Introduction
Building on recent work supporting a sociolinguistic approach to orthographic choice (e.g. Miyake Reference Miyake2007; Bender Reference Bender2008; Sebba Reference Sebba2009, Reference Sebba, Sebba, Jaffe, Androutsopoulos and Johnson2012; Jaffe Reference Jaffe2012; Robertson Reference Robertson2017), this study analyzes the indexical potentialities of script variation in Japanese mass media discourse. The stylistic and visual resources that construct the speaking ‘voice’ are many, ranging from aural features such as voice quality (e.g. Teshigawara Reference Teshigawara and Kinsui2007; Redmond Reference Redmond2016) and structural ones like syntax and semantics (e.g. Sadanobu Reference Sadanobu and Kinsui2011) to more content-oriented ones like lexical variation (e.g. Kinsui Reference Kinsui2003). In writing, languages with more than one orthographic system find that script variation also carries with it the potential for indexing sociocultural meaning, a topic that has until recently been overlooked in broader discussions of indexicality and the construction of ‘voice’. Utilizing the lens of audience design (Bell Reference Bell1984, Reference Bell, Eckert and Rickford2001), this article engages with the potentialities of script variation relative to (i) the writer's own language ideologies and (ii) that writer's assumptions about the linguistic expectations of an imagined readership.
This study explores script variation in the context of the Japanese comics magazine Comic Yuri Hime, which has been in publication since 2005 and is the only commercially owned manga magazine dedicated to yuri ‘lily’, a genre that focuses on same-sex romantic and erotic relationships between female characters. The genre owes its conventions to a number of origins, including mainstream shōjo manga ‘girls’ comics’, which is traditionally aimed at young girls, and ‘S’ narratives, a literary genre centered on hierarchical, homoromantic friendships between girls. The works that appear in Comic Yuri Hime are written by a primarily female authorship for a majority female readership, but as a means of appealing to the minority male contingent, from 2007 to 2011, the editorial board published a quarterly male-targeted version of the magazine, Comic Yuri Hime S. In 2011, the two magazines merged once more into one magazine under the name Comic Yuri Hime and attempted to rebrand for a broader audience.
With these changes regarding imagined readership in mind, this study analyzes text outside the stories that run within the magazine from 2005 through 2011. Referred to by Genette (Reference Genette, Newman and Doubinsky1982/1997) as paratext, this term refers to any text that occurs in situations where the author of the text can be understood as speaking more directly to the reader than in the main text (e.g. in titles, subtitles, author commentaries, or forewords). More specifically, this analysis deals with the paratextual use of second-person pronouns as they are used to address the reader. In Japanese, because the use of a second-person pronoun is optional, its utilization is inherently marked. Further, there are multiple different options at a speaker's disposal (e.g. anata, kimi, etc.), each of which vary in such ways as level of politeness, gender-related indexicality, or dialect association.
By analyzing the function of script variance for the purposes of creating or dissolving a gendered space, as well as in delineating a target audience on the part of the writers and publishers, this study contributes to a deeper understanding of the indexical potential of Japanese pronouns as intertextual cues in popular media. Additionally, because present-day Japanese has four script systems at its disposal, not to mention a degree of variation that occurs specifically at the visual level rather than the aural, it is possible to explore socially salient script-selection practices that may be more limited in other languages. While the topic of script variation has long been an area of exploration in Japanese, the bulk of literature in this area engages predominantly either with frequencies of certain script styles across the language as a whole (e.g. Iwahara, Hatta, & Maehara Reference Iwahara, Hatta and Maehara2003; Tomoda Reference Tomoda2009), or across a particular written genre or collection of genres more broadly (e.g. Reference SatakeSatake 1989b; Franks Reference Franks, Campbell-Kibler, Podesva, Roberts and Wong2002; Robertson Reference Robertson2015). Meanwhile, studies that engage with script variation specifically in the paratext are virtually nonexistent, illustrating a need to examine the way that script choice intersects with authorial intention and imagined audience. To that end, this article begins by giving a summary of script practices and variation in Japanese. Following that is a brief review of the concept of mass media paratext, how it relates to audience design (Bell Reference Bell1984, Reference Bell, Eckert and Rickford2001), and the role that second-person pronouns play in this context. Then, I give an abbreviated history of the yuri genre, explain the nature of the dataset, and finally analyze and discuss the data.
Script, social meaning, and Japanese media
Basics of Japanese scripts and their functions
Traditionally, Japanese is considered to consist of three different script systems: kanji, or characters of Chinese-origin, and the two syllabaries hiragana and katakana. This list may also be expanded to four scripts if one considers rōmaji (the Roman alphabet) given its occasional appearance within Japanese sentences, not to mention the small number of kokuji, or characters of Japanese origin that are based on kanji (Shibamoto Smith & Schmidt Reference Shibamoto Smith and Schmidt1997; Seeley Reference Seeley2000). In today's Japanese, each of the scripts may be said to occupy a particular role in the representation of different components of the written language, and as such they regularly co-occur. Example (1) below exemplifies how these scripts may be used in conjunction with one another. In the transliteration, the script that is used to represent that item is indicated in parentheses. (Kn) refers to kanji, (H) to hiragana, and (Kt) to katakana.
Kanji are typically utilized to represent nouns of Japanese or Chinese origin, as well as the roots of verbs and adjectives; this may be seen in (1) above in which the words tegami ‘letter’ and the root of kaite (ka(k)- ‘write’) are both written with kanji. Hiragana, by contrast, is used for the representation of grammatical units such as inflectional morphemes (the stem of kaite, -ite), case particles (object marker o, instrumental marker de), and other units like the deictic pronoun kono ‘this’. Katakana most commonly represents loanwords from languages other than Chinese such as pen ‘pen’, which is of English origin (Konno Reference Konno2014a).
Previous studies that take a sociolinguistic approach to script choice have revealed a number of different possible sociocultural associations with each script. Shibamoto Smith & Schmidt (Reference Shibamoto Smith and Schmidt1997) discuss how when viewed broadly from the perspective of the popular imagination, kanji are more associated with male, middle-aged and older writers/readers, and their use may index ‘erudition’, particularly in situations where they are used densely within a given text. Hiragana, by contrast, may evoke childishness or cuteness due to its status as the primary script of children's works (Miyake Reference Miyake2007). At the same time, because of this usage it may also be associated with young female writers/readers as an index of ‘softness’ or ‘femininity’ (Shibamoto Smith & Schmidt Reference Shibamoto Smith and Schmidt1997). In contrast to hiragana’s association with a kind of soft femininity, katakana may be associated with young male writers/readers as an index of ‘modernity’ or ‘popular culture’, or even a sense of ‘foreignness’ as a whole due to its heavy usage as the script of recent loanwords (Tomoda Reference Tomoda2009). Finally, rōmaji, much like hiragana, may be associated with young female writers/readers, but rather than softness, connotes a kind of ‘commerciality’ (K. Nakamura Reference Nakamura1983; Satake Reference Satake and Mizutani1989a).
Given the number of scripts at one's disposal, it is crucial to remember that the script used to represent a given linguistic unit is not absolute; rather, it is more accurate to think of script variation as a matter of ‘uncommon’ versus ‘typical’ rather than necessarily correct or incorrect (Konno Reference Konno2014b; Robertson Reference Robertson2017). In other words, it is possible to represent the same denotational content of a given unit while altering its connotational meanings. For example, there exists a commonly used organizational hashtag on the website Twitter that states that writing something in hiragana rather than its typical script yields a cuter result (#hiraganadekakutokawaikumieru, which translates as ‘#when you write things in hiragana they look cute’). An apt, albeit tongue-in-cheek example that compares the three scripts also originates on Twitter. With regard to different representations of the word chintai ‘rent; lease’, an anonymous Twitter user writes: “The fact that chintai (賃貸) seems formal, chintai (ちんたい) seems lewd, and chintai (チンタイ) seems like some kind of Chinese cuisine is one of the amazing charms of Japanese” (Hugki11 2012). With this example, it is even possible to see different interpretations of both hiragana and katakana that deviate from the broader connotations of these scripts as described in previous studies. Despite hiragana’s associations with ‘cuteness’, for example, in the case of certain lexical items (e.g. chintai above) it is possible to see that this ‘cute’ image does not necessarily prevail. Instead, it is the script that is hypothetically associated more with ‘modernity’ or ‘popular culture’ that evokes the image of a Chinese dish, not the script that finds its origin in China.
Script variation and the paratext
While the majority of lexical items may be represented hypothetically in any one of the scripts discussed, script choice and variation are especially salient with regard to the representation of personal pronouns. Though such linguistic units may be homophonous, the high degree of variation in stylistic representation allows for a great deal of creativity on the part of the author, allowing writers to incorporate a level of intertextuality that exists at a strictly visual level. For example, in Robertson's (Reference Robertson2017) corpus analysis of a popular comic series and subsequent interview with the author, it was found that with stereotypically ‘masculine’ first-person pronouns such as ore and boku, katakana was most often used when the speaker was a teenager, whereas the kanji equivalent was restricted to adult speakers. With the more normatively gender-neutral pronoun watashi, the author chose hiragana as the script of choice when the speaker was female, while its kanji counterpart was utilized to indicate formal, more careful speech regardless of gender (Robertson Reference Robertson2017:514). This is not to say that this exact kind of variation and awareness extends to every author, but rather, Robertson's (Reference Robertson2017) data illustrates the kind of language ideologies at work with the representation of fictionalized speech in the area of personal pronouns.
As evidenced by previous studies as well as casual Twitter commentary on Japanese script indexicality, it is apparent that there are numerous, even contradictory, meanings associated with a particular script. Such a multivalent nature with regard to sociocultural meaning indicates that it is perhaps analytically inappropriate to ascribe a single, overarching meaning to the usage of a given script. Instead, the variety of possible meanings suggests that a given script exists as part of an indexical field, by which one of many meanings may be activated with the situated use of that script (Eckert Reference Eckert2008). In other words, opting for the use of hiragana for a single linguistic unit does not necessarily index ‘childishness’ or ‘femininity’ on its own, but widespread hiragana use in accompaniment with other linguistic features, such as opting for certain lexical items or syntactic structures, can create such a ‘voice’. Given Robertson's (Reference Robertson2017) study of the role of authorial ideology in the use of a particular script, it is critical to emphasize the importance of what Robertson refers to as the local norm or the typical orthographic representation in a given context. Adapted from Ukita, Sugishima, Minagawa, Inoue, & Kashū (Reference Ukita, Sugishima, Minagawa, Inoue and Kashū1996), Robertson uses this approach to account for the fact that while popular conceptions exist about the indirect indexical associations of certain scripts, not every author may adhere to these and, indeed, authors may exhibit their own personal script utilization tendencies in their work. By approaching script variation at a local level, rather than as a broader, language-wide phenomenon, it is possible to pursue a more precise understanding of a script in the context of the community that finds it meaningful.
This study expands on Robertson's (Reference Robertson2017) approach by analyzing the local norm of script variation not at the level of the author, but rather at the level of a magazine publication as a whole. More specifically, this study analyzes paratext, as explained in the introduction. Recall that in paratextual areas, the author of that text can be said to be speaking more directly to a reader compared to text in other locations. However, in the case of a magazine publication, the author of that text is not necessarily the author of any specific work in the magazine, but is instead a group of individuals tasked with writing the copy that occurs throughout the magazine. As a point of departure, this study approaches script variation in this paratext through the lens of audience design theory (Bell Reference Bell1984, Reference Bell, Eckert and Rickford2001), which is a framework that engages with the idea that speakers alter their linguistic style relative to their audience. While outside the scope of Bell's original study, this concept finds theoretical power in the context of mass media discourse, whereby the language of a given work ‘indexes an imagined target audience on the assumption that this audience will find this particular style acceptable and attractive within genre constraints’ (Androutsopolous Reference Androutsopoulos2012:304). To put it differently, this study is interested in the usage of script variation as macrocosmic communication (Yamaguchi Reference Yamaguchi and Kinsui2007). Yamaguchi (Reference Yamaguchi and Kinsui2007, Reference Yamaguchi and Kinsui2011) defines its counterpart, microcosmic communication, as the communication that occurs purely between characters for the purposes of advancing the plot, while macrocosmic communication functions as a kind of linguistic stereotype meant to be referenced intertextually between author and imagined audience. By using such frameworks to analyze paratextual occurrences of script variation, this study engages with the way that a publication itself (or rather, that publication's various paratext authors) communicate with and navigate the expectations of their imagined target audience.
Japanese second-person pronouns
With audience-dependent variation in the paratext in mind, as mentioned earlier this study engages with script choice as it relates to second-person pronouns rather than first-person pronouns as in the case of Robertson (Reference Robertson2017). The reason for this focus relates to the paratext's previously mentioned function as a means by which a given author may communicate directly with the reader outside the main text, a circumstance that makes the use of second-person pronouns relatively common, unlike other kinds of Japanese discourse where the use of any personal pronoun is inherently marked. In fictionalized or other mass media discourse, the increased presence of second-person pronouns is likely due to the inherent ‘facelessness’ of a text's addressee, making address by typical means (e.g. one's name, title, etc.) impossible. For example, in the case of female-targeted heterosexually romantic texts, Shibamoto Smith (Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005:121) found that there was a much higher frequency of second-person pronouns than in naturally occurring speech among all characters, but especially among the male love interests. This markedness combined with the degree of stylistic variation possible in its representation make this linguistic feature a prime candidate for analysis within the framework of audience design.
As with first-person pronouns, Japanese has no shortage of potential second-person pronouns at its disposal. Previous literature has traditionally categorized them along the lines of (i) the gender of the speaker using the pronoun and (ii) that pronoun's relative degree of formality with regard to the speaker's gender. According to Shibamoto Smith (Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005), the normative utilization of the most common of these pronouns may be understood as follows.
Of the pronouns Shibamoto Smith (Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005) highlights in Table 1, those that may be expected to appear in magazine paratext are anata and kimi. The reason for this is not only the fact that they are generally the most ‘formal’ based on Shibamoto Smith's (Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005) descriptions, but also because they are each the least marked relative to the others. Of these two, anata carries the least sociocultural indexical baggage, a fact that is reflected in its relatively early appearance in Japanese language instructional materials (e.g. Banno, Ikeda, Ohno, Shinagawa, & Tokashiki Reference Banno, Ikeda, Ohno, Shinagawa and Tokashiki2011; Hatasa, Hatasa, & Makino Reference Hatasa, Hatasa; and Makino2015). This low degree of markedness does not extend to its less formal variant anta, however. This form is especially associated with lower to lower-middle class speakers from the major urban areas of Tokyo or Osaka, and depending on its context of use, can range in characterological association from simply being a more ‘intimate’, less ‘classy’ form of anata, particularly when used by female speakers (Takahara Reference Takahara1992; Shibamoto Smith Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005), to a middle-aged barfly (Fujimoto Reference Fujimoto and Kinsui2014). Kimi, by contrast, is traditionally and ideologically classified as a second-person pronoun typically used by male speakers. Takahara (Reference Takahara1992) classifies this pronoun as the ‘familiar’ male counterpart to anata as it is used by female speakers (compared to, for example, the ‘intimate’ male counterpart omae). However, kimi has a history of use among female speakers as well, especially as a subversive alternate to anata, which has often been imposed top-down by authority figures seeking to further reinforce hegemonic gender ideologies (Inoue Reference Inoue2004, Reference Inoue2006; Endō Reference Endō2006; M. Nakamura Reference Nakamura2014). Additionally, M. Nakamura (Reference Nakamura2007) notes that despite their ideological association with male speakers, both kimi and the first-person pronoun boku have recently come to take on a more gender-inclusive usage in popular music. The remaining pronouns illustrated in Table 1, omae, kisama, and temee, carry with them other-deprecatory connotations, particularly in the case of the latter two, making them unlikely candidates as a means of addressing a readership either directly or indirectly.
The normative usages of Japanese second-person pronouns relative to gender and formality (Shibamoto Smith Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005).
Comic Yuri Hime and its genealogy
The data under analysis in this study are from the magazine Comic Yuri Hime and its short-lived sister publication, Comic Yuri Hime S. For most of its existence, the flagship magazine Comic Yuri Hime has been the only mainstream publication owned by a major publisher that is dedicated specifically to the yuri genre. This position makes Comic Yuri Hime a prime resource for yuri-oriented research, and the existence of a sister publication with a different imagined target audience makes these magazines valuable for inquiries into stylistic practices as they interact with audience design theory in popular media. In order to analyze script variation within the paratext of yuri publications such as Comic Yuri Hime, however, it is first necessary to understand some of the history of yuri as a whole to give historical context to present day stylistic tendencies.
Despite current interpretations of yuri as a genre that concerns same-sex relationships between female characters, such a definition has not always been applicable, and indeed, it does not convey the trajectory by which yuri became what it is today. According to Nagaike (Reference Nagaike2010), the genesis of a so-called ‘pure’, or unified yuri genre came about in 2003 with the publication of the magazine Yuri Shimai (lit. ‘lily sisters’), the predecessor to Comic Yuri Hime. The term itself, however, was introduced to the popular culture sphere as a way to refer to queer female sexuality in 1976 by Itō Bungaku, the editor of the gay men's magazine Barazoku (lit. ‘rose tribe’), who utilized it to delineate a space within the magazine for the minority female readership (Welker Reference Welker, Martin and Jackson2008). Fluctuations in what ‘counts’ as yuri media have been inevitable since its introduction, including appearances in the titles of male-targeted pornography in the 1980s, but the flexibility of the term has decreased in recent years due to the continuous publication of Comic Yuri Hime (Kawasaki Reference Kawasaki2014).
The genre conventions typical of yuri can be traced back to several places, in particular mainstream shōjo ‘girl’ comics, but in terms of linguistic convention, the origin that has had the greatest influence is twentieth century ‘S’ culture (Nagaike Reference Nagaike2010). ‘S’, which stands for the English word ‘sister’, is a term that was first used as early as 1926 to describe a strong homosocial bond or romantic friendship between girls. That being said, according to Hiruma (Reference Hiruma2003), the so-called ‘discovery’ of homosexuality among women can be dated back to the high-profile double-suicide of two girls in 1911 who were said to be in an ‘S’ relationship, illustrating that these relationships were not necessarily platonic. ‘S’ relationships are portrayed in several yuri predecessor works, such as Yoshiya Nobuko's Hana monogatari ‘Flower tales’, a serialized novel that appeared in the girls’ magazine Shōjo gahō ‘Girls’ illustrated magazine’ from 1917–1924 and depicted emotional and often overtly sexual bonding between girls at an all-girls’ school. A slightly later work is Otome no minato ‘Maiden's harbor’, originally attributed to Kawabata Yasunari but found in 1989 to have been originally written by his assistant Nakazato Tsuneko (Maser Reference Maser2015:42). More recent is the work Maria-sama ga miteru ‘The Virgin Mary is watching’ by Konno Oyuki, which was published across thirty-two volumes between Reference Konno1998 and 2012. While none of these works were considered yuri in name at the time of writing, what they each have in common is their setting of an affluent all-girls’ school, a motif that extends into the works of Comic Yuri Hime where the majority of school-aged characters also attend such institutions.
Comic Yuri Hime has been in publication since 2005. It is written by a majority female authorship for a primarily female readership, a fact that is reflected in its slogan on every issue until 2011: danshi kinsei ‘no boys allowed’ followed by two exclamation points as illustrated below in example (2).
(2) 男子禁制!!:少女と少女のヒミツの純愛コミック
Danshi kinsei!!: Shōjo to shōjo no himitsu no jun'ai komikku
‘No boys allowed!!: A comic about the secret, pure love between girls’
Though it initially began as a quarterly magazine, it became a bimonthly publication in 2011, and eventually monthly in 2017. In 2008, a survey conducted by the magazine requested that readers volunteer information about their gender and age through mail-in postcards. While this form of survey is certainly not the most reliable, it still provides an estimation about who is reading, or at least purchasing, the magazine. In 2008, Comic Yuri Hime had a readership that was 73% female and 27% male. In terms of age, 27% of readers were under nineteen, 27% between twenty and twenty-four, 23% between twenty-five and twenty-nine, and the remaining 23% was above thirty (Sugino Reference Sugino2008). This suggests that Comic Yuri Hime, at least at the time of the survey, was consumed primarily by women who were over twenty years of age.
In an effort to appeal to their minority male readership, the editorial board for Comic Yuri Hime decided in 2007 to also publish Comic Yuri Hime S. Its male-targeted nature is reflected in the slogan on the front, danshi kinsei ‘no boys allowed’ followed by an exclamation mark and a question mark, making a direct reference to Comic Yuri Hime’s slogan and calling its validity into question.
(3) 男子禁制!?:君のハートを狙い撃ちするCUTE x LOVELYコミック誌
Danshi kinsei!?: Kimi no hāto o neraiuchi suru CUTE x LOVELY komikkushi
‘No boys allowed!?: A CUTE x LOVELY comic magazine that snipes your heart!’
The magazine ran seasonally from 2007 through 2010, merging once more with Comic Yuri Hime in 2011. While age data is no longer available for this magazine, the self-reported gender make-up was 38% female and 62% male (Sugino Reference Sugino2008).
Because the focus of this study concerns the way that each magazine addresses its readership, this analysis includes 2005 and 2006, when only Comic Yuri Hime existed, as well as 2011, when the magazines merged and attempted to rebrand for a broader audience. Part of this rebranding was changing the slogan of the magazine from a variation on ‘no boys allowed’ to ‘justice for girls’. Additionally, in the second volume of the new magazine, Comic Yuri Hime began featuring the serialized manga Yuri Danshi ‘Yuri boy’, a story with a male protagonist who is learning to interact with the mainstream yuri fandom, which he perceives as being dominated by women (Welker Reference Welker2014).
From its inception through 2011, Comic Yuri Hime and its sister magazine underwent changes as its editorial board attempted to market the magazine in a manner that would not only attract a greater readership, but increase the visibility of the yuri genre as a whole (S. Nakamura Reference Nakamura2011). Over the course of this period, the editorial board that managed both stayed largely unchanged, and the head editor, Nakamura Seitarō, has held his position since Comic Yuri Hime’s inception in 2005. Additionally, a number of authors published works in both magazines. The fact that the editorial board attempted to broaden the scope of their intended readership, combined with the fact that the board remained stable during this period, suggests that any linguistic differences apparent in the magazines would be due to ideological changes within the board, rather than caused by any in- or out-flux of key personnel. Therefore, because it is unlikely that any variation in the use of second-person pronouns and their scripts could have been the result of writing staff instability, it is possible to consider variation across these magazines from the perspective of ideological shifts on the part of the editors.
Methods
The collected data consists of paratextual usages of second-person pronouns observed between the years 2005 and 2011. Examples of paratext that incorporate second-person pronouns are like that in example (4) below, which is the introduction to Miura Shiwon's Yuri no kafun wa ochinikui ‘The pollen of lilies is not easy-to-fall’, a column that analyzes popular culture works from a yuri perspective.
(4) 抱腹絶倒エッセイでおなじみ・三浦しをん氏が貴女へ送る、名作百合漫画の世界!!
Hōfukuzettō essei de onajimi Miura Shiwon ga [female-coded] anata e okuru, meisaku yuri manga no sekai!!
‘Through her hilarious essays, Miura Shiwon sends you the world of famous yuri manga!!’
Another is example (5) below, a call for amateur manga submissions from readers of the magazine.
(5) 憧れの先生に君の原稿を見てもらおう
Akogare no sensei ni [kanji-written] kimi no genkō o mite moraou
‘Have an artist you admire look at your manuscript!’
Both of these examples, which address the reader directly, may be considered examples of direct address. In addition to such instances, I also include second-person pronouns that appear in sections of indirect address. This kind of text falls into the category of what may be referred to as flavor text. The term flavor text, while most often found in the context of role-playing games or collectible card games, refers to snippets of text that are unrelated to the actual rules or gameplay and are present only for ‘atmosphere’. Those that appear in manga magazines serve a similar purpose. Appearing predominantly on the opening and closing pages of a given chapter of a work, these are phrases that are not quotes from the story, but simply add to the ‘atmosphere’ of that work. Instances of indirect paratext in manga can be identified by a lack of reference to events or characters in the story, and visually they are often presented in a different typeset or font from the dialogue or narration that appears in the story. To illustrate indirect paratext in practice, example (6) below comes from the fourth chapter of the manga Ren'ai joshika ‘Women's love department’ by Morimoto Akiko, which ran in volume 21 of Comic Yuri Hime. The title page of this chapter depicts a woman in a wedding dress with another woman resting her head on that woman's knees. In the magazine version, this page includes a variety of different text, including the title of the chapter, the name of the author, a small advertisement for the upcoming release of the first volume of the trade paperback, and the phrase in example (6) below.
(6) 貴女に捧げる永久の誓いを…
[Female-coded] anata ni sasageru eikyū no chikai o…
‘The eternal vow that I offer to you…’
The first volume of the trade paperback for Ren'ai joshika is a compilation of the first four chapters sold as a single unit. Each of the chapters’ title pages are included in this volume, but with the exception of the titles themselves, the text on the first page of each chapter is gone. This is common practice not only with yuri, but across manga genres. The fact that the text is present in the magazine version, but not in the trade paperback, indicates that that text is specifically for consumption by the magazine audience rather than necessarily part of that work. In terms of content, example (6) is a phrase that is never uttered in Ren'ai joshika, and the second-person pronoun anata does not actually appear anywhere in this work's ten chapters. Rather, this phrase is directly related to the image upon which it occurs. This anata, which the portrayed characters do not use in reference to one another, is an indirect way for magazine writers to invite the nameless, imagined readers to place themselves within the context or ‘mood’ of the story.
Flavor text in manga commonly makes use of second-person pronouns despite the fact that these pronouns may see only limited use in the work itself. Similar to the presence of anata on the title page for Ren'ai joshika in example (6) above, second-person pronouns that appear in flavor text may be completely unrelated to those that appear in the work, a tendency that was readily observable in both Comic Yuri Hime and Comic Yuri Hime S. This lack of verisimilitude between the work itself and the flavor text on its title pages is further indication of their separability, justifying the analysis of the flavor text as a property of the magazine itself and not any individual work.
For analysis, I divided the works into four categories: ‘O’, ‘F’, ‘M’, and ‘N’. ‘O’ refers to the original version of Comic Yuri Hime than ran from 2005 through half of 2007 before the magazines split. ‘F’ is the female-targeted magazine that coexisted alongside male-targeted version ‘M’ from the latter half of 2007 through 2010. Finally, ‘N’ is the renewed version of the magazine than began in 2011 after the two merged. These aggregation categories may be seen below in Table 2 along with the total number of volumes and pages in each category.
Description of analysis categories and number of volumes and pages in each category.
Comparison of the description for author Miura Shiwon's column Yuri no kafun wa ochinikui before and after 2011.
Figure 1 below is a visualization of the variety and proportion of second-person pronouns that appeared in the magazines. The data has been collapsed across years into their relative categories of analysis.
Proportion of second-person pronoun usage relative to other appearing second-person pronouns.Footnote 1
Two different pronouns appeared in Comic Yuri Hime and Comic Yuri Hime S with six different script variations: female-coded anata (貴女), neutral-coded anata (貴方), hiragana-written anata (あなた), katakana-written anata (アナタ), kanji-written kimi (君), and katakana-written kimi (キミ). The two most prominent are female-coded anata and hiragana-written anata. Katakana-written anata also occupied a large proportion of pronoun use, but it disappears from the data in 2011. There is also a notable difference in the usage of the pronoun kimi in male-targeted categories M and N compared to the female-targeted categories O and F. However, this study focuses analysis on female-coded anata (貴女) and its hiragana counterpart (あなた), as they make up between 60% and 85% of second-person pronouns used across all issues.
In written Japanese, hiragana-written anata can be understood to be the most standard orthographic representation of this pronoun. To compare the proportion of female-coded anata with hiragana-written anata, the general proportion of anata’s different orthographies were analyzed in the Balanced Corpus of Current Written Japanese (BCCWJ) from the National Institute of Japanese Language and Linguistics (NINJAL). This corpus consists of various written materials from 1970 to 2005 sourced from a large variety of genres, including novels, message board posts, and descriptions of laws, allowing for an estimation of the degree of use of a given linguistic feature across general written media.
As is observable in Figure 2, hiragana-written anata is by far the most common variant across written material, occupying well over 90% of all representations of anata. The female-coded one, however, occupies a little under 2% and, as such, it is possible to see that the proportion of female-coded anata is much higher in Comic Yuri Hime than in general written Japanese as viewed through the BCCWJ. Female-coded anata, which has the character for ‘woman’ as the second-character, has a history of use in yuri due to its association with the hyper-feminine fictionalized register o-jōsama kotoba ‘young lady-speak’, making its high degree of presence in Figure 1 unsurprising. This register is commonly used in narratives that take place in affluent all-girls’ schools, and because of the frequent use of all-girls’ schools as a setting in yuri narratives, there exists an additional association between female-coded anata itself and female romantic pairings more broadly. This association can be seen not only in the titles of transformative works inspired by Maria-sama ga miteru, the serialized novel mentioned earlier (example (7)), but transformative works that feature a noncanonical romantic female pairing from non-yuri works (example (8)). Example (7) is the title of a work by Pawafuru Kokoa (Reference Kokoa2006), and example (8) is the title of a work written by The Earth ~Kono taichi o fumishimete~ (2016) for the Love live animation series fandom.
(7) ずっと貴女が好きだった
Zutto [female-coded] anata ga suki datta.
‘I always loved you’
(8) 貴女が隣にいるだけで
[Female-coded] anata ga tonari ni iru dake de
‘Only with you next to me’
As illustrated in Figure 1, the proportion of female-coded anata relative to hiragana-written anata decreases with Comic Yuri Hime S, and it is even lower in the newly merged Comic Yuri Hime in 2011. This decrease suggests that as the magazine aimed to expand its appeal to a broader, less overtly gendered audience, the presence of female-coded anata decreases.
Proportion of different script representations of anata in the BCCWJ (NINJAL) compared to the overall proportion of different script representations of anata in Comic Yuri Hime.
Results
In addition to the proportion of the featured pronouns in relation to one another, the density of their use is also a key point of analysis in this study. Due to the locations in which paratextual second-person pronouns appear, namely in advertisements, on the magazine covers, and so forth, the density was calculated by analyzing the number of occurrences relative to the number of pages in a given volume. To calculate the proportion, the number of occurrences of a given pronoun was divided by the number of pages within that volume. The data was then normalized for statistical calculations by using the inverse log of the density of pronoun appearance relative to magazine type. A one-way Analysis of Variance Test (ANOVA) was used to determine whether or not there exists a statistically significant (α = 0.05) relationship between the density of pronoun appearance and magazine type.
The densities of female-coded anata and hiragana-written anata relative to magazine type are illustrated in Figure 3. Based on the results of the one-way ANOVA, there is a statistically significant relationship between density of female-coded anata and magazine type ([F(3,34) = 13.58, p < 0.001], R2 = 0.5308). The highest density of female-coded anata is in category O, the editions of Comic Yuri Hime between 2005 and the middle of 2007. In category F, that density decreases, while category S has an even lower density of usage. Finally, the lowest density is seen in Comic Yuri Hime of 2011 after the two magazines merged. In comparison, the density of hiragana-written anata does not change across magazine type. This is confirmed by the lack of statistical significance between type and density of hiragana-written anata.
Density of female-coded anata and hiragana-written anata by number of pages in volume according to magazine type.
Qualitatively, this decrease in usage of female-coded anata can be seen in several places in the magazine. Most visibly, there is a change in description of the column Yuri no kafun wa ochinikui ‘The pollen of lilies is not easy-to-fall’ written by the author Miura Shiwon in the transition from pre-2011 to post-2011 Comic Yuri Hime (Table 3). The column, which only ran in Comic Yuri Hime, was preceded in each issue by a passage describing the general purpose of the column, and this description only rarely changed between issues. Prior to 2011, the anata that appeared in this description was the female-coded one. After 2011, however, that anata was changed to hiragana while the rest of the description remained the same.
While Miura's column represents one of the more visible changes, the shift is also evident in the situations in which anata was used. In both categories O and F, female-coded, hiragana-written, and katakana-written anata were each regularly used in the flavor text before and after comics, in advertisements for other works, and when addressing the reader. In category M, female-coded anata was near nonexistent as a means of directly addressing the reader, but it still appeared in flavor text and advertisements. But, beginning with Type N, female-coded anata nearly disappears; thereafter, its primary appearance is restricted to flavor text, resulting in a comparatively lower density compared to earlier years of the magazine. In its place, hiragana-written anata becomes the preferred script variant.
Discussion
Based on the proportion of second-person pronoun appearances shown in Figure 1 and the results of the one-way ANOVA in Figure 3, it is possible to see that the degree of utilization of female-coded anata in Comic Yuri Hime decreases over time. That being said, given that female-coded anata is homophonous with its other orthographic representations, it is not strictly necessary to use it at any point. Certainly, in magazines belonging to categories O and F, the readership is expected by the editorial board to be primarily female, but it is neither grammatically nor pragmatically necessary to choose between the female-coded anata and the hiragana-written anata to address such an audience. Rather, to use the term posited by Yamaguchi (Reference Yamaguchi and Kinsui2007), the female-coded anata functions as a kind of macrocosmic communication device. Though Yamaguchi's (Reference Yamaguchi and Kinsui2007, Reference Yamaguchi and Kinsui2011) use of this concept refers to character-spoken dialogue within a given narrative, in the case of paratextual linguistic phenomena like those analyzed in this study, there are no specific characters that speak these words. Rather, the conversation is between a faceless speaker on the part of the magazine and the reader who is consuming its content. Through the situated use of female-coded anata, the magazine invokes the use of that same anata in ‘S’ narratives, the primary genre that contributed to modern yuri. Because of its history of use before Comic Yuri Hime, linguistic features like female-coded anata became part of yuri outside of only ‘S’ narratives through the process of narrative accrual, whereby the repetition of certain linguistic features in ‘S’ narratives have resulted in the generation of genre conventions more broadly (Bruner Reference Bruner1991:18). In the case of Comic Yuri Hime, it is possible to see how the adoption of a linguistic feature common to ‘S’ narratives for use in the magazine paratext (i.e. female-coded anata) allowed the magazine's writers to construct a visual relationship between ‘S’ narratives and Comic Yuri Hime. Indeed, at the time of Comic Yuri Hime’s inception, Konno Oyuki's light novel series Maria-sama ga miteru ‘The Virgin Mary is watching’ (1998–2012) was enjoying immense popularity. According to Nagaike (Reference Nagaike2010), it was the popularity of a co-occurring ‘S’ narrative like Maria-sama ga miteru that allowed Comic Yuri Hime to survive at all.
These findings are supported by comments made by Comic Yuri Hime’s editor in chief Nakamura Seitarō in an interview with the monthly magazine Cyzo. He stated that the success of yuri of a genre (i.e. whether or not there will be a ‘yuri boom’) depended on the ability to create another Maria-sama ga miteru (Yasuda Reference Yasuda2008). Though Nakamura was likely referencing the story and setting of Maria-sama ga miteru with his comment, by using female-coded anata, the editors were able to market themselves to the same consumers, who were largely female. In this way, this anata served the dual purpose of explicitly addressing a gendered readership while simultaneously referencing a popular work outside of Comic Yuri Hime that belonged to a genre whose readership the editors were looking to incorporate into their magazine's community. As such, through its usage, female-coded anata was both a way of speaking to a community of fans as well as a kind of intertextual marketing tool; in other words, the magazine presented itself in a way that the editors hoped their target audience would find meaningful and desirable. In the same interview with Cyzo, Nakamura makes a remark that hints at the magazine's eventual decreased utilization of female-coded anata. In addition to stating the need for a work with the popularity of Maria-sama ga miteru, Nakamura attributes that work's success not to the female audience, but to its popularity among male readers who, according to him, comprised about 30% of the total consumer base (Yasuda Reference Yasuda2008). It is not hard to imagine, then, that in order to further appeal to a male demographic, Comic Yuri Hime would include fewer female-coded anata in direct paratext so as to avoid alienating male readers and to pursue a kind of inclusivity that is more easily attained with hiragana-written anata. It is important to remember, however, that female-coded anata stayed present in the flavor text, maintaining its presence as a semiotically meaningful genre convention without directly addressing the audience.
When considering these writing systems more broadly, the necessity of an approach that places at the forefront the idea of the local norm is apparent. We saw earlier that kanji is generally associated with a sense of ‘erudition’ or ‘formality’ and its characterological association with middle-aged male speakers (Shibamoto Smith & Schmidt Reference Shibamoto Smith and Schmidt1997). Hiragana, by contrast, in many ways have associations that are completely at odds with kanji, including ‘childishness’ or ‘cuteness’ (Miyake Reference Miyake2007) as well as ‘softness’ or ‘femininity’ (Shibamoto Smith & Schmidt Reference Shibamoto Smith and Schmidt1997). These associations do not translate to the script variation observed in the paratext of Comic Yuri Hime. Of the available variants of anata, for example, it is the female-coded anata in kanji that may be considered the most strongly associated with ‘femininity’, though to categorize it as strictly ‘feminine’ would do a disservice to its utilization as a genre convention, and such an interpretation ignores any intersection it has with categories of class and education that are brought about through the common occurrence of these narratives in affluent all-girls’ schools. The hiragana-written anata, however, is the one that promotes a more gender-inclusive interpretation by omitting kanji altogether, falling back on the unmarked representation of that pronoun that is observable in Figure 2.
On the subject of the pronouns themselves, there also exists a degree of difference in their utilization as it is reflected in Comic Yuri Hime compared to conventional views of their indexical associations. According to Shibamoto Smith (Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005) in Table 1, while anata may take a gender-neutral reading, this usage is relatively ‘formal’. Anata as it is used in mass media paratext, however, seemingly has little to do with the issue of ‘formality’ versus ‘informality’, as the accompanying sentences in Comic Yuri Hime portray an otherwise friendly speaking voice. Because of anata’s prevalence not only in flavor text, but also as a means by which the magazine directly addresses the readership, it is possible to see the use of anata as an unmarked means by which a variously gendered audience may be addressed. Moreover, this usage is also evident in the low use of the pronoun kimi in the data. Though it occupied less than 20% of all paratextual second-person pronouns, what occurrences there were did not have an overtly male-targeted context. Indeed, they were present in both flavor text and direct address, though in the case of the latter, the instances of direct address seemed to occur only in advertisements sponsored not by the magazine itself but by the publisher as a whole. The use of kimi at all in this data set, particularly in instances where the audience was imagined to be female, is a reflection of what M. Nakamura (Reference Nakamura2007) observes as an emergent gender-neutral reading for kimi (and its associated first-person pronoun boku) in mass media discourse, despite the traditional categorization of these pronouns as ones for exclusively male usage. To document this emergent neutrality with regard to gender, further research needs to be carried out on kimi’s place in the Japanese mediascape.
Conclusion
By placing paratextual linguistic phenomena as the subject of analysis, rather than only those that occur within the marketed texts themselves, it is possible to see the more direct ways that media genres communicate with their consumers. In the case of Japanese, one of these communicational tools is the script in which information is presented to the consumer. Previous research has discussed the indexical associations that these scripts may have more broadly, but more recent academic works (e.g. Robertson Reference Robertson2017, Reference Robertson2018) have revealed the importance of the local norm (Ukita et al. Reference Ukita, Sugishima, Minagawa, Inoue and Kashū1996) of script utilization within a given context. With this idea of the local norm at the forefront, this study approached the means by which a magazine represented second-person pronouns in its paratext, how these representations changed with the imagined audience of that magazine, and how both these representations and the pronouns themselves compared with traditional conceptions about their usage.
In the case of early Comic Yuri Hime, the predominant paratextual second-person pronoun anata was represented most commonly as its female-coded variant (貴女). This variant has a history of use in yuri due to the genre's common utilization of affluent all-girls’ schools as a setting, a setting that often entails the use of the hyper-feminine fictionalized register o-jōsama kotoba ‘young lady-speak’. In addition to its overall history of use within the genre, female-coded anata is also represented with kanji, a script that in Japanese is commonly described has having more ‘erudite’ or ‘formal’ associations with a prototypical user being a ‘middle-aged man’ (e.g. Shibamoto Smith & Schmidt Reference Shibamoto Smith and Schmidt1997). This is an association that does not hold up in the case of female-coded anata, however, a variant that has more indexical ties to young, wealthy women than anything else. While it is perhaps possible to extend that sense of ‘erudition’ to the students of these prestigious institutions, the gender- and age-oriented association of kanji more broadly is brought into question with female-coded anata. Indeed, despite a linguistic reality in which female speakers were historically discouraged from utilizing kanji or Chinese-origin loanwords (Endō Reference Endō2006), this usage nonetheless finds itself as index of a certain kind of female speaker, not necessarily as one of femininity or learnedness, but perhaps of the intersection of both of these categories with class and class-aspirations more broadly.
In their efforts to expand the reach of Comic Yuri Hime to a broader, more gender-inclusive audience, the female-coded variant of anata appeared less in paratext that directly addressed the reader relative to its hiragana-written counterpart (あなた) with the publication of male-targeted sister magazine Comic Yuri Hime S and the renewed Comic Yuri Hime in 2011. Generally speaking, anata is traditionally categorized as a gender-neutral pronoun in formal situations, but normatively associated with female speakers more than male in informal ones (Shibamoto Smith Reference Shibamoto Smith, Okamoto and Shibamoto Smith2005). Given this pronoun's utilization as the primary means of address in a gender-inclusive sense, especially in direct and indirect examples of otherwise informal address, there exists a need to re-evaluate the categorization of this pronoun, not necessarily on a scale of formality, but on one of markedness. Taking its representation into consideration, the increasingly lower proportion of female-coded anata relative to hiragana-written anata reveals that rather than hiragana’s necessarily ‘feminine’ or ‘childish’ associations, the invocation of what is the most common representation of anata removes the gender-related indexical baggage that accompanies female-coded anata, allowing anata to address a mixed-gender audience.
As the magazine's expectations for its audience changed, so, too, did the choice and representations of second-person pronouns to address that imagined audience. Moreover, these changes occurred with regard to the local norm of Comic Yuri Hime, as well as yuri and ‘S’ narratives as a whole, illustrating the difficulty of applying language-wide generalizations to the use of linguistic features within a given media context. Through consideration of these localized areas of use, as well as how these change over time, it is possible to approach a more nuanced understanding of the way that information is written for and marketed to consumers and communities.
