Japanese Sociolinguistics Author(s): Janet S. Shibamoto
Source: Annual Review of Anthropology, Vol. 16 (1987), pp. 261-278 Published by: Annual Reviews
Stable URL: http://www.jstor.org/stable/2155872 Accessed: 10-03-2017 13:16 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/2155872?seq= 1 &cid=pdf-ref erence#ref erences_tab_contents You may need to log in to JSTOR to access the linked references.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://about.j stor.org/terms
	
0	
m	
Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review of An thropology
STOR
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
Ann. Rev. Anthropol. 1987. 16:261-78
Copyright © 1987 by Annual Reviews Inc. All rights reserved
JAPANESE SOCIOLINGUISTICS
Janet S. Shibamoto
Department of Anthropology, University of California, Davis, California 95616
INTRODUCTION
Japanese sociolinguistics, wholly independent of trends in American sociolinguistics, got off to a quick start immediately after World War II in the late 1940s under the rubric "language life research" (71).
Japanese sociolinguistics officially began in December, 1949, with the establishment of the Kokuritsu Kokugo Kenkyujo (National Language Research Institute, hereafter NLRI). This government-formed institute was created for the purpose of "conducting scientific studies regarding the national language and the linguistic life of the Japanese people and building a sound basis for the rationalization of the national language" (Article 1, NLRI Establishment Act). Primary among the charges to the fledgling institute was the description of the "language life" (gengo seikatsu) of the Japanese people and of their "linguistic culture" (gengo bunka). Findings from the Institute's various research projects were to be used as bases for the formulation of official policy on language and language education (16). The research program of the NLRI was characterized from its inception by the use of sociological survey methods, by the statistical manipulation of quantitative linguistic data, and by an emphasis on large-scale group projects rather than individually conceived and executed research. In sum, the Institute has from the beginning been devoted to questions of the relations between language and society and to the quantitative analysis of language form in its context of use (35, 50, 62).
The study of the interrelations between language and society in Japan, then, has a long and well-established history. Despite this, Japanese work in this area differs from Western sociolinguistics both in theoretical motivation and in its relation to the concerns of applied linguistics. Neither do Japanese sociolinguists share with their Western counterparts assumptions about the
261
0084-6570/87/1015-0261 $02.00
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
262 SHIBAMOTO
abstract nature of language form that underlie Western linguistics in general.
There are two separate schools of language study in Japan, each characterized by separate topics, methods, and interpretive principles. The first of these, kokugogaku (national language studies), subsumes gengo seikatsu (language-life studies), under which term most of the work considered here has been done. The second school, gengogaku (linguistics), subsumes the much smaller, newer field of Western-influenced shakaigengogaku (sociolinguistics).
An understanding of the distinctive characteristics of kokugogaku will facilitate discussion of how gengo seikatsu studies, or the investigation of language-society relations, has proceeded in Japan. This branch of language study treats the Japanese language. It stands in contrast to the study of the universal principles of language in general, and in this way remains independent of the field of gengogaku. Both groups study language, so it is natural that considerable overlap, particularly with regard to description, be found. Whereas Western linguistics is characterized by interest in cross-language investigations aimed at discovering the underlying similarities between linguistic systems and by a concern for the links between the universal aspects of language and the nature of the mind, kokugogaku concerns have centered around immediate descriptive goals. This tendency to strict de-scriptivism is intensified by the narrow focus of interest on Japanese alone.
Also evident is a lack of concern for intersubjectively reliable methodologies at the levels both of data collection and analysis. The literature contains few descriptions of the principles of linguistic analyses. Indeed, it often seems the case that little thought concerning these matters has preceded analysis. Readers are left to guess at the methods used by the analyst to extract portions of sentences for investigation. The analysis rests upon the assumption, well founded for the most part, that the audience is as competent in Japanese as the investigator, and as able to make the same sorts of sentence analyses.
Studies in gengo seikatsu developed from within the kokugogaku school and adopted its basic orientation toward detailed description and its relative lack of interest in cross-linguistic comparison. The study of bilingualism and of minority or ethnic group vernaculars has also received virtually no attention in Japan until very recently (6, 49). Language-life studies have instead focused on problems that are natural extensions of the traditional interest of national language grammatical research and have to a large degree adopted their descriptive, particularistic focus as well. This is most true of the first period of investigation of the relation between sociocultural forms or processes and language, 1948-1963, which saw the establishment of the NLRI and the commencement of numerous large-scale jittai chosa (survey of actual conditions) aimed at determining the facts of language use (35). During the 1960s, generative-transformational interests dominated Japanese language
This content downloaded from 147.251.99.135 onFri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 263
studies, and less attention was devoted to language-life or sociolinguistic concerns. In the 1970s, however, there was a renewal of concern for the language life of the Japanese, this time under the influence of a "new" field introduced from the west—sociolinguistics. Under this influence, some Japanese scholars have adapted their research to address the issues of identifying the contextual and speaker-identity features that influence language production, of defining the individual speaker's relationship to his or her speech communities, and of delineating the process of ongoing language change, all issues of central concern to sociolinguistics as practiced in North America, Europe, and elsewhere in the West. The progress made by each group toward fuller understanding of language form and function in its sociocultural context are discussed below. Preceding that discussion are brief comments on the demographics of the field, how it is treated in introductory linguistics courses, and on the use of linguistic findings in the formulation and maintenance of Japanese identity myths.
It was noted above that the field shakaigengogaku was not only much newer but much smaller than gengo seikatsu. In the 1982 membership roster of the Linguistic Society of Japan, only 15 scholars appear who list themselves as sociolinguists; the National Language Research Institute alone has had, since the 1960s, at least 30 permanent staff conducting research on language life. Moreover, non-NLRI researchers in the area of gengo seikatsu are concentrated in a small number of institutions around the country, and their projects, accordingly, tend to be large-scale group projects. Sociolinguists are scattered and the scale of their projects more tailored to the individual conduct of research.
Students were traditionally introduced to the study of language through the classic kokugogaku text of Saeki et al (50); this text incorporated a discussion of gengo seikatsu, which covered primarily an early history of the field and an introduction of the notion that language form and usage may both affect and be affected by sociocultural considerations. More recently, introductory linguistics texts have discussed language and societal relations under the term sociolinguistics. Coverage of the field has grown from a mere three-page discussion in one representative text from the mid-1970s to an entire chapter in another text by the same authors just a few years later (66, 67). This chapter covers language variation accounted for by age, status, setting, occupation, and sex differences, this last particularly focusing on women's use of hon-orifics. It also covers the sociolinguistic emphasis on investigation of vernacular forms, pidgins and Creoles, and bilingualism; but these are introduced without reference to any Japanese examples or possible Japanese applications.
Finally, I wish to point to the existence of a large body of pseudoscientific literature in what Miller terms "contemporary sociolinguistic speculation" (33; for examples, see 15, 65, 69, 70). This speculation centers on beliefs
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
264 SHIBAMOTO
about the uniqueness of the Japanese people. One of the major sources of evidence for this uniqueness is linguistics; an extensive literature cites "unique" characteristics of the Japanese language which are taken as direct evidence for cultural and genetic uniqueness. This literature makes an interesting addition to our understanding of how language and culture are intertwined. Two excellent reviews of this intriguing subfield of language-society/language-culture studies are available in English (33, 34).
MAJOR AREAS OF SOCIOLINGUISTIC RESEARCH IN JAPAN
Shibata (61) identifies four primary areas of interest within Japanese sociolinguistics. These are (a) issues relating to the national writing system, (b) loan word usage, (c) the process of language standardization, and (d) honorifics. These topics were the central concerns of the early large-scale projects conducted under NLRI auspices in various regions of Japan (17, 18, 20, 21). Immediately following the publication of his overview of the field, Shibata himself, along with others in the field, began noting in print that gengo seikatsu was undergoing redefinition, incorporating elements from both sociolinguistics and speech act theory/language symbolism (10, 62). They pointed to shifts in focus with some elements of earlier studies newly coming to the fore and others simply representing new interests. Among these are surveys of language use in urban, specialized occupational settings with an emphasis on politeness and honorifics (36, 64), language and sex or gender (7, 11, 29, 48, 59), and a small but growing interest in issues of minority languages (14, 49) and bilingualism (6).
National Writing
The complexities of the Japanese orthography are well known. Japanese is written in a mixture of Sino-Japanese characters (kanji) with multiple pronunciations determined by the lexical context, plus two syllabaries used for different purposes. The syllabaries, like alphabets, have a fixed number of symbols, are tied to Japanese phonology, and are easily learned; kanji have none of these properties. Further, in addition to the burden imposed on the prospective reader or writer of Japanese by the need to memorize thousands of kanji and their various readings, Japanese writing has been plagued by instability in the choice of a particular kanji to represent a given word, instability in the portion of a given word that was covered by the kanji as opposed to the portion encoded in hiragana (one of the two syllabaries), and instability in spelling conventions when using one or the other of the syllabaries. A literate populace was essential to the government policy of rapid reconstruction after World War II; how to accomplish that widespread,
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 265
effective literacy became an issue both in governmental circles and among national-language/language-life scholars. Clearly, the orthography required some "fixing" (63). What fixing is necessary is left in the hands of language scholars—many of them working in the area of gengo seikatsu.
The problems raised by the Japanese writing system are reducible to the following four principles: economy (whether a small number of elements in the system is sufficient or not), clarity (whether each element can be clearly distinguished from all others or not), systematicity (whether each element is in a balanced relationship to all others or not), and stability (whether the symbol and its pronunciation are fixed or not). The interactions of these principles are sometimes harmonious and sometimes in conflict; where they are in conflict, the system as a whole is unstable and irregular.
Numerous questions could be formulated which sociolinguistic investigations might properly address. By investigating the changes occurring in kanji and syllabary usage over time, for example, the relative importance of reduced numbers of readings per character (stability) as opposed to balance in the inflectional system (systematicity) could be measured. The process and direction of change over time could be investigated for any pair of these principles when they conflict. These are parts of linguistic structure that have hitherto received little attention, and principles that have not been foci of sociolinguistic research on language change. Investigations of this sort would be of use in formulating public policy with regard to the teaching of the orthography in the schools and the efficient diffusion of literacy throughout the population (the primary charges to the committees and agencies involved with orthographic research). It would also allow Japanese sociolinguists to make major contributions to the understanding of language change in progress in entirely new ways—i.e. with different and less numerically limited populations expressing themselves in a formal, standard language.
In the case of orthographic issues, investigators were charged with providing the official committees in charge of formulating national policy information about the writing system such that they would be able to balance the need for a simple, easily learnable and unconfusing writing system with adequate recognition of the merits of the well-established and widely spread tradition of using Sino-Japanese characters.
The first information required was an accurate assessment of just what tradition had been and currently was. Not surprisingly, therefore, language-life scholars approached the national writing problem in the manner for which their national language research background had prepared them. They began quickly and efficiently to prepare detailed descriptions of sentences, in this case taken as samples of orthographic usage, that existed in printed form. Through the description of what "was," then, it was assumed that those in charge of policy would be able to determine what should be.
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
266 SHIBAMOTO
Throughout the 1950s and early 1960s surveys of orthographic usage and competence appeared one after another; the largest of the major surveys undertaken was a survey of the literacy of the populace conducted in the summer of 1948 and reported three years later (72). Surveys conducted in 270 locations throughout the country with a sample of 21,008 selected by random sampling techniques tested the literacy levels of the Japanese citizenry. In this survey, subjects ranging in age from 15 to 64 were tested for their ability to read newspapers, notices, handbills, and personal letters. Not surprisingly, literacy was found to be linked to a combination of age and level of education; that is, younger adults tended both to have higher average levels of education and to be more literate than older respondents. Additional surveys of Sino-Japanese character use in magazines and newspapers also appeared (19, 23). From these it is possible to determine (a) the frequency of use of a given character and (b) its function—that is, in what word or words it is used. The same sorts of information are available for words (or inflections) written in the kana syllabaries. Since the surveys have some historical depth, changes in use may also be charted for particular forms. These facts were taken as proper criteria by which to determine what constituted a tradition, and these traditions then became the basis of policies about what was to be taught in the compulsory education system, how official documents were to be written, and so forth.
Hayashi (5) reports on the results of these surveys in a typical fashion, indicative of the research aims of the field. After describing the surveys, he notes that the number of times each Sino-Japanese character was used in a particular survey (that is, in a woman's magazine, in a general magazine, or in a newspaper) was counted and the characters ranked by the number of times they were used in each medium. Calculations were then made to rank their rates of use across the three different media, and each character was assigned a number between 0 and 6, with 0 indicating that the character was between the most frequently used and the 500th most frequently used, 1 indicating that it was between 501st and 1000th most frequently used, and so on. Category 6 is composed of characters not found at all. These rankings were then matched against the various lists of characters used in the education system: education kanji (881), bikoo kanji (115), hosei kanji (candidates for addition to the common-use list) (28), and other (the remaining common-use kanji and name characters).
This cross-matching allowed Hayashi to categorize characters into sets of those that are commonly used, but not on the list for teaching in the period of compulsory education; those that are on that list but are not commonly used; those that are on (or off) the toyo kanji (general-use character) list but are (not) frequently used; and so on. These patterns are taken to reflect changes in usage that should be taken into account in the ongoing process of reformulat-
This content downloaded from 147.251.99.135 onFri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 267
ing the official lists of kanji with their approved readings, to be taught in the national language education system and used in an official documents.
Work in this area continues as Japanese orthography is adapted to the requirements of word processing, as the dissemination of computers in Japanese society is seen to have major potential effect on how individuals use the orthographic variability available to them (40, 53).
Loan Word Usage
Postwar interest in this area comes from the language contact between Japan and Europe and the United States under conditions that encouraged borrowing from English and other European languages. Research focuses on how to nativize the resultant "inundation" (hanran) of new lexical items. Issues in this area center on determination of what proportion of loan words appear in various types of written and spoken discourse and what a desirable proportion would be (24, 25), development of transliteration conventions (9), and investigations into the interaction between the newly introduced loan words and regular inflectional morphological and morphophonemic processes (60). This last is of particular interest to sociolinguists because it involves certain morphological markers of politeness and beautification shown to be highly sensitive to speaker identity and contextual factors.
Standardization
Japanese researchers on language standardization assumed a smooth progress toward standardization once exposure to the standard language had occurred; no discussion of possible reasons for retention of dialect such as have been prominent in Western sociolinguistics under the terms overt and covert prestige have emerged in the Japanese literature. The early jittai chosa (survey of actual conditions) and subsequent follow-up studies (20, 26) do not, however, support the hypothesis of smooth progress along a continuum away from dialect toward the national standard.
The major work on language standardization began in 1951 with an NLRI large-scale survey of 9 phonetic and phonological variables in the speech of Tsuruoka, Yamagata Prefecture. Respondents in the study, selected by random sampling procedures, were scored for degree of language standardization depending on the rates of production of standard variants of the 9 variables in a list of 31 items. The scoring procedure was simple: The respondent received a score of 1 on an item judged to have the standard variant, and a score of 0 on items judged not to have the standard variant. No intermediate scores were given. Degree of standardization was held to be the average of all scores for a particular respondent. In this first study, investigators found that standardization was linked to age; the younger respondents exhibited more standard patterns of phonetic production than did older respondents, with a peak in
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
268 SHIBAMOTO
standard variant productions occurring in the 20-24 year old group. Twenty years later, production of standard variants was higher in all groups and had almost reached completion among 15-19 year olds. An attempt was made in this second study to reexamine the speech of the original respondents. A match of their 1971 productions with their 1951 productions showed an interesting result: Although standardization was proceeding across age at a fairly steady rate, individuals over the age of approximately 25 showed little change in their speech patterns over 20 years. This suggests that phonological form is fairly well fixed by that age and not highly vulnerable to changes in context of use, a fact of considerable interest to those attempting a theoretical account of language change in progress. Egawa (3) notes also that although all phonetic forms were grouped in these studies, in fact not all nonstandard forms are proceeding to standardization at the same rate: The replacement of [kwa] by standard [ka] is nearly complete, but resolution of the dialect merger of [i] with [e] had hardly begun even by 1978. Unfortunately, little detailed investigation of the linguistic and/or extralinguistic reasons for the differences in rate of standardization of individual forms has followed these observations.
Another interesting observation emerged from the second Tsuruoka study upon a multivariate analysis of the data: Speakers fell into two categories exhibiting different patterns of standardization. One set of speakers uses a combination of dialect and standard forms in most if not all contexts; the other group uses dialect or standard forms distinctively across different settings. What sort of speaker falls into each group is not, however, made clear, so that sociolinguistic generalizations about the relation of speaker identity and strategies for language use under conditions of ongoing phonological change cannot be addressed within the context of these studies.
In addition to a number of smaller-scale studies that focus primarily on the link between age and standardization (13, 27), studies on language standardization have been made that focus on generation (63), setting (4), and the interaction of national and regional standards in setting directions for language change in local-dialect communities (12). A particularly interesting example of such a small-scale study was conducted by Mase (32). He demonstrates that the influence of television is very strong in promoting the standardization of pitch-stress pattern among the generations of young Japanese exposed to television during early language acquisition, even when the pitch-stress pattern of the language heard on television (that of standard Japanese) conflicts with a more proximate regional urban standard. The interplay of phonological process with specific lexical items leading not to standardization but to the emergence of shinhogen (new dialects) is also noted (46). New Dialect forms are nonstandard, used among young people, and regarded as informal; even in Tokyo, the "stronghold of standard Japan" (46:39), new expressions are seen not only to flow outward from city center to
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 269
the suburbs and beyond but also toward city center from the suburbs. Ogino et al explain this phenomenon by pointing to the increased contact between city and suburban residents, but it seems plausible that the notion of covert prestige used to mark solidarity within an urban youth culture also influences the emergence of these New Dialect forms.
A great deal of work on language change has been done in Japanese under government auspices, with a view toward understanding and speeding the process of standardization, and many aspects of language change in Japan are thoroughly documented as a result. One area has been neglected but is cited in Nomoto (37) as a rich field for future work: Even though phonologically, morphologically, and syntactically, local dialects may merge with standard Japanese, dialect-like differences may remain in "ways of thinking about things" (37:162). Nomoto is referring to such differences as that between the Kansai affirmative pattern of request in utterances such as Totte kuremasu ka "Will you pass it to me" as opposed to the Kanto negative pattern Totte kuremasen ka "Won't you pass it to me?" As Nomoto notes, it may be supposed that these differences between dialect and standard obtaining at levels above the sentence have considerable effect on standardization processes, but just what these effects might be is as yet entirely unexplored.
Honor ification
The existence of various speech levels indicating formality and politeness is a well-known characteristic of Japanese. An early English-language description of the respective dimensions of Japanese formality and politeness (honorifica-tion) is available (30), as is an expanded discussion (31). Hayashi (5) also presents a detailed discussion of both the psychological and sociological characteristics of speakers and the contextual factors that affect honorific usage; in this collection of articles first published in the 1960s and early 1970s, Hayashi treats honorific usage from a distinctly sociolinguistic perspective but does not provide reference for empirical studies supporting his analysis.
Empirical support does exist. Two early studies were conducted by the NLRI in the early 1950s in Ueno, Mie Prefecture and Okazaki, Aichi Prefecture. Additionally, a large-scale jittai chosa was conducted nationally, also under the auspices of the NLRI, in 350 rural and urban sites, including Ueno and Okazaki (22). This study was conducted in three stages. First, sociological and biographical information was collected from all informants, who were then surveyed about their actual honorific usage and their attitudes toward the use of the various honorific forms. Finally, the specific honorific forms identified in the second stage of the survey were matched with other linguistic characteristics (such as length of utterance), specific honorific forms were correlated with situational and speaker-identity features, and all of these were
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
270 SHIBAMOTO
correlated with reports of attitude toward honorific usage. All data was quantified and subjected to statistical analysis, with the result that 35 specific patterns were identified in postwar honorific use. These ranged from the highly predictable finding that the politeness of utterances increased as addressees were identified as more distant, to some unsuspected and hence more interesting observances. As investigators moved from east to west across Japan, for example, they found increasingly finer distinctions being made in the choice of particular honorific form both across addressee and across situation. While it is hardly surprising to find that sex of speaker is one of the most salient features influencing degree of honorification, moreover, the survey revealed a bifurcated pattern wherein women used higher levels of honorific form across all addressees in all contexts but men exhibited a much finer-grained sensitivity to hearer identity and setting, using a wider range of honorific (including humiliative) forms. Finally, the attachment of the honorific o- to nominals was consistently reported as being a misuse of honorification by respondents but was not noted or responded to negatively in actual conversation. Discussion of o- attachment and its use by middle-class urban female speakers is available in English in Shibata (60) and Shibamoto (59). This early study is extremely valuable in providing a baseline for the study of honorification in Japanese; but it is marred, as are most later studies, by the almost exclusive reliance on self-report of usage, with little observational or experimental support.
Separate reports on Okazaki are particularly interesting since the second is a follow-up survey conducted 20 years after the first, providing information concerning changes that have occurred in honorific use and attitude in the postwar era. The Okazaki studies also yield some quantified support for the hypothesis that female speech is more polite than male speech.
Until recently, work on honorifics has had a rural focus. In line with this focus is the work of Sanada (51, 52, 53), whose work in Maki Shuraku in the Gokayama district of Toyama Prefecture examines (using the entire community as respondents) the effect of sex, age, and relative status on honorific choice. He found that in the earliest stage of the research (1971), choice of an honorific verbal or pronominal form was influenced primarily by the rank of the family of the addressee (irrespective of the individual addressee's status relative to the speaker) but that by 1983 honorification was affected primarily by age, indication of the breakdown of the old ie (household) system and the prewar household ranking system commonly obtaining in rural areas. Sanada also notes that, particularly among the young men of the community, overall frequency of honorification is decreasing.
From the 1970s, honorification studies have had an increasingly urban focus. Two volumes representing broad-range surveys of urban language use conducted by the NLRI (28) include sections on honorific use, focusing as do the rural surveys on verbal and pronominal honorific forms, but this time with
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 271
a new attention to aisatsu or formulaic expressions. The large group project on the creation of a new "common" language in the late-developing Hokkaido city of Sapporo has focused in large measure on honorifics. Members of the Linguistics Department, University of Tokyo, carried out a field survey on honorifics at Sapporo in 1978; data were obtained from 503 informants and computer-processed, giving rise to the Ogino Quantification Method for mathematically expressing the politeness of word forms and speaker evaluation of address situations simultaneously (41). Six hypothetical address-situations were established on two axes: one of familiarity and one of vertical relation (elder vs younger, etc). Informants were asked what they would say in each situation. Differences across speakers were interpreted in the following four ways: (a) how wide a range of honorific forms was used, (b) how frequently humiliative forms were used, (c) whether an addressee rated as [+familiar] received plain or polite forms, and (d) which of the two axes had more effect on an informant's choice of honorific form. Findings (41, 42, 43, 44, 45, 63) were (a) the broadest range of honorific and humiliative forms were used by women, the elderly, and residents born in the Kanto region; (b) women used the most humiliative forms; (c) the factor most influential in determining whether a [+familiar] addressee receives plain or polite forms was age, with older speakers tending to use polite forms even with close friends, followed by residence type, with residents of single-unit dwellings more likely to use plain forms than residents of co-ops (manshon). This is held by Shibata et al (63) to indicate that residents in neighborhoods of single unit dwellings remain closely tied to the notions of social structure and organization that obtained in traditional regional society (chi'iki shakai), while residents of more modern urban type dwellings are closer to a national urban culture. Finally, (d) it appears that men respond more to a vertical axis of hierarchical social relations in deciding to use honorifics, whereas women are more evenly responsive to both hierarchical relations and familiarity.
A replication of the methods used in Hokkaido is reported by Ogino (44). A field survey on honorifics conducted in the Nezu and Nishikata areas of Tokyo showed that sex, age, and education were, in descending order, the three most important factors in determining politeness of utterance. Again, the method used was to present each respondent with an utterance or situation and ask what they would say in the different contexts or when addressing different types of hearers. As in the other studies, the analytic framework and methods of statistical manipulation used are highly sophisticated, and the work is carried out with the thoroughness, accuracy, and attention to detail characteristic of Japanese linguistic work in general; but the data collection procedure is so problematic that any findings must be viewed with caution.
A final recent focus in honorifics research is the examination of their use in the urban workplace. In 1975, the NLRI began a three-year investigation of honorific usage at a Hitachi production plant in Tokyo. In the first stage of the
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
272 SHIBAMOTO
investigation, all employees were sent a questionnaire; in the second stage, respondents to the questionnaire were grouped by section and interviewed individually for 30 minutes concerning honorific use at work. Finally, a subset of sections was selected for tape recording of all verbal activity, including telephone conversations, that occurred over a specified period (36). In general, indications are that occupational status is the major determiner of honorific use, with sex of hearer next in importance in shaping the speech of the Division Chief (bucho) class and above toward lower-ranking addressees. Speech directed toward low-ranked women employees is somewhat more polite than that directed toward low-ranked men. Few humiliative forms are used in the workplace, although this may be an artifact of the location of the particular production plant chosen. Humiliative use is more strongly associated with the western than with the eastern dialect areas of Japan. A follow-up study in a second Hitachi plant located in a rural area will link the study of honorifics in the workplace to the earlier studies of honorifics in regional society.
In these and other studies of the same type (8), a welcome addition to data collection methodology is the observation of actual speech, although little attention is as yet devoted to how the conditions of observation may affect analysis and the drawing of conclusions. One fact that emerges from a review of the type of work that has formed the backbone of Japanese sociolinguistics is that Japanese sociolinguists do not share the feelings of many Western sociolinguists concerning the existence of a single level of language that forms the locus patterning and regularity—the vernacular—and that they do not, either, share the conviction that empirically based studies on language use are only manageable in closed-network communities. Rather, as we have seen, Japanese sociolinguists have routinely addressed issues of language use across all levels of formality and in all kinds of communities, with the later, urban studies, in particular, focusing on middle-class groups of speakers in open-network speech communities. That this is not viewed as problematic in any way is, I believe, closely tied to the view that information on language use is as accessible to introspection and speaker judgment as is information on language form. Should the ample demonstrations in the Western sociolinguis-tic literature [beginning with Blom & Gumperz (1)] come to have an impact on data collection procedures in Japan, smaller-scale studies in more tightly bounded speech communities would be anticipated.
Language and Sex
Japanese is often cited as differing from English by virtue of having a "true" women's language. Its roots extend far back into the history of Japanese, but linguistic attention to sex-based differences in speech begins with the study of nydbd-kotoba, the language of the ladies of the court from the mid-Kamakura Period to the early Muromachi.
This content downloaded from 147.251.99.135 onFri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS
273
Jorden (11) enumerates these characteristics of women's speech: special self-reference and address terms, special sentence-ending particles and exclamations, a particular pitch range and set of intonations, frequent use of the honorific style, avoidance of kango (Sino-Japanese lexical items), and avoidance of vulgar language. Repeated mention of all these characteristics is found in the literature. The existence of special terms of self-reference and address, particularly the different sets of first- and second-person pronouns used by men and women, is one of the most frequently asserted sex-differentiated characteristics of Japanese.
Ide cites pronouns as "the first factor to catch one's eye" when seeking sex-differentiated forms in Japanese (7:37). Although there are numerous first- and second-person pronominal forms available to Japanese speakers, they are not used with great frequency. Ide claims, moreover, that when they are used, at least as subjects, they are selected on the basis of the sex of the speaker and hearer and the formality of the situation and topic.
For both first and second person, there are pronominal forms used by both sexes (watakushi, watashi T; anata, anta 'you') and forms used exclusively by men or by women (boku, ore 'Im') atakushi, atashi 'If'; kimi, omae, kisama 'youm'). Other forms exist, such as atai 'I,' used by small girls being babyish and was hi 'I,' used in standard Japanese by men of approximately 50 years of age and older with intimates and addressees of lower status; but these are clearly much more restricted in use and are not considered by Ide. Clearly, the forms used exclusively by one sex or the other are a marker of sex of speaker; Ide notes, however, that even those forms used by speakers of both sexes are used differently by men and women; her claim is that the difference resides in the interpretation of events along the dimension of formality, with women consistently producing speech that is more formal than that of men. Ide concludes that this differential use of pronouns stems from the fact that women are expected to be modest even where there is no difference in status among the participants in a conversation. It is appropriate, therefore, that they use more formal speech. Her argument seems, in the main, correct; she fails, however, to provide any more than anecdotal evidence in support of her assertions.
Lee (29) interviewed Japanese couples residing in the United States to elicit common terms of address and reference for spouses. She reported the following terms of address used between husband and wife, listed in descending order of frequency.
HU -> WI first name
kimi, omae
mama
okaasan 'mother'
WI -> HU
first name + san
otosan 'father'
papa
anata
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
274 SHIBAMOTO
Her study too, however, is flawed by lack of attention to the collection of supportive performance data: She has relied exclusively upon the self-report of native speakers who were, moreover, residing outside Japan at the time. She did not delineate the contexts in which these terms were used, nor indicate whether informants exhibited variability in choice of terms across various situations.
Nomoto (37) reports the proportions of kango (Sino-Japanese lexical items) used in interviews with men and women. The female interviewees employed kango 10.6% of the time, the male interviewees 13.7% of the time. Larger differences are observed in interviews conducted by Tsuchiya (cited in Nomoto 37:137). Tsuchiya taped conversations with men and women in formal and informal situations; proportions of kango used were calculated for each. In formal conversations men employed kango 22.5% of the time, women only 15.5% of the time. A possible explanation of this phenomenon is that a large proportion of kango are specialized terms, and relatively few women have the specialized knowledge that would authorize use of the terms. These figures are particularly interesting in light of the opinion commonly voiced by native speakers that sex-differentiated language and language use is restricted to informal contexts.
The use of honorifics is closely associated with stereotypes of Japanese women's speech. A typical characterization of Japanese women's speech is that it is more polite—that is, that women more frequently use honorific and humiliative forms than men do. As with most other syntactic and discourse phenomena, little sociolinguistic work has yet been done to support this claim. An exception is the work of Peng and his group on sex differences in turn length and type, verbal morphology, sentence final particles, word order, etc (48). Another is Shibamoto's work on sex-related variation in a number of deletion and word-order phenomena and on interactions between discourse and syntax in the production of sex-differentiated Japanese speech (55-59).
Bilingualism and Minority Languages
". . .in Japan, which is one nation, one people, one language, problems of the contact of Japanese with other languages have not been extensively treated" (38:88). In the United States, problems of bilingualism have provided an impetus to the development of sociolinguistics. In Japan, on the other hand, the contact of standard Japanese with dialects has been the main force toward development of the field. But, with the internationalization of Japanese business, increasing numbers of Japanese families are spending some years in foreign countries, and the language problems of the children in these families upon returning to Japan have stimulated some interest, leading to the publication in 1980 of a preliminary report on bilingualism in Japan (6). This report begins by presenting a breakdown of the types of bilingual speakers most commonly found in Japanese society, then turns to an analysis of the linguis-
This content downloaded from 147.251.99.135 onFri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 275
tic and educational problems faced by the children of returning families; it closes with a section on the development of testing instruments to ascertain degree of bilingualism. The existence of bilingual communities within Japan (93% of which are Korean-Japanese speakers) is noted, but neither the actual conditions of bilingualism nor the problems speakers from these communities may face in the national education system are addressed in this document.
Even less attention has been paid to minority languages in Japan. The little material available on Ainu focuses more strongly on its structural characteristics than on the organization of its use in modern Ainu communities (47, 49), and information on the use of other languages comes primarily from politically oriented materials (14).
CONCLUSION
Japanese sociolinguistics is a mature discipline. Forty years of active, carefully conducted investigation into the language life of regional communities, the process of language standardization, and the use of honorific language has yielded results that tell a great deal about the structure of Japanese society as it is reflected in language, about the changes that have taken place in that structure in the postwar decades, and about the characteristics of speakers and hearers of Japanese that influence choice of language forms. The work of Japanese sociolinguists is commendable for its thoroughness, its sophisticated analytical methods, and its early use of quantification and advanced statistical techniques. The thrusts of the field, moreover, are entirely independent of Western sociolinguist interests and accordingly provide rich new fields of investigation for our consideration, just as Japanese sociolinguists have in recent years begun to look at Western sociolinguistics for new direction. It is unfortunate that at the core of this achievement lies a data-collection methodology so singularly unsuited to the quantitative investigation of language use—the method of self-report—that these achievements must be viewed with great caution. It is to be hoped that an early happy product of the current cross-fertilization of Japanese with Western sociolinguistic thought will be a stronger interest in the problems of data collection. We may then anticipate supportable statements about language-society relations from the only Asian country to have a linguistic discipline comparable to our own.
Literature Cited
1. Blom, J. P., Gumperz, J. J. 1972. Social meaning in linguistic structures: code-switching in Norway. In Directions in Sociolinguistics: The Ethnography of Communication, ed. J. J. Gumperz, D. Hymes, pp. 407-34. New York: Holt, Rinehart and Winston
2. Egami, T. 1976. Taishu bunka no naka
no nihongo (Japanese in popular culture). In Nihongo Köza 3: Shakai no Naka no Nihongo (Series on Japanese 3: Japanese in Society), ed. Y. Haga, pp. 155-94. Tokyo: Taishükan Sho-ten
3. Egawa, K. 1977. Chi'iki shakai no gen-go seikatsu no henka/hensen (Changes in
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
276 SHIBAMOTO
the language life of regional society). See Ref. 39, pp. 1-30
4. Hata, H. 1983. Bamen to kotoba (Speech situation and speech variation). Kokugogaku 133:55-68
5. Hayashi, S. 1978. Gengo Ködö no Shosö (Aspects of Language Behavior). Tokyo: Meiji Shoin
6. Higa, M. et al. 1980. Nihon ni okeru Bairingarizumu (Bilingualism in Japan). Tuskuba: Tsukuba Univ. Press
7. Ide, S. 1979. Onna no Kotoba, Otoko no Kotoba (Men's Speech, Women's Speech). Tokyo: Nihon Keizai Tsüshin-sha
8. Inoguchi, Y. 1977. Shokugyö shüdan to kotoba (Occupational groups and speech). See Ref. 39, pp. 177-212
9. Ishino, H. 1976. Gairaigo no hyögen (Loan word expressions). In Nihongo Köza 4: Nihongo no Goi to Hyögen (Series on Japanese 4: Japanese Lexical Items and Expressions), ed. T. Suzuki, pp. 227-62. Tokyo: Taishükan Shoten
10. Iwabuchi, T. 1980. Gengo seikatsu (Language life). Kokugogaku 121:115-20
11. Jorden, E. H. 1974. Female speech: persisting myth and persisting reality. In Report of the Second U.S. Japan Joint Sociolinguistics Conference, pp. 103— 18. Tokyo: Jpn. Soc. Promo. Sei.
12. Kamata, R. 1981. Kansai ni okeru chihö kyötsügoka ni tsuite (On the standardization of a local Kansai dialect). Kokugogaku 126:1-12
13. Katö, N. 1975. Chi'iki shakai to kotoba no henka (Regional society and language use). In Shin Nihongo Köza 10: Kotoba to BunkalShakai (New Series on Japanese 10: Language and Culture/ Society), ed. E. Iwabuchi, I. Nishio. Tokyo: Sekibunsha
14. Kim, C. 1977. Sokoku o Shiranai Sedai (The Generation That Doesn't Know Its Fatherland). Tokyo: Tahata Shoten
15. Kindaichi, H. 1957. Nihongo (Japanese). Iwanami Shinsho, 265. Tokyo: Iwanami Shoten
16. Kokugogakkai, eds. 1969. Kokugogaku Jiten (National Language Study Dictionary). Tokyo: Tökyö-dö Shuppan
17. Kokuritsu Kokugo Kenkyüjo (National Language Research Institute). 1950. Hachijöjima no Gengo Chösa (The Linguistic Survey at Hachijojima). Tokyo: Shuei Shuppan
18. Kokuritsu Kokugo Kenkyujo (National Language Research Institute). 1951. Gengo Seikatsu no Jittai—Shirakawa-shi oyobi Fukin Nöson ni okeru (The Actual State of Language Life—In Shir-akawa City and the Neighboring Farm
Communities). Tokyo: Shuei Shuppan
19. Kokuritsu Kokugo Kenkyujo. 1952. Okurigana Shirydshu (Collected Materials on Okurigana Rules). Tokyo: Shuei Shuppan
20. Kokuritsu Kokugo Kenkyujo (National Language Research Institute). 1953. Chi'iki Shakai no Gengo Seikatsu— Tsuruoka no okeru Jittai Chosa (The Language Life of Regional Society—A Survey of Actual Conditions in Tsuruoka). Tokyo: Shuei Shuppan
21. Kokuritsu Kokugo Kenkyujo. 1955. Danwago no Jittai (The Actual State of the Colloquial Language). Tokyo: Shuei Shuppan
22. Kokuritsu Kokugo Kenkyujo (National Language Research Institute). 1957. Keigo to Keigo Ishiki (Honorific Language and the Consciousness of Honorific Language). Tokyo: Shuei Shuppan
23. Kokuritsu Kokugo Kenkyujo. 1959. Meiji Shoki no Shinbun no Yogo (Early Meiji Period Newspaper Vocabulary). Tokyo: Shuei Shuppan
24. Kokuritsu Kokugo Kenkyujo. 1962. Gendai Zasshi Kyujusshu no Yogo, Ydji (1) Soki oyobi Goi-hyd (Lexical Items and Orthographic Forms Used in Ninety Current Periodicals (1) Text and Lexical Table). Tokyo: Shuei Shuppan
25. Kokuritsu Kokugo Kenkyujo. 1964. Gendai Zasshi Kyujusshu no Yogo, Ydji (3) Bunseki (Lexical Items and Orthographic Forms Used in Ninety Current Periodicals (3) Analysis). Tokyo: Shuei Shuppan
26. Kokuritsu Kokugo Kenkyujo. 1974. Chi'iki Shakai no Gengo Seikatsu— Tsuruoka ni okeru 20nenmae to no Hika-ku (The Language Life of Regional Society—A Comparison with 20 Years Ago in Tsuruoka). Tokyo: Shuei Shuppan
27. Kokuritsu Kokugo Kenkyujo. 1974. Gengo Shiyd no Hensen (1)— Fukushima-ken Hokubu Chi'iki no Men-setsu Chosa (Changes in Language Use (1)—Interview Survey in the Northern Region of Fukushima Prefecture). Tokyo: Shuei Shuppan
28. Kokuritsu Kokugo Kenkyujo. 1981. Daitoshi no Gengo Seikatsu: Bunseki-hen Shiryo-hen (The Language Life of Urban Areas. Volume One: Analysis; Volume Two: Data). Tokyo: Sanseido
29. Lee, M. 1976. The married women's status and role as reflected in Japanese: an exploratory sociolinguistic study. Signs 1:991-99
30. Martin, S. E. 1964. Speech levels in Japan and Korea. In Languages in Culture and Society: A Reader in Linguistics
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
JAPANESE SOCIOLINGUISTICS 277
and Anthropology, ed. D. Hymes. pp. 407-15. New York: Harper & Row
31. Martin, S. E. 1975. A Reference Grammar of Japanese. New Haven: Yale Univ. Press
32. Mase, Y. 1981. Gengo keisei ni oyobo-su terebi oyobi toshi no gengo no eikyö (The influence of television and urban language on language formation). Kokugogaku 125:1-19
33. Miller, R. A. 1977. The Japanese Language in Contemporary Japan: Some Sociolinguistic Observations. Stanford: Hoover Institution on War, Revolution and Peace
34. Miller, R. A. 1982. Japan's Modern Myth: The Language and Beyond. New York/Tokyo: Weatherhill
35. Minami, F. 1982. Nihon no shakaigen-gogaku (Japanese sociolinguistics). Gengo (Language) 11(10):74-81
36. Nomoto, K. 1977. Kigyönai ni okeru keigo ködö (Honorific language behavior in industry). See Ref. 39, pp. 103-40
37. Nomoto, K. 1978. Nihonjin to Nihongo (The Japanese People and the Japanese Language). Tokyo: Chikuma Shobö
38. Nomoto, K. 1982. Gengo seikatsu (Language life). Kokugogaku 129:84-91
39. Nomoto, K., Nobayashi, M., eds. 1977 Nihongo to BunkalShakai 2: Kotoba to Shakai (Japanese and Culture/Society 2: Language and Society. Tokyo: Sanseido
40. Nomura, M. 1983. Kanji no genzai (Kanji today). Gengo Seikatsu 378:
41. Ogino, T. 1980. Keigo ni okeru teineisa no süryöka (Quantification of the politeness of word forms). Kokugogaku 120:13-24
42. Ogino, T. 1981. Kinjo no hito ni taisuru keigo (Honorific expressions to neighbors). Kokugogaku 124:28-38
43. Ogino, T. 1981. Keigo ni okeru tsu-kaiwake no taikei to sono közö (The system and syntagmatic relation of honorific expressions). Kokugogaku 125:1-12
44. Ogino, T. 1983. Keigo shiyö kara mita kikite no ichizuke no tayösei (Diversity of speaker-evaluation of address situations from the viewpoint of honorific usage). Kokugogaku 132:1-11
45. Ogino, T. 1984. Keigo tsukaiwake no dankaisü (The number of degrees differentiated in the expression of honorifics). Kokugogaku 138:1-13
46. Ogino, T. et al. 1985. Shühen chi'iki kara Tokyo chüshinbu e no "shinhögen" no ryünyü ni tsuite (The flow of new dialect forms into central Tokyo from the suburbs). Kokugogaku 143:15-24
47. Peng, F. C. C. 1977. The Ainu, the Past
in the Present. Hiroshima: Bunka Hyoron
48. Peng, F. C. C. et al. 1981. Nihongo no Danjo-Sa (Male/Female Differences in Japanese). Tokyo: East-West Sign Language Assoc.
49. Pon, F. 1976. Ainu-go wa Ikite Iru (The Ainu Language Lives). Tokyo: Shin-sensha
50. Saeki, U. et al. 1961. Kokugogaku (National Language Studies). Tokyo: Sanseido
51. Sanada, S. 1977. Ichi-shüraku ni okeru keigo ködö (Honorific language behavior in one community). See Ref. 39, pp. 31-60
52. Sanada, S. 1979. Chi'iki-go e no Sekkin (Approach to Regional Speech). Tokyo: Akiyama Shoten
53. Sanada, S. 1983. Saikin jünenkan no keigo ködö no henyö (Some changes in honorific expression during the last ten years). Kokugogaku 133:69-82
54. Satake, H. 1984. Moji, hyöki (riron, gendai) (Orthographic symbols, writing practices (theory, current)). Kokugogaku 137:74-78
55. Shibamoto, J. S. 1981. Sex-related variation in the production of predicate types in Japanese. Lang. Sei. 3(3):257-82
56. Shibamoto, J. S. 1982. Contributions of sociolinguistics to the language sciences. Lang. Sei. 4(2):115-20
57. Shibamoto, J. S. 1984. Subject ellipsis and topic in Japanese. In Studies in Japanese Language Use, ed. S. Miyaga-wa, C. Kitagawa, pp. 233-66. Carbon-dale: Linguistic Res. Inc.
58. Shibamoto, J. S. 1985. Japanese Women's Language. New York: Academic
59. Shibamoto, J. S. 1987. The womanly woman: manipulation of stereotypical and non-stereotypical features of Japanese female speech. In Comparative Perspectives on Gender, Sex and Language, ed. S. U. Philips et al. New York: Cambridge Univ. Press. In press
60. Shibata, T. 1972. The honorific prefix "o-" in contemporary Japanese. Pap. Jpn. Linguist. 1:29-69
61. Shibata, T. 1975. On some problems in Japanese sociolinguistics: reflections and prospect. In Language in Japanese Society: Current Issues in Sociolinguistics, ed. F. C. D. Peng, pp. 159-74. Tokyo: Univ. Tokyo Press
62. Shibata, T. 1978. Shakaigengogaku no Kadai (Topics in Sociolinguistics). Tokyo: Sanseido
63. Shibata, T. et al. 1980. Toshi no Keigo no Shakaigengogakuteki Kenkyü (Soci-
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms
278 SHIBAMOTO
olinguistic Study of Honorific Expressions in a Japanese City). Tokyo: Tokyo Univ. Press
64. Sugito, K. 1977. Shokubakeigo no ichi-jittai (A case study of honorifics in the workplace). Gengo Seikatsu 328:30-44
65. Suzuki, T. 1975. Tozasareta Gengo— Nihongo no Sekai (The Closed Language: The World of Japanese). Tokyo: Shinchösha
66. Tanaka, H. et al. 1975. Gengogaku Nyümon (Introduction to Linguistics). Tokyo: Taishükan Shoten
67. Tanaka, H. et al. 1978. Gengogaku no Susume (Invitation to Linguistics). Tokyo: Taishükan Shoten
68. Toyama, S. 1977. Nihonjin no dokusho
(Japanese reading). In Iwanami Köza Nihongo 2: Gengo Seikatsu (Iwanami Series on Japanese 2: Language Life), pp. 177-202. Tokyo: Iwanami Shoten
69. Tsunoda, T. 1978. Nihonjin no Nö (The Japanese Brain). Tokyo: Taishükan Shoten
70. Watanabe, S. 1974. On the Japanese language. Jpn. Echo l(2):9-20
71. Watanabe, S. 1979. Review of Shibata, T. Shakaigengogaku no Kadai, San-seido, 1978. Kokugaku 118:81-83
72. Yomikaki Nöryoku I'inkai, eds. 1951. Nihonjin no Yomikaki Nöryoku (The Literacy Levels of the Japanese). Tokyo: Tokyo Univ. Press
This content downloaded from 147.251.99.135 on Fri, 10 Mar 2017 13:16:25 UTC All use subject to http://about.jstor.org/terms