5.1.1 Construction of women’s speech norms

In Section 5.1, we address the construction of linguistic gender norms in the modern period, first focusing on prescriptions for women’s and then on men’s language use. As noted above, women’s speech has historically been much more regulated than men’s, and this is not only true for the modern period. In fact, this is not unique to Japan. Constraints on women’s speech have been “a recurring theme” (Cameron Reference Cameron, Ehrlich, Meyerhoff and Holmes2014: 282) in many societies and in many historical periods. Feminists note that such constraints, when offered in mainstream narratives of the social order, often serve to rationalize women’s subordination and uphold patriarchal privilege. Here, we see that this has largely been the case in Japan as well.

Official policy. We begin, as we have begun each previous set of topics, with the official or semi-official pronouncements on female speech behavior as they developed in the Meiji period. As we saw in Chapter 1, the Meiji government strove to modernize various aspects of Japanese society, which included linguistic matters as well, such as the establishment of Standard Japanese, the genbun itchi movement, and script reform. Under these circumstances of momentous social changes, it would be reasonable to expect that prescription norms for women’s behavior would undergo changes as well, and in fact, they did in some important respects. But at the same time, documents also indicate that some fundamental similarities to the prescribed social roles of “women”Footnote ¹ in the premodern eras were retained, as we will see shortly. One difference from premodern times, though, was that women’s life trajectories and their behavior came to be defined and regulated at the national level.

One of the important events at the dawn of modern Japan was the 1872 start of a “compulsory” national educational system (gakusei), when the elementary, middle, and high school system mentioned in Chapter 1 was established. The class system of the Edo period was abolished, and (co-) education made available both to young boys and young girls by the 1872 Gakusei ‘Education System Order’. Although this might suggest that boys and girls were headed toward educational equality, there were, however, several impediments. First, this compulsory and co-educational requirement only specified sixteen months of schooling for children starting at age sixFootnote ². Higher education was not compulsory and was sex-segregated. But more importantly, for the first months (ultimately years), parents had to pay for their children’s schooling; many parents sent their boys in preference to their girls, so there was from the outset an extra impediment for the young female citizens-to-be (see Chapter 1, footnote 5). Further, by 1879, when the Imperial Rescript on Education was promulgated, it included a complaint against excessive Westernization and called for a return to Confucian ethical ideals. These included separate education for girls and boys (M. Nakamura Reference Nakamura2004, Reference Nakamura2007a).

Meiji reformers quickly noticed Western conceptions of the role of women as part of a modern state’s citizenry and it is arguable that “a major impetus behind modern [Japanese] women’s education was to establish a certain kind of equivalency between Japanese women and their female counterparts in the advanced nations of Europe” (Levy Reference Levy2006: 54). But too much equivalency had to be avoided if the patriarchal structures of the Meiji state were to be preserved (Murakami Reference Murakami1970). Education, thus, proceeded differently for girls and boys. Female citizens were nationalized alongside their male counterparts, but nationalization was, from the outset, gendered (Kitamura Reference Kitamura2009; Wakakuwa Reference Wakakuwa2001). Women’s education was guided by the ideology of ryōsai kenbo ‘good wife and wise mother’ – a good wife who supports her husband, who in turn serves the nation as its workforce and its soldiers, and an educated wise mother who raises children who will become superior adult Japanese citizens.Footnote ³ And this made sense in the era of a return to Confucian ethics and ideas about the proper social order. So, although the 1889 Meiji Imperial Constitution recognized equal rights for women and men as Imperial subjects, it was assumed that since women and men were naturally different, their social roles should also be different (S. Koyama Reference Koyama1991; Kurazumi Reference Kurazumi2008). Thus, while provisions were made for education for both sexes, in practice, different pedagogical trajectories were established, a good example of the justification of gender role differentiation by way of naturalized binary categories of women and men as “instruments of regulatory regimens” (Butler Reference Butler and Fuss1991: 13) deployed in order to uphold a gender order seen as serving the dominant group in the society and a gesture of recognition that Japan needed to convince the Western powers of their ability to stand as an equal and equally “modern” state. Although the idea of ryōsai kenbo was new in that women were nationalized and, therefore, universally expected to be well educated so that they could serve the nations as kokumin, in reality, the subordinate nature of the social role of the new category of “women” (as, in principle, all women) was little changed. And pressures on women to sacrifice themselves for the sake of the nation only increased as the nation became increasingly militaristic and reactionary.

The task of telling all the nation’s women how they should behave was greatly facilitated by the new print media. Women were taught how to live and behave through school shūshin kyōkasho ‘ethics textbooks’ and women’s conduct books (T. Nakamura Reference Nakamura2003; M. Nakamura Reference Nakamura2014a), and articles in the women’s magazines that became available in the 1880s (Bohn and Matsumoto Reference Bohn and Matsumoto2008; Endō Reference Endō1997, Reference Endō2006). These materials often repeated the instructions drawn directly from the Confucian ethics tracts for women popular during the Edo period, referring to such concepts as jotoku ‘feminine virtues’ and fugen ‘feminine speech’. School ethics textbooks, for example, emphasized the importance of filial piety and the primacy of the husband; they recommended modesty, obedience, gentleness, politeness, elegance, and taciturnity, urging women “to be gentle and graceful in all things … not only in manners, but also in speech” (quoted in Endō Reference Endō2006: 61). Another elementary school shūshin kyōkasho, Shūshin Joku (1893), emphasized that women’s speech should be reserved (tsutsushimu beshi), gentle, and adorable (yawaraka ni airashiku), and not argumentative (rikutsuburanu) (quoted in M. Nakamura Reference Nakamura2007a: 121). Similar instructions and comments were also repeated in articles in many women’s magazines and women’s conduct books. Particularly, with regard to language use, they recommend speaking styles that are gentle, polite, reserved, graceful, beautiful, and refined. Few specific linguistic forms that would manifest these qualities are offered, and the most prominent advice to be found in these texts as a whole is that women speak as little as possible. And that when they do, to avoid dialect, rough or vulgar words, and popular slang.

Textbooks also gave young girls very little in the way of concrete advice as to “feminine” terms. Recall, from the discussion in Chapter 1, that textbooks began to circulate the new (or, more properly, the emergent) Standard Japanese. But even these offered girls little in the way of guidance as to the specific forms they, as opposed to their male classmates, should use. Materials used in the national education system played an important role in delineating and authorizing what constitutes, or should constitute, Standard Japanese. In her analyses of one national language reader series, the 1886 Nihon Tokuhon ‘Japan Reader’, a set of two volumes for first-graders followed by six volumes of readers for second through fourth grades, M. Nakamura (Reference Nakamura2005) found many mentions of pronominal and other forms thought to be neologisms derived from the urban and trendy “schoolboy speech” of the day presented as part of the new Standard Japanese. However, new urban forms of “schoolgirl speech” were either not mentioned or listed as “exceptions” and notated as not Standard Japanese.Footnote ⁴ In a later study of eleven grammar books produced from 1873 to 1922, M. Nakamura (Reference Nakamura2007a: 172–184) found the same asymmetrical presentation of gendered forms, leading her to conclude that in its early years of development and dissemination, Standard Japanese could more appropriately have been termed Men’s National Language.

After the Meiji period, attempts to regulate women’s behavior, including language use, intensified as Japan became increasingly ultranationalistic and militaristic. The nationalization of women as Imperial subjects was reinforced during the years of colonial expansion and WWII, when they were expected to serve the nation as a vital part of the jūgo no mamori ‘guardians behind the guns’ (Ueno 1998). “Women’s language” played a significant symbolic role for the construction of women’s national identity during this period. Everyone’s conduct was regarded as indispensable to upholding the national spirit and was under heavy surveillance, but gendered language – which largely meant the language of women – came in for particular attention. In 1941 the government issued Reihō yōkō ‘Essentials of Etiquette’, a set of guidelines for good manners (M. Nakamura Reference Nakamura2007a; Washi Reference Washi2000, Reference Washi, Okamoto and Shibamoto Smith2004). This tract included a section on language use, which emphasized the importance of gender-differentiated language and gave quite specific instructions about such topics as gender-differentiated honorific use, gendered self-reference and address terms (for example, kimi ‘you’ for men versus anata or omae-san for women). Women were also instructed to speak gently using feminine (onnarashii) words.

Non-state actors. To this point, our focus has been on official governmental and educational policies regarding the gendering of the standard language; as Gluck (Reference Gluck2011: 683) reminds us, however, the actual process of modernizing the language included a more diverse set of actors. For a more complete understanding of how modern gendered language forms and styles were drawn into and became a key element in the sociolinguistic story of today’s Japanese language, we need now to turn to the contributions of writers, journalists, and linguists working to highlight and promote new gendered forms.

One of the linguistic styles normally brought up in the discussion of women’s language in the Meiji period is jogakusei kotoba ‘schoolgirl speech’,Footnote ⁵ also called teyo dawa kotoba ‘teyo dawa speech’ due to observers (including linguists and social commentators) attributing to schoolgirls the use of particular sentence-final particles, such as te yo, da wa, and no yo. Note that it is unclear exactly which final particles were considered part of teyo dawa kotoba,Footnote ⁶ as that was not in fact the point; the point, rather, was to create a linguistic category for social commentary,Footnote ⁷ and as such te yo, da wa, and no yo seem to have played a useful symbolic role for this category (M. Inoue Reference Inoue2003). (Indeed, sentence-final particles, albeit a somewhat different set, still do.) In addition to these final particles, schoolgirl speech was also known for the use of words, such as boku ‘I’, kimi ‘you’, tamae ‘do (imperative)’, appropriated from shosei kotoba ‘schoolboy speech’ – another new speech variety that arose among elite boys in the higher levels of the new school system in the Meiji period, kango ‘Sino-Japanese compound words’, and English loan words. Jogakusei kotoba, in its various aspects, is a stereotypified – or enregistered (Agha Reference Agha2007) – speech variety that schoolgirls, who were from privileged families, were believed to have started using as a form of resistance to their ryōsai kenbo-based education; and hence as an index of their own identity (Bohn and Matsumoto Reference Bohn and Matsumoto2008). As is often noted, it was widely criticized in the media as a language of frivolous (keihaku na) or depraved (daraku shita) schoolgirls, or young women elites, or as speech that was vulgar, rough, and impolite. Yet those very same “characteristic” sentence-final particles eventually came to be part of the stereotypical feminine women’s speech forms – an interesting process of resemiotization of indexical meanings that we discuss below.

While it is unclear to what extent schoolgirls actually used it, schoolgirl speech was represented as reported speech in the dialogue in novels – a new genre of literature adopted from the West and popularized through the newly available print media. Moreover, such representations in dialogues were greatly assisted by the genbun itchi movement of the time, as discussed in detail by M. Inoue (Reference Inoue2002, Reference Inoue2003, Reference Inoue2006) and in Chapter 1 of this volume. In particular, katei shōsetsu ‘domestic novels’, often first published serialized in newspapers and later made into books, were popular among women. It has been argued that “Japanese women’s language” is a modern creation severed from premodern women’s speaking styles, as the emergence of newly gendered sentence-final particles, such as te yo and da wa in schoolgirl speech, started to index “modern Japanese woman” as a new social category. Discussing the schoolgirl speech forms utilized repeatedly in novels by male Meiji writers, M. Inoue (Reference Inoue2003: 179), for example, argues that “Japanese male intellectuals at the turn of the twentieth century … heard and cited the schoolgirls’ voice and in doing so, gave rise to the new metapragmatic category of ‘schoolgirl speech’, as well as ‘the schoolgirl’ herself as a new social category. This was, in fact, the epistemic birth of ‘the modern Japanese woman.’” Whether or not the deployment of urban, elite “schoolgirl speech” forms was central to the construction of a national imaginary of the modern woman, despite the ill-repute that these forms enjoyed in the metropolis itself, it is certainly the case that these forms captured the attention of readers in the periphery. And they – or their descendants (in the form of different sentence-final particles) – remain a central part of the contemporary joseigo ‘women’s language’ style.

They are, then, worth a somewhat more detailed look. While schoolgirls may have epitomized “the modern Japanese woman,” they did not resonate well with the state-sanctioned ideology of ryōsai kenbo. They were heavily criticized on the grounds that they were frivolous, depraved, and rude. And it has been observed that not all novelistic representations of schoolgirls were the same: some were characterized as “good” women who embodied the virtues of ryōsai kenbo, while others were “bad” women who did not. Particularly in the early Meiji period, some writers are seen to have assigned teyo dawa kotoba only to the latter, representing the “good” women as avoiding teyo dawa kotoba and instead using high-level honorifics, often used non-reciprocally, and the formal self-reference term watakushi (Bohn and Matsumoto Reference Bohn and Matsumoto2008; M. Nakamura Reference Nakamura2007a, Reference Nakamura2014a). These representations of “good” women’s speech was consistent with what Meiji school textbooks, women’s conduct books, and magazine articles endorsed – that is, a style of women’s speech linked to ryōsai kenbo. Furthermore, young female characters, even the “bad” ones, tended to use teyo dawa kotoba only in certain contexts, particularly those related to romantic relationships; in formal situations, they too typically used very polite language.

These observations about the distribution of teyo dawa kotoba are supported by our own examination of well-known novels published in the Meiji period (and also in the Taishō and Shōwa periods).Footnote ⁸ In fact, according to our analyses, these novels include many female characters of diverse ages, social classes, and regional backgrounds. They were represented through writers’ use of diverse speech styles, including “masculine,” “rough,” or “vulgar” language; for example, older women characters from higher social classes in Tōkyō were represented as using very polite language in formal situations, but as using plain, or informal, speech to lower-status persons; women in countryside regions were represented as using their home dialects, which often included forms considered “unfeminine” from the Standard Japanese perspective. Teyo dawa kotoba as represented speech thus at this time can be considered one of these speech “varieties”– a stereotyped style of speech for a certain kind of schoolgirl, or young urban women in higher social classes, rather than generic Japanese women’s language.Footnote ⁹

To summarize, in contrast to previous eras, during the Meiji period the construction and dissemination of norms for women’s speech took on a more systematic and global nature in three respects: (1) the way norms were nationalized, then rationalized in relation to the ideology of ryōsai kenbo, which specified how women were to serve the nation effectively as (female, or gendered) citizens; (2) the more systematic way in which norms were disseminated and enforced through the newly formed national education system and facilitated by the print media; (3) and the addition to earlier prescriptive dictates about appropriate female speech of sets of particular linguistic forms defined in terms of the emergent variety of Standard Japanese. At the same time, the basic prescripts for women’s speech – that there be not too much of it, first, and that what there was should be polite, soft, charming, and non-argumentative (M. Nakamura Reference Nakamura2004) – remained intact.

Scholarship. What is particularly noteworthy concerning the construction of speech norms for women is that it was not until the pre-war and wartime period that terms such as fujin no kotoba, fujingo, joseigo, onnna kotoba, all meaning ‘women’s language’, came to be used actively by linguists and intellectuals (Hoshina Reference Hoshina1936; Ishiguro Reference Ishiguro1943; Kikuzawa Reference Kikuzawa1929; Kindaichi Reference Kindaichi1942), who increasingly came to comment on women’s language as it related to women’s symbolic role in Japan’s increasingly expansionist nationalism and aggression. In relation to this development, three issues merit special attention from the interwar period: (1) “women’s language” came frequently to be referred to as an important and unique Japanese tradition that makes the language “beautiful,” (2) “women’s language” was established as a “descriptive” linguistic category, and (3) women’s language was characterized in terms of (the female version of) Standard Japanese.

With regard to the first point, many national language scholars, such as Hoshina (Reference Hoshina1936), Ishiguro (Reference Ishiguro1943), Kikuzawa (Reference Kikuzawa1929), and Kindaichi (Reference Kindaichi1942), deemed Japanese women’s language to be an important and unique Japanese “tradition (dentō)” either by “tracing” women’s language back to “graceful (yūga na, yūbi na)” nyōbō kotoba ‘court lady’s language’ or by emphasizing women’s deft use of complex honorifics. Ishiguro (Reference Ishiguro1943: 227), for example, listed the following as the characteristics of women’s language: beautiful, refined, indirect, and polite; he also noted the avoidance of Sino-Japanese words (which are more associated with male speakerhood and public speaking) – characteristics consistent with what Kikuzawa (Reference Kikuzawa1929) presented as the features of nyōbō kotoba, which he claimed as the origin of women’s language. Kindaichi (Reference Kindaichi1942: 309) emphasized that “in order to become an adult Japanese woman … [a woman] should acquire this traditional Japanese women’s language along with honorifics the exquisiteness of which are rarely seen in other languages.” In particular, “tracing” the origin of women’s language to nyōbō kotoba served to link it to the Imperial family (Washi Reference Washi2000, Reference Washi, Okamoto and Shibamoto Smith2004), which may have been considered an effective strategy for enhancing the status of women’s language as an important part of the national language and, thus, as a symbol of national identity during these critical years.

This reminds us of the arguments for a standard Japanese language, or kokugo, as an essential part of Japanese tradition and identity, as discussed in Chapter 1. The “Japanese women’s language” discourse was developed less during the early years of nation-building but rather during this period as an extension, or part, of the national language which had hitherto focused on issues more relevant to male-centric public domains. By presenting women’s language as though all women in the past used nyōbō kotoba, which they did not, and had mastered all the complexities of honorifics, which they had not, a “tradition” was discursively constructed. This had previously received less attention than the “tradition” of a unified Japanese or the “tradition” of keigo as a more “civilized” kind of linguistic politeness than Western nations could offer.

The second issue concerns the notion of women’s language as a descriptive linguistic category. Linguists, or national language scholars, during this period increasingly started using the notion of women’s language (fujingo, joseigo, onna kotoba) as a linguistic category in grammar books and public commentaries. Kikuzawa Sueo (Reference Kikuzawa1929) seems to have been the first scholar to consider “female speech” a worthy object of national language research (Washi Reference Washi, Okamoto and Shibamoto Smith2004: 84), holding up nyōbō kotoba as an ideal of spoken femininity due to its qualities of grace (yūbi) and refinement (jōhin), in contrast to the more vulgar yūjogo of the pleasure quarters. He was the first, but hardly the last, scholar to romanticize the connections of modern linguistic femininity to highly restricted and elite female speech styles of the distant past, and to promote this “special” and “elegant” characteristic of Japanese language as a marker of the “special” (and superior) qualities of the Japanese culture in its totalizing essence.

Subsequently, scholars began to add specific, modern linguistic features to their more general “descriptions” of women’s language and its elegance. Those characterizations were all based on Standard Japanese. Although this issue has not been pointed out previously, the scholarly presentation of “women’s language” is a significant qualitative shift from the earlier discourse surrounding this topic. Up to this point, women’s speech norms had been presented as prescriptions wherein women were instructed that they should use language in certain, feminine, ways. Such didactic prescriptions implicitly presuppose the existence of many women who, in fact, do not speak “properly.” On the other hand, more recent scholarship presents lists of characteristic features identified as constituting fujingo, joseigo, or onna kotoba in grammar books and articles, at which point “women’s” language effectively becomes a “descriptive” (socio)linguistic category, at least on the surface. That is, the deontic “should” in prescriptions becomes the epistemic “should” in descriptions, suggesting a “flexible” line between the two senses of “should.” Such presentations of “women’s language” typically do not refer to the empirically observable diversity in women’s speaking practices, but rather “erase” all the speech patterns that do not fit the normative women’s language from the descriptive record (Irvine and Gal Reference Irvine, Gal and Kroskrity2000). By avoiding reference to diversity, “women’s language” is presented as a fixed “fact,” creating the impression that all Japanese women speak one single language variety, thus adding the authoritative voice of the linguist to the normativizing pressures of the dominant gender ideology.

This brings us to the third issue – the characterization of women’s language as a female version of Standard Japanese in the presuppositionally “neutral” terms of descriptive linguistics. In terms of general styles, its characteristics were hardly different from those recommended in conduct books written in premodern Japan; that is, “women’s language” was said to be, like nyōbō kotoba, polite, gentle, elegant, beautiful (utsukushii or kirei), and refined. In terms of specific linguistic forms, references were made most often in relation to particular self-reference and address terms, honorifics (including the prefix o-), sentence-final particles, all, not coincidentally, corresponding to Standard Japanese forms, as well as the tendency to avoid kango. And this is not so different from the scholarhship on “women’s language” today. See Table 5.1, which offers our own analysis of the stylistic characteristics and the specific linguistic forms which are discursively attached to them today.

Interestingly, we see that those particles attributed to “schoolgirl speech” in the Meiji period and repudiated as “not standard language” are now included, or even promoted, as part of Standard Japanese-based joseigo in linguistic grammar books from the early-to-mid Shōwa period (1925–1989) onwards. School textbooks followed suit. Readers used as school textbooks gave example dialogues that were gendered in regard to self-reference and address terms, honorifics, sentence-final particles, and interjections (M. Nakamura Reference Nakamura2007a). What is notable is that these characteristics were presented as those of “women’s language” now included as a part of kokugo ‘national language’, which was synonymous with Standard Japanese.

As we saw in Chapter 1, when Standard Japanese was established, it was defined as the speech presumed to be used by Tōkyō residents from the middle class or above. “Women’s language,” then, came to be construed as the speech (supposedly) used by middle-class women in Tōkyō, or women living in yamanote ‘uptown’ Tōkyō. This class-based construal is highly ideological. In other words, scholars’ characterizations of “women’s language” as a linguistic category from the early-to-mid Shōwa period served as one of the most powerful forces shaping the discourse of a special, “Japanese” linguistic femininity that occludes the empirical diversity of female voices or subjects them to public critique.

And this continued to be the case even after WWII, when Japan entered an intensive period of democratization across a broad range of social institutions. Women’s legal status began to improve, beginning with the 1945 acquisition of the right to vote and the right to form (political) associations. Explicit and active state interventions such as those seen in shūshin kyōkasho ‘ethics textbooks’, could not be continued, nor could the discourse of women’s language as part of a “long” tradition of Imperial Japan be rationalized as being a way for women as Imperial subjects to serve the nation. These changes, however, did not lead to radical reform in ideas about how women should speak. Even after the war, the (re)presentation of “women’s language” as polite, gentle, and refined continued both in “descriptive” linguistics (and in the media representations we turn to below). Scholarly characterizations of “women’s language” as a (socio)linguistic category continued the practice of listing general stylistic features, such as politeness, gentleness, and refinement, along with specific gendered linguistic forms such as the gendered self-reference and address terms, honorifics, sentence-final particles, and interjections in Standard Japanese outlined in Table 5.1.

Furthermore, throughout the post-war period, we encounter metapragmatic commentaries on women and women’s behavior at critical junctures when social change seems to be moving too fast; as M. Inoue (Reference Inoue, Okamoto and Shibamoto Smith2004) notes, at such moments the notion of “women’s language” is seen as being in danger of becoming corrupted and, thus, in need of increased scrutiny and regulation. Inoue cites the early 1990s as one such time, when the political-economic transformations of the bubble economy made entry into public spheres previously open only to men possible for women. In the mid-2000s, “gender-free” rhetoric triggered a backlash of “gender bashing” wherein conservative forces essentially equated gender theory with the denial of any sex differences. This was considered a clear danger to the health of the nation, and indeed, to the health of women (Ueno et al. Reference Ueno, Shinji, Tamaki and Mari2006; Wakakuwa et al. Reference Wakakuwa, Shūichi, Masumi and Chieko2006). Within this backlash, women were exhorted not to forget their “essential” womanliness and to act – and speak – accordingly (Shibamoto-Smith Reference Occhi, SturtzSreetharan and Shibamoto-Smith2010).

Our review of the historical norm construction of women’s speech illustrates the complex indexical process of linking linguistic forms and femininity. We see that while the way in which normative women’s speech was rationalized changed according to the political and social climate of the time and while many of the specific linguistic forms regarded as feminine have changed, the idea of feminine speech as reserved, polite, and gentle is retained (see examples from media discourse in Section 5.2). These characteristics have historically regimented Japanese onnarashisa ‘femininity’ and, over time, have been linked to a variety of specific linguistic forms. In fact, such restrictions on women’s speech are not cross-culturally unusual (Sherzer Reference Sherzer, Philips, Steele and Tanz1987; Talbot Reference Talbot2010), although the silent and deferential woman idealized among the eighteenth and nineteenth-century bourgeoisie in the West is hardly a universal either. Despite Danish linguist Otto Jespersen’s biologically deterministic characterization of women as exercising “a great and universal influence on linguistic development through their instinctive shrinking from coarse and gross expressions and their preference for refined and (in certain spheres) veiled and indirect expressions” (Jespersen Reference Jespersen1922: 246, emphasis added), the ethnographic record proves that these stylistic qualities are far from universal (for particularly striking counterexamples to these prescripts, see Keenan [Ochs] Reference Keenan [Ochs], Bauman and Sherzer1974; Kulick Reference Kulick1993).

Clearly, the linking of particular linguistic forms to an idealized set of feminine “styles” is not a given but rather is ideologically mediated. Nonetheless, these characteristics are often presented as natural traits of women due to the differences women and men are believed to innately possess – as in the gender bashing debates in 2006 – and serve to support underlying assumptions in the gender role differentiation seen in the education of women before the end of WWII. These assumptions are also found in the present day, as one young woman expressed in a letter to Asahi Shinbun in 2003. On July 19 of that year, a sixteen-year-old high school girl from Aomori prefecture wrote to complain about her treatment in her world history class. See Example (1) for a complete translation of her complaint.

The issue of “special” speaking practices, if not “special” language forms for Japanese women (and “special” contexts and contents for speaking) is clearly one that remains a focal point for understanding the social life of language today and we will examine more metapragmatic commentary about its effects on speakers in Chapter 6. But as this single example suggests, it is hardly surprising to find language and gender issues a central arena for sociolinguistic research.

We close this section with a brief look at contemporary language and gender research in Japan. There have been, over the last forty or more years, two central questions, which we summarize as: 1) questions of whether modern Japanese linguistic femininity is continuous with the past or is something newly born along with the modern Japanese state and 2) questions of how and why specific language forms are enregistered and re-enregistered as manifesting the underlying stylistic characteristics of modern Japanese femininity, and what this implies for the full (and diverse) range of speakers who may or may not align with that social construct.

Debates about the “origin” of modern Japanese women’s language. The first of these questions was addressed, although not resolved, in the form of a series of position pieces outlining differences of opinion that exist between scholarship that comes down on the side of positing the ahistorical existence (and persistence) of a special Japanese gender order, special in part precisely because of the concrete linguistic manifestations that can be drawn upon as evidence, and scholarship claiming that modern joseigo is just that – a modern phenomenon. Scholarship arguing for an ahistorical joseigo, one that has always been part of Japanese language and culture, notes, for example, that from the earliest periods in Japanese history, women were expected to speak differently from men in certain ways. Heian period (794–1192) noblewomen were expected, for example, to use kana syllabary script, wago ‘native Japanese words’, and wabun ‘Japanese writing style’, while men were permitted kanji ‘Chinese characters’, kango ‘Sino-Japanese words’, and kambun ‘classical Chinese writing style’ (Endō Reference Endō1997: 26; Reference Endō2006). Women were to speak quietly (hikui koe de) and in a reserved manner (tsutsumashiku); they were also adjured not to use explicit expressions (Endō Reference Endō1997; Jugaku Reference Jugaku1979; Sugimoto Reference Sugimoto1997). Remarks on these expectations appear in such monuments of the period as Murasaki Shikibu’s Genji Monogatari ‘The Tale of Genji’ and Sei Shōnagon’s Makura no Sōshi ‘The Pillow Book’. Note, however, that these expectations for “women” were in fact expectations for noblewomen; expectations for the speech of women in other social categories are not mentioned in extant texts.

Medieval (1192–1603) texts (for example, Niwa no Oshie ‘Domestic Teaching’ (1283) and the early fourteenth century Menoto no Sōshi ‘A Nurse’s Book’) also instructed aristocratic and bushi ‘warrior class’ women to speak with reserve and without emotion, not to use rough language, and to speak quietly (Endō Reference Endō1997: 44–48; M. Nakamura Reference Nakamura2007a: 44–46). During this period, gender differences concerning specific linguistic forms also became more salient, as illustrated in gendered self-reference terms – soregashi ‘I’, watakushi ‘I’, warera ‘we’, and mi ‘self’ were prescribed as suitable for men, while warawa ‘I’, mizukara ‘I’, wagami ‘I/myself’ were mentioned as suitable for women (Endō Reference Endō1997: 52; Sugimoto Reference Sugimoto1997: 176–177).

Different expectations for women’s language use were part of more general constraints on women’s behavior in a society dominated by feudalism and Buddhism, which regarded women as inherently flawed and inferior (Endō Reference Endō1997: 41–43). Nyōbō kotoba ‘court ladies language’ also emerged during this time. Originally a type of occupational jargon, it gradually spread to women in military-class families and to commoners. Most of the word formation processes in nyōbō kotoba resulted in indirect or vague expressions (for details, see Endō Reference Endō1997, Reference Endō2006; Horii Reference Horii1993; Sugimoto Reference Sugimoto1997), and is frequently adduced as evidence of continuities between today’s joseigo and that of the past.

The Edo period (1603–1868) saw a return to an interest in Confucianism, with its danson-johi ‘male superiority, female inferiority’ ideology that emphasized gender difference, women’s domestic roles, and the imperative for women to subordinate themselves to parents and husbands. Women’s demeanor came under intensified scrutiny, especially as urbanization and economic changes took place that gave women more opportunities to support themselves and thus achieve some measure of independence from male approval. Social commentary fueled in part by Confucian principles and in part by the moral panic at the possibility of female independence appeared in the form of numerous conduct manuals or ethics books, such as Jokunshō ‘Collection of Instructions for Women’ (1642), Onna Shisho ‘Four Books for Women’ (1656), and Kaibara Ekiken’s famous Onna Daigaku Takarabako ‘Treasure Chest of Higher Learning for Women’ (1716),Footnote ¹⁰ aimed at reminding women how they should live and behave. These books included instructions for language use. Stylistically, they emphasized the importance of speaking in a reserved (tsutsumashii or hikaeme na), polite (teinei na), gentle (yasashii or yawarakai), and refined (jōhin na or hin no aru) manner. As specific linguistic forms that could achieve these styles of speaking, women were told to use forms such as the polite prefix o-, lexical items similar to the words used in nyōbō kotoba, and native Japanese words rather than Sino-Japanese alternatives. Women were also, in these and other conduct tracts of the time, told not to speak like men; not to be harsh, rough, or vulgar; not to be pretentious or arrogant in their speech; and not to laugh at people.Footnote ¹¹

Again, however, while the instructions in these Edo period discipline books may have had important implications for the construction of women’s speech norms in modern Japan, the extent of their influence over actual practice of real women during this period is unclear. First, although the authors of these tracts implied that their instructions should apply to onna, that is, all women, their actual scope of influence was far from the totalizing language standardization policiesFootnote ¹² of modern Japan, as the intended readership was limited (primarily from bushi ‘warrior’ class to chōnin ‘merchant’ status women). And second, these policies were not issued through a central governmental structure supported by a nationwide and compulsory education system. That is, ideologies with an all-inclusive scope, one that included all female members of the Japanese population, could not have developed in the Edo period since identities were not conceived of in terms of Japaneseness but rather in terms of social status (mibun) categories (Howell Reference Howell2005) and enacted in the context of those categories.Footnote ¹³

And these are precisely the arguments made by scholarship claiming that modern Japanese joseigo is a product of modernization severed from premodern times (M. Inoue Reference Inoue, Bucholtz, Liang, Sutton and Hines1994, Reference Inoue2006; M. Nakamura Reference Nakamura2007a, Reference Nakamura, Ehrlich, Meyerhoff and Holmes2014b). This scholarship tends to stress the totalizing effects of any discourse of linguistic (or other) femininity in national cultures as creating a distinct line between the premodern and the modern. It stresses as well the differences in the scale – and coercive capacity – of the forces that define and distribute prescriptions for modern gendered speech based on Standard Japanese. Powerful centripetalizing forces do, of course, create a more definite line of demarcation between the premodern and the modern in terms of a public culture of uniformity. The line is not drawn, however, automatically; it takes a lot of ongoing work to create a modern totalizing homogeneity (Heinrich and Galan Reference Heinrich and Galan2011a). Nor is it uncontested by late- or postmodernist scholars’ arguing that such homogenizing policies occlude ongoing negotiations and challenges to it in on-the-ground interaction and interpretation (for a concise summary, see Cameron Reference Cameron2005).

Our own thinking leads us, then, away from an either-or vision of Japanese projects of speaking-while-female. Through our examination of the indexical process of historical norm construction, we are drawn to the conclusion that women’s language, or joseigo, as a totalizing ideology of linguistic femininity is neither an exclusive product of modernization nor a mere continuation of a premodern discourse about Japanese women – although at any given time, only some kinds of Japanese women – as gentle, reserved, polite, and refined (see Okamoto 2008c). Rather, where we stand today (as speaking and writing women) is the result of a more complicated process through which gendered language styles historically deemed “appropriate to [certain] women” is woven into the modern ideology of standard language with its “feminine” accessorization (with specific linguistic forms) for all. Our argument is based on a consideration of how sets of certain linguistic forms have been selected to be linked to femininity, how they are recontextualized and resemiotized at various historical moments, how ideologies have shaped this process of norm construction, and how norms are disseminated in society.

Enregistering linguistic femininity. This leads us to the second of the two major topics in Japanese language and gender research today: how and why specific language forms are enregistered and re-enregistered as manifesting the underlying stylistic characteristics of modern Japanese femininity, and what this implies for the full (and diverse) range of speakers who may or may not align with that social construct.

Let us begin with the most commonly “described” stylistic characteristics in the scholarly literature we reviewed in 2008 (see Table 5.1). The very most common qualities were politeness, gentleness or softness, or refinement (elegance), followed by indirection, empathy, and unassertiveness – which we provisionally take as positively evaluated qualities – and verbosity and emotionality – which we, again provisionally, take as negative qualities. The task, then, is to examine how these link to the various linguistic forms offered as manifestations of these qualities: first and second-person pronominal forms specific to women, special sentence-final (and medial) particles, and extensive use of honorifics.Footnote ¹⁴

Linkings of some other forms, such as avoidance of rough language and sloppy articulations (the phonological reductions from Table 5.1) seem unproblematic (and very similar to the comments about American English “women’s” language in Lakoff Reference Lakoff1975); the morphosyntactic prescript to use indirect directive forms or leave sentences incomplete also seems unproblematically linkable to the stylistic characterization of, well, indirection. Not all traits that are readily relatable to the most commonly cited stylistic characterizations in the scholarly literature are, of course, positive. The large pitch range and varied intonational contours attributed to women’s speech links readily to women’s presupposed emotionality, as it does in other languages around the world. But overt expressions of emotionality are not, today, taken as a sign of true social maturity in Japan (Matsumoto Reference Matsumoto1996).

And women’s verbosity, as well, is related to another “Japanese” communication feature: the ideal of taciturnity (which verbosity violates). Taciturnity, which linguistically means simply not speaking much, can, in fact, be interpreted in a variety of ways (anger, fear, power, powerlessness; see Tannen 1994), but in Japanese communicative culture it is interpreted as an index of reserve – a “natural” trait of Japanese people in general (A. Tsujimura Reference Tsujimura and Kincaid1987). Women, urged throughout the premodern period to be reserved, however, were so admonished precisely because they were thought to be naturally chatty – and indiscriminately so. That characterization has carried over into Japanese modernity, at least according to scholarly reports. But this turns out to be a double-edged sword for women (and girls). When men are taciturn, it is stereotypically viewed positively, while their volubility is often criticized. It often turns out that women – perhaps because of their “natural” verbosity – are criticized for not being taciturn (kamoku, chinmoku kagen), which are good traits, but they are also criticized for being sullenly silent (mukuchi, muttsuri shita), when they, apparently, are judged to have gone too far. Men are not susceptible to this latter constraintFootnote ¹⁵. These linkings are, to be sure, ideological, and the differences illustrate the variability of indexical meanings depending on the context (in this case, the speaker’s gender). But modern Japanese language and gender scholarship has taken the most interest in the surface-segmentable forms. These are highly salient, but more problematic in terms of their links to the general stylistic characterizations. The greatest attention has gone to the sex/gender-differentiated pronominals and the sentence-final particle sets, along with honorifics.

The scholars who defined “women’s language (fujingo, joseigo, or onna kotoba)” as a linguistic category in grammar books and linguistic literature, particularly from the 1920s on (Mashimo Reference Mashimo1948; Matsushita Reference Matsushita1924; Sakuma Reference Sakuma1983/1956) characterized it as a “women’s version” of Standard Japanese, attesting to the effectiveness of the language standardization project. In other words, originally, this set of speech forms was an index of regionality and class status (a first-order indexical). But when the style constituted by these linguistic forms was compared with the speech of women in other regions and from other social classes, it came to be reinterpreted as an index of modern urban femininity (a second-order index).Footnote ¹⁶ Again, the specific linguistic forms they selected to describe as feminine did not inherently possess such characteristics as the “feminine” qualities of politeness, gentleness, and refinement, but because their users – elite women in the metropolis – were believed to have such characteristics, so was their speech held to index them. That is, the mid-century normative women’s language based on Standard Japanese was informed by a modernist ideology explicitly based on the eradication of regional language differences, but implicitly incorporating class and gender-based stratifications as well. By the mid-twentieth century, those latter, gender, stratifications had become highly explicit.

And this brings us to the most intriguing change that occurred in the indexical meanings of teyo dawa speech (M. Inoue Reference Inoue2002, Reference Inoue2006; M. Nakamura Reference Nakamura2014a; Okamoto Reference Okamoto, Sato and Doer2014). As we saw earlier, sentence-final particles such as te yo and da wa were first-order indexicals of a particular social group, schoolgirls, young educated women from privileged families, that is, a social group identified in terms of age, class, gender, and locality (yamanote Tōkyō). But although this was a highly desirable complex of social characteristics, they were a highly controversial group of “new” women. Their behavior was subject to considerable critique, and so was their speech (or their putative speech, see M. Inoue Reference Inoue2003), which was harshly criticized as vulgar, impolite, and unfeminine. Thus, in the Meiji period, the second-order indexical value of this speech style associated it not with proper modern femininity but with a resistant (and unfeminine) stance toward traditional understandings of women’s place in the social order as contained within ryōsai kenbo policy and educational practice. The forms such as te yo and da wa were not themselves inherently vulgar or impolite,Footnote ¹⁷ but their supposed users were evaluated as such. Interestingly, however, these particles eventually came to be perceived as polite, gentle, refined, and feminine, and in the mid-twentieth century were “promoted” as part of the set of linguistic forms that constituted the wartime and post-war “women’s language.”

Part of this process of the elevation of these forms came about through representation; when linguistic forms such as te yo and da wa began to be used in represented speech in novels, their interpretations beyond the Tōkyō center were not necessarily negative. Because their users (the characters in novels who were constructed as using these forms) were urban, educated, sophisticated, and modern women, it is reasonable to suggest that readers also perceived their speech as such, and gradually the style came to be used by some, although not necessarily all, real speakers, including older women in higher social classes who were, or were presumed to be, speakers of Standard Japanese and by women readers in other parts of Japan. This suggests that at some point, both negative and positive meanings constituted the “indexical field” (Eckert Reference Eckert2008) of these linguistic forms. Gradually, these final particles were incorporated into the “women’s version” of Standard Japanese in the national context and their associations with normative (polite, gentle, refined) feminine speech for all female citizens were regularized.

These mutations of social meanings associated with nyōbō kotoba, yamanote women’s Standard Japanese, and teyo dawa kotoba thus serve as excellent illustrations of the variability and multiplicity of indexical meanings across space and through time. Particularly interesting in this process is the way the speech associated with a particular group of speakers, such as those naughty Tōkyō schoolgirls, came to be selectedFootnote ¹⁸ as an ideal model of women’s speech. This example offers an excellent illustration of the resemiotization of linguistic forms in the indexical order through recontextualization (Agha Reference Agha2003; Silverstein Reference Silverstein2003).

Heteronormativity and linguistic femininity. Another important issue that deserves attention in the discussion of women’s speech norm construction concerns the regimenting norms of sexuality, that is, the notion of heteronormativity that underlies the construction of women’s speech norms, or “women’s language.” Women who use “women’s language” (and men who use normative “men’s language”) are assumed to be heterosexual. As we saw above, expectations of social roles have been gendered based largely on the belief that women and men are “naturally” different because they are biologically different. In the case of women, their roles as mothers and wives – essential roles for the heteronormative marriage system – have been emphasized, as we saw in the idea of ryōsai kenbo and their wifely and motherly behaviors expected to be naturally feminine. This includes their language use. And the assumption of heteronormativity for women’s (or men’s) language is taken for granted to the extent that it often escapes critical examination in thinking about the construction of gendered speech norms.

However, this assumption serves as a powerful constraint on women’s (and men’s) behavior and has significant implications. Its effects become particularly salient when gendered speech norms are violated. For example, lesbian women may struggle with the choice of self-reference terms and other gendered linguistic forms (Lunsing and Maree Reference Lunsing, Maree, Okamoto and Shibamoto Smith2004), and if they do not use forms expected of heterosexual women, or use forms expected of heterosexual men by “crossing” (Bucholtz Reference Bucholtz1999; Rampton Reference Rampton1995) the gender boundary, they are likely to be criticized or seen as “not normal.” On the other hand, the gender boundary may be intentionally crossed in order to construct a lesbian identity, challenging the normative gender order (Abe Reference Abe, Okamoto and Shibamoto Smith2004, Reference Abe2010; Lunsing and Maree Reference Lunsing, Maree, Okamoto and Shibamoto Smith2004), a point that will be addressed in Chapter 6.

Heteronormative assumptions are also evident in the general speech styles women are expected to use. As we saw above, the adjectives used to characterize women’s “ideal” speech styles include reserve, politeness, gentleness, elegance, grace, and refinement. Why these styles, if they are not innate traits of women? It may be that they serve to maintain the hierarchical gender order in society by limiting women’s proper scope of activity to that of being subordinate and attractive. Normative expectations exclude women who are strong, assertive, rough, or unrefined from the gender model for womanliness. M. Nakamura (Reference Nakamura2007a: 158–167) claims that teyo dawa kotoba also became an index of sexuality through its use in represented speech in Meiji novels, which often involved romantic or sexual relationships. Analyses of contemporary media also suggests that in novels, films, and televisual broadcasts, female characters who use the gendered pronouns and women’s sentence-final particles of Standard Japanese tend to be heroines in romantic stories (Mizumoto Reference Mizumoto2006; K. Satake 2004; Shibamoto-Smith 2004; Shibamoto Smith and Occhi 2009).Footnote ¹⁹

What we can conclude from the preceding discussion is that today’s normative women’s speech, or “modern women’s speech,” that is identified based on Standard Japanese is not simply a modern invention – an argument initiated in Okamoto (2008c, 2014) and further developed here. Rather, it is a result of the more complicated process described previously. The “tradition” of polite, gentle, and refined femininity, a femininity “discovered” through a highly selective, or ideological, reading of the premodern and modern record and put into service by a series of moves to create indexicalities between these qualities and contemporary language forms. The selective and ideological nature of this “discovery” must be stressed, since the premodern record itself actually included diverse female speaking practices, ranging from the desirable refined and elegant (nyōbō kotoba ‘court ladies language’) to the richly expressive but hardly refined speech of old women in regional Japan, or working-class women, or even curses of the yamanba ‘mountain witch’ (Copeland Reference Copeland, Miller and Bardsley2005). But it was out of this process that the narrative of timeless Japanese femininity was born.

This perspective is in line with the argument forwarded in Bauman and Briggs (Reference Bauman and Briggs2003: 5) that two elements that simultaneously “make modernity work and make it precarious” are precisely language and tradition. We argue that in order to construct the social category of “modern Japanese women” who speak “modern women’s language,”Footnote ²⁰ it is necessary both to regiment their language, and then to tie it to the ideologically constructed category of “traditional Japanese women” who instantiated the same traits desired in the modern version. Without this tradition, the force of the prescriptions to women would be vitiated; without the modern women, the “beautiful traditions” of the Japanese past – which validate its modernity – would lose meaning, at least in the realm of feminine gender constructs.

We argue, then, that what today is stereotypified in Japanese as joseigo (and in English as “(Japanese) women’s language”) is neither fully continuous with past traditions of womanly behavioral ideologies nor a newly minted construct first seen in the building of the Meiji state, but rather the result of complex processes in which the modern ideology of standard language has incorporated the premodern femininity regimes of status elites from the immediately preceding era, thereby giving the effect of a Japan that has always been, at least with regard to femininity, civilized (Anderson Reference Anderson2009). From this point on, characterizations of linguistic femininity served the construction of modernity in Japan by simultaneously linking – no matter how selectively – those characterizations to a timeless past and creating a womanly woman fit for civilized modernity.

5.1.2 Construction of men’s speech norms

Historically, men’s language behaviors in general do not seem to have been as explicitly regulated as women’s, which may give the impression that men, and masculinity, constitute the unmarked, or “invisible” categories (Benwell Reference Benwell, Ehrlich, Meyerhoff and Holmes2014). This underwent some change in the Meiji modernity projects of language standardization, which in many ways were aimed in their earliest iterations at fitting male subjects into their new roles as male citizens, but did not result in men being defined as masculine, a point we elaborate on shortly. But it is important to recognize that for women to be constructed as a social category, there must be a category to which they stand in opposition (K. Iwabuchi Reference Iwabuchi1994; Talbot Reference Talbot2010) – namely, men. Like the category of women, that of men assumes heteronomativity; the two categories complement each other and are used to maintain not only the gender order but also the social order more generally (Butler Reference Butler and Fuss1991; Eckert Reference Eckert, Campbell-Kibler, Podesva, Roberts and Wong2002; McElhinny Reference McElhinny, Campbell-Kibler, Podesva, Roberts and Wong2002; Milani Reference Milani, Ehrlich, Meyerhoff and Holmes2014). Constructing norms for linguistic behavior according to these binary identity categories is one of the most important means of upholding the gender order.

In the case of Japanese language and gender, the category of (normative) “men’s language,” or stereotypical masculine speech, has been characterized as much by catalogs of specific linguistic forms of Standard Japanese, such as first-person boku ‘I’, sentence-final particles, such as zo, ze, and da that stand in contrast to counterpart “feminine” forms, masculine interjections (ō ‘oh’, oi ‘hey’), phonological contractions (e.g., ē for ai, as in takē for takai ‘tall’); and certain lexical items that are considered rough or very informal (e.g., kū for taberu ‘eat’), as by stylistic features, such as forcefulness, directness, and roughness. But in neither premodern nor modern Japan has men’s language been policed as much as, or as explicitly as, women’s language. Whereas we find a large number of books written in premodern Japan – as well as in modern Japan (see Okamoto and Shibamoto Smith Reference Okamoto and Shibamoto Smith2008 and the following text) – to teach women proper demeanor and manners, including how to talk properly, we find virtually none written on this topic specifically for men.Footnote ²¹ This, however, does not mean that there were no normative expectations for men’s language use.

The place to begin when considering the construction of modern speech norms for men is with their role in the making of the national language.Footnote ²² Within the early Meiji language standardization project, the relation of male speakers to Standard Japanese was assumed, and both then and to the present the notion of “men’s language” has received little attention (M. Nakamura Reference Nakamura2007a, Reference Nakamura2014a; Occhi et al. Reference Occhi, SturtzSreetharan and Shibamoto-Smith2010). As we saw in Chapter 1, the Meiji government’s language planning focused on the establishment of Standard Japanese and the writing system reform in terms of the development of script conventions and genbun itchi style. Addressing gender differences in speech did not seem to have been a major concern for the development of either. This may appear odd given that school education was gender-based, as indicated by the ryōsai kenbo education.

However, Standard Japanese was established based on what was thought to be the speech of Tōkyōite men who were educated, and from the middle class or above. National language scholar Okano Hisatane argued in 1902 that, while the language spoken in Tōkyō exhibits a great deal of diversity across class, occupation, age, and gender, Standard Japanese for the use of genbun itchi style should be the language that is generally accepted throughout Tōkyō, that is, the language of the middle-class men of the city (Tōkyō no shakai ippan ni tsūyō suru gengo, sunawachi chūryū-shakai no danshi no gengo; quoted in A. Tanaka Reference Tanaka1983: 74–75, emphasis added). This was echoed by others. Because the building of “men’s language” was effectively isomorphic with the building of Standard Japanese – just as the building of the male citizen seemed to be isomorphic with the building of the generic citizen – the modern language of men seems to have been considered unmarked and not in need of special explanations based on some totalizing characterization of male nature. That the notion of “educated Tōkyōites” meant “educated men” is also suggested by the fact that the most visible (or notorious) educated women during this period, namely, schoolgirls, were negatively evaluated, as we saw above, and their speech style (teyo dawa kotoba) was strongly criticized.

The various official and scholarly discourses reviewed here reveal the asymmetrical positioning of men and women vis-à-vis the developing Standard language, a positioning which highlights the male voice as the authoritative voice of the nation (and, thus, voided of any gendered quality). This process is hardly unique to Japan. Black and Coward (Reference Black, Coward and Cameron1990/1981: 128–131) provide a review of how the development of the state in the West came increasingly to direct its representations of a “generalized ‘citizen’,” who was sexless and classless, but somehow universally male.

The fact that there were few explicit references to “male language” in relation to Standard Japanese, however, does not mean that men’s behavior was not regulated in early modern Japan. Shūshin Kyōkasho emphasized gender role differentiation in terms of the husband/man’s responsibility as the head of household, and taught boys to behave courageously (isamashiku) and actively (kappatsu ni) (Endō Reference Endō1997: 120–124). Instruction in male gender roles and behaviors heated up as Japan became more ultranationalistic in the period leading up to and during WWII. The 1941 government-issued Reihō Yōkō Kaisetsu ‘Commentary on the Essentials of Etiquette’ included instruction on language, emphasizing that men should use masculine language, while women should use feminine language (Washi Reference Washi, Okamoto and Shibamoto Smith2004: 79) and gave quite specific instructions about what (Standard Japanese) linguistic forms to use. This was also the period when scholarly publications and textbooks began contrasting men’s language with women’s language in their descriptions of such features as self-reference and address terms, sentence-final particles, honorifics, interjections, and lexical items.

After the war, this trend continued. Catalogs of contrastive male and female forms have come to be a standard presentational mode in grammar books and other linguistic literature (Ide Reference Ide and Kunihiro1982; Shibamoto Reference Shibamoto Smith, Hellinger and βussmann2003; Shibatani Reference Shibatani1990; Sugimoto Reference Sugimoto1997). Comparisons of gendered speech patterns are also made by presenting lists of contrastive example sentences, as in Atashi shitteru wa (yo) versus Boku shitteru ze (or zo) ‘I know’ for a female versus a male speaker, respectively (Sugimoto Reference Sugimoto1997: 222).

Men, like women, thus seem to be constrained and regulated by what circulates in official and popular narratives about manly behavior and manly language. Their behavior, including speech, may be evaluated negatively if they are not perceived as manly, as suggested by the existence of evaluating expressions like memeshii ‘sissy/effeminate’ and onna no kusatta yō na ‘(a man who is like a rotten woman)’.Footnote ²³ These are very strong “accusations” to offer a man, and ones with some gesture toward a man’s possible sexual orientation. A very important part of the notion of masculinity is the assumption of heteronormativity. Accordingly, heterosexual men may be quite concerned about not appearing to be speaking like women so as not to be seen as a gay man, as observed by Cameron (Reference Cameron, Johnson and Meinhof1997) and Kiesling (Reference Kiesling2005) with regard to English-speaking heterosexual men. On the other hand, gay men may intentionally cross the gender boundary and use stereotypical women’s speech patterns for constructing their complex gender and sexuality identities (see Chapter 6).

It is well recognized today that Japan abounds in social diversity that subdivides the macro-categories of women and men along fault lines of regionality, age, social class, sexual orientation, and the like. It is equally well recognized that many men and women do not use “gender-appropriate” forms, either due to particular aspects of their social identities or simply by personal choice, as we detail further in Chapter 6. The norms for gendered language covered in this section, however, serves to flatten that diversity and effectively to erase it from the narrative of the “authentic” Japanese women and her partner, the “authentic” Japanese man.

5.2.1 Media representations of normative linguistic femininity

Conduct manuals for women. As in the case of helpful manuals for mastering “correct” keigo, a visit to any bookstore (or Amazon Japan) will turn up a large number of self-improvement books instructing women how to speak. These offer a notable contrast to self-improvement books for men, in which one finds virtually no counterpart instructions on male speaking practice (for more on these latter books, see below). This asymmetry indicates how important it is – at least to book publishers and, one presumes, to book purchasers – that women speak in a socially appropriate manner.

A 2015 search on Amazon Japan for josei no hanashikata ‘women’s ways of speaking’ yielded sixty-seven books published between 1987 and 2014.Footnote ²⁴ In what follows, we present an analysis of these sixty-seven titles plus analysis of the contents of twenty-one of them. In particular, we focus on questions of how or whether the books in our corpus offer instructions on the use of gendered language as a set of rules or whether – again, as in the case of similar books on “correct” keigo – they help women to use gendered language as a creative index, or a resource, to construct femininity. We also examine to what extent these books focus speakers’ attention on the need to coordinate their use of a variety of linguistic and extra-linguistic expressions. Note that in our analyses, we consider not only general stylistic features and specific linguistic forms, but also extra-linguistic features, such as bowing and facial expression, as they offer an important clue to understanding the indexical processes involved in the construction of linguistic femininity.

There are many similarities between scholarly characterizations of joseigo and those found in self-improvement books. With regard to general stylistic features, both the “descriptions” of joseigo and the instructions about how to instantiate it in self-improvement books repeatedly mention characteristics such as politeness, gentleness, and refinement, and warn against linguistic rudeness, roughness, and vulgarity, as illustrated in Example (2). As pointed out earlier in the chapter, these characteristics serve to maintain the hierarchical gender order as well as for making personal evaluations of individual women on the basis of their attractiveness to men. Let us start, ladies, by not being unattractively “sharp”:

General styles are then linked to specific linguistic forms, such as honorifics, kushon kotoba ‘hedges’ (for example, osoreirimasu ga ‘excuse me’), aizuchi ‘backchannels’, indirect speech acts, and other softeners, as illustrated in Example (3):

This example suggests that honorifics index refinement, which can contribute to creating a beautiful woman through beautiful language.

One difference between the scholarly joseigo characterizations and the self-improvement books’ instructions is the latter’s extraordinary emphasis on speaking beautifully (utsukushiku or kirei ni) – the contribution of feminine speaking practices to female beauty receives far more emphasis in these books than in the linguists’ joseigo characterization. For example, words related to beauty are used in twenty-four of the sixty-seven books examined (36%), as illustrated in (4):

Words related to beauty are repeated numerous times in instructions in the texts of these books as well as in online lessons, as shown in Example (5):

Moreover, there are many compound words that include the word bijin (a beautiful woman), such as kotoba bijin ‘language beauty’, denwa bijin ‘telephone beauty’, kaiwa bijin ‘conversation beauty’, hanashikata bijin ‘speech beauty’, koe bijin ‘voice beauty’, and even keigo bijin ‘honorifics beauty’. Such emphasis on the importance of beautiful ways of speaking suggests the underlying assumption that a woman’s appearance is a core element of her value. It is difficult to imagine a book entitled Beautiful Ways of Speaking that Help You to Become a Beautiful Man.Footnote ²⁶

A second question is whether these sorts of conduct books treat particular linguistic forms as inherently feminine or as social constructs.Footnote ²⁷ The premise of instructions in the books we examined is that not all women are beautiful nor that they naturally speak beautifully, so they need to make efforts to learn to speak beautifully in order to become attractive. Many instructions take forms such as those in (6):

These examples indicate that linguistic forms such as honorifics and speech acts such as requests (versus commands) are assumed to contribute to the creation of general styles that in turn lead to femininity.

One of the points that these books underscore is that the linguistic forms they recommend must be accompanied with appropriate extra-linguistic elements, such as facial expressions and gestures, as illustrated in (7):

There are numerous instructions of this sort that emphasize the coordination of language and extra-linguistic manners of delivery. If the latter are not appropriate, they can “cancel” the meanings stereotypically associated with a linguistic form. In other words, these instructions point to the underlying assumption that linguistic forms are indirectly and ideologically linked to certain other qualities, and that it is the very detachability of the form from the stereotypified (or enregistered) linkage that allows indexical meanings in particular co-textual or contextual settings to be so variable.

And from this it follows that linguistic femininity is achieved not simply by the application of individual “feminine” forms of speech, but by the concerted and coordinated use of a variety of linguistic forms accompanied by appropriate extra-linguistic semiotic gestures. Scholarly characterizations of joseigo tend to focus on individual linguistic forms such as the final particles wa and kashira, treating them, however tacitly, as if they could be linked directly to female gender constructions. Instructions in self-improvement books, on the other hand, are much more complex. They teach women to become deft at strategically deploying a variety of linguistic and extra-linguistic forms in concert. For example, instructions given frequently talk about speech acts, as illustrated in (8):

This example offers women a model of how to decline an invitation, consisting of a number of components, including an apology, giving reasons for declining, offering an alternative idea, in order to appear sincerely sorry. Language, we see, or women’s ways of speaking, is “linguistic capital” (Bourdieu Reference Bourdieu1977) that can be used to raise a woman’s social value. These instructions treat language as a creative index (Bucholtz and Hall Reference Bucholtz and Hall2005; Eckert Reference Eckert2012; Silverstein Reference Silverstein, Basso and Selby1976, Reference Silverstein, Clyne, Hanks and Hofbauer1979) or a semiotic resource. This is a fundamentally different approach than the referential approach underlying the scholarly characterizations. Self-improvement books like the ones in our sample promise readers that, by reading them, they can acquire effective ways of speaking, which will turn them into beautiful women. In other words, they sell “expert” knowledge about “ideal” women’s speech as a commodity. Not all readers, of course, adopt the instructions given by these books. And by no means all women are readers. But some women may be led by these books to conclude that their speaking style is not properly feminine and that this detracts from their general social attractiveness.

Emphasis on improving one’s individual persona focuses attention on the function of indexical signs for the construction of the speaking individual rather than on that speaker’s interpersonal relationship with others, which is in part at least structurally shaped by elements outside individual identities (for example, status differences based on workplace relations, generational asymmetries, and the like). Agha (Reference Agha2005: 55) notes that “to be able to speak the [here, joseigo] register is to be able to perform an image of social personhood as one’s own image and to perform it in a register-dependent way.” In the case of self-improvement books aimed at a Japanese female audience, “a register-dependent way” of speaking is the stereotypical Standard Japanese-based feminine speech, although these books emphasize that women should apply these linguistic features appropriately accompanied by para- and extra-linguistic features.

Representations of linguistic femininity in the fictional worlds of television dramas. But women need not visit a bookstore to find models of normative femininity. They will come right into your home through televisual media. There have been numerous studies on media representations of Japanese women’s speech (M. Itō Reference Itō and Iwabuchi2004; Kinsui Reference Kinsui2003, Reference Kinsui and Nakamura2010; S. Kumagai Reference Kumagai2011; Mizumoto Reference Mizumoto2006; Mizumoto et al. Reference Mizumoto, Sugako and Kyōko2008; M. Nakamura Reference Nakamura2007b; K. Satake Reference Satake2004; see as well as our own studies: Okamoto Reference Okamoto1996a,Reference Okamotob, forthcoming; Okamoto and Shibamoto-Smith Reference Okamoto and Shibamoto Smith2008; [Shibamoto] Smith Reference Shibamoto Smith1992; Shibamoto Smith Reference Shibamoto Smith, Palmer and Occhi1999, Reference Shibamoto Smith, Okamoto and Shibamoto Smith2004, Reference Shibamoto Smith and Santaemilia2005, Reference Okamoto and Shibamoto Smith2008; Shibamoto-Smith and Occhi Reference Shibamoto Smith and Occhi2009). One of the notable findings of previous studies on this topic is that stereotypical feminine speech, or Standard Japanese-based joseigo, has been produced and reproduced in a wide variety of media genres, including films, television dramas, manga, anime, novels, and magazines. For example, comparing the speech of female characters in sixteen television dramas with that of real women (Standard Japanese speakers), Mizumoto (Reference Mizumoto2006) found that the former used stereotypical joseigo sentence-final particles (wa, kashira) in most sentences, whereas the latter hardly used them at all. Shibamoto (Reference Shibamoto1986) reported similar findings in data collected in the 1970s. Shibamoto Smith (Reference Shibamoto Smith, Okamoto and Shibamoto Smith2004) also reported a similar discrepancy in her comparative analysis of the speech of women in real conversations and that of female characters in romance novels with respect to the use of gendered personal pronouns and sentence-final forms, as did K. Satake (Reference Satake2004) in the use of sentence-final forms and request expressions in her examination of anime. Satake also observed that gentle, cute, and well-behaved female characters used more feminine language.

These discrepancies may indicate that in fiction, women’s speech is exaggerated, or “stylized” (Bucholtz and Lopez Reference Bucholtz and Lopez2011; Coupland Reference Coupland2001). Such exaggerations can cumulate to form linguistic stereotypes: “a given form, or repertoire of forms, is regularly treated as indexical of a social type by a given social domain of persons” (Agha Reference Agha2005: 45). Calling such linguistic stereotypes yakuwarigo ‘role-playing language’, Kinsui (Reference Kinsui2003) claims that such stereotypes tend to be used for creating a variety of characters, because they fit expected patterns of speech styles associated with particular social personae and are easy to understand. In other words, these stereotypes can be considered “mediatized artifacts” (Agha Reference Agha2011: 311–312), serving as linguistic commodities effective for creating fictional characters that are easily recognizable and about which certain qualities can be presumed (see also Kinsui Reference Kinsui2003 for a related discussion). The extensive use of stereotypical women’s language in represented speech found in these studies, however, does not mean that it is the only speech style used by female characters. Socially diverse female characters appear and their diverse speech forms are also represented, often however, along a dimension of relative centrality to the narrative, with the most central characters most likely to use Standard Japanese gendered forms. Example (9) is drawn from an interaction between the two most central characters, heroine Machiko and her closest friend Aya, in a many-times revived version of Kimi no Na wa ‘What’s Your Name’, an NHK drama that originated as a 1952 serial radio drama; our data are from the 1991 serialized television version.Footnote ²⁸ Here, feminine sentence-final particles are bolded; honorifics are underlined; feminine pronouns are indicated with a wavy underline.

Here we have two Standard Japanese speakers, close friends, speaking in a fairly intimate fashion in private; their speech is, however, a veritable showcase of strongly feminine sentence-final particles and honorifics. Lots of honorifics. And one occurrence of atashi, a strongly feminine first-person pronoun. Note, however, that as strongly feminine as Aya’s speech is, she cannot match Machiko’s expert, if depressed-sounding, deployment of honorifics. One speculation might be that all linguistic femininity is not equal, that it is not the femininity of sentence-final particles (or feminine pronoun choices) alone that construct the “real” heroine, especially in romantic dramas. It may be the case, but perhaps it should be those plus the much trickier deployment of honorific forms if a girl is to attract the romantic hero (Shibamoto Smith Reference Shibamoto Smith2006). Space precludes an extended examination of this possibility through subsequent decades, but as social realist dramas have reduced the use of sentence-final particles in the dialogue of the heroines from their peak in the 1980s (Mizumoto et al. Reference Mizumoto, Sugako and Kyōko2008), or diverted them into the dialogue of the close-but-non-heroine female characters (Shibamoto-Smith Reference Shibamoto-Smith2013), it is noted that heroines still tend to be able to use proper honorifics,Footnote ²⁹ and to use them frequently.

The orientation to standard language ideology becomes even more complex – and more interesting – when we consider issues of gender in the context of regionality in contrast to the urban center, or the Tōkyō metropolitan area, by examining the representation of the two social identity categories of gender and regionality. In our analysis of television dramas set in regional Japan (Okamoto and Shibamoto Smith Reference Okamoto and Shibamoto Smith2008; Shibamoto Smith and Occhi Reference Shibamoto Smith and Occhi2009), we found that heroines, especially young and middle-class women living in regional areas, tend to use stereotypical women’s speech, or Standard Japanese-based “Japanese women’s language,” even if that would be unlikely in real life situations. An example can be found in the speech of the heroine of NHK’s 2004 morning drama series Tenka, which is partly set in Sendai, Tōhoku. Tenka, the 17-year-old heroine after whom the series is named, grew up and lives in Sendai and is strongly loyal to the region.Footnote ³⁰ Throughout the series, Tōhoku dialect features are used to help create a sense of the Tōhoku locale, but they are used selectively and, critically for our purposes here, not by Tenka herself. Tenka and her friend Kaoru, in fact, use only Standard Japanese, including many joseigo features, such as the self-reference term atashi ‘I’ and the final particle wa.

The same pattern is seen in the 2005 morning drama Wakaba, set in Miyazaki and Kōbe (Shibamoto Smith and Occhi Reference Shibamoto Smith and Occhi2009). Wakaba, the heroine, was born in Kōbe but moved as a young child to her mother’s home in Nichinan, Miyazaki prefecture, after her father was killed in the Kōbe earthquake. She grew up in Miyazaki and in the early episodes of the series, is attending college there. In Example (10), she is talking to her best friend, Miyazaki native Eri, about her possible future with a local farmer’s son. Here, Miyazaki dialect forms are double-underlined, and Standard Japanese feminine forms are bolded. Kōbe dialect forms, as the urban, modern counterpart to eastern Japan’s Standard Japanese, are unmarked.

True to our promise that the virtual flood of strongly feminine forms would abate in the decades after 1991, when Kimi no Na wa was filmed, Wakaba and Eri use none of the most highly stereotypified forms, with the sparse exceptions of Wakaba’s use of the Standard Japanese feminine feature first-person pronoun atashi and both girls’ use of the highly feminine nominal + yo form (Eri’s Anta ∅ yo!, Wakaba’s hontō ∅ yo) (Shibatani Reference Shibatani1990). Analysis of longer segments of the full text, however, led Shibamoto Smith and Occhi (Reference Shibamoto Smith and Occhi2009) to conclude that, as the heroine, Wakaba was constructed as most linguistically lovable by using a) no Miyazaki dialect features, b) primarily cosmopolitan Kōbe features with c) a sprinkling of the Standard Japanese feminine features. And until very recently indeed,Footnote ³¹ this has been the typical representational model for heroines. What the future brings for Standard Japanese-based linguistic femininity in the world of media representations is yet to be determined.

Of course, in a televisual text language use is supported by a myriad of other “messages” offered through the multiple semiotic channels available to the viewer, such as dress, bodily posture, gestures, and the like (Silverstein Reference Silverstein2003: 14–15). And these multiple contrasts effectively serve to stratify language varieties as well as their speakers. Nonetheless, our examples here and in previous publications suggest that different varieties of Japanese come to be defined as “worthy” or “peripheral” through repeated media representations of stereotypes. In the Japanese case in question, Standard Japanese-based joseigo is likely to be interpreted as indexing its user as being an urban sophisticate, or linguistically equipped to become so, polite, middle-class, and feminine. In other words, speaking practice can point to which character(s) is (or are) qualified to play a maximally central role in the narrative. Many of the speaking practices that do so are consistent with the scholarly characterizations of joseigo we saw earlier. Use of a peripheral dialect, on the other hand, is likely to be construed as indexing its user as being variously rural, unsophisticated, old-fashioned, rough, or working-class, and not, thus, feminine enough for a starring role. In other words, rendering herFootnote ³² ineligible to be central to the textual narrative.Footnote ³³

The preceding discussion indicates that mediatized and commoditized Standard Japanese gendered forms index not only an orientation to the national centerFootnote ³⁴ (a first-order indexical), but also are involved in higher levels of indexicality, in particular in marking femininity. It has been pointed out that Japanese attitudes toward dialects have become increasingly positive in recent years (T. Kobayashi Reference Kobayashi2004; Y. Tanaka Reference Tanaka2011). Although dialects may be seen relatively more positively than in the past, however, the higher status of Standard Japanese as the normative language has not changed fundamentally and our analysis here illustrates a lingering linguistic inequality by indicating, albeit tacitly, which women should reap the rewards of the heroine – those who speak in stereotypically Standard Japanese-based “feminine” ways – and which should not – those who use stigmatized dialects. It also shows how closely gender is intertwined with the standard language ideology.

Televisual dramas do not tell their viewers explicitly that women should use the joseigo style of Standard Japanese or draw upon it for emphasizing linguistic femininity, although the implicit message is generally clear in the distribution of the types of characters who use different varieties. This implicitness, however, does not mean that such representations are ineffective. On the contrary, the very implicitness can be highly effective when the same kind of (re)presentation is repeated, because the repetition, as a “technique of normalization” (Foucault Reference Foucault1990/1978: 89), can work to naturalize the underlying ideology, whereby the receivers come to take the message for granted.Footnote ³⁵

5.2.2 Media representations of normative linguistic masculinity

As previously noted, Japanese society’s strong emphasis on women’s speech does not mean that men are free to speak in any way they please. Attempts to regulate men’s behavior, including their speech, directly and indirectly, abound in the popular media. In the remainder of this section, we examine the representations of masculinity and masculine speech in the fictional world of print and televisual text.

Before embarking on this, however, a discussion of the concept of masculinity in the Japanese case is in order. While dominant, or stereotypical, linguistic femininity images seem quite clear, the landscapes of linguistic masculinity are today not so clearly focused. Issues surrounding men and masculinity have been put forward in the media and in the emergent field of masculinity studies (T. Tanaka Reference Tanaka2009: 20). Traditional male-dominant social structures began to be compromised in the 1980s’ bubble economy, and more traditional ideas about masculinity destabilized (Watanabe Reference Watanabe1986: 12). The post-bubble 1990s, a decade of deflation and economic stagnation, threw sarariiman masculinity into crisis, and many men became uncomfortably aware that the old ideas about patriarchal dominance and its form of otokorashisa ‘masculinity’ were fast being discredited or disdained (Kimio Itō Reference Itō2009/1993: 75; Fujimura Reference Fujimura, Tsunehisa, Sumio and Masako2006). Overall, it has become difficult to find a singular common idea of masculinity for the twenty-first century (Kijima Reference Kijima, Shinji, Izumi and Takayuki2009: 140–141).

As early as the 1980s, phrases such as Asshii-kun, Messhii-kun, Mitsugu-kun, referring, respectively, to men who were willing to drive their girlfriends around, men who would take women out to meals, and men who bought women expensive gifts, circulated through the popular media. Other terms suggestive of new, and non-dominant, roles for men appeared later: femio-kun from the 1990s denoted men interested in women’s fashions, and in the new century, sōshokukei danshi ‘herbivore men’Footnote ³⁶ has come to refer popularly to gentle men uninterested in the active pursuit of women (Kimio Itō Reference Itō2009: 18); surveys show that many young men identify themselves as sōshokukei danshi.Footnote ³⁷ Although many deplore the “effeminizing” of the new man (Tanimoto and Nishiyama 2009: 50–51), the older normative model of a silent, powerful, and emotionally distant manly man absorbed only in work or career seems to be encountering some counter-discursive competition.

Today, then, multiple forms of masculinities compete in a field in flux, with older dominant models competing in a crowded field of alternatives, from ikemen ‘good-looking man’ to ikumen ‘fathers actively involved in child-rearing’ and beyond, to those herbivore men who are overtly uninterested in the concept of the masculine. Despite the somewhat bedraggled state of the sarariiman image (Bardsley Reference Bardsley, Bardsley and Miller2011), however, no other singular masculinity has managed to supplant it either in political-economic status or in dominant discourses.

This indeterminacy has consequences for language and masculinity. Although the majority of research on gendered language in popular materials focuses on linguistic femininity, there is a small but growing body of work on the construction of masculine personae in print and in televisual representations (Kinsui Reference Kinsui and Nakamura2010; M. Nakamura Reference Nakamura2007b; Occhi et al. Reference Occhi, SturtzSreetharan and Shibamoto-Smith2010; Shibamoto Smith Reference Shibamoto Smith, Okamoto and Shibamoto Smith2004, Reference Shibamoto Smith and Santaemilia2005, Reference Okamoto and Shibamoto Smith2008). Here, we build on these studies to explore what kinds of masculinity or masculinities are represented through male characters’ speech in the fictional world.

Although the evidence is very preliminary, it appears that at least two (perhaps, overlapping) models of masculinity emerge from our analyses. The first is the older model of rugged masculinity accompanied by forceful or rough language practices (Fujimura Reference Fujimura, Tsunehisa, Sumio and Masako2006), and the second seems to correspond more to a newer, gentler image of masculinity associated with language that is neither rough nor particularly indexical of physical prowess.

Conduct manuals for men. There are few conduct manuals specifically designated for men aimed at helping them with language issues. While the presentation of linguistic masculinity in the fictional world is implicit as in the case of presentation of linguistic femininity, there are also more direct comments on men’s behavior, including speech, as expressed in such public places as letters to the editor in newspapers and magazines and in online blogs (see Chapter 6). But there are only a few self-improvement books for men that focus centrally on speech, as indicated by our 2015 search for dansei no hanashikata ‘men’s ways of speaking’ on Amazon Japan, which yielded only four books. They are all about ways of speaking that help attract women, such as, for example, Naze ka josei ni moteru hanashikata to okane no himitsu ‘Ways of Speaking and Secrets of Money that Make You Popular among Women’ (Uekusa Reference Uekusa2014).

However, although the titles may not explicitly suggest that men are the target readership, there are many other books on speech to which men can resort. For example, a book entitled Hito o ugokasu riidā no hanashikata ‘Ways of Speaking of Leaders who Effectively Manage People’ (Sasaki Reference Sasaki2014) uses the gender-neutral term hito ‘person’ in its title (‘leaders’ – hito o ugokasu), but the book cover features a photograph of a middle-aged businessman. As a follow-up to his book Josei wa hanashikata de 9-wari kawaru ‘A Woman [Can] Change 90% (of her persona) by Her Way of Speaking’ (Fukuda Reference Fukuda2008), the same author published another book Hito wa hanashi-kata de 9-wari kawaru ‘A Person [Can] Change 90% (of his or her persona) by His or Her Way of Speaking’ (Fukuda Reference Fukuda2012). The second book includes useful advice for men as well as women. Overall, however, none of the books we were able to identify emphasized the importance of speaking practices that serve the construction of masculinity per se.

Representations of linguistic masculinity in the fictional worlds of television dramas. Fortunately, the fictional world has more to offer. An analysis of print texts published between 1970 and 2000 demonstrated that even by the end of the twentieth century the romantic heroes in popular romance novels rarely succeeded by aligning with older models of aggressive masculinity (Shibamoto Smith Reference Okamoto and Shibamoto Smith2008), although other, less central, male characters may have done so. Successful romantic heroes, then and now, tended to use moderately masculine Standard Japanese.Footnote ³⁸ The lover who, upon hearing that his longtime girlfriend has given up waiting for him to propose and has gone out with another young man, bursts out with Bukkoroshite yaru. Soitsu mo omae mo bukkoroshite yaru … Omae wa ore no onna da, dare ni mo watasanai zo ‘I’ll kill you. Him, you, I’ll kill you both. You’re my woman; I won’t hand you over to anybody’ (Mori Reference Mori1985: 12, masculine forms bolded) doesn’t tend to get the girl in romance fiction. But that was romance; other dramatic representations may follow different trajectories.

In historical reviews of the representational politics of Japanese masculinity, it is suggested that the forceful or rough language practices of the manly man of the past is passé, at least for the romantic hero (Fujimura Reference Fujimura, Tsunehisa, Sumio and Masako2006), but the older model of powerful or rugged masculinity does not seem to have disappeared completely from popular media representations. Although this is genre-dependent, the remnants of the ruggedly masculine, “manly man” that appear in televisual representations may be in part a reaction to the weakening of the status of the older model in real life in contemporary Japan. With regard to language, both boku and ore are used in social realist televisual representations, although more often than not without the accompanying strongly masculine sentence-final particles, the “vulgar” lexicon, and the phonological reductions of stereotypified “Japanese men’s language.” Stereotypically strongly masculine forms do appear in social realist dramas in particularly emotionally charged scenes, as in Example (11), an excerpt from Nihon Terebi’ s 2007 evening drama series Haken no Hinkaku ‘The Dignity of the Temp’. The excerpt features romantic hero (of sorts) Shōji Takeshi, who is goaded into a fit of temper by his love interest, Ōmae Haruko. We introduce him as he arrives at work just as Ōmae does, and he attempts to assess her reaction to his (sudden, unannounced) kiss the last time they met. What starts out very mildly escalates into stronger (and increasingly “masculine”) language as the impact of her overt indifference sinks in. Strongly masculine forms are in bold.

While there is an asymmetry in politeness (he uses plain forms versus Haruko’s desu/~masu forms), there are very few strongly masculine forms in Takeshi’s speech – at first. He does, from the outset, use ore, the strongly masculine first-person pronoun. But for the most part, his language is relatively neutral. That is, until she goads him into shocked anger – it is only then that we see a significant increase in forcefully masculine forms, with tomeru na (the bald imperative ‘don’t stop [me]’), da zo (the plain copular form, associated with men’s rather than women’s speech + the strongly masculine sentence-final particle zo), and second-person pronoun omae (also associated with sexist men’s usage toward women) all concentrated at the end of an encounter which leaves him virtually spitting with outrage as temp Haruko calmly walks off toward the elevators.

Stereotypically strongly masculine speech forms that are even rougher also appear in these dramas, but they are typically reserved for more-or-less private scenes between two men who are close friends or, on the other end of the spectrum, are arch enemies. Beyond the use of ore ‘I’ instead of boku, these scenes frequently include the phonological reduction of ai and oi to ee in forms such as itee ‘it hurts’ for itai; lexical items such as kuu ‘eat’ instead of taberu, and so on. These appear to be largely related to the construction of strong homosocial bonds (Kiesling Reference Kiesling, Campbell-Kibler, Podesva, Roberts and Wong2002, Reference Kiesling2007, Reference Kiesling2011) rather than a generic display of normative masculinity (see more evidence of this kind of man-to-man talk in Chapter 6).

The films and TV dramas we have examined also included strongly masculine heroes, in particular relatively young male characters portrayed as tough, rough, non-conformists, maverick, and/or rebellious.Footnote ⁴⁰ And here we will step out of our social realist, non-manga-derived set of data to look into the language lives of two young men, who offer – if not a realistic model for young men to adopt –a place where we can, in fact, still see the “norms” for strong masculinity made manifest. Hana yori dango ‘Men rather than Flowers’ (TBS, 2005 and 2007) and Doragon Zakura ‘Dragon Cherry Blossoms’ (TBS, 2005), very popular TV drama series,Footnote ⁴¹ both have young and handsome heroes who are portrayed as non-conformist, even rebellious men. These heroes are constructed in part through their use of strongly masculine speech forms. Examples (12) and (13) show this. Masculine forms are bolded.

The protagonist Dōmyōji is the son of a wealthy family; he is a high school student – that is, not a fully fledged shakaijin ‘adult out in the “real” world’ – and leader of a group of four powerful and at times violent bullies who dominate and prey upon their classmates. Example (12) is a representative example of his speech. It includes numerous masculine forms (bolded in the example), some of which are highly stereotypified as strongly masculine or rough. In this excerpt, Dōmyōji is browbeating heroine Makino because she won’t do what he wants. This situation may require the use of strong speech, but even in more normal, friendly conversations, Dōmyōji uses strongly masculine expressions, such as in the phonological reduction of diphthong ai to ee in Nani mo nee yo ‘Nothing in particular’, used in an ordinary conversation with a friend’s girlfriend in response to her question Dō shita no? ‘What’s the matter?’

In Doragon Zakura, hero Sakuragi Kenji is an individualist, a lawyer – young, but a full “adult” – and a substitute high school teacher. As a young man, he passed the entrance exam to Tōkyō University but decided not to go. This character is also constructed through dialogue in a style similar to that of Dōmyōji, although he shifts to a more polite style when talking to students’ parents and supervisors at school. His typically strongly masculine speech is illustrated in (13).

It has been argued that an individual man cannot be regarded as “having” one form of masculinity, but rather constructs different forms of masculinity to fit different situations (Kimio Itō Reference Itō2009/1993; Kijima Reference Kijima, Shinji, Izumi and Takayuki2009; T. Tanaka 2009 in relation to Japanese men) – a view that has been gaining increasing attention in relation to other languages as well (Bucholtz and Hall Reference Bucholtz and Hall2004, Reference Bucholtz and Hall2005; Connell Reference Connell2005). Sakuragi, more than student Dōmyōji, seems to support this notion. That said, no matter how realistic either of these characters (or the dramas in which they play central roles) may be, in terms of alignment with the dominant discourse of linguistic masculinity, these two characters offer a glimpse into what their speaking practices might be.

The dominant models of femininity and masculinity are repeatedly presented in the popular media through instructional programs in the education system, through scholarly writings, popular how-to books, and print and media fictional representations. In the context of the more explicit messaging about how to be a womanly woman or a manly man, the media representations that circulate through leisure time consumption may have considerable effects on men’s and women’s actual speaking behavior as well as their evaluations of themselves and of each other, as our analyses above suggested. However, as we will see in the next chapter, our analyses also suggest the existence of alternative femininities and masculinities. These competing ideologies and their linguistic reflexes are the focus of the next chapter.

6.1.1 Linguistic femininity: Metapragmatic commentary from blogs

We collected blogs posted to Yomiuri’s online site Hatsugen Komachi over a one-year period from January 17, 2014 to January 17, 2015 to see how the following three words were used in the initial post of each of the eighty-four threads of posts collected, as well as the response posts in one thread: onnarashii (4 posts) and joseirashii (41), both meaning ‘feminine’,Footnote ¹ and the word joshiryoku ‘(lit.) woman’s power’ (39). We examined the characteristics associated with these words. For onnarashii and joseirashii, the most frequently noted attributes concerned fashion (22 posts), followed by women’s physical beauty (13), their gentleness and reserve (12), refinement, elegance, or sophistication (6), bearing and mannerisms (6), and caring personae (4).

The word joshiryoku literally means ‘a woman’s power’. It is a recently coined word that was selected as the number-one new word of 2009.Footnote ² The word ryoku ‘power’ may evoke an image of a strong woman who does not conform to the dominant femininity norms of her time. According to Wikipedia, however, it refers to the power of a woman who exemplifies a “shining” (kagayakashii) way of living and who clearly exhibits her beauty and good taste; a woman popular among men.Footnote ³ We examined the thirty-nine posts that included this word, as well as twenty-five posts responding to an initial post asking what joshiryoku means. The most frequently mentioned characteristics concerned fashion (15 instances), followed in descending order by references to age (10), domestic qualities such as being a good cook (9), gentleness and concern for others (9), sensitivity (7), cuteness (7), beauty (6), and refinement or sophistication (6). These characteristics conform quite closely to the dominant ideology of gender roles and expectations for women discussed in the preceding chapter – an ideology that sustains the heterosexual gender arrangement as part of the normative social order (McElhinny Reference McElhinny, Ehrlich, Meyerhoff and Holmes2014; Milani Reference Milani, Ehrlich, Meyerhoff and Holmes2014).

In the same Hatsugen Komachi data set, we found that many writers’ ideas about femininity – being physically beautiful and having a gentle and caring persona – were quite normative. However, differing views were also expressed. For example, in the same data set we examined, nine women and one man wrote that they do not find a woman who expresses too much femininity attractive. Such a woman strikes them as trying too hard to look feminine or young, comes across as being too sexy, or appears cold and too focused on external factors. Six women reported that they considered themselves unfeminine for reasons such as being fond of sports, not wearing feminine clothes, or not being domestic, but they articulated themselves as content because they were able to attract a nice man (a husband or boyfriend). A number of women (in their 30s, 40s, and 50s) expressed concerns about their deteriorating physical features and hence not being seen as attractive women by their husbands or boyfriends. But in response, many writers wrote that there is age-appropriate femininity and that it is not the appearance but the persona that matters. Note, however, that the standard for being accepted is one related almost uniformly to being able to establish a successful heterosexual relationship.

The Hatsugen Komachi blogs we examined included very few direct comments on how women should talk, but there were some references. For example, one mother-in-law criticized her son’s wife for lacking onnarashisa ‘womanliness’, and supported her critique by saying that her daughter-in-law calls her husband by his first name without the honorific suffix -san, even in front of her in-laws. One 40-year-old woman, on the other hand, reported that a man to whom she was attracted criticized her for not being gentle, calling her speech kitsui ‘sharp’ and not onnarashii. In sum, our analysis of blogs suggests that as far as the majority of writers of these blogs are concerned, dominant ideas about femininity are shared.Footnote ⁴

With regard to online comments about the linguistic aspects of femininity, we supplement our data from Hatsugen Komachi with blogs from two other sites (Chiebukuro ‘Knowledge Bag’, a Yahoo! Japan blog site and NanJ Taimuzu ‘AnyJ Times’, a twitter site)Footnote ⁵. Several threads on each site concerned women who use otoko kotoba ‘men’s language’. A 26-year old woman, for example, wrote (on Hatsugen Komachi, February 10, 2006) that she often uses men’s language such as suggee ‘awesome’ and shiro yo ‘do it’ because it is straightforward, but that some men have criticized her, saying things like “If you were quiet, you would be cute” and so forth. She asked if men did not like women who use men’s language. In response, a couple of writers noted that it is a sign of nakama ‘group membership’, and that they do not consider her usages to be men’s language. Their comments indicate that it is not simply a matter of gender but also of the nature of the interpersonal relationship. On the other hand, four people thought that she should not use that kind of language because it is vulgar (gehin), indicative of the lack of good upbringing or intelligence, and not attractive to men who wish to protect (mamoritai) women. These comments suggest that linguistic femininity is closely related to class status (and, yes, to normative femininity). There were also responses that implied the relevance of age to linguistic femininity. For example, one person noted (posted February 27, 2006) that many young men seem to like women who use rougher language because it is cute, but added that an adult woman should try to use polite and correct Japanese so that she can become beautiful naturally.Footnote ⁶

6.1.2 Linguistic masculinity: Metapragmatic commentary from blogs

We now turn to masculinity and the metapragmatic comments on men and masculinity expressed in the Yomiuri online blog site Hatsugen Komachi. We searched for five expressions related to men and masculinity that were used in the initial post of each of the sixty-three threads of posts over the one-year period between January 17, 2015 and January 17, 2014. The five expressions and their number of tokens/posts (in parentheses) are: otokorashii ‘manly’ (29 instances/posts), danseiteki ‘manly’ (4), memeshii ‘effeminate’ (20), otoko no kuse ni ‘even though he is a man’ (5), and sōshokukei danshi ‘herbivorous man’ (5).

We start with otokorashii and danseiteki, both meaning ‘manly’. We have tallied all the tokens of the attributes the writers characterized as otokorashii. Those attributes that had more than three tokens were: dependable or economically secure (22 tokens/posts); active, decisive, or proactive (20); good-looking: tall, muscular, not bald (10); gentle (5); broad-minded (3), and timid (2). There were three posts written by men, who all expressed their concerns about appearing unmanly because of their indecisiveness or their appearance (short, skinny). The remaining thirty instances of danseiteki and otokorashii were mentioned by women, many of whom characterized otokorashii men as being dependable, economically secure, as well as decisive and proactive. These characteristics seem to reflect their beliefs about traditional heteronormative gender roles. Note also that some women mentioned gentleness as an element of masculinity, but at the same time noted other traits such as decisive and dependable. There were also quite a few references to men’s physical features, suggesting that men may also be evaluated for their appearance more than is generally imagined (see Miller 2003 for a discussion of the beauty demands on Japanese men in late modernity).

The word memeshii ‘effeminate’ is a highly evaluative word with negative connotations. Among the twenty instances we found, half of them were written by men, self-evaluating their own behavior as (possibly) memeshii; and the other half were written by women, evaluating a man, usually their spouse or (former or current) boyfriend. Particularly interesting is the former set of posts, in which male writers expressed concerns about appearing memeshii for thinking or writing about a past girlfriend or for behaving indecisively. For example, one male writer wrote that he kept thinking about his ex-girlfriend and experiencing all sorts of feelings such as anger, regret, and sadness. He added: “I know I’m memeshii, but I can’t help it” (posted July 11, 2014). Female writers, on the other hand, complained about men for being memeshii for traits such as lacking self-confidence, being passive, worrying about their appearance (e.g., fretting over thinning hair), not being dependable, and complaining about small things. The five instances of otoko no kuse ni ‘even though he is a man’ also mentioned by male and female writers involved comments similar to those on memeshii.

Among the five instances of sōshokukei danshi ‘herbivorous man’, four of them were written by women; one of them evaluated such men negatively as unmanly, but three expressed positive views. The sole sōshokukei danshi post written by a (24-year-old) man revealed that he wanted to become a nikushokukei danshi ‘(lit.) a man oriented toward meat-eating’ because he had been hurt when a woman he liked told him that he was too gentle, so not otokorashii. There were twenty-eight responses to his post. Thirteen of them said he need not change but just try to be more self-confident; twelve said that the term nikushokukei suggests an image of a tough guy, which is not necessarily good. Some added that a truly otokorashii man is one who is dependable, decisive, and competent at work.

Although the metapragmatic comments examined above are limited, they indicate that men are also subjected to evaluation. And the above observations indicate that men themselves seem highly constrained, and sometimes confused or conflicted, by the demands of the ideologies of masculinity currently competing for space in the Japanese social field. Clearly, the idea that men and masculinity are unmarked and invisible cannot be sustained (Benwell Reference Benwell, Ehrlich, Meyerhoff and Holmes2014: 240).

Not many posts commented on men’s ways of speaking per se, but there were eleven posts that did, noting that an otokorashii man speaks decisively and assertively or that he neither complains about small things nor is unnecessarily talkative. For example, one 26-year-old man expressed his worries about his new girlfriend, writing: “I truly love her. Whether I’m at work or at home, I always think I want to see her, but I can’t say that to her, because she might be put off if I show her this memeshii side of me” (posted on June 11, 2014). A wife reports that her husband said decisively (kippari to), in an otokorashii manner: “I will pay for the hotel, so let’s continue with the trip,” when they encountered trouble while driving to a resort and had to decide whether to return home or continue the trip.

One important issue common to the comments concerning both femininity and masculinity is that they all assume heteronormativity, as they largely concern evaluations of women and men involved in a marital or heterosexual romantic relationship. No writer mentioned homosexuality or even heterosexuality in talking about feminine or masculine behavior. As noted by Land and Kitzinger (Reference Land and Kitzinger2005: 387), social heteronormativity is still largely naturalized and taken for granted. In other words, sex, gender, and sexuality are ideologically consolidated (Milani Reference Milani, Ehrlich, Meyerhoff and Holmes2014: 261). How then are speakers of alternative sexuality and their speech perceived? Blogs we examined talked about memeshii ‘effeminate’ (heterosexual) men but not the behavior of homosexual men and women. In other words, non-heterosexual individuals do not surface in this mainstream public blog site. There are a number of terms used to refer to homosexual women and men, such as rezu, rezubian, gei, okama, and homo (Abe Reference Abe2010; Lunsing and Maree Reference Lunsing, Maree, Okamoto and Shibamoto Smith2004), and there is also a speech variety, onee kotoba, literally ‘older sister speech’, characterized by an exaggerated (or perhaps even parodic) deployment of stereotypically feminine speech. This variety is used by some gay men in some contexts (Maree Reference Maree2013), but its use is highly contested and largely limited to certain gay-friendly domains.Footnote ⁷

We saw earlier that most writers in Hatsugen Komachi expressed the newer model as an otokorashii ‘masculine’ man, that is, a dependable, economically secure, and active man with leadership qualities. However, there were expressions of diverse views. For example, there were writers who saw sōshokukei danshi ‘herbivorous man’ as weak and unmanly, but there were also many who expressed favorable opinions about them, because they were gentle or not domineering. One woman, for example, wrote that she is a strong assertive woman and cannot stand men who wish to dominate her. Quite a few of them noted that a truly otokorashii man is not a rough and tough guy (who fits the older model of masculinity), but one who is dependable, proactive, and competent at work – characterizations consistent with the newer model of masculinity.

With regard to the linguistic aspect of masculinity, there were blogs that concerned men’s use of women’s language. They also illustrate diverse views. For example, one writer (posted on Hatsugen Komachi on August 28, 2011) criticized her/his male colleague who used onna kotoba ‘women’s language’, such as shite minai ‘shall we try?’ or desho ‘right?’ because it is kimoi (a slang word for kimochi warui ‘weird’) and irritated her/him. Obviously, this negative perception stems from the writer’s judgment that her/his colleague’s speech does not meet the normative expectation that men should talk forcefully and decisively. Of the three responses to this post, one writer completely agreed with this writer. On the other hand, the other two said it was not women’s language, but a variety used in Tōkyō that may be perceived as feminine – a perception not uncommon among speakers of regional dialects. Here, we can see the intersection of gender, or linguistic masculinity, and regionality (see below).

One woman wrote (posted on Hatsugen Komachi on April 18, 2008) that her husband used women’s language with a gentle voice and prolonged sentence endings, and that she told him that she wanted a divorce if his language could not be “cured (naoranai).” Of the forty-eight respondents, seventeen wrote that they understood her feelings; three of the seventeen said that they were also irritated by their husbands when they talked like a woman; two said that, just as men wanted women to speak in a feminine way, it was natural for women to expect men to speak in a manly manner. These comments thus endorse the dominant linguistic gender norms and expectations. On the other hand, seventeen other writers expressed positive views toward men’s use of women’s language (onna kotoba). Many of them wrote that they liked a gentle and calm way of speaking; one woman was fond of her husband’s onee kotoba (stereotypically and exaggeratedly feminine speech used by gay men) because it was cute; another woman noted that men’s use of masculine speech was unnatural because they tried too hard to look masculine. These comments suggest women’s alienation from the image of a strong, tough man as the preferred object of (heterosexual) attractiveness but preferred their masculinity in a gentler form. Three writers noted that the initial poster’s husband might be a speaker of Tōkyō dialect because it tends to sound gentle and feminine to dialect speakers. Fourteen posts mentioned that the husband could not or need not change his speech style because it was difficult to do so and/or because it reflected his individuality.

Considerations of masculinity cannot be complete without taking into account the speaker’s sexuality (Kiesling Reference Kiesling2005; Levon Reference Levon2012; McElhinny Reference McElhinny, Ehrlich, Meyerhoff and Holmes2014; Wong Reference Wong2005). For example, Kiesling (Reference Kiesling, Campbell-Kibler, Podesva, Roberts and Wong2002, Reference Kiesling2005) observed the importance of heterosexuality in his study of US fraternity students, who would engage in evaluating women’s sexual attractiveness or treat weak men as women in homosocial contexts in order to construct their masculine identity linguistically. Cameron (Reference Cameron, Johnson and Meinhof1997) also found that young men in Britain engage in “queer bashing” or label weak heterosexual men as gay as a way to separate themselves from homosexuality and weakness to project their desired masculine image. These discursive activities contribute to constructing and reinforcing the speakers as normatively strong heterosexual men. As we will show in this section, in the Japanese context, too, men who speak in a feminine way may be criticized not only as being weak, but also as gay, which is considered an even greater deviation from the norm than a heterosexual man being weak. Many heterosexual men thus make great efforts to avoid being seen as homosexual.

Comments in Japanese blogs also reveal the importance of heterosexuality for masculinity. As can be seen in some of the comments above, the feminine speech men use may be regarded as onee kotoba ‘drag queen speech’ rather than simply onna kotoba ‘women’s speech’. For example, in one of the threads regarding this issue, the initial writer wrote (in Chiebukuro on February 21, 2011) that she knows many men who speak gently, but that one of them sounds not just gentle, but speaks like an okama ‘feminine gay man (a gay man who plays a feminine role, or a queen)’ because he uses the sentence-final particle no, which strikes her as incongruous for a man. It may be that gentle speech by men is fine as long as it does not cross the linguistic gender boundary that she believes to exist, that is, speech that does not suggest the speaker to be homosexual. The writer of the initial post in another thread (June 19, 2012) claimed that Ogi Naoki, a professor in education and a popular figure in the media, is a failure as an educator because he uses onee kotoba.Footnote ⁸ One respondent agreed, but another (a woman) wrote that it is not onee kotoba Ogi uses but just a gentle way of speaking, which is preferable to rough speech. Clearly, people differ in the way they conceptualize masculinity by assigning different linguistic gender boundaries.

Some of the comments related to sexuality also concern dialects versus Standard or Tōkyō Japanese. As some of the comments above noted, dialect speakers may perceive male speakers of Tōkyō dialect, or Standard Japanese, as being gentle and feminine. There were also several threads that concerned the question of whether men who use this variety sound like okama and are weird. For example, one thread was started on a tweet site (NanJ Timuzu, June 10, 2012) with the post “KantōFootnote ⁹ dialect is an okama-like way of talking (okamappoi shaberikata) and is weird.” In response, twenty-nine writers posted agreements with this statement, and twenty-three disagreed. The former included the following comments: It’s galling (mukatsuku); there are many homosexuals in Kantō; sentence-final endings like da yo ne and jan (both meaning ‘right?’) are disgusting. Two writers (one from Kansai and the other from Kyūshū) remembered that when they were students, male students who came from Kantō were bullied because they spoke like okama and were weird. In contrast, comments from those who disagreed included the following: Such a perception is something only countrified folks have; they are jealous of Tōkyō; it is just a perception of Kansai people and normally people do not like the vulgar, or low class (gehin na), Kansai dialect; it is rather Kansai dialect, not Kantō dialect, that is feminine or okama-like, because it uses vague expressions. While this debate can be seen as an instantiation of a Kansai and Kantō rivalry, these tweets put down the Kantō (or Kansai) variety by linking it to homosexuality as the most effective insult to someone’s masculinity. And the evaluations of Kansai dialect by (presumably) Kantō speakers imply their feelings of superiority in terms of such aspects as modernity, urbanity, and class status. This Kantō-Kansai debate illuminates the way the conceptualization of masculinity can differ depending on how linguistic forms are ideologically linked not only to gender, but also to other social aspects of individual’s identities, including sexuality and regionality or urbanity.

The comments examined above tell us how varied the perceptions of gendered linguistic forms are, and how subjectively and ideologically based. For one thing, these comments assume heteronormativity. Many writers used the word kimoi ‘disgusting’, suggesting that they believe that men should not “cross” (Rampton Reference Rampton1995) the linguistic gender boundary because it indicates that they are gays, that is, “non-normal.” Another important issue that emerges from these comments is that the intersection of gender, sexuality, and region (or modernity) requires more attention. We saw, for example, repeated references to Kantō dialect being feminine or okama-like – perceptions that seem particularly common among regional dialect speakers. The evaluation of Kantō (aka the Standard variety of Japanese) as unmasculine may originate in its stereotypical image as formal and polite, which are, cross-culturally, also features of normative feminine speech, but which in this case are associated with dominant norms for sarariiman stereotypical masculinity. These perceptions regarding standard and non-standard varieties are, of course, not peculiar to Japanese. For example, Trudgill (Reference Trudgill1972) and others regarded non-standard language as indexing covert prestige associated with working-class masculinity, while standardness is associated with femininity (see also Pujolar Reference Pujolar, Sally and Meinhof1997). According to Bucholtz and Lopez’s (Reference Bucholtz and Lopez2011) analysis of Hollywood films, Standard English-speaking white male protagonists try to adopt (stereotypical features of) African American Vernacular English to enhance their (heterosexual) masculinity, as Standard English is not adequately masculine.

We also observed threads that concerned the language used by gay men. In two threads (posted on February 21, 2011 and October 8, 2010 on Chiebukuro), the initial posters wrote that they are gay men but do not use onee kotoba, and that they feel irritated by the use of onee kotoba by okama men, especially by okama gei tarento ‘gay talents’ who tend to be vulgar (gehin). Two responses to these posts pointed out that, while those who use onee kotoba may not be gay, they tend to draw attention because of the way they dress and use language. Two people wrote that they do not have any negative feelings about gay men who use onee kotoba. The initial posts in two other threads (February 11, 2013 and August 16, 2013 on Chiebukuro) asked why the posts to “gay threads” on another blog site, 2 channeru ‘Channel 2’Footnote ¹⁰ are all in onee kotoba. One poster added that, although he is a gay man, he uses futsū no kotoba ‘ordinary language’ and not onee kotoba. Two responses noted that many gay men use otoko kotoba ‘men’s language’ and not onee kotoba in their daily life, but on gay blog sites it may be used to heighten the sense of group membership (nakama ishiki) and to reinforce solidarity (rentaikan) among themselves. Both also added that gay men who usually do not use onee kotoba tend to use it in gay bars, again to indicate group membership. These comments reflect the tensions within the gay community around onee kotoba, which lead to substantial inter- and intra-speaker variation in its use, as well as its interpretation as a variety of value for marking gay identity (Abe Reference Abe2010).

6.1.3 Thoughts on linguistic femininity and masculinity: A preliminary survey

Our last piece of evidence for diverse views, or metapragmatic commentaries, concerning linguistic femininity and masculinity is drawn from the results of a survey on mediatized language we conducted in January, 2015. We were particularly interested in what ideas about linguistic femininity and masculinity speakers of regional dialects would report, since gendered speech norms are based on Standard Japanese. We distributed a questionnaire containing six questions, English translations of which are shown in Table 6.1.

The total number of respondents was 124, but we eliminated 14 who did not supply demographic information and two who were from Tōkyō.Footnote ¹¹ Thus we examined a total of 108 responses – 42 from men and 66 from women. The respondents’ geographical areas spanned over twenty-four prefectures from Hokkaidō to Kyūshū.Footnote ¹² We divided the respondents into two age groups for each gender, as shown in Table 6.2. The sample size is too small for statistical analyses and no statistical analysis is presented here. A larger survey with more controls on age, social class, and regional distribution of respondents would make a significant contribution to the as-yet sparse audience response literature.

Because of the small number of older respondents, we do not refer to age differences unless there are noteworthy differences.

There was hardly any gender difference in the answers to the first question, as shown in Table 6.3. Of the 108 respondents, 32% thought that stereotypical feminine speech based on Standard Japanese is a model or ideal of women’s speech, while 64% did not share this perception.

In Chapter 2 we noted that Kansai dialects have been viewed as having higher prestige than other regional dialects and that speakers of this dialect often do not hold Standard Japanese in high regard. Question 1 in our survey specifically referred to Standard Japanese; accordingly, we separated the Kansai respondents (from Kōbe, Kyōto, and Ōsaka) from respondents from other regions to see if distinctive differences between the two groups emerge (Table 6.4).

Although the sample size is small, Table 6.4 shows a difference that is consistent with the previous observation regarding Kansai dialect speakers’ opinions of Standard Japanese. This result is consistent with some of the survey results discussed in Chapter 1.

Table 6.5 shows the reasons respondents gave as to why they answered affirmatively or negatively to Question 1. Only reasons mentioned by three or more respondents are included.

Respondents who reported viewing the normative variety of feminine language as a model also offered reasons highlighting the normative association of Standard Japanese-based feminine speech with politeness, refinement, and beauty (see Chapter 5). Most who responded negatively noted that this normative or “enregistered” variety of feminine language is not a type of language used in actual practice, but one only used in fiction, or that it is only used by older women, was used by women in the past, or by women in the upper class. Some thought it was not a model or ideal because women should be able to use language styles that fit their individual personae. Together, these results suggest that although many people may react to Standard Japanese-based joseigo as ideal, there are also many who appear to have different understandings of how women should speak, particularly in actual everyday practice.

Table 6.6 shows the answers to the second question.

As can be seen in Table 6.6, many more women than men reported that they have been told to speak in a gender-appropriate manner, supporting feminist and others’ assertions that women are more closely surveilled and their behaviors more strictly regimented than are men (see Chapter 5). That notwithstanding, when men or women reported having been instructed about what language forms to use, what they were told to use was quite normative. That is, men reported that they were told to use masculine forms like ore ‘I’ or boku ‘I’ and to avoid feminine forms like kashira ‘I wonder’. They were also told to speak in a loud voice and not to cry or make excuses. Women reported being told to use gentle and kirei na ‘beautiful’ language and not to use rough or kitanai ‘dirty’ language or not to use masculine language forms, such as boku ‘I’, umai ‘delicious’ (instead of oishii), kuu ‘eat’ (instead of taberu), hara hetta ‘I’m hungry’ (instead of onaka suita) or slang words like uzē na ‘annoying/bug off’.

Table 6.7 shows the answers to Question 3, listing attributes given by five or more respondents. Since, with one exception, there were no noticeable gender or age differences, the list covers responses from the whole group.

The last attribute, kirei na, was exceptional in that it was mentioned only by female respondents, which reminds us of the emphasis on women’s appearance seen in Chapter 5 in relation to self-help books and comments on the Hatsugen Komachi blog site. It is notable that all the attributes in Table 6.7 are consistent with characteristics attributed to normative feminine speech, including the nine responses that referred to specific Standard Japanese forms.

Table 6.8 shows the answers to the first part of Question 4.

As can be seen in Table 6.8, there is virtually no gender difference in the responses. Overall, there are slightly more people who think women should speak in a feminine way, although the significance of this slight difference cannot be assessed. Thus, even though respondents reported similar normative concepts about what feminine speech is (in their answers to Question 3), their views about whether or not women should use it are quite evenly divided.

Table 6.9 lists the reasons for the respondents’ answers to the first part of Question 4; only reasons offered by more than five respondents are included.

All three reasons listed above for supporting the use of feminine speech align with the dominant discourse of linguistic femininity. Moreover, they indicate that feminine speech is understood as a resource for constructing women’s (gendered) personae as attractive, gentle, and refined, and not simply a reflection of her sex. The reasons for not supporting the use of feminine speech vary, but also suggest that language style is seen not as a direct reflection of one’s sex, and that it can be used to index and also bring about various social meanings pertaining to individual personae, specific situations, or, indeed, to a speaker’s gender construction.

Table 6.10 shows a summary of the answers to Question 5, listing the attributes mentioned by five or more respondents.

These attributes are quite consistent with the characteristics of stereotypical masculine speech. Many of them (the first six) align with the older model of masculinity. Note, however, that there were quite a few people (nine respondents: four men and five women) who noted that a gentle and calm way of speaking is preferable to rough ways of speaking. The last attribute in the table indicates that heterosexuality is expected or assumed in language use.

Table 6.11 lists the answers to the first part of Question 6.

According to Table 6.11, the number of women who answered affirmatively (38%) was fewer than those who answered negatively (55%), indicating that in this small sample, the majority of female respondents did not report a preference for men to speak in a masculine manner. In contrast, about half of the male respondents (49%) thought men should speak in a masculine manner, suggesting that many men either endorse the dominant norms or are constrained by the dominant norm to report that they do.

Table 6.12 lists the reasons mentioned by five or more respondents for their answers to the first part of Question 6.

As in the case of reasons offered for preferring women to use feminine speech, the reasons for or against the use of masculine speech appear to indicate that respondents understand language as a linguistic resource, or “semiotic capital” (Agha Reference Agha2005). Five men and one woman wrote that if a man does not speak in a masculine manner, he might be seen as effeminate, again highlighting the importance of heterosexuality in the dominant masculinity discourse. On the negative side, three of the four reasons (numbers 1–3) given for not privileging masculine speech styles are the same as the answers to Question 4 regarding feminine speech. Responses noted the importance of choosing language appropriate to particular situations – a point also made in many posts to the blogs examined earlier. Furthermore, some respondents noted individual variation in the use of language, suggesting a view of masculinity as a diverse construct, with language as one resource useful in its construction. The last reason offered here also suggests that some respondents see language forms associated with the dominant notion of linguistic masculinity as a creative index, but also hint at a preference for the newer model of masculinity rather than the old, rugged stereotype.

The observations made in this section – admittedly drawn from a very small sample – offer a preliminary glimpse into public responses to the circulation of models of linguistic femininity and masculinity. The responses to our pilot survey suggest that, while the culturally dominant formations of “women’s language” and “men’s language” are recognized, in the everyday world of Japanese speakers, stances toward those models vary. Normatively feminine linguistic forms may be construed as indexing politeness and refinement, gentleness, modernity, and/or beauty, on the one hand, or aloofness, awkwardness, and/or excessive sexiness, on the other. Likewise, normatively masculine linguistic forms may be interpreted as variously indexing attractive traits such as forcefulness and reliability, or less attractive traits such as roughness, vulgarity, and bad upbringing. This is consistent with the contemporary notion in language and gender research that the conceptualizations of linguistic femininity and masculinity must be understood not as uniform, but rather as variable and multiple (Bucholtz Reference Bucholtz, Ehrlich, Meyerhoff and Holmes2014; Bucholtz and Hall Reference Bucholtz and Hall1995, Reference Bucholtz and Hall2004; Eckert and McConnell-Ginet Reference Eckert and McConnell-Ginet2013). These findings indicate that the interpretation of the normatively gendered linguistic forms may vary widely among individuals and across contexts, that linguistic forms are indexically variable and multiple, in ways similar to the variable interpretations of Standard Japanese and “correct” keigo seen in the previous two parts of this volume. We are not surprised, thus, to find it to be the case here. Indeed, social meanings in the indexical fields of linguistic forms are multiple, fluid, and open-ended (Eckert Reference Eckert2008, Reference Eckert2012). Such variable interpretations in turn may have an important bearing on women’s and men’s actual speaking practices, a topic addressed in the following section.

6.2.1 Face-to-face private conversations

We start with the intersection of regionality and femininity. From the perspective of the dominant narrative of linguistic femininity, based firmly in Standard Japanese, regional dialects tend to be marked as unfeminine, as illustrated in the normativizing official and scholarly characterizations examined in Chapter 5. However, not everyone shares such a perception. As an example, we present excerpts from conversations of female speakers of Tōhoku dialect, one of the most negatively evaluated dialect groups (see Chapter 1). Example (1), drawn from our data set collected in 2006 in Sakata CityFootnote ¹³ in Yamagata prefecture in Tōhoku, illustrates the use of stereotypically masculine forms by female speakers of a regional dialect. Example (1) is an excerpt from a conversation between two elderly women. (In the examples, forms considered normatively strongly masculineFootnote ¹⁴ are bolded.)

Throughout the thirty minutes of their recorded conversation, both Speakers I and J used forms normatively characterized as strongly masculine, including those bolded in Example (1) – these include the lexical items umai and umee ‘delicious’ in lines 2 and 3 (instead of oishii), kutte ‘eat’ (instead of tabete), and yazu ‘thing’ in line 2 (instead of mono); the self-reference term ora in line 6 (instead of watashi or atashi);Footnote ¹⁵ and the phonological contraction of ai to ee (nee in line 1 and umee in line 3). Not only did they use these forms, their speech was also replete with other dialectal features, for example, the voicing of intervocalic /t/, as in midee ‘try’, and /ts/, as in yazu ‘thing’ in line 1; the merging of /i/ and /u/, as in sūzun ‘season’ in line 5;Footnote ¹⁶ the use of the presumptive auxiliary verb –bee in lines 5 and 6; and lexical items, such as gozzō ‘good food’ instead of gochisō, bari ‘only’ instead of bakari, and honte ‘really’ instead of honto ni), which may also be considered unrefined and hence unfeminine.

However, in using these (from a Standard Japanese perspective) “masculine” forms, Speakers I and J are clearly not trying to be unfeminine, or deliberately trying to cross a recognized linguistic gender boundary, in order to convey some particular meanings that would be associated with such deliberate “crossings” (such as subversiveness, creating an image of a strong woman, or marking an alternative sexuality). In this local context, for elderly speakers using these forms it is simply the normal mode for conversations between elderly female friends. Another study of rural women speakers, this time in Ibaraki prefecture,Footnote ¹⁷ found that the women she observed used forms considered to be men’s forms from the standpoint of Standard Japanese, but argued that such use may serve to reinforce solidarity among them (Sunaoshi Reference Sunaoshi, Okamoto and Shibamoto Smith2004). We, too, see the two elderly Tōhoku women’s use of “masculine” forms, as well as other dialectal forms, as not only “normal” for the place, the interlocutors, and the setting but also as serving to indicate friendship or solidarity between them. Their speech forms, that is, are not about femininity. Indeed, if they used Standard Japanese-based feminine speech, especially given all the dialect forms not implicated in gender constructions contained in their utterances, it would be very awkward and likely to trigger negative reactions. As suggested by some of the metapragmatic comments we observed in the preceding section, each would open herself up to being seen by her interlocutor as distant, pretentious (and possibly ridiculous). The uses “counted” as masculine in Example (1), then, may simply signify the speakers’ regionality, closeness, and age.

Standardization has, however, been advancing in this region (Yoneda Reference Yoneda1997). When we compare Speaker J’s speech with that of her granddaughter, who is in her early 20s, we see that the younger woman’s use of dialectal features is somewhat less extensive. The granddaughter (but not J), for example, used Standard Japanese forms such as the final forms na no ‘it is’ (a normatively feminine form) and da yo ne ‘right?’ (which, however, is a normatively masculine form) as well as the self-reference form atashi; she also used both Standard Japanese and dialectal variants of the same variables (nai and nee ‘not’; honto and honte ‘really’) in the same conversation. The speaking choices of the granddaughter are thus likely to index a modern, relatively more urban persona while the speaking choices of the older women seem to index a more local and rural persona.Footnote ¹⁸

We emphasize, however, that although Speakers I and J may not use stereotypically feminine sentence-final forms or self-reference and address terms, they did use many other features that may contribute to the construction of femininity. These features include extensive uses of politeness strategies, including both positive politeness strategies such as offering compliments, making (other) offers, agreeing, and sharing frequent laughter, and negative politeness strategies, such as expressing reservation and modesty, as seen in Example (1). The conversation is fast-paced with many (cooperative) overlaps and backchannels, and the intonation patterns are not blunt but rather give the impression of gentleness. It is clearly, in terms of cooperative and polite conversational style choices, possible to express femininity at the discourse level using a regional dialect, even a generally stigmatized dialect. We see in this example speakers who are concerned to be polite, reserved, and empathetically oriented to their interlocutor – all qualities found in the dominant discourse of linguistic femininity at the stylistic level. While more research is required, this example suggests that the speakers negotiate linguistic gender norms vis-à-vis their immediate local context and use a variety of linguistic and extra-linguistic features to construct personae they think appropriate for elderly women interacting in a specific social situation.

The language choices of the two women in the above conversation suggests that they are not simply performing gender but doing so in ways that are complexly intertwined with regionality, age, and their own interpersonal relationship. This is not to say, however, that some people – listening in on this conversation from a different social positionality, in particular people who uphold the standard language ideology – will not perceive their speech as being unrefined and unfeminine. They certainly may do so.

Let us further consider the interaction of language, gender, and age, this time focusing on the speech of young middle-class women who are speakers of Standard Japanese. Although the speech of this group is relatively well studied, we will examine it from a new angle. It has been reported that young women use fewer honorifics, fewer stereotypically feminine sentence-final particles, and more stereotypically masculine forms than older women (Okamoto Reference Okamoto, Hall and Bucholtz1995, Reference Okamoto1996a; Okamoto and Sato Reference Okamoto, Sato, Hall, Bucholtz and Moonwomon1992; Takasaki Reference Takasaki2002). And they have been criticized by many as being unfeminine, as mentioned in Chapter 5. According to the interviews with young women (college students) in Okamoto’s studies, however, women claimed to be using masculine forms in order to avoid the use of normatively feminine forms, particularly “strongly” feminine forms such as sentence-final da wa and kashira, in order not to sound like older women (see Section 6.1 for this interpretation).

A second look at the same data, however, shows that in addition to masculine sentence-final forms, these young women quite frequently used stereotypically feminine forms, such as the self-reference terms watashi and atashi. In addition, they used intonation patterns characteristic of young women, that is, a prolonged level or fall-rise intonation, as well as forms that have not been well studied in relation to linguistic femininity in Japanese such as hedges (toka ‘and so on’, mitaina ‘like’), overlaps, backchannels, and sentence co-constructions. Many of these are used particularly frequently by young women (see below). Furthermore, young women, along with their older counterparts, used speech act strategies that are often associated with stereotypical femininity’s stylistic characterization of feminine speech as gentle and reserved, such as indirect speech acts and agreements.

Young women may negotiate gender norms in ways that are today recognizable as feminine by their age peers, at least, but that clearly mark them off from the older women in the data who are seen as more closely embodying the dominant gender norms for both behavior and for speaking. This may represent an alienation from mid to late twentieth-century gender constraints, or it may simply be a way to mark youthfulness. More research is clearly needed to determine whether – as has happened before with, for example, the recuperation of the once-transgressive schoolgirl sentence-final particles from the Meiji period – new trends (and thus, new norms) are on the horizon for linguistic femininity in Japan.

In order to pursue this point further, we look more deeply into the details of how gender interacts with other aspects of the social context. Fine-grained analyses of the conversations presented below help uncover ways in which speakers negotiate linguistic gender norms simultaneously with social membership categories such as regionality and age, along with the particulars of interpersonal relationships in reference to each interactional context. As our first example, we analyzed two dyadic same-sex conversations, one between women and one between men. Both were recorded in 2010.Footnote ¹⁹ The participants in the two conversations are shown in Table 6.13.

The four participants are from Gunma prefecture or nearby areas in the Kantō region; all used Standard Japanese except for a few dialectal forms used by the male speakers. Each conversation lasted for about forty minutes. Both conversations were transcribed. Following Okamoto and Sato (Reference Okamoto, Sato, Hall, Bucholtz and Moonwomon1992), we examined 130 utterances (an approximately 15-minute segment of each conversation) for each speaker’s use of normative gendered features, including not only sentence-final particles and self-reference terms – the two most commonly examined features in the literature – but also many other features that are stereotypically and non-stereotypically gendered. All features examined are listed in Table 6.14.

Our analyses show wide inter- and intra-speaker variation within each conversation. In what follows, we first present the findings concerning each feature and then discuss their implications.

The first feature is sentence-final forms. Following the classifications used in Okamoto and Sato (Reference Okamoto, Sato, Hall, Bucholtz and Moonwomon1992), we examined the use of sentence-final forms by the four speakers; our findings are summarized in Tables 6.15 and 6.16.

The two female speakers’ use of gendered sentence-final forms showed distribution patterns quite similar to those of the young women in Okamoto and Sato’s (Reference Okamoto, Sato, Hall, Bucholtz and Moonwomon1992) and Okamoto’s (Reference Okamoto, Hall and Bucholtz1995, Reference Okamoto1996a) earlier studies. That is, Speakers FA and FB used more masculine forms than feminine forms, although they both used neutral forms the most frequently. The two male speakers used neutral forms the most frequently, but they also used masculine forms, including strongly masculine forms, quite frequently. Although it is impossible to generalize from such small samples, these findings are suggestive of a somewhat greater alienation from the gender norms of linguistic femininity on the part of the young women than we see in the men’s conversation. Here, and as we will see in terms of other features as well, FA used masculine forms less frequently than FB, contributing perhaps to an impression of somewhat different femininity orientations or slightly different interpretations of the indexicalities of the forms themselves.

With regard to honorifics, a feature often associated with femininity, the four speakers used neither addressee nor referent honorifics for each other as the two speakers in each conversation are close friends. However, some of them used the polite, or beautification, prefix o- (Chapter 5), as shown in Table 6.17, which shows the presence or absence of the prefix in all places where it is possible to use o- before a noun.

While the polite prefix o- is often associated with femininity (Chapter 5), Okamoto (Reference Okamoto1996a) found that the ten younger women (ages 18–20) used it less frequently (on average 43%) than the ten older women (67%). The current data set does not allow us to compare the age differences, but we see that FB used it only infrequently, and that FA used it more frequently than FB, although its use was not that extensive, either. The greater use of the feminine prefix o- by FA as compared to FB is consistent with the difference between the two speakers with regard to the use of the sentence-final forms we saw above. For example, FA used o-mise ‘store’ and o-bentōya-san ‘lunch vendor’, which includes both the prefix o- and the suffix -san, while FB used mise and bentōya. On the other hand, the two men did not use the o- prefix (for example, niku ‘meat’; hashi ‘chopsticks’, with the sole exception of MA’s o-hirudoki ‘lunch time’).

Table 6.18 shows the use of self-reference and address terms.

With regard to the use of self-reference terms, all four speakers align with normative expectations. Both men used ore, a term considered more masculine than boku (Miyazaki Reference Miyazaki, Okamoto and Shibamoto Smith2004), while the feminine form atashi was used by both women. However, only FA used atashi exclusively (in the entire 45-minute conversation), while FB used atashi and watashi, which is normatively defined as less feminine than atashi.

With regard to the use of address terms, FA used none in the 130 utterances examined (nor in the entire 45 minutes of the conversation). FB’s use of the addressee’s first name and -chan, a suffix often used by children and women, is normative, but she also used two masculine forms, omae and omee. MA used two different forms of address (omee and the addressee’s last name without a polite title suffix), which are normatively masculine. MB used an address term only once (MA’s last name + suffix-chan), which is, put most simply, odd.

Table 6.19 shows the use of normatively masculine or “vulgar” lexical items.

Here, the two men used “masculine” lexical items more than the two women. There was only one lexical item coded as feminine (chitchai ‘small’); both FA and FB used it once. Neither of the two men used “feminine” lexical items.

In addition, we examined the use of slang as part of wakamonogo ‘youth language’ (Yonekawa Reference Yonekawa1998). Women are, and, we recall from Chapter 5, have long been, normatively expected to avoid slang. See Table 6.20.

While MA’s use of these words in terms of the number of tokens is larger than that of the others, it is mainly because he used maji de ‘seriously?’ ten times. Overall, the four speakers, in particular FB and MA, used slang terms relatively frequently. According to Lauwereyns (Reference Lauwereyns2002), slang terms are used much less frequently by older women than younger women. The use of slang by FA and FB also seems to suggest their young age. There are no comparable numbers for older men versus younger men.

We now turn to the use of phonological features. Table 6.21 shows the use of the phonological reductions of /ai/ and /oi/ to /ee/ (shinee ‘do not do’ for shinai) – a phonological set of variants widely seen (in Standard Japanese) as (strongly) masculine and “low class.” (We examined all cases in the 130 utterance segments in which /ai/ and /oi/ could be contracted.)

FA did not use any contractions, but FB used them 7 (of 27 possible) times. The two men, however, used this feature quite frequently.

The intonation patterns were found to be clearly gendered. Both the two women used a phrase- or utterance-final prolonged level or fall-rise intonation considered characteristic of young women. (In the data sets for Okamoto and Sato (Reference Okamoto, Sato, Hall, Bucholtz and Moonwomon1992) and Okamoto (Reference Okamoto1996a), the younger participants used it extensively, but not the older women.) Both FA and FB used it, but FA more frequently than FB (e.g., FA: Gurando ga ate::, sono mawari ga:: ma, yama na n da kedo:: ‘There is a field track::, around there:: are mountains but::’; see also Example (2)). The two men did not use these intonation patterns. As for the pitch height, the use of a high-pitched voice is normatively associated with femininity, and FA used a very high-pitched voice in both utterances and laughter. FB’s pitch height was much lower. MA used a relatively higher pitch than MB, who used the lowest pitch among the four.

The interactional/pragmatic features are a very little-studied arena of Japanese language and gender research, although they have received sustained attention in other, largely Western, languages. The first feature examined was the use of hedges, which are associated with young women (Lauwereyns Reference Lauwereyns2002); examples are toka ‘and the like’, te yū ka ‘or something’, mitai na ‘like’, and kanji ‘something like’. Table 6.22 shows the results of the analysis.

Although the numbers of tokens are small, the two women appear to make greater use of hedges than the men. This is consistent with previous observations that hedges are associated with young women (Talbot Reference Talbot2010). Data from an earlier study (Okamoto Reference Okamoto, Hall and Bucholtz1995, Reference Okamoto1996a) showed that younger women in Tōkyō also used more hedges than older women did.

Table 6.23 shows the four speakers’ use of backchannels (for example, un, aa, hūn) and occurrences of laughter as a form of backchannel – a feature often interpreted as a means of showing a cooperative stance, which is cross-culturally, but largely on the basis of studies done in the West, stereotypically associated with women’s conversational styles (Chapter 5; Coates Reference Coates2004; Fishman Reference Fishman, Thorne, Kramarae and Henley1983).

Among the four speakers, FA stands out in her frequent use of backchannels. In contrast, FB used them quite infrequently. The two men in fact used backchannels more than FB did, suggesting that they are engaging in a cooperative and supportive conversation, unexpected in terms of the cross-cultural understanding of men’s conversations as less collaborative and more oriented toward status positioning strategies.

Table 6.24 shows the occurrences of overlaps (of words or parts of words, and not of backchannels or laughter) and co-constructions that are supportiveFootnote ²⁰ – again a feature associated with femininity (Coates Reference Coates2004).

The two women’s conversation was fast-paced, while the two men’s was slower and full of pauses. Example (2) illustrates the two women’s supportive use of overlaps and co-constructions.

In line 1 in (2), FB says “I take small change (to buy lunch) and,” and in line 4, FA takes over FB’s utterance and says “go and,” which is an instance of co-construction. Meanwhile, FB continues in line 3, so FB’s and FA’s utterances partly overlap in lines 3 and 4. FB continues her utterance in line 5, which is partly overlapped by FA’s utterance in line 6, which is a supportive repetition of FB’s utterance (kaette kite ‘come back’) followed by an elaboration of what FB is saying in line 5 (kau ‘buy’) describing how FB buys lunch – an instance of co-construction.

Table 6.25 shows the numbers of repetitions of the other speaker’s utterance or part of it.

The repetitions observed are all cooperative or supportive ones – qualities resonant with the normative prescriptions for linguistic (or discursive) femininity described in Chapter 5. The two women used repetition quite frequently, as did one of the male speakers, MA. Example (3) illustrates MA’s repetitions.

MA’s repetition in line 2 offers empathetic support for MB’s thoughts about what to do next after taping the conversation.

The last feature we examined concerns the delivery of speech acts. As noted earlier, both speakers in each conversation were engaging in an informal and very friendly conversation. Both pairs made extensive use of positive politeness strategies (Brown and Levinson Reference Brown and Levinson1987), using direct speech acts in offers (FA: Jibun de nomi na yo:: ‘You drink it’; MA: Kuu? ‘Do you want [to eat] this?’), requests (FB: Yaite ‘Bake it’; MA: Akete ‘Open it’), in (joking) criticism of the addressee (FB: Meiwaku na yatsu ‘annoying girl’; MA: Mendokusee na, omee wa ‘You are a real pain’), in making proposals or suggestions (FA: Dare ka no tsute de shika nai ‘[Using] a personal connection is the only way’; MB: Yameta hō ga ii ‘You’d better not do it’), and the like. In addition, both conversations were full of expressions of agreement, as indicated by the use of the repetitions seen above. The use of positive politeness strategies is associated with femininity (Holmes Reference Holmes1995), but the two men in these data also used them extensively.

On the other hand, direct speech acts are considered more masculine, but the two women used them extensively. In older Western research based on assumptions of essential differences in the ways in which girls and boys were socialized to the “rules of friendly conversation” (for a good review of this research trajectory, see Wodak and Benke Reference Wodak, Benke and Coulmas1997), this might be surprising; it might, from this perspective, be hard to explain why FB’s discursive strategies shift from “normative” (hedges, overlaps and co-constructions, and repetitions) to “partly normative” (speech acts) and “non-normative” (speech acts). More recent research, which is cautious about totalizing narratives (such as Japan’s dominant language and gender norms) (Cameron Reference Cameron2005), this – along with many of the findings with respect to the other features we marked in these conversations – is more suggestive of a need to rethink the norms for expressing femininity and masculinity linguistically than it is of a need to rethink the femininity or masculinity of these speakers.

Although the use of these “gendered” featuresFootnote ²² needs to be further investigated with much larger samples, these two female speakers show an interesting pattern of mixing “feminine” forms with “masculine” ones as a discursive strategy. In much Japanese research, specific linguistic forms were considered to be tightly linked to a culturally normative femininity or masculinity. Thus, it is unsettling to see that both FA and FB use “non-normative” sentence-final particles with “normative” self-reference terms and only “partly normative” or “non-normative” use of the bikago ‘beautification’ prefix o-.

Our look at these four speakers’ conversations also tells us that the indexical meanings of gendered linguistic forms vary depending on the context. For example, both FA and FB used forms that are not “gender normative,” including the greater use of neutral and masculine sentence-final forms than feminine forms, the infrequent use of the prefix o- (especially by FB), and the use of masculine lexical forms and slang. Furthermore, FB also used other “masculine” forms, including address terms, the contracted form /ee/, and low voice pitch. However, just as in the case of Speakers I and J, the two elderly women from Yamagata, these observations do not mean that these women are ignoring gender norms or aiming at presenting themselves as manly. Their speech also included normatively feminine forms of other features. FA used many feminine forms, including the stereotypically feminine atashi ‘I’, a high-pitched voice, intonation patterns associated with young women, hedges, backchannels, overlaps, co-constructions, and repetition. FB also used some of these as well, although overall, she deployed them less frequently than FA.

Furthermore, the interactional functions of the strongly masculine forms used by the two young women differ from the ways their male counterparts used them. That is, these young women’s use of strongly masculine forms was largely restricted to contexts such as joking, teasing, expressing a strong emotion (anger), criticizing a non-present third person, or quoting their own speech; and their delivery was often accompanied by a mitigation of the force of such expressions by deploying hedges such as mitaina ‘it’s like’ or accompanying their remark with laughter. Example (4) illustrates FB’s use of strongly masculine sentence-final forms.

FB works part-time for a car wash stand and is criticizing customers who bring in extremely dirty cars. She used strongly masculine forms such as the overtly rude Fuzaken na ‘Stop fucking around’ and the bald imperative Mainichi arae ya ‘Wash it every day’, as though she were talking to a customer. Although she would not use such strong or rude expressions with her customers,Footnote ²³ in this context she seems to think that these masculinity-associated forms best express her annoyance. However, she then qualified her remarks with the mitigator tte omou yo ne ‘you would think, right?’ Such qualification suggests that she is aware of and constrained by the normative expectation, whether it be the normative expectation that she meet the feminine norm to avoid strong language or the normative expectation that employees don’t talk to customers that way. FB used strongly masculine forms of other features in a similarly restricted manner. For example, she used omae ‘you’, but used it when addressing herself in a lightheartedly self-deprecating way over some mistake she had made, and omee, but again, when she was teasing FA about something.

In the case of the two male speakers, the use of strongly masculine forms is not restricted to special contexts, but rather used such unmarked offerings as Mizu motte koyō ze ‘Let’s bring water’ (MA), Denki ga tsukanee ‘The light won’t turn on’ (MA), and Gamen kawatta zo ‘The screen changed’ (MB). Moreover, the utterances that contain these strongly masculine forms are not accompanied by any qualifiers or laughter, as in the case of FA’s and FB’s use of these forms. The same applies to the use of masculine lexical items and the contracted /ee/, which is quite regularized in both young men’s speech. Clearly, both FA and FB present themselves as more constrained by dominant gender norms when balancing using strongly masculine forms against an awareness that on some level they are violating the linguistic gender boundary. But these usages also indicate that the same masculine form may be interpreted quite differently. That is, when used by these women, it may be seen as an index of a certain speech act or emotion and also as a sign of friendship or solidarity, which is generated by the act of crossing, or breaking the norm. It may also be a way of indexing their youth – a way of creating an age-appropriate femininity distinct from that associated with older Standard Japanese-speaking women. These two young women’s use of masculine forms could also be interpreted as a way to construct attractiveness in terms of “cuteness” (Matsumoto Reference Matsumoto1996), but this requires further examination.

On the other hand, the two young men’s use of masculine forms requires less in the way of analytic justification; it is likely that they are more straightforwardly “talking while being men.” The differences between them are minor, and the farthest they “stray” from the normative is to violate the “norms” (from, we remind readers, older research models largely employed in Western language and gender research) of men as competitive conversationalists. MA and MB both participate in their conversation in very friendly and collaborative ways, as exemplified by their use of supportive backchannels, repetitions, and speech acts. The images of masculinity that emerge from MA and MB’s conversation are, thus, far from the old model of forceful and rough masculinity. The overall friendly tone of their conversation also may suggest that, while they used strongly masculine forms quite frequently, these may in fact not be deployed as indexes of alignment with the old model of rugged masculinity but rather of homosocial friendship. Note also that their use of strongly feminine features was quite limited and did not include use of any of the stereotyped feminine pronouns or sentence-final particles, which may represent the kind of “heterosexuality display” that men perform, both in Japan and elsewhere, not so much as to stress their masculinity but to forestall potential questions about their heterosexuality (Kiesling Reference Kiesling, Campbell-Kibler, Podesva, Roberts and Wong2002).

6.2.2 Talk on television

Issues of sexuality intertwine with gender to add further complexity both to the Japanese language and gender norms and to language practices. The need to move beyond the bounds of heteronormatively gendered discourse is increasingly recognized outside the Japanese sociolinguistic arena, as sexuality is seen to be one of the social aspects of a speaker that cuts across gender in particularly salient and complicated ways (Bucholtz Reference Bucholtz, Ehrlich, Meyerhoff and Holmes2014; Cameron and Kulick Reference Cameron and Kulick2003; Eckert Reference Eckert, Campbell-Kibler, Podesva, Roberts and Wong2002). There has, of course, been some interest in the area of Japanese language and sexuality, starting with the very early explorations of Ogawa and Shibamoto Smith (Reference Ogawa and Shibamoto Smith1996, Reference Ogawa, Shibamoto Smith, Livia and Hall1997), Shibamoto (Reference Shibamoto1986), and Valentine (Reference Valentine, Livia and Hall1997). Valentine’s study tended to focus on special lexical terms for various alternative sexualities (gei ‘gay’, homo ‘homosexual’, resubian or rezu ‘lesbian’, okama ‘feminine gay men’); Ogawa and Shibamoto Smith’s, and Shibamoto’s studies focused on issues of the appropriation by gay men of stereotypical Standard Japanese “women’s language” forms. The field has expanded its focus since then to include considerations of the construction of diverse personae through the selective appropriation of particular cross-gender forms, often in specific (delimited as “gay”) contexts (see Abe Reference Abe, Okamoto and Shibamoto Smith2004, Reference Abe2010; Lunsing and Maree Reference Lunsing, Maree, Okamoto and Shibamoto Smith2004; Maree Reference Maree2011, Reference Maree2013; Shibamoto-Smith Reference Shibamoto Smith2002). It has also expanded its focus away from particular usages (first-person pronoun choices) by gay and lesbian speakers in gay-friendly settings to include considerations of such non-stereotypical features as script choice in lesbian versus mainstream women’s magazines and linkages of lesbian-oriented script usages (that is, the mix of hiragana, katakana, and kanji) to such notions as modernity and cosmopolitanism (Franks Reference Franks, Campbell-Kibler, Podesva, Roberts and Wong2002).

In order to explore the possibilities for better understanding the relationship between language, gender, and sexuality, we emphasize the importance of approaching this issue without resorting to static identity categories such as “gay” and “straight,” as the linguistic practice of speakers in each category is complex and manifold. Rather, we need to employ a dynamic approach in which speakers are seen to negotiate the dominant linguistic gender norms in specific contexts – an approach consistent with the approach taken above with regard to the speech of (supposedly) heterosexual speakers. In so doing, we highlight the fluid and multiple indexical meanings of gendered linguistic forms that contribute to the construction of a particular persona in a given interaction.

Examining the speech of gei tarento, Maree (Reference Maree2011, Reference Maree2013) distinguishes onee kotoba and onee kyara kotoba. She characterizes onee kotoba as a parody of stereotypical Japanese women’s language that challenges the heteronormatively dichotomous gender that often symbolizes a strong community sense of queerness, and onee kyara kotoba as a stereotypical media-hyped onee kotoba used by gay men or transgender/transsexual tarento ‘talents’, or celebrities, for entertainment. In terms of specific linguistic forms, the two “varieties” overlap considerably, but onee kyara kotoba may be more exaggerated or stylized than onee kotoba. Here we further examine the speech of gay tarento to demonstrate the complexity and multiplicity of its indexical meanings.

The data we examine here are drawn from a variety show called Sanma no Honto no Koi no Kamasawagi ‘Sanma’s Empty/Gay Men’s Fuss about True Love Affairs’Footnote ²⁴ (broadcast on April 13, 2011 and on August 23, 2011) to which the host Akashiya Sanma (a 58-year-old man) invited ten gay tarento and two straight men and one straight woman as guests. Among these guests, several, those who are most well known, had frequent speaking turns, while the others seldom spoke. We focus first on the speech of Matsuko Derakkusu ‘Delux’, because compared to the other guests he consistently used an informal style with plain (or non-honorific) forms, which generally include gendered speech forms more than honorific-based formal styles. Matsuko is a cross-dressing gay tarento as well as an essayist and columnist; heFootnote ²⁵ is known for scathing commentaries on social issues and also for his hugely overweight body. Example (5) illustrates Matsuko’s use of stereotypical feminine forms, which are underlined:

The stereotypical feminine forms Matsuko used in Example (5) include the self-reference term atashi in line 4, the final forms no (after the plain form of a verb), da wa in line 3, and no yo and da wa in line 4, the word-final prolonging level intonation in line 6, as well as a phrase-final prolonged level intonation pattern. Matsuko thus uses many stereotypically strongly feminine forms, the sheer number of which constitutes a stylized (and exaggerated) form of linguistic femininity. Moreover, there are other indexes of stylized femininity, including Matsuko’s long black dress, his long hair, and the fact that he has all kinds of Chanel perfumes.

The use of a stylized, or exaggerated, form of feminine speech may serve to indicate that Matsuko is not trying to be like an “ordinary” heterosexual woman. There are also other features that indicate this. For example, Matsuko uses a low-pitched and very deep voice. When he conveys negative information (for example, expressing criticism or disliking someone), he speaks in a scathing manner, as Maree (Reference Maree2011, Reference Maree2013) also observed in her data. Moreover, Matsuko occasionally uses stereotypically strongly masculine forms in certain contexts, such as when expressing anger, disagreeing, protesting, or strongly criticizing someone or something. For example, in Example (6), Matsuko is teasing a straight female guest Tanaka Minami, saying that he does not like her.

Before line 3 in Example (6), Matsuko had been using stereotypical feminine forms, teasing Tanaka. But in line 3, he suddenly switches to strongly masculine forms (in bold) using a very deep and strong voice. The audience bursts into laughter. Three other gay tarento guests also make use of mixtures of feminine and masculine forms in similar ways, although their speech tends to be in a formal style using addressee honorifics, which forecloses the opportunity for many sentence-final gender markers. For example, Haruna Ai, a transgender tarento, singer and actor, who appears wearing a short pink dress, uses addressee honorifics with some feminine features, including a high-pitched voice and frequent use of the interactional particle ne ‘right?’ with rising intonation to solicit interlocutor agreement. But when sheFootnote ²⁶ was interrupted by one of the guests who disagreed with what she was saying, she says, Urusai n da yo, omae wa! ‘Shut up, you!’ in a deep and low-pitched voice.

Matsuko’s (and Haruna’s) use of feminine speech illustrates linguistic gender boundary crossing. It evokes a “possibility of disrupting the normative alignment of sex assignment, gender identity, and sexual identity through such practices as drag performances, in which gay men appropriate femininity in order to challenge heteronormativity” (Bucholtz Reference Bucholtz, Ehrlich, Meyerhoff and Holmes2014: 37). That is, drag performances reveal the artificial or “imitative” nature of gender, which may be otherwise taken for granted as natural (Butler Reference Butler1993: 175). Furthermore, Matsuko’s use of exaggerated, or stylized, feminine speech delivered with some unfeminine features (in particular, a low-pitched deep voice) serves to “deauthenticate” his speech as being not his own genuine voice. In fact, it is a parody that is designed to induce laughter from the audience. However, while it is a parody (cf. Maree Reference Maree2011) and appears to challenge the heteronormative norm, it is at the same time acknowledging the existence of the norm. In other words, it amounts to reinstantiating, however ironically, the very heteronormative discourse it is subverting. This is parallel to the case of the use of regional dialect by Standard Japanese-speaking actors (Chapter 1).

The fact that the use of stereotypical women’s speech by gay tarento in public is “allowed” may suggest a degree of tolerance toward alternative sexualities being manifested in speaking practices, but in fact any tolerance is very limited to the realm of entertainment, where deviations from macro-sociologically developed and circulated norms can be suspended. These speech practices gesture toward a kind of aesthetic of deviation; they are not meant to be taken seriously. And they do not link to political action aimed at broadening the bases for understanding gender. This is, rather, reminiscent of the limited acceptance of regional dialects that we observed in the media (Chapter 1).

On the other hand, we cannot ignore the fact that gay tarento like Matsuko exploit gendered speech as a resource in their construction of successful media personae. Their language may contribute to the construction of their identity as gay men, but it seems quite different from their use of feminine speech in private interactions within the gay community. Those interactions are not intended to be taken as humorous or funny. Gay tarento, on the other hand, utilize both stereotypically feminine and masculine speech to entertain an audience. For “men,” whether merely cross-dressing or actually male-to-female transgender, crossing the normative gender boundary by using a highly feminine speech style is “unexpected” or incongruous, and this has been utilized by the television industry to create humorous effects, although some may not share in the “fun” but instead – for a wide range of reasons – find these performances unpleasant or odd, as seen in the metapragmatic comments examined earlier. Likewise, the use of stereotypically strongly masculine forms seems intended to be perceived as funny due to the “reverse crossing” by supposedly very feminine gay men. In this way, both highly stereotypical feminine and masculine speech varieties are objectified and “sold” to audiences as linguistic commodities.

We also note that even though we used the term “gay tarento” to refer to all ten guests, they themselves were quite a mixed crowd, differing in the way they dressed – ranging from men’s suits to a most becoming dress to more adult women’s dresses to kimonos. They also differed in their speech, as indicated by the differences between Matsuko and Haruna in their voice quality and their use of plain versus honorific verbal forms.

Another issue related to the inter-digitation of gender and sexuality is the intra-speaker variation in gay male speech. A gay man does not always talk in the same way. As we saw in the preceding section, some gay men are reported to use onee kotoba in a gay bar or in online blog sites for gay men for the sake of building solidarity, but not when they are interacting in private. We can see this kind of style shift, or intra-speaker variation, for example, in Matsuko’s speech. As a popular gay tarento, he appears in many different TV shows. One of these shows is his own talk show Matsuko no shiranai sekai ‘The World that Matsuko Does Not Know’, in which he interviews a guest each time. He changes his speech style depending on the guest in terms of the use of addressee honorifics, but in all cases, he uses exaggerated feminine speech styles infrequently, unlike his speech in the Sanma no Koi no Kamasawagi excerpt we saw above. In one of the shows in which the guest was a male expert in the making of meishi ‘business cards’ (broadcast March 3, 2013), Matsuko basically used addressee honorifics throughout. In the first twenty minutes of the show, therefore, he used feminine sentence-final forms only three times (as, for example, in okane kakete n no ne ‘they are putting a lot of money [into this], aren’t they?’), while he used normatively masculine forms six times, although only one was an exaggerated strongly masculine form (temee nameten no ka ‘are you kidding?’ addressed to an imagined person, and not to the guest).

In another show (Go-ji ni muchū ‘Immersed in five o’clock’), where he serves as a news commentator, Matsuko basically speaks in plain forms with an occasional use of addressee honorifics. Compared to his speech in Sanma no Koi no Kamasawagi, however, strongly feminine final forms are much less frequently used. In contrast, he used strongly masculine forms more frequently as he offers scathing comments on the issues in question (Dete ike! Nani yatten da yo, omae ‘Get out! What are you doing?’, addressed to the (non-present) people he is criticizing). These examples indicate that Matsuko uses gendered linguistic forms in a complex manner in order to construct different personae appropriate for each situation (a persona as one of the guests invited as gay tarento, that of an interviewer, and that as a critical news commentator); and these are only examples from his speech in public, or on television, not from his speech in private. Style shifts such as these involving speakers of alternative sexualities have been attested in other languages (Podesva Reference Podesva2011) and its importance is increasingly coming to be recognized (Queen Reference Queen, Ehrlich, Meyerhoff and Holmes2014: 214). As additional support for this research, our analysis demonstrates how the speaker’s gender and sexuality interacts with other aspects of the context, including the speaker’s current role, the genre, interlocutors, and speech acts.

As indicated above, interactional contexts are also important for understanding the diversity in the use and interpretation of gendered speech. We further consider this issue by examining the speech of a (supposedly) heterosexual male speaker. The data are drawn from the same data set from the talk show Sanma no Manma examined in Chapter 2. We analyze the speech of Akashiya Sanma, the interviewer of the show, in four interviews. See Table 2.3, reproduced here as Table 6.26).

In Chapter 2, we saw that Akashiya changed the ratio of Standard Japanese and Ōsaka dialect variants according to who his interviewee was, with the use of Standard Japanese morphological and lexical features ranging from 16 to 85%. With respect to Akashiya’s use of gendered forms, we examined the following features: self-reference terms, address terms, honorifics, reactive tokens, sentence-final forms, lexical items, and speech acts. His use of gendered speech also varied widely in the four interviews. Table 6.27 shows Akashiya’s use of self-reference and address terms.Footnote ²⁷ The table also includes the percentage of Standard Japanese morphological and lexical forms and the number of addressee honorifics used in seven-minute segments of each interview to see if this is related to the use of gendered forms.

As seen in Table 6.27, Akashiya used three different self-reference terms: ore, boku, and watashi. Two of them, ore and boku, are normatively regarded as masculine, ore being characterized as more masculine than boku. Watashi can be used by both women and men, although it is considered relatively feminine.Footnote ²⁸ Akashiya used only ore in Interviews 1 and 2, where the interviewees are younger men. These two interviews are more informal than Interview 3 and 4, where the interviewees are older, well-established women. In Interview 3 Akashiya used boku four times and ore three times. The use of ore may be due to the fact that he has known the interviewee (Wada Akiko) for many years and that they are good friends. In Interview 4, with an equally well-known and older women (but one he did not know as well), he used boku five times and watashi twice; he did not use ore.

As for the use of address terms, Akashiya used the masculine form omae in Interviews 1 and 2 and the last name of one of the guests without an honorific suffix in Interview 1. On the other hand, he did not use omae in Interviews 3 or 4 except once in Interview 3 when he was jokingly protesting something Wada said about him, to which he responded Omae yaro! ‘It’s you, isn’t it?!’ Otherwise, he used Akko-san, her first name plus the honorific suffix -san, in Interview 3; and in Interview 4 he used Agawa-san, the interviewee’s last name plus the same honorific suffix, and also obāchan ‘grandma’, which he uses when teasing Agawa about her age. Table 6.27 also shows that Akashiya’s (non-)use of masculine forms is related to his use of Standard Japanese and addressee honorifics. Most overt markers identified as strongly masculine or strongly feminine are neutralized in formal speech. Akashiya did not use verbal or other honorifics in interviews in which he used ore and omae, which is precisely reflective of this point. The use of ore, boku, and omae not only index the speaker’s gender, but also the degree of formality and the nature of the interpersonal relationship.

Another feature that exhibited clear differentiations in Akashiya’s use of gendered forms is the use of reactive tokens, or forms used to respond to another speaker’s conversational offering. The reactive token ō ‘yeah’ is a normatively masculine form; un ‘yeah’ is informal but not gendered; hai and ee, both meaning ‘yes’, are more formal and not gendered. In Interview 1, Akashiya used ō thirteen times; he did not use un, hai, or ee. In Interview 2, he used ō 6 times and un once; he did not use hai or ee. In Interview 3, he did not use ō; he used hai nine times, ee six times, and un six times. In Interview 4, he used only ee six times and did not use ō, un, or hai. In other words, Akashiya used masculine tokens only in Interviews 1 and 2, a result similar to his use of masculine self-reference and address terms as well as honorifics.

With regard to sentence-final forms, Akashiya did not use many normatively masculine forms, but when he used them, they were restricted to Interviews 1 and 2. In Interview 1, there were only two instances (ie yo ‘say it’; chigau zo ‘it’s different’). This may be in part due to the fact that he used Kansai dialect, which has fewer gendered forms than Standard Japanese, more uniformly in these interviews. In Interview 2, he used normatively masculine forms eleven times (yū na yo ‘don’t say that’; ee ya nai kai ‘isn’t it fine?’; hayo sē yo ‘hurry’), including even some Standard Japanese forms (kiiteta yo na ‘you were asking, weren’t you?’; zubutoin da ‘You’re pretty bold there’). In none of the interviews did Akashiya use normatively feminine final forms except for desho ‘isn’t it?’ (rather than daro), which appeared in Interviews 3 and 4. Desho, however, is not a strongly feminine form and is widely used by men, as we saw with Speakers MA and MB above.

Akashiya also used a few normatively masculine lexical items, but only in Interview 2, where he used umai ‘delicious’; koitsu ‘this guy’; aitsu ‘that guy’; yatsu ‘guy’; oi ‘hey’. This may have to do with the fact that the interviews are televised, that is, public, which may cause speakers to avoid expressions considered ‘vulgar’ or ‘coarse’. Akashiya did, however, prefer direct approaches to “face-threatening” speech acts (Brown and Levinson Reference Brown and Levinson1987), using bald on record strategies when making requests, criticisms, and protests, and giving negative evaluations, mainly in Interviews 1 and 2, where he says things like Omiyage ‘(Give me) a present’ when asking for presents from the guests or Yū na yo ‘Don’t say that’. A bald on record FTA did appear, but only once in Interview 3, when he told interviewee Wada Sore mitara ii n desu yo ‘You should look at that’, and even then, the addressee honorific desu served to soften the imperative force of the directive.

As we see, Akashiya did not use masculine speech in the same way across the different interview contexts. Rather, he varied the use of a variety of gendered speech forms in order to construct a complex repertoire of personae shaped to meet current interactional needs; this set of personae encompasses his roles as a male talk show interviewer and as a comedian from Ōsaka. These role inhabitances interact with those of his guest interviewees. Furthermore, contextual variation occurred even within interviews, as he switched his language use from a neutral, polite style to a strongly masculine style depending on the interactional needs (or content) of the moment.

In summary, our analyses in this section demonstrate the importance of interactional contexts for the use of gendered linguistic forms. They indicate that speakers are neither completely conforming to nor completely ignoring linguistic gender norms. Rather, they negotiate such norms vis-à-vis the relevant features of the local context, including regionality, age, sexuality, the interpersonal relationship between speaker and interlocutor(s), the pragmatic work being done through particular speech acts, and other social variables that we have not, in these small samples, examined here. Through this negotiation, speakers use gendered forms by combining many surface-segmentable linguistic features and discourse patterns in a complex manner in order to meet their on-the-ground, moment-to-moment interactional needs and according to her or his stance toward the dominant norms, just as in the use of Standard and regional variants (Chapter 2) and the use of honorific and plain forms (Chapter 4).

This diversity in use is intimately related to diversity in the interpretation of gendered forms. Our discussion in this section illustrates the polyindexicality of linguistic forms by demonstrating how the same or similar linguistic forms, in this case, stereotypical women’s and men’s language, may be interpreted differently depending on the context, with different social and political implications. For example, strongly masculine forms, such as ore ‘I’ or the contracted form /ee/, may index masculinity when used by heterosexual men as well as friendship or informality. When used by heterosexual women, they may be interpreted as unfeminine from the normative point of view, but for (elderly) women in a rural area in Yamagata, it may be a sign of friendship and also their age and regional origin without giving any impression of breaking the gender norms. When used by young, even regional dialect-speaking, women, it may be construed as breaking the norms, which in turn may index friendship, solidarity, emotions, and certain types of speech acts. When used by gay men in a TV variety show, it may index their (alternative) sexuality as well as their current role as an entertainer. As mentioned earlier, issues of polyindexicality are seldom raised explicitly in the Japanese sociolinguistic literature, but we argue that it is crucial to examine them closely in order to better understand the relationship between norms and situated practice.