Because social models are usually constructed explicitly or implicitly for a particular
issue with a particular readership in mind -- e.g. in the case of Japanese, to plan the
Allied Occupation of post-war Japan, to explain Japan's rapid economic growth in the 60s,
to solve US-Japan trade imbalances in the 80s -- the adequacy of each model is relative to
the objectives of the investigation. It is, therefore, justifiable, in principle, for
researchers to use linguistic data selectively to meet their needs. However, when a
model is applied to unintended areas, this selectivity of linguistic data can lead to a
distorted view of the speech community. This, in fact, is the case with the Japanese
language. The present study (i) reviews several major works covering both the Japanese
language and Japanese society, and (ii) demonstrates that the prevailing group model of
Japanese society is inconsistent with the notion of self as encoded in the Japanese
Groupism vis-a-vis individualism refers to such characteristics as 'the individual's identification with and immersion into the group, conformity and loyalty to group causes, selfless orientation towards group goals, and consensus and the lack of conflict among group members' (Yoshino 1992:19). The Japanese are said to be 'extremely sensitive to and concerned about social interaction and relationships' (Lebra 1976:2). 'Even in intimate groups there are strong pressures to conformity, which many have seen as the source of the deepest psychological malaise in Japanese society' (Smith 1983:56). To provide an anecdotal example of Japanese group consciousness, Nakane (1970:2-3) points out that the Japanese commonly introduce themselves with their affiliation, rather than with their personal attributes. For example, a Japanese person is more likely to say 'I'm from X University' or 'I belong to Y Company' rather than 'I'm a psychologist' or 'I'm an engineer.'1
Since Ruth Benedict's 1946 book, The Chrysanthemum and the Sword, Japanese society has often been characterized as hierarchical: it involves vertical stratification by institution or group of institutions, rather than horizontal stratification by class or caste, and each group is vertically organized based on the relationships between paternalistic superiors and their subordinates (Nakane 1970). The society assumes loyalty from below and benevolence from above.
Underlying this vertical society is said to be the Japanese personality called amaeedependence/indulgencef(Doi 1973). Amae is 'the feelings that normal infants at the breast harbor toward the mother -- dependence, the desire to be passively loved, the unwillingness to be separated from the warm mother-child circle and cast into a world of objective "reality"' (Bester 1973:7). This attitude of dependence is reported to be carried into adulthood, and dependence on others' benevolence is encouraged during socialization processes of the Japanese (DeVos 1985:165). This type of dependency is considered to occur in group settings: subordinates, who play the child role, can seek dependence on their superior, and the superior, who plays the parent role, is expected to display benevolence (Yoshino 1992:18).
Closely linked to Japanese group orientation is the notion of uti. The translational approximation of uti is 'inside', but uti is also used to refer to the speaker's own home, house, or household. Uti is also commonly associated with 'insider(s)'. Uti and its antonym soto 'outside/outsider' are said to constitute 'a major orgaizational focus for Japanese self, social life, and language' (Bachnik 1994:3), and without these concepts, much of Japanese behavior is 'inexplicable (at least from a Western perspective)' (Wetzel 1994:74). Uti and soto are also the key words to connect groupism and contextualism -- the latter concerning the notion of Japanese self.
Characterized as relational and social, Japanese self is said to be situationally defined (Araki 1973). Advocates of this view claim that in Japan, relationships between individuals are prioritized over individual self -- which is more central to the Western notion of self (Bachnik 1994:18), and that 'the identification of self and other is always indeterminate in the sense that there is no fixed center from which, in effect, the individual asserts a noncontingent existence' (Smith 1983:81). Furthermore, 'proper use of Japanese teaches one that a human being is always and inevitably involved in a multiplicity of social relationships. Boundaries between self and other are fluid and constantly changing, depending on context and on the social positioning people adopt in particular situations' (Kondo 1990:31).
To recapitulate, compared with 'the Western notion of self,' Japanese self is claimed to be unstable, constantly shifting, and context dependent. It is imperative to discuss here what the term 'Western notion of self' is used to refer to. Because most works I consulted use 'Western self' as a reference point to which Japanese self is compared and explained, I could not find elaborate discussions of the Western self. However, it appears that what is widely assumed is the Cartesian notion of self: 'I am transparent to myself: My self-knowledge is mediated neither by inference nor by any teleological element such as a passing purpose or project. I know my own identity directly and completely, whereas others know it only inferentially and relative to certain sets of purposes' (Boër & Lycan 1986:139). Like most authors cited in the present work, I use the term self in this sense.
The group model with fluidity of self is customarily utilized to account for linguistic phenomena in Japanese, most notably for address/kinship terms, extended use of donatory verbs, and the polite/honorific systems.
Regarding kinship terms, the unmarked choice for the word corresponding to mother is okaasan. It can be used to refer to the addressee's as well as a third-person's mother. It can also be used for the speaker's own mother in a conversation with in-group members or in informal conversations with outsiders. In formal conversations, on the other hand, haha must be employed rather than okaasan to refer to one's own mother. The same distinction is made for father, grandparents, siblings, and other close relatives. This demonstrates that an appropriate choice of kinship terms depends on context -- whether or not the addressee is an insider and the formality of the conversation.
An example to illustrate fluidity of boundaries between the self and others is drawn from the usage of donatory verbs. There are two kinds of verbs corresponding to give: kure- and age-. With kure-, the inherent destination of the transfer is the speaker, i.e. the self. To indicate that Okada gave money to her, the speaker would say:
(1)a. okada-san ga okane o kasite kure-ta.
Lit. 'Mr. Okada gave [me a favor of] lending money.'
Kure- can also be used when the recipient of the transfer is regarded as an insider, as shown in (1b).
(1)b. okada-san ga haha ni okane o kasite kure-ta.
Lit. 'Mr. Okada gave [my mother a favor of] lending money.'
Thus, a common analysis treats insiders as extended self, i.e. the boundary between the self and others is considered to be shifted.2
Fluidity of in-group/out-group boundaries can be observed in the proper usage of honorific/humble forms of the predicate. Speaking with a colleague about their company president, an honorific form should be used when the president is encoded as the subject (2a), whereas a humble form should be used with the self as the subject (2b). However, when the addressee is a client, the speaker is conventionally considered to be representing the company, and a humble form must be used even when the subject referent is the company president, as if the president is part of, or belongs to, the (extended) self. Suppose that the president's name is Tanaka. Because syatyoo 'company president' can be categorized as an honorific referential term, rather than a neutral descriptive term, the use of syatyoo in this context is inappropriate, much less the honorific form of the predicate. In this situation, a sentence such as (2c) is mandatory.
(2)a. syatyoo wa syusseki nasai-masu.
'The President will attend [the meeting].'
(2)b. watasi wa syusseki itasi-masu.
'I'll attend [the meeting].'
(2)c. tanaka wa syusseki itasi-masu.
'Tanaka (i.e. the president) will attend [the meeting].'
From these characteristics of the Japanese language, Wetzel (1984, 1994) concludes that
the 'I' of Indo-European languages is not the universal deictic ground, and that the
deictic anchor point in Japanese is a collectively defined vantage point represented as
(3)a. watasi wa samu-i.
'I feel cold.'
(3)b. *haha wa samu-i.
'My mother feels cold.' [Intended]
(3)c. haha wa samu-gatte iru.
Lit. 'My mother is showing signs of feeling cold.'
(3)d. haha wa samu-soo da.
'My mother appears to be feeling cold.'
This restriction on psych predicates and their potential subjects is so inflexible that when the predicate is polysemous, the semantic role of the subject necessarily shifts to conform to the restriction. In (4), kanasi- 'be sad' can indicate either the subject's being sad (subject = experiencer) or some entity causing a sad feeling in the speaker's mind (subject = stimulus/source). When the subject is the speaker, the former construal automatically applies; otherwise, the latter does.
(4)a. watasi wa kanasi-i.
(4)b. haha wa kanasi-i.
'My mother makes me sad.'
Not: 'My mother is sad.'
Omow- 'think' is another example of psych predicates, and, as such, it cannot be used to describe a third-personfs mental state.
(5)a. watasi wa haha wa byooki da to omo-u.
'I think my mother is ill.'
(5)b. haha wa byooki da to omo-u.
'I think my mother is ill.'
Not: 'My mother thinks she is ill.'
(5)c. haha wa (zibun wa/ga) byooki da to omo-tte-i-ru.
'My mother thinks she is ill.'
In (5a), the matrix subject, watasi 'I', of which omow- is predicative is overtly present; in (5b), by contrast, watasi does not appear and the overtly-present potential subject of omow- is haha, and yet this construal is impossible: the subject of omow- must still be watasi. For a third-person subject, the auxiliary phrase te-i-ru must be added to omow-. (If the embedded subject, zibun 'self' in (5c), were absent, it would be construed with the matrix subject or an arbitrary referent.)
This constraint is not on the potential subject per se, but on the accessibility of information. For example, the author of a novel can use psych predicates freely with third-person subjects because s/he, as creator, can be omniscient, and therefore has direct access to a character's mental state. In fact, expressions like (6) are commonplace in narrative literature.
(6) akiko wa haha wa byooki da to omo-tta.
'Akiko thought her mother was ill.'
The relevant notion of self here is absolute and cannot be relative or context
dependent. Furthermore, the boundary between the self and others cannot be fluid. That
is, as shown in (3b, 4b, 5b), typical in-group members such as haha 'Mother' cannot
be regarded as extended self. This very fact indicates that the Japanese language is
extremely self-conscious, however primordial such a notion of self may
be.4 And it is difficult to envision that such a language has been
formed by a speech community lacking the concept of individual self, akin to the Western
notion of self, contrary to those who contend that Japan is a 'selfless' society. This is
a well-known fact in Japanese linguistics, but it has never received serious attention in
research on Japanese society. The prevailing relational model of self is inconsistent with
this aspect of the Japanese language.
First, linguistic expressions based on the absolute self are learned by children automatically and unconsciously, whereas expressions requiring the notion of relational self must be taught explicitly by adults. First graders are not confused as to whose opinion is represented in (7), with a psych predicate omow- 'think'. It is the speaker's own, not her mother's or teacher's or whoever is present in the discourse.
(7) aki-tyan kuru to omo-u.
'[I] think Aki will come.'
NOT: 'Aki thinks [she/someone else] will come.'
On the other hand, the use of haha to refer to one's own mother is usually taught in the upper grades of elementary education.
Second, most dialects of Japanese provide neither dual kinship terms nor honorific/humble forms of verbals. The elaboration of the honorific system began in Late Old Japanese by members of the Japanese aristocracy living in and around Kyoto, the capital city at the time. Unlike the Tokyo dialect, which borrowed a considerable number of honorific expressions from the Kyoto dialect, most other dialects do not have the elaborate honorific systems that are frequently cited in the literature on Japanese society (cf. Shibatani 1990:123-26).
Third, expressions that presuppose shifting (relational) self as illustrated in (2) are artificial, not a natural part of language as linguists conceive it. Manipulation of honorific and humble forms according to the situation is not learned naturally. Many companies provide their employees with honorific-language lessons as part of on-the-job training. Those employees, most of whom hold college degrees, have not acquired such a language in their twenty-plus years of life. Books teaching how to use honorifics properly are always in great demand and found in virtually all book stores in urban communities.
While it is indeed amazing that the Japanese continue to maintain their elaborate honorific language at such a high cost, as far as the Japanese language is concerned, the presence of individual (absolute) self is unquestionable.
With a few exceptions,5 the researchers who advocate a multiple or relational notion of Japanese self have attempted to derive it based on behavioral observations; as Bachnik (1992:152) asserts, 'Japanese choose appropriate behavior situationally, from among a range of possibilities, resulting in depictions of the Japanese self as "shifting" or "relational."' This statement is a non sequitur, for holders of individual self can and should also behave appropriately in various situations.
Bachnik's idea of relational self is not as simple as it sounds, however. She argues that the crucial question for defining self is 'to look away from "fixing" a unified typology of self [i.e. individual (absolute) self vs. relational self] to the process of "fixing" a series of points along a sliding scale for a self which is defined by shifting. If movement between different modes is central to the organization of self, then the question of how movement is initiated and defined at a given "point" along the sliding scale is the central organizing factor for the double continuum [e.g. uti and soto] which social behavior is identified, not as a general set of behaviors which transcends situations, but rather as a series of particular situations which generate a kaleidoscope of different behaviors which are nonetheless ordered and agreed upon' (155, emphasis in original). Backnik's notion of self appears to be constantly shifting, but at any given moment, it must be identifiable at some point along a sliding scale. Let us now examine whether or not this is the case.
As mentioned earlier, the verb kure- 'give' is used when the favor is made for the benefit of the speaker, and the speaker is grateful for it. It can also be used when the beneficiary is an insider, e.g. Mother. In the theory of relational self, this is made possible because the boundary between the self and others is shifted: a third-person beneficiary (the speaker's mother in this case) is included in the notion of relational self. We have also observed that the use of psych predicates, such as omow- 'think', is restricted to the first-person subject. Let us examine the case in which these two kinds of predicates co-occur in a single sentence.
(8) okada-san ga haha ni okane o kasite kure-ru to omo-u.
'I think Mr. Okada will give my mother [a favor of] lending money.'
NOT: 'My mother thinks Mr. Okada will give [her a favor of] lending money.'
In (8), the destination of the favor of lending money is the speaker's mother, but the
subject of omow- remains constantly the first person, watasi. That is, the
use of psych predicate is still impossible with a third-person subject. This fact
indicates that the self and the mother must belong to different conceptual categories --
one permits a direct expression about a mental state based on the guaranteed accessibility
to the source, but the other does not, due to the lack of such
Since Basil Hall Chamberlain (1850-1935), Western observers have depicted Japanese national cultural traits in an Orientalist tradition, which shares ontological assumptions about the West as the universal reference point and the exotic and inferior other (Minear 1980:508). Those observers were fascinated with some exotic characteristics of Japan. Benedict (1946:2) depicted Japan as fundamentally paradoxical: 'the Japanese are both aggressive and unaggressive, both militaristic and aesthetic, both insolent and polite, rigid but adaptable, submissive and resentful of being pushed around, loyal and treacherous, brave and timid, conservative and hospitable to new ways.' (This is like saying that Americans are so rich that they even own private air planes, and yet live on the street!)
While Japanfs constructed unity was created by Western observers, such an imaginary community has been supported and extended by the Japanese themselves, for representation and dissemination of various ideologies and myths are necessary parts of construction of a unified nation. This strategic 'Japaneseness,' characterized by Miller (1982) as 'self-Orientalism,' maximizes national interests and minimizes individualism.
Iwabuchi (1994) analyzes that in the process of Japan's self-Orientalization, which has been done most extensively in the last fifty years, images of the 'West' have been discursively created:7 Western nations were considered as superior to be emulated, but they were also condemned as individualistic and selfish. He goes on to assert that it was in this context that self-Orientalisation was utilized as a strategy to construct and self-assert Japan's national cultural identity, and 'the West' was necessary for Japan's invention of tradition, the suppression of heterogeneous voices within Japan, and the creation of a modern nation whose people are loyal to Japan.
Both the West's Orientalist characterization of Japan and Japan's self-characterization tend to use the other to formulate the self and to repress the heterogeneous voices within: heterogeneous voices of people within the nation have been repressed through the homogenizing discourses of an imaginary 'us' versus 'them,' although Japan is neither static nor homogenous nor is it closed as a society (Iwabuchi 1994).
In the late 1970s, researchers began casting doubt upon the validity of the group model of Japanese. In comparison with people of other societies, are the Japanese more group-oriented and accustomed to vertical organization? Do they place more emphasis on consensus and social harmony, and value more deeply group membership or social solidarity? Do they have underdeveloped egos, and lack an autonomous sense of self-interest? (Mouer & Sugimoto 1986:11-14).
Rosenberger (1992:13) asserts that the Japanese are 'not essentialized individuals despite economic and social changes. ... Japanese self emerges as neither entirely collective nor completely individualistic. In fact, [the essays in the same volume] imply that whether or not Japanese are becoming individualistic in an American sense is the wrong question, one itself rooted in Western dichotomies. The more appropriate question is, what shifts occur as Japanese people make Western lifestyles and concepts of individuality part of their own processes of self and social relationship.' I counter her opinion with the contention that the real questions here are (i) what concepts the term individuality signifies in each work on the relationship between Japanese society and its language, and (ii) what evidence the advocates of the group model provide to support their claims.
The most essential feature of self is self-awareness, which is generated and fostered through self-other interaction and the symbolic processing of information (Lebra 1992:105). Symbolic processing of information in Japanese requires a clear demarcation between the self and others, whereas such a demarcation is less significant in English, which permits sentences like 'Mom wants to go shopping,' without any evidential expressions. Therefore, there is no basis to assume that the Japanese have a less-developed concept of self than do Westerners. And, if the Japanese were 'much more likely than Westerners to operate in groups or at least to see themselves as operating in this way' (Reischauer 1977:125), the ultimate cause should be sought elsewhere than in the cognitive domain.
Concerning evidence for the group model, there is a fundamental problem in methodology:
the claims are frequently, if not always, based on anecdotes, without providing a clear
argument to what extent such anecdotes are representative of a larger population, as
eloquently discussed by Mouer and Sugimoto (1986:130-33). For example, it has been pointed
out that the Japanese travel abroad in groups; they bathe together in groups; in Japanese
festivals a group of young men carry a portable shrine (ibid.). In the presence of
countless counterexamples, such anecdotes cannot be used as evidence.
Bachnik, Jane. 1992. Kejime: Defining a shifting self in multiple organizational modes. In N. Rosenberger, ed., Japanese Sense of Self, 152-72. Cambridge: Cambridge University Press.
Bachnik, Jane. 1994. Uchi/Soto: Challenging our conceptualizations of self, social order, and language. In J. Bachnik and C. Quinn, eds., Situated Meaning: Inside and Outside in Japanese Self, Society, and Language, 3-37. Princeton: Princeton University Press.
Benedict, Ruth. 1946. The Chrysanthemum and the Sword. Boston: Houghton Mifflin.
Banfield, Ann. 1982. Unspeakable Sentences: Narration and Representation in the Language of Fiction. Boston: Routledge and Kegan Paul.
Boer, Steven, and William Lycan. 1986. Knowing Who. Cambridge, Mass.: MIT Press.
DeVos, George. 1985. Dimensions of self in Japanese culture. In A. Marsella, G. DeVos, and F. Hsu, eds., Culture and Self, 141-84. New York: Tavistock.
Doi, Takeo. 1973. The Anatomy of Dependence. Trans. by John Bester. Tokyo: Kodansha International.
Iwabuchi, Koichi. 1994. Complicit exoticism: Japan and its other. Continuum: The Australian Journal of Media & Culture 8:49-82.
Kondo, Dorinne. 1990. Crafting Selves: Power, Gender, and Discourses of Identity in a Japanese Workplace. Chicago: University of Chicago Press.
Kunihiro, Masao. 1976. The Japanese language and intercultural communication. The Japan Interpreter 10: 267-83.
Kuroda, S.-Y. 1973. Where epistemology, style and grammar meet: A case study from Japanese. In S. Anderson and P. Kiparsky, eds., Festschrift for Morris Halle, 377-91. New York: Holt, Rinehart and Winston.
Lebra, Takie S. 1976. Japanese Patterns of Behavior. Honolulu: University of Hawaii Press.
Lebra, Takie S. 1992. Self in Japanese culture. In N. Rosenberger, ed., Japanese Sense of Self, 105-20. Cambridge: Cambridge University Press.
Miller, Roy A. 1982. Japan's Modern Myth: The Language and Beyond. New York: Weather Hill.
Minear, Richard H. 1980. Orientalism and the study of Japan. Journal of Asian Studies 30:507-17.
Mouer, Ross, and Yoshio Sugimoto. 1986. Images of Japanese Society: A Study in the Structure of Social Reality. London: KPI.
Nakane, Chie. 1970. Japanese Society. London: Weidenfeld and Nicolson.
Reischauer, Edwin. 1950. The United States and Japan. Cambridge, Mass.: Harvard University Press.
Reischauer, Edwin. 1978. The Japanese. Cambridge, Mass.: Harvard University Press.
Rosenberger, Nancy. 1992. Introduction. In N. Rosenberger, ed., Japanese Sense of Self, 1-20. Cambridge: Cambridge University Press.
Shibatani, Masayoshi. 1990. The Languages of Japan. Cambridge: Cambridge University Press.
Smith, Robert. 1983. Japanese Society: Tradition, Self and the Social Order. Cambridge: Cambridge University Press.
Wetzel, Patricia. 1984. Uchi and soto: Social deixis in Japanese. Ph.D. dissertation, Cornell University.
Wetzel, Patricia. 1994. A movable self: The linguistic indexing of uchi and soto. In J. Bachnik and C. Quinn, eds., Situated Meaning: Inside and Outside in Japanese Self, Society, and Language, 74-87. Princeton: Princeton University Press.
Yoshino, Kosaku. 1992. Cultural Nationalism in Contemporary Japan. London:
1 This generalization on Japanese preference of group over individual attribute neglects the crucial aspect of self-introduction; one selects introductory statements according to the presumed knowledge of the addressee and the purpose of the subsequent conversations. Usually a Japanese introduces him-/herself in the way Nakane describes when it is already known that the speaker is a scholar or a company employee. Only then, stating onefs affiliation becomes natural and relevant. (Because being a company employee is a default in modern urban Japan, the phenomenon mentioned by Nakane is commonly observed.) It sounds strange, however, to say 'I'm from XYZ University' unless the addressee already knows that the speaker is a college student or professor or somehow affiliated with a university.
Smith (1983:82) reports that many Japanese men have two different cards, the business card that includes rank in his organization, business address, and telephone number, and the personal card that bears only his name, home address and telephone number. The latter is said to be given out only to those in the closed circle of one's acquaintances. Although I myself do not know anyone practicing this, it indicates that one's affiliation is not an automatic part of identification.
2 Another, equally common, analysis posits three concentric circles: a fixed self as the innermost circle, in-group surrounding the self, and out-group representing the rest of universe. In this analysis, kure- is considered to be used for an inward transfer, e.g. from an outsider to the self or an insider, or from an insider to the self.
3 The direct representation of subjective experiences (represented consciousness) other than the speaker's own yields what Banfield (1982) refers to as an unspeakable sentence, i.e., one which cannot naturally occur in spoken language. Kuroda (1973) calls the style where such expressions appear only with first-person subjects the reportive style. In reportive style, only the speaker is entitled to express one's own psychological state. In the nonreportive style, by contrast, a third-person subject is permitted for such expressions.
4 Naturally, keen observers have recognized that the 'individual Japanese [is] a very self-conscious person' (Reischauer 1950:143).
5 Lebra (1992) proposes three dimensions of self: the interactional self, the inner self, and the boundless self. The interactional self is relative, multiple, and variable in accordance with where and how self stands with respect to others. The inner self is less relative, more stable, fixed, and purely subjective. The boundless self is embedded in the Buddhist version of transcendentalism, disengaging from dichotomies between subject and object, self and other, inner and outer realms, existence and non-existence, and so forth.
6 One may wonder why the use of kure- is possible in (8) when the destination of the favor of lending money is the speaker's mother, not the speaker herself, because in the theory of relational self, the mother is considered to be included in the speaker's extended self, and yet she is excluded from the potential subject of omow-. As mentioned in note 3 above, it is possible to characterize the use of kure- differently: a fixed self as the innermost circle, in-group surrounding the self, and out-group representing the rest of universe. This latter analysis does not conflict with the fact that the first person alone can be the subject of psych predicates.
7 An example of discursively created images of the 'West' can be drawn from Kunihiro (1976), who claims that the cognitive behavior of the Japanese is different from that of Westerners. According to him, Westerners employ the dualistic Aristotelian logic, but the Japanese do not rely on such reasoning processes.
Copyright (c) 1998 Yoko Hasegawa