law),
in case they cause a collision involving other party property damage.
Segment 21:
Female
Anchor: Onul cungkwen sihwang ala-po-pnita. = Samseng cungkwen
-i-pnita.
Today securities market know-see-DEF Samsung
Securities COP-DEF
(We'll) find out [DEF] today's stock market (activity).
(This) is [DEF] Samsung Securities. ((The story is then continued by an
on-site reporter from Samsung.))
With respect to the polite form in dataset 1, the earthquake data,
we find an almost identical match between speech level production and
traditional accounts. The earthquake data consist of spontaneous
conversation and narratives exchanged between four graduate student
dyads in which speakers recount their experiences during a strong
earthquake which occurred in their neighborhoods just a few weeks prior
to the recording session. In all four cases, the participant pairs did
not previously know each other. Example (5) is an excerpt from a
female-female dyad:
(5) [Earthquake, dyad 4]
Sumi: cicin-i cheum na-ss-ul ttay (*) e:::
Earthquake-SM first come out-PST-ATTR-when uh:::
Hankwuk-ey-nun cicin-i an-na-canha-yo::.
Korea-LOC-TM earthquake SM NEG-come out-MDL-POL
Kulayse koyngeang-hi nolla-kwu, (*) mak (*) huntul-linuntey (*)
so quite-ADV surprise-CONN just shake-PSS-CONN
Cheum-ey-nun mak ku-key mwe-n-ci-to moll-ass-eyo.
First-LOC-TM just that thing what-NML-even not know-
PST-POL
Cicin-i-n-ci-to molu-kwu.
Earthquake-COP-ATTR-NML-even not-know-CONN
'When the earthquake first hit, uh::, we don't have [POL]
earthquakes in Korea, so I was quite surprised and it just shook, but at
first (I) had no idea what it was [POL]. I didn't even know it was
an earthquake, and'
Eun-Young: ung.
uh huh
Sumi: nemwu nollay-kaci-kwu (*) ku-ttay pam ca-ko iss-ess-canha-yo.
too much surprise-have-CONN that time night CONN-be-PST-MDL-POL
'I was so surprised and (*) I'd been sleeping [POL] that
night (i.e. the night before).'
The interactants in all four dyads are adults (graduate students)
of non-intimate, but relatively equal status; as such, the use of the
polite form is predictable, just as the use of the deferential form is
predictable in news broadcasts.
However, as noted in Table 1, it is not always the case that these
forms pattern with such consistency. The datasets exhibiting mixed
usages in Table 1, but which represent the strongest skew toward the
deferential are Sermon 2 by Pastor Cen Byeng Wuk (dataset 8, ratio of
0.43 to 1), and Sermon 3 by Pastor Ha Yong Co (dataset 9, ratio of 0.2
to 1). Moreover, each of these datasets was produced by a single speaker
within the context of formal public address where the congregation or
intended audience remained constant. In each case, while choice of
speech level in this genre of "sermon" tends to favor the
deferential, as one might expect on the basis of traditional
discussions, there still remains a significant number of tokens of the
polite form, that is, 49 in a 20-minute delivery and 57 in a 40-minute
delivery, respectively. Even more striking is the fact that Sermon 1
(dataset 7), also produced by Pastor Cen, contains a nearly equal ratio
of both forms (258 tokens of polite and 259 tokens of deferential, or a
1 to 1 ratio). Datasets 4 (English instruction television program) and 6
(television commercial data) also exhibit an almost equal distribution
of the two speech styles, or 0.96 to 1 and 1.41 to 1, respectively.
Thus, it is clear that speakers in various contexts do mix
honorific speech styles within a variety of discourse genres.
5. An alternative analysis: deferential and polite--semantic
signals of +/- BOUNDARY
As an alternative to the strictly politeness/deference-based
explanations, our analysis provides a perspective which views these
target markers as more fluid and contextually dynamic than that put
forth by traditional accounts. Speakers may or may not be aware of style
alternations in their own speech of what underlies these alternations.
Further, the fact that alternations occur back and forth from polite to
deferential and vice versa within the same extended tuna at talk or
within contiguous tunas reveals that accounts relating uniquely to
social variables and formality are neither sufficient nor accurate as
predictors for this phenomenon.
With respect to the deferential form, we propose that the semantic
feature +BOUNDARY emerges from the combination of morphemes which
comprise it. Recall from Figure 2 that the sentence ender -(s)upnita is
composed of three identifiable morphemes:
(6) DEFERENTIAL: mek-sup-ni-ta eat-AH-IN-DC
As seen in (6), the form is a complex one, including an addressee
honorific infix (-sup-), together with both the indicative (-ni-) and
declarative (-ta) markers. These morphemes combine to set the addressee
apart from the speaker; the form at once acknowledges the addressee as
socially distinct from the speaker and frames the utterance as an
objective assertion of declaration. Given this perspective, it becomes
all the more clear why -(s)upnita-marked utterances ate both expected
and appropriate in interactions between persons of markedly distinct
status. Further, utterances marked by the deferential verb ending are
detached and objective in nature, hence the predominant use of the form
in such formal discourse genres as news broadcasts as we observed in
example (4). Thus, it is from the original semantics of the compound
suffix, consisting of the three bound morphemes, that we claim the
feature of +BOUNDARY derives. (15)
In contrast, the semantic content of the so-called polite suffix
-yo is not as readily decomposable. (16) Its function, however, is also
essentially one of designating the concept of deference as discussed
earlier. That is, its use marks a status/power differential between
speaker and addressee, which becomes especially clear when it appears in
opposition with the four non-honorific forms noted in Figure 2. To
illustrate, we have isolated two of the entries from Figure 2 and
juxtapose the polite and intimate forms in (7).
(7) a. Honorific:
POLITE: mek-e-yo
eat-INT-POL
b. Non-honorific:
INTIMATE: mek-e
eat-INT
'I, you, she/he, we, they eat, are eating, etc.'
Here, the (a) and (b) elements differ only in the fact that the (a)
version contains the -yo suffix while the (b) version does not. Adult
speakers of Korean are exceptionally sensitive to the presence or
absence of -yo in direct interaction. That is, when non-intimate, nearly
equal status speakers interact, the lower status participant is expected
to use -yo-marked utterances to the higher status one. The use of the
intimate speech level, without -yo, in such a context could easily stir
negative feelings on the part of the addressee.
6. Language and indexicality
Our discussion of the signals +/-BOUNDARY, respectively, for the
deferential and polite verbal suffixes in Korean centers on the notion
of indexicality as put forth by Peirce (1955), Lyons (1977), Silverstein
(1976), Hanks (1990, 1992), and Ochs (1990, 1996), among others. Lyons
provides the following as the essence of Peiree's (1955) general
definition of "indexical": "There shall be some known or
assumed connexion between a sign A and its significatum C such that the
occurrence of A can be held to imply the presente of existence of
C" (Lyons 1977: 106).
According to both Peirce (1955) and Lyons (1977), what renders this
AC relationship an indexical one (and not a symbolic of iconic one) is
the fact that A does not mean C through an arbitrariness of
conventionality of the relationship between form and meaning (symbolic),
and that A does not resemble C in a natural and non-arbitrary manner
(iconic). An index of indexical, then, is distinct from a symbol or icon
in that the relationship between sigla and meaning is indirect, and one
that requires a contextual and/or spatio-temporal association--in
contrast with direct symbolic or iconic relationships where the sign
carries encoded meaning (symbol) on the one hand, or the sign and
meaning resemble each other through likeness, similarity, or analogy
(icon) on the other.
Prototypical indexicals include such language-specific deictic terms as demonstratives, pronouns, temporal adverbials, and directional/
locational adverbials, since the interpretation of their meanings is
inherently contingent upon the context and/or spatio-temporal
orientations of the participants at the time of production. This type of
indexical is referred to as "shifters" by Jespersen (1965) and
Jakobson (1971), and as "referential indexicals" by
Silverstein (1976). "Shifters" and "referential
indexicals" require attendant contextual information in order to
express meaning at all or to express the degree of precision intended in
their meaning by the speakers or writers who produce them.
Context-based studies on demonstrative reference in various
languages illustrate the richness and complexity of the sign-meaning
relationship within the subset of referential indexicality. Traditional
treatments of demonstratives generally appeal to the domain of physical
space and center on the speaker as the primary locus of referential
information, with "this" (and its functional equivalents in
other languages) denoting an entity that is 'near the speaker'
and "that" (and its functional equivalents) an entity that is
'far from the speaker' (Halliday and Hasan 1976; Lyons 1977;
Levinson 1983; Quirk et al. 1985; among others). However, discourse-and
context-based research has demonstrated dually that physical proximity
is not the sole determinant for speakers' choice of form and that
the speaker is not a priori the single and central locus for analysis.
Rather, demonstrative choice as analyzed within and across various
discourse contexts has been shown to transcend static space-relational
criteria and, more importantly, to reveal scalar relationships
concerning cognition, focus of attention, sharedness of information,
alignment with co-participants, and so forth (Kirsner 1979, 1993
[Dutch]; Hanks 1992 [Yucatec Maya]; Strauss 2002 [English], in prep.
[English and Korean]; Enfield 2003 [for Lao]; among others).
Enfield (2003), in his treatment of demonstratives in Lao from the
point of view of space and interaction, addresses the form-meaning
relationship surrounding semantic encoding and pragmatic inference in
the following way:
The SEMANTIC INVARIANT of a linguistic sign is the minimal meaning
that is always derived from the signa1 regardless of the context in
which it is used ... [I]t is EFFECTIVELY the case that linguistic signs
have stable and context-independent meanings ... The domain of
PRAGMATICS concerns the process whereby meanings richer than the encoded
conventional meanings of signs arise in real contexts. Pragmatic
meanings are semiotically accessible and/or logically derivable, but are
not semantically encoded. They include contingent aspects of context and
common ground (things one knows and things one can see and hear), on the
one hand, and derived inferences on the other. (Enfield 2003: 83-84)
Enfield's discussion of the semantic-pragmatic interplay is
useful to the current study on indexicality in that it solidifies the
distinction between semantic encoding on the one hand, and pragmatic
inference (in a Gricean sense) on the other. According to this approach,
the semantic relation of indexicality is distinct from encoded meaning.
If a meaning is part of a sign's encoded content, that meaning is
not indexed. Indexicality is set into motion through context-bound
interaction and involves an array of associations and expectations
deriving from inferences of what "a speaker 'could have
said' but didn't" (Enfield 2003: 84).
The distinction holds for "nonreferential indexicals"
(Silverstein 1976) as well. Whereas referential indexicals identify and
individuate entities and locations relative to the immediate context
and/or interaction, non-referential indexicals single out variables
within the sociocultural context--variables such as social rank (Duranti
1997; Platt 1986), gender (Ochs 1990, 1992), and epistemic/affective
stance (Ochs 1990, 1996). Excerpts (8) and (9) from Ochs (1996) will
illustrate. Ochs presents this example pair to depict a parallel
linguistic and socio-interactive phenomenon that occurs in English and
Samoan, respectively. The target form in each example involves an
epistemic marker of uncertainty.
(8) ((Mother, Father, and two children [Susan and Artie] are eating
dinner: Susan talks with food in her mouth))
1 Mother: ((deliberately to Susan)) finish chewing and then you may
talk.
2 Artie: ((takes a noisy gasp for air))
[right arrow] 3 Mother: ((continuing in the same tone of voice to
Susan)) maybe. (from Ochs 1996: 422)
Here, the target lexical item is 'maybe,' which,
according to Ochs (1996: 422) is an "epistemic indexical of
relative uncertainty," (17) which also serves to express a warning
or threat for Susan not to speak with her mouth full. In other words, by
virtue of the mother's addition of the increment "maybe"
in line 3, the conditional permission granted Susan at line 2 becomes
even more conditional and less absolute. Now, even after Susan has
complied with her mother's initial directive, she mayor may not be
able to speak, depending upon her mother's predisposition at the
time of Susan's compliance.
The next excerpt presents a nearly identical context and
interactional exchange, this time in a Samoan household.
(9) ((In a Samoan house, a mother is talking with one of her three
children who is acting selfishly toward his siblings.))
Mother: e le koe f'akau aa mai
TA NEG again buy EMPH DEICT.PRT
[right arrow] aa sau fagu e!
EMPH ANY. YOUR GUN EMPH
'She won't buy again any water pistol for you (unless
you shape up)!'
(from Ochs 1996: 423)
In example (9), the target item is the turn-final particle e, which
Ochs (1996) characterizes as "a future world that might come true
if certain behaviors continue that the speaker does not condone"
(Ochs 1996: 423). Hence, it, too, serves as a threat. In the latter
case, the speech act derives from the potential negative consequences
which might result from some specific type of unacceptable behavior.
What both of these examples share, according to Ochs, is the fact
that each highlights a particular lexical item which simultaneously
indexes an epistemic stance of uncertainty in addition to a speech act
pragmatically parallel to one of a warning or threat.
However, another interpretation of (8) and (9) is that it is not
true "indexing," but semantic encoding, according to
Enfield's discussion above. True indexing, though, does emerge
through pragmatic inference, because what is also implicit in both
examples is the fact that warnings and threats are usually uttered by
persons in a position of higher status or control. In other words, what
constitutes the index here is the inference arising not from what is
encoded in the semantics, but from the association of that encoding with
other contextually relevant and available information.
This brings us to the discussion of honorific speech levels in
Korean, where we appeal to the notion of indexicality and the multiple
domains of interaction and cognition within which it operates. It is our
claim that the use of honorific speech in Korean in general is indeed
linked to aspects of fixed social status and degrees of formality
related to such status, as we have illustrated. However, we will
demonstrate that when speakers use the deferential form, they are
selecting the semantic feature of +BOUNDARY and indexing a core stance
of EXCLUSION vis a vis their interlocutors or audience. Conversely, when
speakers use the polite form, they select the feature of --BOUNDARY, and
thus index in their on-going interaction a core stance of INCLUSION.
Based on this analysis, we also elucidate the fact that the grammar
of any spoken language is not constrained by formal and static
linguistic or generic properties, but is complexly indicative of a
system of variation and choice whereby speakers, both consciously and
subconsciously, project a range of stances with respect to their
interlocutors and to the information that they are conveying.
7. Honorific form alternations--beyond the prototypical
As we have discussed and demonstrated, both the deferential and
polite forms are used in contexts that require speakers' attention
to social contexts and hence to issues of deference. We have examined
prototypical instances of each, where honorific speech level
distributions accurately match the traditional approaches, namely,
example (4) illustrating the use of the deferential form in news
broadcasts, and example (5) illustrating the use of the polite form in
conversations among non-intimate and socially equal adults--the one
unilaterally imparting information with objectivity and detachment, the
other sharing stories about common experiences of worry, fear, and
surprise. However, as we will see in the excerpts that follow, the
majority of our instances do not fit the prototypical mold. And on the
basis of those excerpts, we propose that an alternative analysis of
+/-BOUNDARY and its concomitant indexing of EXCLUSION and INCLUSION
serves to capture more precisely the interactional and cognitive
dynamics underlying the use of honorific speech in Korean--an analysis
which becomes all the more clear in contexts where alternations between
both honorific forms occur within the same discourse.
Example (10) below is excerpted from the children's television
show "Ppo Ppo Ppo" (dataset 5 from Table 1). Recall that the
ratio of polite to deferential usages here is 155 to 6, or 26: 1. What
is interesting about the dataset is that it contains a rich mixture of a
variety of speech levels --predominantly the intimate and plain for the
non-honorific styles and the polite and deferential for the honorific
styles. This is because the program centers around an equally rich
mixture of interactional situations: children interacting with each
other, with adolescents, and with adults, in addition to cases in which
these characters address the viewing audience. When children interact
with one another, and when adults address children, their speech style
tends to be non-honorific; when adults interact with one another and
when children address adolescents and adults, the polite form tends to
be used. The six tokens of the deferential form were uttered by speakers
who otherwise used any one of the other speech levels. The excerpt in
(10) occurs at the very moment of a shift in the ongoing activity.
(10) "Ppo Ppo Ppo," 23 June 1999
((Ppomi enni, one of the main characters of the show, is sitting
next to Mrs. Wuahay. They have just watched a fashion show put on by the
children and the show will now be judged by both Ppomi enni and Mrs.
Wuahay))
Ppomi enni: Ca kulem simsa kyelkwa-lul al h o ha-keyss-supnita.
So then judge result-OM announce do-MODAL-DEF
'So, (now) I will announce the results.'
Wuahay acuma: camkkan simsa wiwencang-uy han malssum
-i-iss-keyss-supnita.
Wait a minute, judge head-GEN one word-
HON-SM exist-MODAL-DEF
'Wait a minute!! The head judge (meaning me)
will make a speech.'
Here, not only is there an activity shift, there is a blatant shift
in participant role, tone, and interactional quality. That is, speakers
who elsewhere used a preponderance of polite, intimate, and plain speech
styles suddenly switch to a speech style that signals their roles as
"official" judges, setting all other interlocutors apart from
the interaction. The feature of +BOUNDARY carried by the deferential
form thus operates on the level of the unfolding interaction by
designating a distinct activity shift, while at the same time boldly
delineating the domain of authority and responsibility as it relates to
these two speakers. A strictly deference-based account of the excerpt
could not sufficiently explain the interaction, since it is the speakers
themselves who invoke the speech level with respect to their own roles
and their own authority vis a vis all other participants. The
grammatical form does not signal a sudden shift upward as a display of
socially/contextually driven respect; rather it circumscribes a domain
of participation within which others are set apart. The two utterances
in (10) constitute an announcement/declaration by speakers who have the
unilateral authority to produce and carry out such a speech act.
We find a similar pattern of speech level mixing in the television
commercial data (dataset 6 from Table 1). The commercials ate structured
with a variety of formats: short, dramatic vignettes, both realistic and
burlesqued; testimonials; spots in which announcers or purported
corporate representatives directly address the viewers; and combinations
of these formats. And here, too, the speech levels in the full dataset
include non-honorific forms (intimate and plain) and a near equal
distribution of polite and deferential forms, or a ratio of 1:41 to 1.
What is rather consistent, however, is that even in commercials with a
predominant use of non-honorific and polite forms, the majority of ads
end with a final comment in which a promise, commitment, declaration, or
other such speech act is made on the part of the company sponsoring the
ad. Excerpts (11) and (12) below illustrate two such instances.
(11) (excerpted from Kaya Tangkun nongcang--carrot juice
commercial)
... kulena payk pheseynthu ku swunswu-ha-m-man-un cikhye
kakeyss-supnita.
But 100 percent that pure-do-NML-only-TM keep go-MDL-DEF.
'But we will always keep [DEF] the 100% purity.'
[PROMISE]
(12) (excerpted from Kwangssangthang--herbal drink for flu or cold)
eps-supnita.
not exist-DEF
pangpwucey-ka eps-supnita.
Preservatives-SM not exist-DEF
'There aren't [DEF] any. There aren't any [DEF]
preservatives.'
[DECLARATION]
The use of the deferential form in these (and many other) examples
from the database clearly expresses a stance of detachment, of apparent
objectivity on the part of the advertiser. The message here is not one
of interpersonal sharing, but a unilateral commitment on the part of the
advertiser to the quality of the product, an expression of unbending
confidence. These examples reflect the use of the deferential as a
persuasive device in the genre of advertising through its creation of an
atmosphere whereby the advertiser demarcates its authority and expertise
within a domain undeniably separate from that of the viewing audience.
While grammatically possible, a--BOUNDARY signal in these same speech
acts significantly weakens the stance of authority, expertise, and
commitment.
Further, if explanations predicated on social variables were
sufficient to account for how and under what conditions the deferential
form is used in discourse, one might wonder about its occurrence when
the direction of address is the exact opposite of that expected in a
deference-based account, namely, from persons of higher status to
persons of lower status. Example (13) is a case in point. Here, we find
a female middle school teacher in South Korea in the midst of presenting
an English grammar lesson on relative clauses.
(13) (Middle school English lesson on relative clauses (18))
Teacher: 1 cikum po-myen-un sensayng-nim-i cwungyo-ha-n mwuncang
twu kay-lul
Now look-COND-TM teacher-HON-SM importance-do-ATTR sentence two
thing-OM
2 koll-a-po-keyss-supnita.=yelepwun-i al swu-iss-nun-ka.
pick-AUX-MDL-DEF, everyone-HON-SM know be able to-exist-whether
3 ches-pen-ccay mwuncang i-pnita.
first-number-CLSF sentence COP-DEF
'Now, watch, I will pick [DEF] two important sentences to see
whether you can understand them or not. (This) is [DEF] the first
sentence.'
The example is significant in two respects. First, as mentioned
above, it involves the use of the deferential form in spite of the fact
that the vector of address is toward lower status interactants. It is
clearly not an instance of the teacher expressing deference or
politeness to her students. Had it been intended as such, the speaker
might have included other honorific markers, such as the honorific infix
-si in the conditional construction in line 1 (po-si-myen instead of
pomyen, literally, 'if you (+/-H) look'), though this would be
contextually odd. Further, if the teacher had intended to create a
hyperpolite/deferential atmosphere in her classroom, one would expect
her to consistently use the deferential form in other contexts of her
teaching. However, this teacher's classroom speech involves a
predominant mixture of polite and intimate forms, even in the dialogues
addressing students as they work with relative clause
structures--interactions which directly follow the teacher's speech
excerpted in (13). The second point of interest here is the fact that
the teacher is female, and recall that the literature suggests that the
deferential form is more naturally used by males.
The deferential-marked utterance in the foregoing example is
typical of what we find throughout our datasets. That is, when a speaker
is in a position of authority or one in which s/he controls the
activities within a speech event, we find frequent occurrences of this
speech level. Example (14) will illustrate further. As in the previous
excerpt, the teacher in this English language instructional television program (this time, a male) is introducing the topic of the day's
lesson; and, as in the previous excerpt, he does not use the deferential
form exclusively throughout his teaching. In fact, in this latter case,
as is evident in Table 1 (dataset 4), the instructors on this program
use the deferential forms and polite forms with near equal frequency
(i.e. 0.96 to 1). What makes the excerpt in (14) so illuminating is the
fact that it evinces a shift from the polite to the deferential by the
same speaker addressing the identical audience within two contiguous
turns at talk: In the first turn, he uses the polite form. In the very
next one, he shifts to the deferential.
(14) (from "Survival English," topic: making purchases,
air date: May 1999) (19)
Host: Macimak-ulo shopping-hay-se mwulken-ul
Last-INSTR shopping-do-and merchandise-OM
kaci-ko-se cip-ey ka-ki cen-ey Hay-ya
take-and-CONN home-DIR go-NML before-DIR do
toyl-il-i kakyek-ul chil-eya-toy-cyo.
have to-ATTR thing-SM price-OM-pay-have--to COMM-POL
'In the end, when we go shopping, before we take the
merchandise home, we have to pay for it.' [POL].
Ca, mwulken kaps-ul chilu-I ttay sayong hal swu iss-nun phyohyen-ul
wuli-ka
so, merchandise price-OM pay-ATTR when use can--ATTR expression-OM
we SM
cwungcem-cek-ulo--cwungcem-cek ulo-onul kongpwu hay potolok
ha-keyss-supnita
concentrate-ADJ-INSTR concentrate-ADJ-INSTR today study do
see-INTEND do-MDL-DEF
'So, today we will be studying [DEF] focusing on expressions
we can use when paying for merchandise.'
The above example elucidates just such a +/-BOUNDARY feature for
the deferential and polite, respectively. In the first utterance, the
speaker provides background for the day's lesson--making
purchases--by providing a common sense statement about shopping. Not
only is the utterance marked with the polite form, it also contains the
particle -ci, discussed in the literature as a sentence ender through
which the speaker expresses his/her stance with respect to the truth of
a proposition (K. Lee 1993; H. Lee 1991, 1999; Sohn 1999; among others).
The use of -ci heightens the interactional quality of utterances in
general and the above utterance in particular, by drawing the addressee
closer into the interaction and by creating the inference that a
response (e.g. an answer, confirmation, or even simply silent agreement)
is directly relevant to the just produced talk. As such, the --BOUNDARY
feature signaled by the polite form is underscored all the more by the
interactional particle affixed to it. (20)
The host's switch to the deferential form in the immediately
contiguous turn signals a shift in the activity, namely, moving on to
the actual topic at hand, in addition to an attendant shift in
interactional stance. The +BOUNDARY signaled by this form moves the host
out of a more personalized type of discourse and into one of authority,
thereby pragmatically repositioning him as the participant in charge.
(21)
7.1. Multi-party interactional public discourse: "Achim
Matang"
The interaction as it unfolds in dataset 2, "Achim
Matang" ("Morning Yard") presents even more compelling
evidence for our analysis. "Achim Matang" is a popular
television program currently airing in South Korea in which guests are
invited to the show to try and locate family members from whom they were
separated during their early childhood. Often it was the burden of
severe financial hardship, illness, or the death of a family member that
caused the family to give the child up. In some cases, children were
abandoned as toddlers and raised by the strangers who found them; in
others, they were raised by more distant relatives or by acquaintances.
The majority of the guests on this program, now mature adults, belong to
low to mid socioeconomic levels; they are generally not educated beyond
high school or college.
Two regular hosts emcee the show: one male and one female. The
premise is that the family members being sought would watch the program,
recognize a guest, and phone in to be reunited with them. It is
structured such that the hosts introduce the guests to the studio and
television audiences by first providing a short narrative about the
guests' lives and the circumstances surrounding their separation.
Each guest then tells his or her own story, presenting an emotional plea
to viewers to help find the relatives s/he is searching for. Since this
program contains an inherent triplicity of participant roles, namely,
hosts, guests, and audience, its discursive, interactional structures
are also inherently complex. In this regard, it will be revealing to
note the distribution of deferential and polite forros among the
speakers. The distribution is indicated in Table 2 below. The
hosts' use of polite and deferential forms exhibits a collective
ratio of 2: 1, contrasting sharply with the collective ratio of the
guests' speech of 9: 1. In other words, the hosts' social
position is generally higher than that of their guests and potentially
ambiguous vis a vis that of the studio and television audiences, yet
they produce many more tokens of the deferential form. Conversely, the
guests' speech, generally involving a consistently upward direction
of address to the hosts and the viewing audiences, contains
approximately one quarter the number of tokens of the deferential
form--a counterintuitive finding if based solely on traditional
accounts.
However, an analysis which rests on the concept of indexicality of
EXCLUSION and INCLUSION seems to solve the puzzle. That is, when the
hosts use the deferential form, they do so in ways that underscore their
roles as hosts--as the participants who lead the entire show. When hosts
engage in dialogue with each other or with guests, they predominantly
use the polite form; when guests converse with family members, they
alternate between non-honorific levels and the polite form, and when
guests and hosts address the audience, they alternate between polite and
deferential forms, depending upon both the content of the discourse and
the interactional stance that they are indexing.
Examples (15) and (16) illustrate the hosts' use of the
deferential form when addressing the audience. The first excerpt occurs
at a relatively early point in the program as Mr. Song announces the
opening of the segment in which family members located as a result of
the previous airing meet on this day's broadcast.
(15) (from "Achim Matang," p. 12 of transcript)
Song: ... onul pankawu-n manna-m sosik-ulo sicak-ha-keyss-supnita.
Today happy-ATTR meet-NML news-INSTR begin-do-MDL-DEF
'... (we'll) start [DEF] today's happy
meeting.'
Excerpt (14), also by Song, emerges in a somewhat a later segment
as he presents another guest.
(16) (from "Achim Matang," p. 63 of transcript)
Song: Iese sey-pen-ccay chulyen-ca mosi-pnita.
Following this three-time-CLSF participate-person invite-HON-DEF
'Now, (we'll) introduce the third participant.'
In both excerpts, through the use of the deferential form, Song is
establishing his stance qua host, in the sense that it is only the hosts
who have the authority and control to designate what will be coming
next.
In fact, examples (10) through (16) involve speakers who are in the
unilateral position to be in charge of the particular activities in
which they are engaged. In all seven instances, the speakers appeal to
detachment and objectivity in order to achieve a certain interactional
goal: making unilaterally authoritative announcements, promises, or
declarations, as in (10), (11), and (12), controlling the sequences of
activities that comprise the overall context of their interactions, as
in (13) and (14), and overseeing the progression of the program
segments, as in the case of "Achim Matang," as observed in
(15) and (16).
The variations in +/-BOUNDARY signals and the concomitant indexing
of INCLUSION VS. EXCLUSION become the most salient in cases where
speakers switch from one level to the other within the same spate of
talk. Excerpt (17) is one such instance. The excerpt highlights the
host's speech during the third guest's appearance, and the
target utterance emerges shortly after he has introduced the guest and
presented her separation narrative. At this moment, viewers begin to
call the studio:
(17) (from "Achim Matang," p. 69 of transcript)
Song: Cham camkkanman-yo. Cenhwa
Oh. Wait a minute-POL Telephone
wa-ss-ta-nikka-n-yo. Pata-po-tolok ha-keyss-supnita.
come-PST-Q-RSN-TM-POL Receive-see-MDL do-MDL-DEF
'Wait a minute [POL], because we have a phone call [POL].
(We'll) take it [def].'
Here, even though the scope of Mr. Song's addressees remains
unchanged, his speech level shifts from the polite, where he expresses
excitement and empathy with his guest, to the deferential, where he
instantaneously resumes his unilateral role as a representative of the
program by indexing an interactional boundary between the hosts and
other participants (i.e. guests and audience alike). The speech style
shift serves as a deictic marker delineating precisely and unequivocally
who is included within the sphere of the action designated by the
utterance (i.e. the hosts) and who is excluded (i.e. guests and
audience). That is, in the first two utterances marked by the polite
form, all participants are framed as equally sharing the sphere of
experience, emotion, and empathy as they receive the news that a caller,
a potential family member, has responded; in the final one, the host
shifts from a personalized, other-inclusive discourse style to a
depersonalized, other-exclusive one.
The next example illustrates honorific speech level alternations
among guests' talk. What is interesting in this regard is that the
guests' narratives are recounted twice on the show--the first time
by a host providing the background of the program contents, and the
second time by the guest him/herself. The details surrounding each
guest's account, for example, story line and episode structure, are
rather parallel, whether presented by host or guest. Where these
narratives differ, however, is that in the hosts' retelling, the
predominant speech style is the deferential; when told by the guests
themselves, the predominant style is the polite. Excerpt (18) below is a
typical example of a guest's narrative opening.
(18) (from "Achim Matang," p. 30 of transcript: 1st
guest, Mrs. Cengok Co]
Co: ce-nun apeci co pyenghwun-ssi-ha-kwu-yo, emeni sengham-un cal
molu-keyss-kwu-yo.
I-TM father Co Pyenghwun-Mr. do-CONN-POL mother name-TM well not
know-MDL-CONN-POL
*hh enni-, oppa co cengpok, enni co cengswun, kuliko
tongsayng-yetongsayng
older sister older brother Co Cengpok older sister Co Cengswun and
younger brother- younger sister
co cengswuk-ul chac-supnita.
Co Cengswuk-OM look for-DEF
'I am looking for [DEF] my father and [POL] my mother, whose
name I don't know and [POL] my older sister- (I mean) my older
brother, Co Cengpok, my older sister, Co Cengswun, and my younger
brother- (I mean) younger sister.'
Note that the single token of the deferential style used by Mrs. Co
emerges as she indicates the names of the relatives she is searching
for. The interactional dynamic is clear: the speaker initiates her
narrative with a tone of objectivity and detachment, and once launched,
the boundary between her and her collective interlocutors recedes as she
recounts the episodes leading up to her estrangement and the experiences
she underwent as a result. (22) The contrast in narrative delivery style
by host and guest, most conspicuously evinced by honorific speech level
choice, underscores all the more just to what degree the semantic
feature of +/-BOUNDARY is at work here--directly experienced events are
narrated from a highly personalized point of view (-BOUNDARY); events
experienced and narrated by and about others are more dispassionate,
distant, and detached (+BOUNDARY).
7.1.1. The deferential form with modal -keyss. It is significant to
note that in addition to the use of the deferential style in utterances
expressing objectivity, detachment, and authority, we also find a
relatively high frequency of co-occurring tokens of the morpheme -keyss.
-Keyss is a modal which typically expresses intentionality and
volitionality when used with agentive verbs. Because of this function,
-keyss has been discussed in the literature as a future marker similar
to English will, even though its semantic core contains no element
relating to tense; it also functions as a modal of conjecture (Sohn
1999; Lee and Ramsey 2000; Suh and Kim 1993). (23)
The collateral occurrence of-keyss as an intentionality marker
serves to advance our analysis of the deferential speech level as a
+BOUNDARY marker and an index of EXCLUSION, whereby speakers both
position themselves as interactants within a contextually situated
position of power and authority, thereby distancing themselves
experientially and/or epistemologically from their interlocutors.
As Table 3 indicates, -keyss emerges in the "Achim
Matang" data as a strong intentionality marker in 23 of the l04
deferential-marked utterances by the hosts. In stark contrast, the
guests' speech contains only one co-occurring use of-keyss, and
there, it is used as a modal of conjecture.
Example (19) represents a typical spate of talk by the two hosts
exhibiting the co-occurrence of the deferential speech level with the
modal -keyss. This excerpt is characteristic of the hosts' talk in
the program and of host or moderator talk in general in the sense that
it is essentially devoid of affective interactional features which draw
audiences or interlocutors closer into their own spheres of interaction.
(19) (from "Achim Matang," p. 107 of transcript) Song: 1
... Ney. Yenlak-ul hamkkey
Yes. contact-OM together
kitalye-po-tolok ha-keyss-supnita.
wait-see-INTEND do-MDL-DEF.
((skipped two lines of transcript))
Lee: 4 ... Sewul cenhwa chil-payk-phal-sip-il-kwuk-ey
sam-chen-o-payk-i-sip-il-pen-eyse
Seoul telephone 700 81 exchange-hyphen 3,000 521 number-from
5 phal-pen-kkaci, naynay pat-keyss-supnita. Onul nao-si-n
pwun-tul-kkey-nun
eight-number-until, always receive-MDL-DEF. Today appear-H-ATTR
person-H-PL DAT-TM
6 cehuy-ka senmwul-lo sayng-kulin hanpang hwacangphwum-eyse
nongsan-mwul
they-SM present-INSTR Sayngkulin Chinese medicine cosmetics-from
agriculture-product
7 sangphwum-kwen-ul tuli-keyss-supnita.
gift certificate-OM giveHON-MDL-DEF.
Lee: '(We'll) be waiting [DEF] together for the call
(from the guests' families).'
Song: 'The Seoul telephone number (is) from 781-3521 through
(781-352)8. (We will) answer [DEF]. For today's participants, [DEF]
them with a certificate for agricultural products sponsored by the
Sayngkulin Chinese medicine cosmetics company.'
Note, for example, that lines 1 and 5, while syntactically
declarative statements, actually function on a pragmatic level as a
request--an appeal by the hosts to have viewers call into the show. This
appeal is delivered neither through an imperative construction nor
through an emotional plea, but indirectly through the declaration by the
moderators that they will be waiting for the calls and that they (or the
program's representatives) will be available to receive these calls
on a 24-hour basis. This type of syntactic-pragmatic interplay is common
throughout the hosts' speech in the program. That is, in spite of
the program's highly emotional nature, in the majority of these
preplanned segments, the hosts remain superficially detached, with the
predominant type of emotional involvement they display tending to
surface at the level of their commitment to the purpose of the program.
In contrast, when Ms. Lee and Mr. Song exhibit empathetic displays of
emotion towards their guests and the hardships they have endured, they
tend to do so in the more spontaneous segments with a mixture of polite
and deferential speech levels.
Beyond the "Achim Matang" dataset, we observed similar
instances of -keyss-plus-deferential in examples (10) from "Ppo Ppo
Ppo," (11) from the carrot juice commercial, (13) from the middle
school English lesson, and (14) from "Survival English." What
is interesting, too, is that in many of the examples presented thus far,
both female and male speakers appeal to the deferential form to achieve
the same interactional goal, and as such, gender in and of itself does
not appear to be a primary predictive factor in the use of the
deferential vs. polite form as previous accounts have suggested.
7.2. Monologic public discourse--religious sermons
In contrast with excerpts from dialogic or multi-party discourse
activities, in this final section, we will focus on the formal,
monologic genre of the religious sermon--an inherently less complex
genre of oral discourse, from the standpoint of interactional features
and participants, than those previously examined. That is, the sermon
(be it scripted, spontaneous, or a combination of both) is produced by a
single speaker to an entire congregation as a whole. And, from the
standpoint of religious faith, training, and conviction, this single
speaker occupies a higher "social"/ spiritual position than
the majority, if not all, of the congregants--it is he or she who
delivers messages of religious and/or moral values for the purpose of
instructing and inspiring the listeners. As noted earlier, honorific
speech style alternations do indeed occur in this genre, affecting a
similar shift in frame of +/-BOUNDARY as we examined in the multi-party
and dialogic examples. However, here we see that the alternations
function on a more subtle level and are squarely related to didactics
and to another level of persuasion, distinct from that observed in the
television commercial data. Further, where we found a preponderance of
the deferential form co-occurring with -keyss in cases where speakers
index their own volition and commitment to an action yet to be
undertaken, the modal -keyss is rare in the sermon data, since the
purpose of this latter speech genre is to preach, not promise.
The next two excerpts are typical of the phenomenon as it patterns
across all three sermons in our database (i.e. datasets 7, 8, and 9).
Example (20) is from a sermon by Pastor Yong Co Ha entitled
"Three principles for marriage." As indicated in Table 1
(dataset 9), the 40-minute sermon contains a total of 343 tokens of
honorific sentence enders--286 deferential and 57 polite, or a ratio of
0.2 to 1.
(20) (Pastor Ha, "Three principles for marriage")
1 Salam-tul-un ilehkey mal-ha-pnita. (.) namca-to yeca-ka
nah-ass-ci (.)
Person-PM-TM like this word-do-DEF man-also woman-SM bear-PST-COMM
2 Mal-un mac-su npnjta. (.)
Word-TM be correct-DEF
3 ani yeca-ka an na-n namca-ka eti-ss-eyo (.)
no woman-SM not bear-ATTR man-SM where-exist-POL
4 Ta:: yeca-ka (.) namca-lul nah-ass-ci-yo (.)
All woman-SM man-OM bear-PST-COMM-POL
5 kelehciman sengkyeng-un kulehkey oh en-ha-ci (.) anh-supnita
however Bible -TM like that expression-do-COMM not-DEF
'People say [DEF] this: "Man also was born by
woman." Literally, (that) is correct [DEF]. Where in the world is
[POL] a man who was not born by woman? Of course all women gave birth
[POL] to men. However, the Bible doesn't express (it) [DEF] like
that.'
Note that these five lines contain two tokens of the polite form
sandwiched between deferential-marked utterances, and this, in spite of
the fact that it is the same speaker addressing the same congregation.
However, what shifts here is the nature of the information being
presented and how such information is framed with respect to the
congregants. That is, as Pastor Ha broaches this topic of his talk, he
first presents a counter argument in line 1 ('people say this: man
was also born by woman') followed in the next line by an objective,
detached, and qualified assessment of that claim ('Literally, that
is correct'). Both utterances, marked with the deferential, are
authoritative in nature and serve to systematically lead to Pastor
Ha's argument that ultimately appears in line 5, also marked with
the deferential (kelehciman sengkyengun kulehkey pyohyenhaci (.)
anhsupnita 'However the Bible doesn't express it like
that')--the upshot of this segment.
What is interesting is that the two instances of the polite form
appear as a development of the assertive counter-claim, essentially as a
re-voicing in simple terms of its apparent commonsensicality. Line 3 is
a rhetorical question ('Where in the world is a man who was not
born by woman?'), the answer to which is obvious: nowhere. But he
doesn't simply leave it at that. He provides an explicit answer in
the immediately contiguous line that contains the most common sense of
statements that 'all men ate born from women.' Note, too, that
this utterance at line 4 is also marked with -ci, which amplifies the
personalization all the more. Thus, in order to first build on the
counter-argument which rests on what could be termed popular belief, he
appeals to a personalized, more congregation-interactive speech style,
only to dismantle it in the final line, where it is the Biblical
interpretation which should prevail.
The interplay of persuasion and didactics in this and the other
sermons in our database is powerful. When appealing to notions of common
sense and of shared beliefs (mistaken or otherwise), we find uses of the
polite form signaling -BOUNDARY, which serves to intersubjectively draw
the congregants into the discourse through associations with what is
pedestrian and familiar--hence its frequent use in the sermon data with
rhetorical questions. When underscoring the authoritative and canonical
messages of the sermon, the +BOUNDARY, deferential marker is used.
In the case of Pastor Ha, the deferential is clearly the preferred
form. His tone is consistently authoritative, official, and detached.
Additionally, his age is between 50 and 60, which may account for the
skewing.
In contrast, Pastor Pyeng-Wuk Cen, who is in his mid-thirties, uses
the polite form with a much higher frequency than Pastor Ha in the two
sermons analyzed for this study. In one 20-minute sermon (dataset 8 from
Table 1), his speech exhibited 49 tokens of the polite form and 115 of
the deferential form, or a ratio of 0.43 to 1. In his 40-minute sermon
(dataset 9 from Table 1), we find 258 tokens of the polite and 259 of
the deferential. And, as we saw with Pastor Ha, the polite and
deferential serve the same purpose of establishing unity, empathy, and
familiarity with the congregants on the one hand, and establishing
authoritative distance on the other. Example (21) below, from dataset 9
entitled "Parents and children" will further illustrate:
(21) (Pastor Cen, "Parents and children")
1 mwe may kac-ko an thong-ha-ko mal kac-ko phokse- ((aborted word))
DM stick have-CONN not work-do-CONN word have-CONN violent language
2 yok-sel kac-ko thong ha-ci-l anh-nun kes-i-pnita::. bad language-
have-CONN work do-NML-OM NEG-ATTR thing-COP-DEF
3 ku ttay-ey ka-se kkaytat-nun ke-cyo. that time-at go-CONN
realize-ATTR thing COMM-POL.
4 May-lo an-toy-n-ta. Kuleko nwa twe-yo. Kunyang ni mes tay-lo
hay-la.
Stick INSTR not become-PLN then let-put-POL. Just you style
according to do-IMP-PLN
5 kulenikka ay-ka icey mes-tay-lo ha-nikka cincca mes-tay-lo
ha-c-yo I mean child-SM now style-according to do-RSN really
style-according to do-COM-POL
6 ta mwuneci-key toyn-ta-nun ke-yey-yo all collapse-ADV
become-QT-ATTR thing-COP POL
7 kulayse hesey pwuli-myen an-toy-pnita hesey. so false power
execute-COND NEG-become-DEF false power.
8 elyess-ul-ttay-nun ohilye may-lul tul-e-ya toy-pnita be
young-PST-ATTR when-TM rather stick-OM take-MDL become-DEF
9 sengkyeng-ey kuleci-anha-yo. Bible-in so-COMM-NEG-POL
10 chotal-ul chama mos-ha-nun ca-nun ai-lul miwe ha-nun ke-la-ko
hit-OM ADV unable-do-ATTR person-TM child-OM hate do-ATTR thing-QT
11 mongtwungi-lo phay-ya-cyo stick-INSTR hit-MDL-COMM-POL
12 an-cwuk-supnita phay-to not die-DEF hit-even if ((congregation
laughter)).
13 ely-ess-ul-ttay-n ttaylye-ya tway-yo. be young-PST-ATTR when-TM
hit MDL-become-POL
14 kulena:: sachwunki-ey tulese-se-nun may kac-ko-n celtay
an-toy-pnita. but puberty-in enter-CONN-TM stick have-TM absolutely
NEG-become-DEF
15 seltuk-hay-ya toy-pnita. Hapli-cek-i-n seltuk. persuasion do-MDL
become-DEF. reason-ADJ COP-ATTR persuasion
'Using a stick doesn't work and using violent lang- bad language
doesn't work [DEF]. At that time, (they, i.e., parents who order
their kids to do things) realize don't the [POL] 'I can't do it
using a stick [PLN].' So they just let it go [POL], (i.e. let kids
do whatever they want) (saying), "do whatever you want." So, because
the child does whatever he wants, really he does whatever he wants
[POL]. I mean, eve thin just collapses [POL]. So, you must not
bluff your power [DEF] false show of power. When (the child) is
young, you have to hold your stick [DEF] (i.e. you should spank
your child). You know, it says so in the Bible, right? [POL]. "Spare
the rod, spoil the child' (Lit. 'He who spares the rod hates his
son.') You have to spank him with a stick [POL]. He's not going to
die [DEF], even if you hit him. ((congregation laughter)) When the
child is young, you have to spank him [POL]. But, when he enters
puberty, using the rod absolutely will not work [DEF]. You must
persuade him [DEF] (using) reasonable persuasion.'
The above excerpt exhibits a similar pattern as we observed in
(20), in which persuasive, familiar scenarios, and commonly accepted
lines of reasoning are interwoven into the discourse as a means of
establishing and supporting the primary argumentation which constitutes
the lessons of the sermons. In this excerpt, we find six tokens of the
deferential and seven of the polite. The opening and concluding lines,
both conveying the crux of Pastor Cen's argument (i.e. that neither
corporal discipline nor harsh language will serve as disciplinary
measures for older children), are marked by the deferential.
Interspersed between these are variable instances of polite and
deferential forms--their uses now appearing to be almost predictable
given the proposed system centering on the signal of +/-BOUNDARY for
each.
That is, immediately after Pastor Cen delivers his initial
statement in lines 1 and 2, he expands and develops the argument by
first exemplifying lenient parents who don't spank or discipline
their children at all, in lines 3, 4, and 5, and the consequences of
this in line 6 ('I mean everything just collapses'), all of
which are marked with the polite form. We soon come to learn that these
parents are, in fact, the negative model, yet they are introduced in a
seemingly innocuous fashion. Pastor Cen frames this as a situation to
which many can relate--spanking children is obviously a controversial
issue in child rearing, and he cites as examples those who don't,
using a more personalized interactional style, only to build, in the
immediately next lines, his own case which condones and advocates the
spanking of young children. In lines 7 and 8, both now marked with the
deferential, his stance suddenly becomes strong, more authoritative
('So you must not bluff your power' and 'when your child
is young, you should spank your child'). Yet another shift ensues
from the deferential (in 7 and 8) to the polite (in 9 and 10) as the
official argument is expanded using a softer, more persuasive style,
with Pastor Cen appealing to the Bible and the congregation's
collective knowledge of it ('you know, it says so in the Bible,
right? "Spare the rod and spoil the child" ') followed by
a reiteration in line 11 of the previous idea that it is necessary to
spank children ('you have to spank him with a stick'), also in
the polite form. His tone shifts once more to the deferential in line 12
as he declares 'he won't die, even if you hit him.' What
is noteworthy here is that this statement, too, is common sense, yet
here it is framed as if it were not--in order to underscore the strength
of Pastor Cen's conviction.
And, so the argument is built, step by step, line by line, with a
subtle yet powerful interlacing of opinions, ideas, and lessons, at
times framed softly as if these are commonly shared, and at times
authoritatively as life's lessons to be learned. In other words,
Pastor Cen consistently uses the polite form when he appeals to issues
that are already well known to the congregation. These include concepts
that experienced parents can easily relate to--either from their own
experiences or from witnessing behaviors of other children, concepts
from the Bible, and concepts that the congregation becomes familiar
with, simply by virtue of listening to this very sermon. That is, the
polite form is used to express the following: how parents react when
they realize that they have lost control of their children, how children
react when they realize that their parents cannot discipline them, a
well-known quote from the Bible, and finally, the reiteration of Pastor
Cen's own previous statement that one must spank one's child
in order to discipline him. Note that when the pastor first makes this
statement about corporal discipline, it appears in the deferential form
(line 8); when the statement is reiterated (line 10), it appears in the
polite form. Crucially, however, while the content (though not the words
themselves) is nearly identical in lines 1-2 and 14-15, we find no
switch to the polite form in the latter statement, since it is this very
message that Pastor Cen has set out to convey in this segment of his
sermon.
Examples (20) and (21) also establish a nice contrast with respect
to how pastors in general, and these in particular, invoke biblical
knowledge as a persuasive preaching technique. In (20), Pastor Ha
indexes iris authority (EXCLUSION) as biblical expert, detached from the
congregation in his interpretation: kelehciman sengkyengun kulehkey
pyohyenhaci (.) anhsupnita 'But the Bible doesn't express it
that way [DEF].' In (21), Pastor Cen draws his congregants into a
shared sphere of this knowledge by framing the passage as one that every
Christian knows and can recite by heart sengkyengey kulecianhayo
chotalul chama moshanuncanun ailul miwe hanun kelako: 'You know, it
says so in the Bible, right? "Spare the rod, spoil the
child."' Such contrast underscores all the more just to what
degree the signal of +/-BOUNDARY and its attendant function of indexing
EXCLUSION or INCLUSION is at work in these data.
Moreover, what the segments in (20) and (21) underscore is that the
semantic encoding of +/-BOUNDARY and its indices of EXCLUSION and
INCLUSION operate on multiple planes. These, and other examples from the
data, illustrate that the phenomenon of "honorific" speech
level alternation in Korean systematically encodes perceived and actual
boundaries at the interpersonal level as well as at the levels of
experience and cognition. In other words, the choice of one form over
the other not only marks boundaries as they pertain to social identities
of and interpersonal relationships between speaker and addressee, but
also boundaries as they pertain to the respective domains of personal
experience, belief systems, and knowledge bases of interlocutors.
8. Conclusion
We have demonstrated that the phenomenon of honorific speech level
alternation in Korean relates more, if not entirely, to the semantic
encoding of +/-BOUNDARY and its attendant indexing of EXCLUSION and
INCLUSION than it does to actual social factors such as age, profession,
and status, on the one hand, and discourse genre on the other.
If such a priori concepts were the primary determining factor, then
the use of polite vs. deferential would clearly be governed by
obligation as noted by Sohn (1994, 1999), Lee and Ramsey (2000), Lukoff
(1982), and Martin (1964) for Korean, and by Matsumoto (1988) for
Japanese, and this, in situations where the vector of address is
unequivocally upward or downward; the model leaves little room for
variability.
However, using a wide variety of spoken discourse data we have
shown instead that honorific speech level alternation between polite and
deferential forms occurs frequently in both monologic and dialogic
speech, and that the alternation exhibits strikingly systematic
patterns. We have also shown that it is the speakers themselves who
invoke the deferential speech level with respect to their own roles,
their own authority, and their own expertise--a notion which runs
counter to the dominant deference-grounded approaches which assume a
strictly addressee-sensitive motivation based on an upward vector of
address.
In conclusion, our analysis proposes that while honorific speech
level choice seems at first glance to be related to issues of deference
and formality, its patterning in discourse reveals a different and more
dynamic underlying system. Rather than a system driven by such static,
macrosociological forces as power and/or status differential, we propose
instead that honorific speech level alternation in Korean is motivated
first and foremost by the essential feature of the semantic encodings of
+/-BOUNDARY. This perspective on the phenomenon as it occurs in natural
discourse makes manifest a multiplicity of interactional and cognitive
variables on the part of the speaker-producer vis a vis his/her
interlocutor. The study also has implications for other languages with
morphosyntactic honorific systems, the dynamics of which might well be
influenced by similar types of social, cognitive, experiential, and
interactional factors as those presented here.
Appendix. List of abbreviations
AH addressee honorific
ATTR attributive
AUX auxiliary
BLN blunt
CLSF classifier
COMM committal
COND conditional
DAT dative
DC declarative
DEF deferential
FML familiar
H honorific
H (LI) honorific lexical item
H (T) honorific title
H (VI) honorific verbal infix
HM hesitation marker
IN indicative
INF informal
INSTR instrumental
INT intimate
INTERR interrogative
MD modal
NML nominalizer
OM object marker
POL polite
PLN plain
PM plural marker
PSS passive
PST past
Q question marker
QT quotative
RETR retrospective
SM subject marker
TM topic marker
Received 10 May 2001
Revised version received
24 September 2003
Pennsylvania State University
Table 1. Ratio of polite to deferential forms according to data type
Dataset Polite Deferential Ratio POL:DEF
Spontaneous speech
1. Earthquake data (dyads 423 2 212:1
1, 3, 4, 5) [1 hour] (12)
2. TV Talk Show (Achim Matang) 446 131 3:1
[1 hour]
3. Radio Talk Show (Alumtawun 247 88 3:1
Seysang) [1 hour]
4. English Instructional TV 173 181 0.96:1
Program [1 hour, 40
minutes]
Scripted speech
5. Children's TV Program 155 6 26:1
(Ppo Ppo Ppo) [4 episodes--
1 hour total]
6. TV Commercials [20 minutes] 69 49 1.41:1
7. Sermon 1 (Cen Byeng Wuk) 258 259 1:1
[40 minutes]
8. Sermon 2 (Cen Byeng Wuk) 49 115 0.43:1
[20 minutes]
9. Sermon 3 (Ha Yong Co) 57 286 0.2:1
[40 minutes]
10. MBC news broadcast 1 107 0.009:1
[20 minutes]
Table 2. Polite and deferential forms
by hosts and guests in "Achim Matang"
Achim Matang Polite Deferential Ratio POL:DEF
Host Song (M) 153 62 2.5:1
Hostess Lee (F) 62 42 1.5:1
Hosts (total) 215 104 2:1
Guests 231 27 9:1
Table 3. The co-occurrence of -keyss with
the deferential form--"Achim Matang"
Achim Matang Deferential -keyss + DEF (%)
Host Song (M) 62 15/62 (24%)
Host Lee (F) 42 8/42 (19%)
Hosts (total) 104 23/104 (22%)
Guests 27 1/27 (4%)
Figure 2.
The six Korean speech levels (excerpted and adapted
from Sohn 1999: 236-237)
Honorific:
DEFERENTIAL: mek-sup-ni-ta
eat-AH-IN-DC
POLITE: mek-e-yo
eat-INT(4)-POL
Non-honorific:
BLUNT (5): mek-sol-uo
eat-BLN
FAMILIAR: mek-ney
eat-FML
INTIMATE: mek-e
eat-INT
PLAIN: mek-nun-ta
eat-IN-DC
'I, you, she/he, we, they eat, are eating, etc.'
Notes
(1.) Research for this project was partially funded by the Penn
State University Center for Advanced Language Proficiency Education and
Research (CALPER) funded by a grant from the U.S. Department of
Education (CFDA 84.229A P229A020010) The authors are indebted to the two
anonymous reviewers who provided thorough and insightful comments on the
original version of this manuscript. Susan would also like to thank Ms.
Soojung Choi for her careful reading and correction of the Korean
examples. Correspondence address: 124 Grandview Road, State College, PA,
16801, U.S.A. E-mail:
[email protected].
(2.) While we attribute the use of all four honorific tokens in
this example to "status differential" in accordance with
traditional accounts of Korean honorifics in general, we will deviate
from this interpretation as ir pertains to the deferential verbal ending
-(s)upnita in the body of this article and propose a significantly
different analysis.
(3.) Scholars seem to concur about the number of speech levels,
however, the naming of these levels and their functional descriptions
are not always consistent. See, for example, Lee and Ramsey (2000) and
Martin (1964).
(4.) Sohn (1994) glosses this morpheme as INF ("informal") and later as INT ("intimate") in Sohn
(1999). We have chosen the more recent categorization here.
(5.) The morpheme -so or -uo referred to as the blunt form by Sohn
(1994, 1999), the semiformal style by Lee and Ramsey (2000), and the
authoritative style by Martin (1964), is said to be disappearing from
the language, and as such, will not be included in the current
discussion.
(6.) Hankwuke, the textbook series published by Korea University (1991).
(7.) Eun and Strauss (2004) propose a preliminary hypothesis based
entirely on the notion of 'information status,' where the
deferential form is used to impart information that is 'new'
or framed as 'new,' and the polite form to impart
'given/shared' information or information that is framed as
such.
(8.) Matsumoto's (1988: 416) use of "face" here
relates to the Japanese conception of face, which, according to
Matsumoto is categorically different from that proposed by Brown and
Levinson (1987). In fact, Matsumoto's entire premise (and one with
which these authors do not agree) is that the concept of negative face,
as proposed by Brown and Levinson, is not applicable in the Japanese
context, since the Brown and Levinson model is based on individual face
wants (both positive and negative), and, for Matsumoto, especially in
the case of negative face wants, individual "territory" does
not exist; rather, Japanese individuals must constantly acknowledge
dependence upon others within the society.
(9.) In both cases, namely, the positive FTA (correction) and the
negative FTA (reminder of a past commitment), the Japanese utterance,
while still containing the polite form of the copula desu (if uttered in
a context requiring that speech level), would probably also be marked
with some type of redressive marker (Brown and Levinson 1987) to
mitigate the force of the FTA, be it an FTA to the addressee's
positive or negative face. A more natural sounding utterance in one of
these contexts might take the following shape:
(i) anoo, kyoo wa doyoobi desu kedo HM today TM Saturday COP-POL
but 'Uhm, but today is Saturday.'
In this case, the hesitation marker anoo ('uhmm') would
soften the initial intrusion of the utterance, and the final particle
kedo ('but, however') would position it as a more "off
record" (Brown and Levinson 1987) correction or reminding. See Park
(1997) for a detailed discussion of Japanese kedo 'but' and
Korean nuntey 'and/but' as "accountability relevance
markers" in conversation.
(10.) With the exception of the graduate student conversations,
these data may be somewhat hybrid in nature, with the spontaneous speech
also containing elements of 'script' and the purportedly
scripted data containing elements of extemporaneous talk. However, the
spontaneous speech is predominantly spontaneous and the scripted speech
is predominantly scripted.
(11.) The polite marker -yo can also occur in conjunction with
isolated nominal expressions, but we excluded these from our counts
since there is no deferential counterpart for this grammatical
structure.
(12.) From the Strauss and Kawanishi (1994) database.
(13.) As pointed out in Eun and Strauss (forthcoming), native
speakers often self report that they do not alternate between honorific
sentence enders, and that non-alternation is standard.
(14.) The single token of the -yo form in dataset 10 appears toward
the end of the broadcast and is uttered in the context of a weather
report delivered by a female. And here, it is affixed to the discourse
marker nuntey ('but'), as part of the construction: nonfinite
V+nuntey+yo, not to a finite verb.
(15.) Although the majority of instances of deferential verbal
morphology in our examples and data occur in the declarative mood, our
claim is that any use of the deferential (e.g. in the interrogative mood -[s]upnikka and the imperative mood -[u]sipsio) carries the semantic
message of +BOUNDARY.
(16.) We have investigated the etymological source of this particle
in Korean reference grammars and dictionaries and have also consulted
with a number of Korean linguists. Our efforts have uncovered nothing
with respect to its source beyond functional descriptions and folk
etymologies (e.g. that it derived from Japanese particle yo, which has
the same surface representation, but which serves an altogether
different purpose).
(17.) While Ochs refers to this marker as an "epistemic"
indexical, its function is also clearly deontic in nature. The same is
true for the excerpt in (9) and Ochs' discussion of its target
particle.
(18.) The authors are grateful to Eunju Kim for allowing us to use
her data here.
(19.) This excerpt appears as examples (10) and (14) in Eun and
Strauss (2004).
(20.) Note that another significant characteristic of the
deferential form is that it typically resists interactional particles
such as -ci, -canha, -ney, -tela, -kwun, etc., all of which serve an
amplificatory function with respect to the personalization of on-going
interaction. See Strauss and Eun (in prep.).
(21.) This phenomenon is reminiscent in part of Gumperz'
(1982: 80-81) discussion of code-switching and the personalization
versus objectivization dynamic that underlies ir in particular contexts.
Gumperz cites a specific example in which a speaker code-switches from
Slovenian to German in countering the previous turn by his interlocutor;
the code switching at that moment in the talk "lends (the
speaker's) statement more authority" (Gumperz 1982: 80).
(22.) Note regarding translation of (18): Korean is an SOV language. The first few lines of Mrs. Co's narrative begin with the
subject pronoun ce (nun) 'I (TM)' and ends with the finite
verb chacsupnita 'look for.' The objects of the verb
'father,' 'older brother,' 'older sister,'
and 'younger sister' are placed between the subject and verb.
(23.) Note that the presence of -keyss is not required to express
future time, as evidenced by the contrast of examples (15), (16), and
(17); -keyss appears in (15) and (17), but not in (16), yet all express
immediate present/future time reference.
References
Agha, Asif (1993). Grammatical and indexical convention in
honorific discourse. Journal of Linguistic Anthropology 3(2), 131-163.
--(1994). Honorification. American Review of Anthropology 23,
277-302.
Atkinson, J. Maxwell; and Heritage, John (1984). Structures of
Social Action: Studies in Conversation Analysis. Cambridge: Cambridge
University Press.
Brown, R.; and Gilman, A. (1960). The pronouns of power and
solidarity. In Style in Language, T. A. Sebeok (ed.), 252-282.
Cambridge, MA: MIT Press.
Brown, Penelope; and Levinson, Stephen (1987). Politeness: Some
Universals in Language Usage. Cambridge: Cambridge University Press.
Duranti, Alessandro (1997). Indexical speech across Samoan
communities. American Anthropologist 99(2), 342-454.
Enfield, N. J. (2003). Demonstratives in space and interaction:
data from Lao speakers and implications for semantic analysis. Language
79(1), 82-117.
Eun, Jong Oh; and Strauss, Susan (2004). The primacy of information
status in the alternation between the deferential and polite forms in
Korean. Language Sciences 26, 251-272.
Gumperz, John (1982). Discourse Strategies. New York: Cambridge
University Press.
--; and Hasan, Ruquiya (1976). Cohesion in English. London: Longman
Group, Ltd.
Hanks, William F. (1990). Referential Practice: Language and Lived
Space among the Maya. Chicago: University of Chicago Press.
--(1992). The indexical ground of deictic reference. In Rethinking
Context: Language as an Interactive Phenomenon, A. Duranti and C.
Goodwin (eds.), 43-76. New York: Cambridge University Press.
Hwang, Juck-Ryoon (1990). "Deference" versus
"politeness" in Korean speech. IJSL 82, 41-55.
Jakobson, Roman (1971). Shifters, verbal categories and the Russian
verb. In Selected Writings of Roman Jakobson, vol. 2, 130-147. The
Hague: Mouton.
Jespersen, Otto (1965). The Philosophy of Grammar. New York: W.
Norton.
Kirsner, Robert S. (1979). Deixis in discourse: an exploratory
quantitative study of the Modern Dutch demonstrative adjectives. In
Discourse and Syntax, Syntax and Semantics Vol. 12, Talmy Givon (ed.),
355-375. New York: Academic Press.
--(1993). From meaning to message in two theories: cognitive and
Saussurean views of the Modern Dutch demonstratives. In
Conceptualizations and Mental Processing in Language, Richard A. Geiger
and Brygida Rudzka-Ostyn (eds.), 81-114. Berlin and New York: Mouton de
Gruyter.
Lee, Hyo Sang (1991). Tense, aspect, and modality: a
discourse-pragmatic analysis of verbal suffixes in Korean from a
typological perspective. Unpublished doctoral dissertation, UCLA.
--(1999). A discourse-pragmatic analysis of the committal -ci in
Korean: a synthetic approach to the form-meaning relation. Journal of
Pragmatics 31, 243-275.
Lee, I. and Ramsey, S. R. (2000). The Korean Language. Albany, NY:
State University of New York Press.
Lee, Keedong (1993). A Korean Grammar on Semantic Pragmatic
Principles. Seoul: Hankwuk Mwunhwasa.
Levinson, Stephen (1983). Pragmatics. Cambridge: Cambridge
University Press.
Lukoff, Fred (1982). An Introductory Course in Korean. Seoul:
Yonsei University Press.
Lyons, John (1977). Semantics, vol. 1. Cambridge: Cambridge
University Press.
Martin, Samuel (1964). Speech levels in Japan and Korea. In
Language in Culture and Society, Del Hymes (ed.), 407-415. New York:
Harper and Row.
Matsumoto, Yoshiko (1988). Reexamination of the universality of
face: politeness phenomena in Japanese. Journal of Pragmatics 12,
403-426.
Ochs, Elinor (1990). Indexicality and socialization. In Cultural
Psychology: Essays on Comparative Human Development, James W. Stigler,
Richard A. Shweder, and Gilbert Herdt (eds.), 287-308. Cambridge:
Cambridge University Press.
--(1992). Indexing gender. In Rethinking Context: Language as an
Interactive Phenomenon, Alessandro Duranti and Charles Goodwin (eds.),
335-358. New York: Cambridge University Press.
--(1996). Linguistic resources for socializing humanity. In
Rethinking Linguistic Relativity, John J. Gumperz and Stephen Levinson
(eds.), 407-437. Cambridge: Cambridge University Press.
Park, Yong-Yae (1997). A cross-linguistic study of the use of
contrastive connectives in English, Korean, and Japanese conversation.
Unpublished doctoral dissertation, UCLA.
Peirce, Charles Sanders (1955). Philosophical Writings of Peirce.
New York: Dover Publishers.
Platt, Martha (1986). Social norms and lexical acquisition: a study
of deictic verbs in Samoan child language. In Language Socialization
across Cultures, Bambi Schieffelin and Elinor Ochs (eds.), 127-152.
Cambridge: Cambridge University Press.
Quirk, Randolph; Greenbaum, Sidney; Leech, Geoffrey; and Svartik,
Jan (1985). A Comprehensive Grammar of the English Language. London:
Longman.
Sacks, Harvey; Schegloff, Emanuel; and Jefferson, Gail (1974). A
simplest systematics for the organization of turn-taking in
conversation. Language 50(4), 696-735.
Schegloff, Emanuel A.; Jefferson, Gail; and Sacks, Harvey (1977).
The preference for self-correction in the organization of repair in
conversation. Language 53(2), 361-382.
Scott, J. C. (1990). Domination and the Arts of Resistance. New
Haven, CT: Yale University Press.
Silverstein, Michael (1976). Shifters, verbal categories, and
cultural description. In Meaning in Anthropology, Keith Basso and H.
Selby (eds.), 11-55. Albuquerque, NM: University of New Mexico Press.
Sohn, Ho-min (1994). Korean. London: Routledge.
--(1999). The Korean Language. Cambridge: Cambridge University
Press.
Strauss, Susan (2002). 'This,' 'That,' and
'It' in spoken American English: a demonstrative system of
gradient focus. Language Sciences 24, 131-152.
--(in preparation). Indexicality, stance, and demonstrative
reference in spoken and written discourse.
--; and Eun, Jong Oh (in preparation). Cognitive realization
markers: a discourse-pragmatic study of the sentence ending particles
-kwun, -ney, -ta, and -tela.
Strauss, Susan; and Kawanishi, Yumiko (1994). Korean earthquake
database.
Suh, Kyung-Hee; and Kim, Kyu-hyun (1993). The Korean modal marker
-keyss as a marker of affect: an interactional perspective. In
Japanese/Korean, Patricia M. Clancy (ed.), 98-114. Stanford, CA: Center
for the Study of Language and Information, Stanford University.
Treichler, P. A.; Frankel, R. M.; Kramarae, C.; Zoppi, K.; and
Beckman, H. B. (1994). Problems and problems: power relationships in a
medical encounter. In Language and Power, C. Kramarae, M. Schulz, and W.
O'Barr (eds.), 62-88. Beverly Hills, CA: Sage.
