Professional Documents
Culture Documents
Shoko Ikuta
Cornell University
ABSTRACT
This paper attempts to analyze the mecSanisms involved in speech level shift in
Japanese conversational discourse that contains the des/-rnas2 forms. The use of
des/-mas is usually regarded as a matter of 'politeness' or 'formalness'. However,
by focusing on Japanese ethnography and cohesion in discourse, this study reveals the
basic meaning of des/-mas to be 'distance' which may be social, attitudinal, or
cohesional. In face-to-face conversations, where frequent shifting is observed despite
social condition remaining unchanged, level shift is often used (1) to signal the flow
of empathy between speakers - the use of ]-Distant] level coincides with a speaker's
attitudinal closeness to his interlocutor at any moment whereas [+Distant] represents
his attitudinal distance - and/or (2) to indicate coherence and the hierarchical posi-
tioning of utterances in a discourse.
The traditional study of des/-mas has usually been such that the forms are
explained in terms of 'formalness' of the social conditions involved in a conversation.
Degree of formalness is determined by (1) the ingroup/outgroup relationships of the
interlocutors, (2) the social setting involved, and/or (3) the nature of the information
being given.
These factors certainly account for a great deal in conversational discourse, but
cannot account for all of the complex shifts of speech level that can be observed in
actual conversation.
When focus is set on Japanese ethnography and cohesion of discourse, it is found
that the meaning potential of des/-mas is concerned not so much with 'politeness'
or 'formalness' as with 'distance' that may be social, attitudinal, or cohesional. This
study will thus show how the shift between the use and the non-use of des/-mas
works as a conversational strategy.
38 Language Sciences, Volume 5, Number 1 (1983)
The selection of an appropriate level, i.e., the use or non-use of des/-mas, for any
given situation requires three factors, viz., social condition, mutual empathy, and the
coherency of discourse. Although these factors seem to be independent of one
another on the surface, a common notion underlies them. It is 'distance'. (See Peng
1974 and 1975 for earlier discussions of this concept.)
The following examples show the structural contrast of the use and the non-use
o f des/-mas; the (a) sentences lack the morphological modification by des/-mas
fomas, the (b) sentences, on the other hand, contain des/-mas forms.
In the traditional approach, the contrast between the (a) and (b) sentences has
been regarded as a matter of 'formalness', and explained in terms of the speaker-
addressee relationship.
According to Hinds (1976), many aspects of "formal speech" in Japanese
depend exclusively on the relationship between the speaker and the addressee(s), as
that relationship is perceived by the speaker; that is, whether a speaker says iku
(non-formal) or ikimasu (formal); sukida (non-formal) or sukidesu (formal) depends
on whether he views the addressee as an intimate friend or not. If the addressee is
an intimate friend, the former expressions are used; if he is not, the latter expressions
axe used.
This argument is plausible but limited. Granted that this may be one basis for
choice of level of formalness, other factors must be taken into account in the study of
actual discourse.
lde (1982) discusses that the des/-mas form is used when the speaker's deference
toward the addressee is expressed, and proposes the social rules which control the
use o f des/-mas as well as other forms o f expressing 'politeness'. The major factors
involved in her rules are (1) social position, (2) power, (3) age, and (4) formalness.
These factors may be sufficient to account for the normative use of des/-mas forms
Speech Level Shift mad Conversational Strategy 39
DATA
The data analyzed in this paper are excerpts from natural conversations
collected for the most part from TV talk shows and interviews; each conversation
carried out was natural in that it was not read from a prepared material, and that the
setting and participants remained unchanged throughout. That natural conversations
were used as data is because we are not interested in the meaning of a level from an
isolated utterance, but rather in the sequential flow of a conversation where the
40 Language Sciences, Volume 5, Number 1 (1983)
social conditions remain constant from beginning to end. The social conditions o f
these conversations, as explained below, require, in a traditional grammar, the use o f
des/-mas. The level observed, however, is not "frozen." It shifts back and forth
between the use and the non-use o f des/-mas. Yet, each o f them is a natural-
sounding conversation, and the occasional non-use of des/-mas does not mean that
the speaker lacks communicative competence. (See Ikuta 1980 for the analysis
o f a larger text.)
Japanese data cited will be romaized, following Jorden (1963). In the text,
des and -mas will be italicized, so as to indicate the location o f the [ + / - Distant]
feature. Each predication which has a [ + / - Distant] feature (except that which
modifies a following noun) will be numbered and analyzed individually. An English
equivalent will be provided after each excerpt for the convenience of the reader.
Word-for-word or morpheme-for-morpheme translations will be avoided, however,
since they are not relevant to our discussion.
EMPATHY
Although we have narrowed the range of data, since the social factors which
more or less determine the choice of speech levels stay put in each of the examples
below, the examples are sufficient to illustrate level shifting within each sequential
conversation.
The first excerpt is from a conversation between two female speakers, K
(Kuroyanagi) and J (Jo). It illustrates level shift from ]+Distant] to [ - D i s t a n f ] and
back to ]+Distant]. The shift occurs in Line (3), where [+Distant] des is momentarily
omitted:
Excerpt 1.
1. K: Sono oheya wa kositu ni natte iru fi desu ka? [+Distant]
2. J: Ee, roku-zyoo to yo-zyoo-hafi to safiruumu ga
taihefi hiroi fi desu no. [+Distant]
3. K: Maa, Zuibufi ii no ne. [-Distant]
4. J: Hitori ni wa tyoodo ii fi desu. [+Distant]
5. K: Otonari nahka wa sizuka na ~ desu ka? [+Distant]
6. J: Ee, boooh ga tyafi to site iru fi desu. [+Distant]
Note that the participants do not change, the setting is stable, and the subject
matter remains the same. What, then, is the effect of [-Distant] in Line (3)?
Before discussing this shift, it is relevant to clarify the social conditions of this
talk. Social factors which do most to determine the choice of a level are sex, age,
occupation, education, social status of participants, their relationship, place, time
and purpose of talk, and the subject matter. These are shown below:
Table 1
Social Conditions of Excerpt 1
K J
Participants: Sex female female
Age 40-45 70-75
Occupation actress writer
Education college high school +
Relationship not close friends
Social setting: Time at dinner (evening)
Place at a restaurant
Occasion informal dialogue
Subject matter: Conversation as a whole J's life
Present excerpt J's apartment
In line with the above information, the level most likely to be chosen for this
dialoque is [+Distant]. It is also predictable that Speaker J will have the option
to use [-Distant] occasionally, while Speaker K will not, since Speaker J is much
older than K, and K is interviewing J, albeit in an informal conversation. In fact,
[-Distant] is often observed in J's speech for the rest of the conversation.
The answer to the question, 'Is the use of [-Distant] in Line (3) by Speaker K
explainable as a haphazard deviation?', therefore, is 'No'. The fact that the excerpt
is a perfectly natural-sounding Japanese conversation indicates that there must be
some other factor which has led to the choice of [-Distant], despite the social
restrictions shown above.
An important factor is speaker's empathy with his dyadic partner. This can be
related to the notion of distance involved in the use of des/-mas. Thus, [+Distant]
indicates that the speaker is maintaining a certain distance while speaking, and
[-Distant] signals that he feels no need to maintain the same distance at that
moment. In other words, [+Distant] gives an utterance a more oblique effect in
discourse, and [-Distant] a more direct effect by means of which the concept of
'distance' is defined. It follows that a speaker can demonstrate momentary empathy
42 LanguageSciences, Volume 5, Number 1 (1983)
1. [+Distant] K ®, :
',
2. '-.. ~ J [+Distant]
3. Maa, zuibufi ii no ne. [-Distant] ~ :
4. ,.--'" q9 [+Distant]
5. [+Distant] ® I
,o
i
Attention must be paid to the role of Line (3) in the context, which indicates
positive evaluation of the preceding c o m m e n t by Speaker J. Alternatively, the
following version might also be possible, thereby carrying the same grammatical
meaning.
Speech Level Shift and Conversational Sl~ategy 43
i
1. [+Distant] ® K i
2. o, J [+Distant]
3a. Maa, zuibufi ii fi desu ne. [+Distant] ® "i
4. [+Distant[
5. [+Distant] @
6. i
[+Distant]
This example also has the potential to be a natural conversation. It is, however,
significantly different from the original one. Because of [+Distant], Line (3a)
sounds more like a flat statement: the admiration is more perfunctory, in contrast
to Line (3) in which [-Distant] reinforces agreement. This [-Distant] in a
[+Distant] environment not only retains the grammatical meaning of Line (3), but
also effectively makes it sound more genuine. In Japanese society, expression of
empathy in the positive evaluation of another party is preferred socially. Empathetic
( [ - D i s t a n t ] ) utterances support more strongly the person receiving the positive
evaluation. 3 On the contrary, Line (3a) conveys less genuineness by employing
[+Distant].
In cases where the social setting is extremely formal - for example, where
Speaker J is socially superior and/or the two speakers are total strangers to each
other - the use of [+Distant] might be more appropriate for the utterance discussed
above. Here, [+Distant] would be considered socially unmarked. However, if
[+Distant] were used instead of [-Distant] in Line (3), the [+Dxstant] would be
considered marked, since it would violate ethnographic expectations of empathy at
the moment.
On the other hand, there are cases where [+Distant] must be used even in a
[-Distant] environment. Failure to choose [+Distant] on such occasions may result
in a loss of effectiveness in communication. These are domains where empathy is
neither expected nor allowed. The use of [-Distant], when empathy is inappro-
priate, may affront the dyadic partner and, accordingly, lead to his reluctance to
continue the conversation or even the relationship with the speaker.
The excerpt below is from the second half of the conversation from which the
first excerpt was extracted. The participants and the setting are identical; however,
the context has changed, so that [-Distant] has become the dominant level.
Excerpt 2.
1. K: Situreedesu kedo, [+Distantl
2. Zyoo-safi wa zutto dokusifi de irassyaru no? [-Distant]
3. J: Iie, ano ne, hi-do oyome ni itta no. [-Distant]
44 Language Sciences, Volume 5, Number 1 (1983)
What is important here is that [-Distant] does not always create a friendly,
sincere atmosphere, even when the relationship between the interlocutors is close.
It can also momentarily destroy attitudinal closeness that has developed during
the preceding conversation. [-Distant] can have the effect of 'too much too soon',
thereby violating the dyadic partner's personal domain, and making him more
cautious and less prone to continue the conversation.
On the other hand, if both Lines (1) and (2) were [+Distant], a different effect
would obtain:
DISCOURSE COHESION
Excerpt 3.
1. S: Soo desyoo ka. [+Distant]
2. Da tte tenisu mo sugoi desyoo? [+Distant]
3. K: Ie, hyaku-ma.fito too, nihyaku-mafi to too, tenisu-
zifikoo wa huete ori masu keredomo, [+Distant]
4. doo site mo toppu ga dame na ~ desu. [+Distant]
5. Mukasi no hoo ga zyoozu datta desu ne. [+Distant]
6. Shimizu-safi, kumagai-safinafite, dehai ni hatu-
safika de ikinari kessyoo made itta. [-Distant]
7. Sore-igo sofina seeseki ageta hito wa hitori mo inai. [-Distant]
This excerpt has been taken from a conversation in which the participants are
considered equal in terms of socml status and are not especially intimate with, or
distant from, each other. In this particular conversation, three women two of whom
are present in Excerpt 3 - all specialists in sports like tennis, golf, and skiing - are
discussing their own experiences as well as criticizing recent trends in the world of
sports. The social setting leads us to predict that reciprocal ]+Distant] will be the
dominant level throughout the conversation. However, a number of utterances
characterized by [ - D i s t a n t ] are observed. Speaker K commences a reply to S with
]+Distant] (as in Lines [3], [4], and [5]), then shifts to [--Distant] in Line (6).
This shift cannot be interpreted as the result of social factors or of empathy.
Discourse structure may be useful in explaining this type of level shift. The
structure of K's sequential utterances (3) to (7) may be illustrated as follows:
Speech Level Shift and Conversational Surategy 47
K: A Topic
3. [+Distant]
4. [+Distant]
5. [+Distant]
Illustrative instances
6. [-Distant]
7. [-Distant]
From a structural point of view, it can be said that Lines (6) and (7) are subordinate
to Lines (3) to (5), since (3) to (5) are autonomous in the discourse, comprising a
reply to Lines (1) and (2), while (6) and (7) are dependent on previous utterances (3)
to (5), and without them would have no contextual link with the entire discourse.
In this example, level shifting from [+Distant[ to [-Distant] coincides with a
structural shift to a subordinate part in the discourse, and implies that there are
correlations between level shift and discourse structure.
To facilitate the identification of those 'units' or 'parts' of discourse and to
better view the construction o f a discourse in terms o f them, an analysis according
to the 'context spaces' proposed by Reichman (1978) may be useful here. s The
term 'topic' is not appropriate as a label for this kind of discourse unit, because it
has been used in various ways in linguistics.
Reichman in her study of conversations groups utterances into distinct context
spaces and formalizes the types of relationships that spaces can have with each other.
A "context space' is defined as 'a group of utterances that refers to a single issue or
episode'. 6 Hereafter, in this paper the term 'space' will be used in this sense. A
conversation is, at a deeper level, a structural entity of hierarchically related spaces,
though it is merely a sequence of utterances on the surface.
She mentions specific linguistic rules that must be employed to enable smooth
context space transitions, and demonstrates five major mechanisms: (a) Clue Word
Shifts and Deictic Expression, (b) Explicitly Labelled Shifts, (c) Mode of Reference,
(d) Repetition o f Words, and (e) Tense Shift. All of these seem relevant to Japanese
conversational discourse, with the addition of at least three more mechanisms,
namely, shift in speech level, extended predication - "no da' construction - and the
use of sentence final particles. Our focus here is on the first mechanism - level shift. 7
As Reichman points out, a conversation is not a mere temporal sequence of
unrelated context spaces. Context spaces themselves relate to one another and form
hierarchical networks. With this perspective, she categorizes six relationships
between context space sequences: (a) Illustrative and Restatement Relations, (b)
48 language Sciences, Volume 5, Number 1 (1983)
Generalization Relation, (c) Interruption and Return Relations, (d) Subissue and
Joining Relations, (e) Respecification Relations, and (f) Total Shift Relation. In the
analysis o f Japanese conversation structure, I have found similar types o f relation-
ships between space sequences, and found that level shift is often used to indicate
a transition between context spaces, and the hierarchical positioning of an utterance
in a discourse.
Speaker K's utterances in Excerpt 3, for example, may be interpreted to have an
illustrative relation. The shift from [+Distant] to [ - D i s t a n t ] in Line (6) signals the
hierarchical relationship of the first space ( S I ) and a subspace ($2) as illustrated
below:
3. K: [+Distant]
4. [+Distant]
5. [+Distant] S1
Shift - -
6. [-Distant]
$2: Illustration o f ( 3 ) - (5)
7. [-Distant]
The correlation between level shift and discourse structure is often found at a
point in time where level shifts between [+Distant] and [ - D i s t a n t ] . The correlation
is not limited to illustrative relation as in the above example, but can be seen in all of
the sex relationships as listed above, a
Excerpt 4.
1. K: Ima demo sono kata to wa otukiai ga aru fi desu ka? [+Distant]
2. J: Ee, watasi no hofi ga deru to sasiagetaxi site ne. [ -Distantl
3. K: Sore wa suteki. [-Distant]
4. Nafiniti-gurai no ryokoo datta ~ desu ka? [+Distant]
Speech Level Shift and Conversational Strategy 49
The shift from [+Distant[ to [-Distant] in Line (3) demonstrates the former
mechanism of level shift. Since Line (3) is a positive comment on the preceding
utterance, the shifting to [-Distant] does not violate social constraints, but gives
more genuine effect to K's admiration. (Line [2] is not discussed here, since Speaker
J's choice of [-Distant] is socially optional as explained under Excerpt 1. However,
at the same time, Line [2] may be interpreted as an illustrative subspace of Line
[1]; therefore, [-Distant] in Line [2] seems to be motivated by both factors, i.e.,
social and cohesional.)
The shift from [-Distant] to [+Distant] in Line (4) is motwated by the space
structure. The focus in the excerpt is on J's trip to Europe. In the first space, the
interlocutors are talking about the relationship between J and another member of
the tour group she traveled with. Here, K shows an empathetic shift to [-Distant]
in Line (3) as mentioned above. Then, in Line (4), she turns to talk about how many
days the trip was. The new space is opened by a question marked as [+Distant]
which is the dominant level in the whole conversation. The shift from [-Distant] to
[+Distant[ here signals both an opening of a new space and the hierarchical
positioning of the utterance in the discourse. The new space is not a subordinate
space of the immediately preceding space. It is independent from the first space in
this excerpt.
The following diagram depicts the interaction of empathy and space construc-
tton as revealed in level shift at the surface.
I. K: [+Distant] SI K ®
Shift ---,
2. J: [-Distant] , oj
$2: Illustration 4
3. K: [-Distant]
i
i Total
Shift '~) :
4. [+Distant[ shift
Q- :
$3 I
I
[-Distant] forms. However, deeper into the conversation level shift can be used to
reflect a relationship as it changes from moment to moment.
Discourse cohesion is a secondary factor in level shift. As long as it does not
conflict with shifting that is motivated by the flow of empathy between interlocutors,
construction of context space also motivates level shift. This mechanism is
illustrated in Figure 5.
• [+Distant]
- [-Distant]
NOTE: ~ = neutral/fuzzy
NOTES
1. The earlier version of this paper was read at the 1981 Annual Meeting of the
Association for Asian Studies, Japanese Socioliguistics Session, Sheraton Center,
Toronto, Canada, March 13-16, 1981.
2. Des and -rnas are used to represent desu, -masu and all of their derivatives.
3. [-Distant] can, thus, work as an expression of 'positive politeness' (cf. Brown
and Levinson 1978).
4. [+Distant] has an effect to express 'negative politeness'.
52 Language Sciences, Volume 5, Number 1 (1983)
5. Grimes (1980) applies the term 'reference space' to the similar notion, pointing
out that the term 'context' might cause a misunderstanding.
6. Reichman also proposes a subdivision of context spaces into 'issue spaces',
which concern a generalized activity or state, and 'event spaces', which concern
a particular episode that has occurred in the past. Here, this subdivision, is
avoided since these two spaces are not clear-cut in practice, and are irrelevant
to the present study.
7. These seven mechanisms other than level shift, though they are not considered
here, may also indicate space transition and help establish the hierarchical frame-
work of a conversation. Speech level shift often reinforces signaling by these
mechanisms, and is reinforced by them, when both are present.
8. See Ikuta (1980) for more detailed analysis of level shift in terms of space
relationships listed here.
9. The range of data analyzed in this paper is narrowed as [+Distant] level is
dominant in the environment where each excerpt was extracted. For extended
data analysis, see lkuta (1980).
10. Our focus is on the right-hand side of the dotted line in the model.
REFERENCES