4.1.1 Survey data

One of the issues often raised in surveys is whether the more complex aspects of honorific usage should be simplified. According to a 2008 Yomiuri Shinbun survey, 61% of respondents to the question “What should the future of honorifics (keigo no shōrai) be?” reported that accurate expressions (seikaku na hyōgen) should be maintained, but 31% thought honorifics should be simpler and easier to understand than they are now. Evidence of such simplification taking place on the ground is seen in survey reports about the domains and genres in which honorifics are still thought to be normatively required. The well-known 1953 and 1972 Okazaki City surveys (Kokuritsu Kokugo Kenkyūjo Reference Kenkyūjo1983), for example, show that the proportion of respondents who agreed that speakers should use honorifics when speaking to older or senior family members decreased substantially in all five age groups (from 50% in 1953 to 20% in 1982). In a 2006 survey in the same city, an even further decrease (to only 6%) was seen (Asahi 2007: 71).

Honorific use in written correspondence may also be undergoing simplification. In the past, personal letters were usually written in a formal style minimally including addressee honorifics even if the addressee was someone with whom one usually spoke in an informal style using plain forms. The change away from this style may, in part, be an effect of new technologies for written communication, such as email. University students in one 2006–2007 study reported using plain forms in emails to their parents, although their parents reportedly responded with addressee honorifics (T. Kumagai Reference Kumagai2007). In fact, meeru buntai ‘email style’, consisting of plain forms and many other informal expressions, is well known (and, by some, greatly lamented) to be common in young people’s messages (H. Satake Reference Satake and Yoshiaki2005). Even in business emails, simpler styles than those used in face-to-face interaction are coming to be preferred, as indicated by the many self-help books that urge honorific simplicity in emails (for example, Mērubun wa keigo mo kanketsu ni ‘In emails, use simpler honorifics’; Shain Kyōiku Kenkyūkai Reference Kenkyūkai2009). The long-term effects of these practices on communicative norms more generally have yet to be explored.

Simplification of honorifics can also be seen in certain aspects of the honorifics themselves, as in, for example, a decline in the reported use of the addressee honorific (de) gozaimasu ‘to be_HUM’’, a more formal variant of desu, except in certain situations, such as service encounters and formal speech events.Footnote ¹ Simpler keigo has been part of government policy since the 1952 Kokugo Shingikai “approved” the use of adjective + desu as a clear and simple expression, and this pattern seems to have taken hold. By 1996, only10% of the respondents to another government survey objected to adjective-desu, while 45% objected to adjective-gozaimasu.

There are other kinds of diversity in what survey respondents react to as (in) correct honorific use. One issue that often comes up is the use of so-called nijū-keigo ‘double honorifics’;Footnote ² another is the use of humble forms. According to government statements, “nijū-keigo is generally considered inappropriate. However, certain words are fixed in custom” (Bunka Shingikai 2007: 30). Examples are offered: o-mie-ni-naru ‘s/he will come_RSP’, including both mieru ‘to come_RSP’ and o- ni naru_RSP, and o-ukagai suru, with both ukagau ‘I will visit [you] _HUM’ and o- suru_HUM kenjōgo I forms. And negative messages also circulate through the popular media such as the self-help books introduced in Chapter 3. For example, in an advice book for women’s speech, Shimodaira (Reference Shimodaira2004: 58) notes that while women are encouraged to be polite and to use keigo, the use of nijū-keigo is not advisable because it doesn’t sound smart. Despite a fair amount of official and popular reluctance to accept nijū-keigo, however, they are widely used and considered acceptable by many people. The National Language Research Institute conducted a survey in Tōkyō (n=306) in 1998 and found that fewer than 20% of respondents objected to sentences that contain nijū-keigo, such as Kōchō-sensei ga osshar_RSPareta_RSP tōri desu ‘It is as the principal said’. Almost 80% of respondents found such sentences to be just fine (Yoshioka Reference Yoshioka2006). The fact that an overwhelming majority of survey respondents report finding nijū-keigo appropriate raises the question of how grammaticality or correctness should be determined. Indeed, this may not be an issue of grammaticality or correctness but rather social appropriateness. And what is deemed appropriate is at the time of writing far from a universally agreed-upon matter.

People also express different views on what to do about “humble” keigo forms. We saw in Chapter 3 that the 2006 Bunka Shingikai’s guidelines for honorifics divided the kenjōgo category of humble forms into two separate categories: kenjōgo I (used for elevating the target of the action by lowering the agent of the action) and kenjōgo II (used for showing politeness toward the addressee). Survey respondents offer widely differing views about humble forms used in the sense of kenjōgo II. The same 1998 survey on honorifics found that opinions divided almost evenly on the appropriateness or inappropriateness of sentences such as X-san, orimashitara go-renraku kudasai. ‘Mr./Ms. X, if you are here, please contact us’ (Yoshioka Reference Yoshioka2006).

Prescriptively, orimashitara is a mistake. Self-help books all treat this as a typical mistaken application of a kenjōgo I form: that is, referring to a higher-status person’s action with a humiliative verb. However, many respondents (42%) found orimashitara appropriate (versus 44% who did not). It is possible that these respondents consider the orimashitara a kenjōgo II form since the referent is also the addressee. But the distinctions between addressee-directed deference through teineigo and addressee-directed deference through kenjōgo II forms has not, to date, been clarified; and the relatively even “split” between approval and disapproval may not be resolved until such clarification is made. That may, however, not be soon. The distinction between kenjōgo I and II seems unclear not only to ordinary speakers, but also to “experts.” The verb mairu ‘go/come’ in a sentence like Sensei no o-taku ni mairimasu ‘I will go to the teacher’s house’ got a thumbs-up in some of the self-help books we examined (for example, Kajiwara Reference Kajiwara2010: 58), but was rejected in others (for example, A. Suzuki Reference Suzuki2010: 214). The linguists have yet fully to weigh in on this issue.

Thus, what is considered correct or incorrect horrific use differs from individual to individual. One of the questions in the 2005 government survey mentioned above asked what kinds of honorifics the respondents think are commonly used incorrectly and gave the following four items to choose from: (a) Sonkeigo, kenjōgo, and teineigo are used incorrectly (chosen by 55% of the respondents); (b) honorifics are not used when they should be (51%); (c) honorifics are used when they are not necessary (36%); and (d) there are many excessive uses of honorifics, such as nijū keigo ‘double honorifics’ (25%) (Bunkachō 2005). These items suggest that the judgments of an honorific as correct or incorrect primarily rests on the following criteria: (a) categories of specific honorific forms; (b) the level of honorifics to be used; (c) the contexts in which honorifics should be used; and (d) how many honorifics should be used in any given stretch of speech (see below for further discussion on these points). Our observations of the survey results indicate that speakers can be expected to differ widely in their assessments of whether any specific honorific use is correct or incorrect. In what follows we examine this point further by examining metapragmatic comments on honorifics from a more popular source.

4.1.2 Blogs

Metapragmatic comments on honorific use are often expressed in letters to the editor in newspapers and magazines (Okamoto Reference Okamoto1999). Another avenue that is increasingly visible is the online blog. Here, we examine threads of posts on the topic of honorifics appearing between 2010–2013 on the blog site Hatsugen Komachi on Yomiuri Online.Footnote ³ The comments expressed in these posts can tell us more specifically than the survey results we examined above about why there is variation in people’s views on honorific use. Furthermore, they also reveal the extent to which the use of honorifics is viewed as a way of actively constructing interpersonal relationships and one’s own social persona, that is, using honorifics as creative indexes to define or change the context (Silverstein Reference Silverstein, Basso and Selby1976).

Our search for the word keigo ‘honorifics’ in posts to Hatsugen Komachi over a one-year period (April 5, 2012–April 5, 2013) yielded forty-one threads. The number of responses to the first post in each thread varied from 3 to 127. In seven (17%) of the forty-one threads, all respondents agreed with the writer of the first post, while in twenty-nine (71%) there were disagreements, which we focus on below.Footnote ⁴ Blog writers from one site in no way represent all segments of Japanese society, but our purpose here is to illustrate how extensive the diversity is across even this limited group of contributors. Following the questions asked in the aforementioned 2005 government survey regarding incorrect use of honorifics, we examined our blog data with regard to the following three issues: (a) the identification of certain forms as honorifics and of the category into which a given form would fall; (b) questions of how many (and what level) of honorifics should be used in a given interaction; and (c) determination of the contexts in which honorifics should be used.

Identification and categorization of honorifics. We start with examples that concern what forms “count” as belonging to the various categories of honorifics. Disagreements over the allocation of particular honorific forms into the taxonomic categories described in Chapter 3 suggest that speakers have different understandings of the functions prescriptively attributed to the various honorific categories. This may be related to the shift away from the old tripartite system for categorizing honorifics, as the negative commentary about the “confusing” nature of the five-part system at the time of its unveiling suggests. In the blogs, the first thread on this topic is about the nominal prefix ‘beautification’ o-. The writer of the first post to one thread wrote:

There were forty-five responses to this post: sixteen of them agreed with the writer, fourteen disagreed, and fifteen said it depends on the inherent qualities of the noun to which it is prefixed or the context. The first writer and those who agreed regarded o- as sonkeigo ‘respect forms’. But some of those who disagreed explicitly treated o- as bikago ‘beautification forms’, as in Example (2):Footnote ⁵

There were also writers who consider the o- in question as teineigo. As noted earlier, the distinction between teineigo and bikago is quite unclear even in the 2007 guidelines, which defines both in terms of politeness and care toward the addressee. Some of those who disagreed with the opinions expressed in the initial post also accorded this prefix other meanings, such as refined (jōhin na) and feminine (onnarashii).

In another thread, focused on the word itadaku ‘receive/eat/drink’, we see a similar range of contradictory responses. The original post, entitled Itadakanai de ~ ~ ~! ‘Please do not itadaku’ is given in Example (3):

This post triggered lively exchanges of agreements and disagreements. There were eighty-five responses (as of July 5, 2012). Approximately 60% of them agreed with the writer, 25% disagreed, and 15% agreed to some degree.

Many writers who agreed with the initial post rationalized their opinions in terms of rejection of the general politeness interpretation of kenjōgo (the new kenjōgo II), as in the following example:

For this writer, self-lowering itadaku should be used when receiving something from someone else; that is, it is kenjōgo I. The use of itadaku in the original post’s examples cannot fit the definition of deference of the other by self-lowering, so comes across instead as an artificial effort to make oneself look refined. Note, however, that this offers the possibility that those who use itadaku in the sense of kenjōgo II may be trying to be polite and refined, but that it strikes this writer as feigned. The writer of the original post also offered her opinion of this use as pretentious.

Many respondents, on the other hand, disagreed with the original writer and explained why, as in the following post:

This writer thus seems to be interpreting itadaku as the generically polite kenjōgo II rather than the self-lowering kenjōgo I.

Itadaku has been traditionally regarded as kenjōgo (that is, kenjōgo I), but, as we see in Example (5), some bloggers’ comments made it clear that they were interpreting it as kenjōgo II. As we saw in Chapter 3, the 2007 guidelines added kenjōgo II as a new category because kenjōgo I forms such as mairu ‘go/come’ and mōsu ‘say’ had, over the years, come to be used to index the interactional context (including some type of “consideration” for the addressee or audience) as one of gentility or cultured politeness (kikite ni hairyo o shimesu) (Kokuritsu Kokugo Kenkyūjo Reference Kenkyūjo2007: 90). But this definition of kenjōgo II overlaps significantly with the definition of teineigo.Footnote ⁶ Accordingly, the perception of itadaku as either teineigo or kenjōgo II is understandable. In fact, treating itadaku as a kind of bikago ‘beautification honorifics’ is not unreasonable, either, as the guidelines note that bikago tend to be used when encoding ‘consideration’ for (rather than encoding deference to) the addressee (Kokuritsu Kokugo Kenkyūjo Reference Kenkyūjo2007: 111).

What level or frequency of honorifics is best? The next point of contention revolves around what level of honorific to use and what proportion of honorifics to be used in a given interaction. Using too many honorifics is often discouraged in prescriptive accounts, whether official grammars or popular self-help books, but what counts as “too many” varies from speaker to speaker.Footnote ⁷ In the itadaku thread, about, twelve posts characterized using itadaku when talking about one’s own beer consumption, whether on the patio or elsewhere, as too polite or pretentious, whereas eighteen respondents argued it was appropriate. The post shown in Example (6) is one that comes down on the side of excessive use (kajō keigo):

This bifurcated set of responses is not limited to the single verb itadaku, of course, but is found in the typically divided support for and critique of other kinds of honorifics as well. Another set of forms that frequently come up for debate, on the grounds that their use is “excessive,” are the nijū keigo ‘double honorifics’ discussed previously, but they too are widely used. In fact, the 2007 guidelines note that some nijū keigo forms are today conventionalized and accepted; they offer five examples. Interestingly, one of the five also has to do with eating (or drinking): o-meshiagari-ni naru ‘eat’. Beyond the five verbs listed in the guidelines, however, just which forms are “acceptable” by virtue of conventionalization is unclear. And the guidelines themselves are inconsistent with an earlier 1998 survey (reported in Kokuritsu Kokugo Kenkyūjo Reference Kenkyūjo2007) that included two nijū keigo forms not included in the 2007 guidelines: osshararata ‘said’ and o-kaki-ni narareta ‘wrote’. More than 60% of the 1,136 respondents in that earlier survey judged both to be acceptable.

When are honorifics needed? The most frequently discussed topic on Hatsugen Komachi, comprising 71% of the forty-one threads posted during the year, had to do with what contexts called for honorifics. Particularly interesting subsets of our blog corpus are posts about how to assess the vertical or horizontal relationship between interlocutors and what stance to take toward that perceived vertical or horizontal relationship (twenty threads) and about how to assess the relationship between the vertical and horizontal relationship and what stance to take toward the complex of vertical and horizontal distances combined (six threads).

The first concern centers on what to make of a vertical, that is, status asymmetric, relationship. One blogger (August 2, 2012) complained about a colleague who was five years younger but who did not use honorifics when speaking to her. There were nineteen responses to this post: six expressed agreement with the opinion that the younger colleague should use honorifics, nine disagreed, and four were ambivalent or ambiguous. Some posts suggested that a five-year age difference was not enough to trigger honorifics, especially since the blogger wasn’t the younger colleague’s supervisor. This post and the responses to it suggest a general preference for constructing egalitarian relationships in the workplace, but this preference is hardly categorical. Other posts clearly supported a preference for the linguistic maintenance of workplace hierarchies.

Another thread addressed the way hospital personnel (doctors and nurses) talk to elderly patients.Footnote ⁸ In terms of age, “elderly patients” are likely to be older than the hospital personnel, but in terms of medical expertise, patients are the status subordinates, even though in recent years the media have taken up the notion that patients should be treated as customers (kanja-sama ‘Mr./Ms. patient’). The following is part of the first post in a thread about how patients should be spoken to (May 8, 2012):

This blogger apparently thinks that elderly people should be addressed with deference encoded in honorifics and that using plain forms toward them is rude, although at the same time wonders if medical personnel are attempting to show friendliness. There were seventy-seven responses to this post (as of May 9, 2012), including many from hospital personnel. About 40% of them agreed with the sentiments expressed in the initial post. Many who agreed with the post expressed the opinion that the age difference should be acknowledged,Footnote ⁹ and that the use of plain forms in this interactional context is rude, as illustrated in Example (8):

Clearly the person who posted this response is interpreting the use of tameguchi (informal style using plain forms) by nurses as talking down and rude, and not as an indication of friendliness. (Note that we do not know how the mother felt about the use of hospital personnel’s tameguchi toward her, which may not be at all the same as her blogging child’s reaction.) Several people made similar comments, including: “When friendliness goes too far, it reverses the relative positions/statuses of the participants (tachiba o gyakuten saseru)” (May 8, 2012); “Isn’t it because they see themselves as superior (ue kara no mesen)?” (May 9, 2012); “These doctors may not understand the difference between friendly talk (shinkinkan o yobu kotobazukai) and overbearing (ōhei na) talk” (May 9, 2012); “Tameguchi sounds self-important (erasō) and cold” (May 9, 2012). These opinions demonstrate that even if a speaker uses plain forms intending them to mark friendliness, the absence of honorifics may be interpreted otherwise, that is, as an index of vertical distance or as a marker of status superiority – a possibility inherently associated with polyindexical forms such as these. A few posts also pointed out that friendliness/closeness (shitashisa) is different from excessive friendliness (narenareshisa), a thin line with an indeterminate location, as the posts in this thread reveal.

In contrast to those who express opinions to the effect that honorifics should be used when talking with the elderly, privileging an age or generational hierarchy, other posts take a different stance, expressing the opinion that hospital personnel should use honorifics toward the patients because they are customers (okyaku-sama).

In response to this blogger’s complaint, the nurses stopped using tameguchi, but not out of sincere politeness, and so the blogger interprets their use of polite or honorific forms as sarcastic and, thus, rude.Footnote ¹⁰ This post is not the only one offering the opinion that hospital staff should use honorifics toward patients because they are customers. Several posts written, not by patients, but by hospital personnel, note that some patients want to be treated as customers (see Example (11)).

There were, however, also many posts to this thread that disagreed with the position that tameguchi were inappropriate. Many argued that tameguchi is a sign of friendliness, while honorifics sound aloof and create distance. See Example (10):

Many hospital personnel also expressed this view. Although a few said that they consistently use honorifics when talking with their patients, the majority said they use both honorific and tameguchi, depending on the details of the specific situation.

Although the post does not explicitly indicate it, the writer clearly works at a hospital and is aware of the potentially different meanings of honorific and plain forms in the medical context. The choice to strategically differentiate their use depending on the patient’s age, expressed or assumed preference for a formal or informal speech style, degree of intimacy, and the like is one of this individual’s ways of dealing with the complexities of these interactions. But even this strategy does not suit all interlocutors. The writer of the following post from a nurse expresses frustrations with the difficulties involved in adequately fine-tuning the use of honorifics and tameguchi:

This post is a good example of how hard it is to please everyone, since her experience has been that different people (not only patients but their visiting relatives and friends) “hear” the same expression differently.

Quite a few hospital personnel remarked that elderly patients often have difficulty understanding them when they use honorifics because the sentences become more complicated. One example of this kind of post offers that thought, along with the opinion that honorifics inherently produce indirection:Footnote ¹¹ “I think honorifics are difficult to hear for elderly persons, because they make utterances roundabout” (May 9, 2012). Several posts linked tameguchi and regional dialects (or honorifics and Standard Japanese) to assert the merits of using tameguchi and regional dialect as being straightforward, easy to understand, and effective in helping the patient relax. Note the presumptive pairing of informal speech with simple, socially and/or geographically peripheral people. This is illustrated in Example (13):

For this person, friendly talk includes non-use of honorifics and use of the regional dialect. Another hospital worker posted that it is often impossible to convey the most important information using honorifics, and that it also makes patients uneasy and guarded; patients are happy to be addressed with expressions “other than honorifics,” such as dialect (May 8, 2012). Both writers appear to assume that honorifics are used together with Standard Japanese, and plain forms with a regional dialect (see also Chapter 2). And this hints at a more common understanding of Standard Japanese and the Standard Japanese-based system of honorifics as formal, creating distance, and making it difficult to speak openly and honestly. One writer simply said: “Using Standard Japanese and honorifics feels cold, lacking warmth” (May 9, 2012).Footnote ¹²

Posters in other threads worry about how horizontal distance (lack of social intimacy or solidarity) should be marked. One woman initiated a thread with a post (November 11, 2012) offering her observation that, regardless of age, the mothers of the other children at her child’s nursery school talked to her using tameguchi even from their first meeting, while she used honorifics. Only one respondent agreed, saying that she too would use honorifics in that situation. Two respondents claimed that they would use tameguchi, and the remaining six posts offered a compromise solution, saying that they would adjust their speech to align with the interlocutor or mix honorifics and tameguchi.

These observations gesture toward the normatively muted reality that individuals differ widely in their assessments, first, of the vertical or horizontal relationship, largely because, if the posts we have examined are anything to go by, individuals draw on different criteria in making their choices. The posts in our corpus also offer evidence of how individuals differ in the stances they take toward the perceived vertical or horizontal relationships due to how they assess various factors outside the simple facts of vertical status asymmetry or a measure of social distance; these may include a sense of interlocutor preference, the need for the efficient transmission of information critical to the interlocutor’s well-being, and response to interlocutors’ use or non-use of honorifics forms on a case-by-case basis.

The second issue involves the complexities that arise when both vertical and horizontal distance must be assessed simultaneously. The blogs we examined demonstrate that how to assess the relationship between the two kinds of distance and what stance to take toward such combinations can be a pesky matter. This is a classic example of the ideological struggles between a primary interest in power (hierarchy) and an emergent orientation toward solidarity (egalitarianism) in changing societies (Morford Reference Morford1997; Silverstein Reference Silverstein, Mertz and Parmentier1985, Reference Silverstein2003). Our bloggers clearly showed us that this concern emerges in a wide variety of social situations or interpersonal relationships. Example (11) touched upon the issue of horizontal distance in relation to vertical distance in the case of hospital personnel and patients. We also found many other threads that discussed this issue as it arises in other social situations. Two threads, for example, addressed the question of whether to use honorifics when talking with one’s in-laws. The original post in one thread was from a writer disturbed by her younger sister-in-law’s use of tameguchi when talking with her, while she has been using honorifics.

In this post, the writer implicitly acknowledges that vertical distance is more important to her than expressing horizontal closeness, and expresses some frustration that her younger sister-in-law does not use honorifics in their interactions, even though she – the writer – has been using honorifics. At the same time, the very fact that this writer has been using honorifics to the younger woman indicates that she considers expressing mutual respect important as well. There were four responses to this post (as of July 5, 2010). One of them (July 3, 2010) supported the writer, saying Nihon ni wa chōyō no jo to yū reigi, dōtoku ga aru no desu yo ‘In Japan, there is an etiquette, an ethics, of respect for age hierarchies’. Here, vertical distance is articulated as the most important basis for choice. But the other three responses disagreed, as illustrated in Example (15):

This writer articulates using tameguchi as an attempt to be friendly. For her, then, the use of honorifics would mark distance or aloofness (yosoyososhii). In other words, she privileges egalitarian horizontal closeness over normative prescripts to acknowledge socially defined status asymmetries (at least, those based on the relative ages of the interlocutors).

Another thread concerns the use of honorific and plain forms in interactions with parents-in-law. Example (16) gives part of the original post:

Here, the use of honorifics is openly described as both a way of acknowledging vertical distance and as a way of denying social closeness (maintaining horizontal distance). The post describes a struggle between two competing needs – the need to show respect for the writer’s parents-in-law and a concern to indicate closeness with them. Among the eighteen responses (as of July 5, 2010), eleven wrote that they also use honorifics toward their parents-in-law. See Example (17):

This writer is clearly aligning herself with the idea that vertical distance is more important than horizontal closeness.

Two other writers, however, claimed to know people who use tameguchi with their parents-in-law. And four wrote that they themselves use both honorifics and tameguchi, sometimes in ways that appear to be highly strategic, as illustrated in Examples (18) and (19):

This begins with the writer claiming that she normally does not use honorifics, which suggests that the effects of increased intimacy over time have overridden the generational status asymmetry. However, she then goes on to admit that she does sometimes use honorifics. Unsurprisingly, this is in cases when she is asking for something or thanking her parents-in-law for something. That is because, on the one hand, it is nice to be polite to anyone when you want something from them and, on the other, formulaic speech acts, of which there are many in Japanese, are known to elicit semi-fixed and formal phrasings. Strategic uses of honorifics are also illustrated in a post from a different thread that addresses the issue of caring for elderly parents-in-law (June 14, 2012). In the initial post, the writer says that she does not wish to look after her parents-in-law when they get old and is making every effort to keep her distance from them, including using baribari no keigo ‘full-scale honorifics’ like the keigo she uses when talking with aka no tanin ‘total strangers’.

As these examples show, how a speaker assesses the horizontal distance between self and interlocutor vis-à-vis the vertical distance between them varies widely. In the examples we have presented so far, the writers have made it clear that they have had a relatively long-term relationship with their parents-in-law. But even when the relationship is not (yet) of very long duration – a situation in which honorific use is expected – various posts indicate that speakers differ in their opinions about honorific use. See Example (20):

Of the forty-one responses to this post, nine respondents agreed with the initial writer’s outrage, nineteen disagreed, and thirteen offered limited or ambivalent agreement. As in other threads, disagreements were founded in arguments for tameguchi’s friendliness and warmth, coupled with claims that the use of honorifics might sound cold or pretentious.

Another thread also involves issues of both vertical and horizontal distance and illustrates another fraught arena of concern: naming practices. In Japanese, it is standard practice to call someone by either their first or their last name plus a title suffix. In Standard Japanese, these title suffixes include -sama, -san, -kun (or -chan) in descending order of level of deference. Use of a bare first or last name without any title suffix is the least deferential possibility. In the initial post to this thread, a mother complains about her son’s elementary school teacher because she does not use the honorific suffix -kun to address her son. She just calls him Masao, without any suffix (called keishō in Japanese) at all. The writer argues that the teacher should be polite toward her son and furthermore, that maintaining a horizontally distant relationship between teacher and student is important. In other words, this mother says in her post that she expects students to be polite toward their teachers, who should also show mutual respect by being polite toward them.Footnote ¹³ The mother includes in her post that the teacher has said she doesn’t use title suffixes because she wants to treat her students like friends (tomodachi no yō ni), but the mother, nonetheless, writes that she thinks it is rude.

There were seventy-six responses to this post (as of July 5, 2010); many of them supported the mother, but others disagreed with her complaint. Among those who disagreed with the mother, many of them simply said yobisute is fine because she is a teacher, that is, a higher-status person. This view regards the non-reciprocal use of plain forms as a sign of occupying the superior position in a status asymmetry; in other words, for people who responded in disagreement with this post, honorifics are used, at least in the context in question, non-reciprocally; they are not for showing mutual respect but for marking status asymmetries. It is, thus, appropriate for the teacher to “talk down,” as it were, to the children in her charge. This view, however, is different from that expressed to the mother by the teacher, who claimed to use plain first name forms to reduce the distance between her and her students by treating them as “friends,” a term which suggests both equal status and social intimacy. This example illustrates how, even if a speaker intends to be democratic and friendly and to indicate that by using plain forms, there is the ever-present alternative indexicality of a status-superior speaking to a status inferior (for the multiple possible interpretations of the much simpler pair of French pronominals tu and vous, see Morford Reference Morford1997).

In the posts discussed above, bloggers offer an interpretive landscape in which plain forms are largely seen as a sign of, alternatively, intimacy, power, or disrespect, while honorifics are variously seen a sign of deference or one of rudeness offered by someone inappropriately holding her- or himself distant or aloof. They are, thus, inherently poly-, or at least, bi-indexical. Honorifics are generally associated with a formal or deferential stance, which is in turn linked to specific social structures, that is, either to status asymmetries or social distance. However, in any specific context, depending on interlocutor uptake, an honorific may trigger a range of interpretations, not all of which are straightforwardly related to politeness, but ones in which the speaker is seen as aloof, distant, unfriendly, or pretentious, as illustrated above. Plain forms tend to be associated with informality or lack of deference, which in turn may be linked to a specific aspect of the social context, that is, the close social relationship or the speaker’s higher status. But again in a given context, it may offend the addressee, who perceives the speaker as being too friendly, overbearing, or rude. This built-in polyindexicality is typically treated as unproblematic when discussing form-meaning relationships out of context, but we argue that in terms of coming to agreement about what counts in any given interaction, this is far from the case, as we hope we have illustrated above. The stereotypical meanings of honorifics as indexes of politeness, deference, status difference, and lack of intimacy may be accepted by everyone in the abstract, but those meanings may take on further construal in specific contexts and may be interpreted even with opposite meanings.

Furthermore, in addition to the polyindexicalities centering on the immediate social relationship between the speaker and the addressee (or referent), that is, the first-order indexicalities, there is yet another order of indexicality complicating speaker choices and hearer interpretations. We refer here to second-order indexicality, the indexicality that associates certain honorific choices with social types of speakers (Silverstein Reference Silverstein2003). A second pass through the blog posts examined above reveals that there coexists with the normative association of specific speaker-hearer relationships with certain patterns of honorific use a second order of indexical relations related specifically to the construction of certain social personae of the speaker. For example, in the thread from Hatsugen Komachi that discussed the word itadaku we observed that the use of itadaku in question can be interpreted variously as kenjōgo I or II or bikago. The interpretation of itadaku as kenjōgo I is said to be based on self-lowering in order to elevate a deference-entitled referent, while the other three categories are determined on the basis of a speaker’s “consideration” for an addressee (Chapter 3). However, many of the blogs indicated that the appropriateness or inappropriateness of the use of itadaku is related to the user’s social persona, as variously characterized in terms of refinement, classiness, or some other criteria rather than simply about the form in some particular setting. As we saw in Examples (4) and (5), there were some writers who characterized the use of itadaku to refer to the eating of food one had prepared oneself as refined, beautiful, and also as feminine. The writer of the post in Example (21) clearly relates itadaku to the (imaged) speaker’s persona:

This post clearly characterizes itadaku as a humble form and then links its self-lowering humility to the class positionality and refinement of the people – in particular, the female people – he or she imagines associated with that kind of language practice. The association of form to person draws on stereotypified notions of women – or of normative femininity – as elegant, refined, and polite, or humble (Okamoto and Shibamoto Smith Reference Okamoto and Shibamoto Smith2008).

Example (22) expresses a similar view, while also relating honorifics to Japanese “tradition”:

This writer relates the use of honorifics to refinement and femininity (see Chapter 6 for further discussion of the indexical relationship between honorifics and femininity; see also Okamoto Reference Okamoto, Sato and Doer2014; Okamoto and Shibamoto Smith Reference Okamoto and Shibamoto Smith2008; Shibamoto Smith Reference Shibamoto Smith2006). “Tradition,” that is, declares that the use of itadaku in question is correct. People like the writer of the initial post are themselves to be criticized for using plain forms like taberu ‘eat’, which are gasatsu ‘rough and unrefined’. At the very least, they should not get in the way of those who are merely trying to uphold the “traditional” order.

The above metapragmatic comments regarding honorifics speak eloquently of how diverse interpretations of honorifics and plain forms can be, challenging notions such as “socially agreed-upon rules of honorifics,” or “shared norms” [for honorifics use], and “correct” honorific usage. Such differing views are not surprising given the changes in the dominant narrative of “proper” honorific use reviewed in Chapter 3. Diachronic changes entail synchronic variation, and here we have offered ample evidence of the synchronic variation in honorific evaluation. Normative, or prescriptive, characterizations of honorifics as indexes of politeness, deference, status difference, or lack of intimacy may be generally accepted in the abstract, but the same honorific form may take on further indexicalities in specific contexts, even being interpreted as aloof or deliberately insulting. Such different interpretations suggest different stances that speakers may take toward the relevant relationships (Du Bois Reference Du Bois and Englebretson2007) or differing stances toward the normative rules. Some speakers, for example, may wish to construct a hierarchical relationship by using honorifics non-reciprocally, while others may prefer to construct an egalitarian relationship through the reciprocal use of honorifics or by the reciprocal use of plain forms. This cross-speaker (and cross-interlocutor) variability may be unproblematic when considering form-meaning relationships in the abstract, but we argue that when it comes to agreeing about what counts in a given interaction, this is far from the case, as our analyses above demonstrated.

Variation in views about proper honorific use indicates that the same honorific form used in equivalent contexts may not always be interpreted in the same way. Speakers who believe that nijū-keigo is appropriate may find those who do not use it to be rude. Or those who use oru as kenjōgo II (or as sonkeigo) may be perceived as impolite by those who believe oru is (or “should” be) used only in the sense of kenjōgo I. Variability in the interpretation of the same linguistic form is, of course, not unique to the interpretation of Japanese honorifics. Morford (Reference Morford1997) offers examples of similar issues with regard to the indexical “meanings” of the second-person pronouns tu and vous in Parisian French. The generalized or wider use of tu may be seen as a sign of “egalitarian sensitivity” or “intimacy,” but it may also be construed as “evidence of a decline in decency” or a loss of “true intimacy” due to its indiscriminate use. These examples as well as the comments from Hatsugen Komachi we saw above illustrate how one’s beliefs affect language use and interpretation in a specific context. Eckert (Reference Eckert2008) draws on the voluminous work on the variable -ing in English to point out that the velar variant may be variously construed positively as articulated and educated or negatively as pretentious and that the apical variant may be interpreted favorably as unpretentious and relaxed, but unfavorably as lazy and impolite. We concur with Eckert (Reference Eckert2008: 466), who argues that “since the same variable will be used to make ideological moves by different people, in different situations, and to different purposes, its meaning in practice will not be uniform across the population.”

Our analysis offers support for our claim that, while many speakers may share some broad or abstract ideas about how keigo should be used, there are no universally shared norms for “correct” keigo use that are mechanically useful in deciding what to say in specific contexts or, for that matter, useful in estimating how one’s utterance will be interpreted by one’s interlocutor. Individuals come to the meaning of words/utterances through a historical chain of experiences, which are always and inevitably different, informed by different pasts (Bakhtin Reference Bakhtin1981). We see that there are many possible interpretations of the meanings of one honorific form or another available within the indexical fields of these forms. And all these interpretations make sense to some people but not necessarily to others, in any given instance, because they are drawing on particular metapragmatic beliefs or cultural models of honorifics. That is, if signs are interpretable, it is only because there are definite sets of beliefs within a community that members draw on regarding the relationship between linguistic forms and social structure (Agha Reference Agha2007; Morford Reference Morford1997; Silverstein Reference Silverstein1992). But such beliefs may not be shared universally in a community, which in turn brings about diverse interpretations in specific contexts. Analysis of the contexts in which honorifics are observed to occur (or not to occur), no matter how fine-grained the consideration of constituent features of those contexts may be, cannot offer a prediction model of honorific use, precisely because of the ever-present tension between the normative model and the individual agency of the speaker. We address this issue further in the following section in which we analyze the use and interpretation of honorifics in actual discourses.

4.2.1 The use and interpretation of addressee honorifics

We first discuss the use and interpretation of addressee honorifics, re-examining data from Okamoto (Reference Okamoto2011b), which consists of six dyadic conversations.Footnote ¹⁶ The participants in these conversations are listed in Table 4.1.

All participants are native speakers of Japanese.Footnote ¹⁷ The two speakers in each conversation were in some kind of hierarchical relationship (vertical distance) in terms of their age, social role, and the like. The degrees of intimacy (horizontal distance) in these six pairs of speakers differed considerably from very close to quite distant. In Tables 4.1 and 4.2, the conversations are listed in ascending order of social intimacy between the participants, with those in Conversation 1 having the most distant relationship and those in Conversation 6 the closest. The use of addressee honorifics and plain forms by each speaker was analyzed for the first sixty relevant tokens.

Generally speaking, both vertical and horizontal distance seems to have had considerable influence on the use of honorific and plain forms by these speakers. In each conversation, we see status asymmetry, and in each case, the lower-status speaker’s use of plain forms was much more limited both in frequency and function than the higher-status person’s in each dyad, as dominant norms would dictate. In general, then, we see substantial effects of the dominant ideology of honorifics. On the other hand, there was also wide variation among the speakers that may be an indication of their different relations in terms of horizontal distance. For example, the higher-status persons differed greatly in the use of addressee honorifics, ranging from 61% by O1 to 2% by O6 (see Table 4.2). This variation seems in part due to the different stances speakers took toward the combination of horizontal and vertical distance. O1, for example, used addressee honorifics 62% of the time, the highest frequency among the six higher-status participants. If we take into consideration the diverse metapragmatic comments in Section 4.1, we could interpret O1’s use of honorifics as a polite gesture toward a newly employed, highly educated interlocutor by offering reciprocal honorific forms in order to reduce the vertical distance between them even though O1 is both a status superior (a supervisor) and older than Y1. However, it is also possible that O1 was instead simply marking the social (horizontal) distance between them, since they were not closely acquainted; that is, O1 may have privileged marking the horizontal distance – which would call for more addressee honorific forms – over the vertical distance – which would not. Note, however, on the interpretive side, that it is also possible that some people would find O1 too polite and formal in using a high proportion of addressee honorifics in conversation with a lower-status person. O1’s use of these forms thus potentially has multiple and ambiguous indexicalities.

O4, on the other hand, used addressee honorifics only 3% of the time. O4 is much older than Y4 and does not know Y4 very well. O4’s very infrequent uses of honorifics may suggest that she considers the vertical distance, here based on age or generational status rather than workplace, hierarchically defined status, as more central than considerations of horizontal distance in making her linguistic choices. But it could also be that she is trying to construct a friendly, or horizontally close, relationship, in part because Y4 is O4’s daughter’s friend and also because the conversation is taking place in Y4’s home rather than her office. There is nothing in either of these texts (O1’s and O4’s speech) that speaks to which of these indexicalities is in play in the minds of the speakers. In this respect, any use of honorific and plain forms always offers polyindexical and ambiguous potential. Furthermore, this potential polyindexicality applies not only on the side of speaker intention but also on the side of interlocutors who interpret the interaction through their own stances toward honorific norms.

We also see the effects of horizontal distance in the speech choices of the status-inferior participant. For example, Y1, a newly employed subordinate of O1, used the greatest proportion of honorifics (72%) of all the six lower-status persons in the data, while Y5, who knows O5 quite well, used honorifics much less frequently (45%) than Y1. Note, also, that Y6 – speaking to his mother – used virtually no addressee honorifics (2%), which is today normative in most families, as indicated by the survey results we saw earlier. However, while it is evident that horizontal distance affects or constrains the speakers, there was also variation among individuals, beyond simply the nuclear family pair (O/Y6). For example, even when the horizontal distance was similar,Footnote ¹⁸ not everyone used honorifics in the same way. O2 and O4, for instance, were not close to Y2 and Y4, respectively, and interacted with them only infrequently, but O2 used addressee honorifics 41% of the time while O4 used them only 3% of the time. Considering the first pair, Y2 is a former student of O2, but is now a graduate student elsewhere and has returned for a visit; the setting is in the office of O2. Again, returning to the metapragmatic commentary in the preceding section, O2’s greater use of honorifics may suggest that he is acknowledging the horizontal distance between Y2 and himself more than marking the vertical distance that still obtains between them. Or, he may be intending to construct a less hierarchical relationship by approximating a reciprocal honorific use (42% to 68%). And, of course, the setting – an office, which would prescriptively call for more formality – is a factor that may be affecting his use of honorifics. In contrast, O4 hardly used honorifics at all. As discussed above, this may be a sign of her acknowledgment of vertical distance based on a significant age difference, a sign of her attempt to create a friendly relationship based on Y4’s relationship with her daughter, or a more private (home) setting. It may indeed be all of the above, and more.Footnote ¹⁹

Furthermore, there are other contextual factors besides the two major interpersonal factors (vertical and horizontal distance) and one contextual factor (setting) that seem to have affected, but not determined, the use of these forms in the conversations above. These contextual factors include speech event and genre, discourse organization, and the linguistic environment. In what follows, we discuss these factors and offer examples, but in doing so, we also stress that the multiple indexicalities that inhere to each linguistic form are available for interlocutor uptake in any given context.

We discuss the linguistic environment first. In Japanese prescriptive grammar, honorifics are said not to be used in dependent clauses, but as our data show, they may be used in very formal conversations. In other words, this is not simply a matter of grammar, but of choosing forms for their social significance (Lavandera Reference Lavandera1978). Speakers may opt to choose honorific or plain forms. Although the numbers of tokens were small in some of the conversations, our analysis shows that the use of plain forms by lower-status persons was more limited in main clauses than in dependent clauses, while higher-status persons used plain forms frequently both in main and dependent clauses. Y3’s utterance in Example (23) illustrates the use of a plain form in a dependent clause by a lower-status person. In this example and in examples 24–32, addressee honorifics are marked by a single underline and plain forms by a double underline.

In the first dependent clause, or relative clause, in line 1, Y1 used the plain form osshatteta rather than osshatte imashita with an addressee honorific, which she could have used; in the second dependent clause, or if-clause, she also used the plain form dattara rather than deshitara with an addressee honorific. She then used an addressee honorific desu in line 3 in the main clause. The greater use of plain forms in dependent clauses, then, may be related to the normative prescriptive grammar convention that plain forms should be used in dependent clauses, or a concern that using honorifics even in dependent clauses may sound too odd, as overly formal or even pompous in the situation. It may also simply reflect the speaker’s strategy for organizing information, as dependent clauses tend to convey background information (Cook Reference Cook2006).

Similar uses of plain forms were often observed when speakers ended a sentence with a structurally non-final or dependent clause marker, as illustrated in line 2 in Example (24):

In line 2, Y2 ended her sentence using the non-final, or continuative, form omotte ‘think and’ rather than ending with an honorific form (omotte ukagaimashita ‘thought and came’ or omoimashita ‘thought’), even though she was in fact ending her utterance. This can be interpreted as an attempt to avoid sounding blunt by ending an utterance clearly with a finite form. Moreover, using a plain form in a structurally dependent clause may have been easier for Y2 than using it in a clearly final main clause (for example, omotta no or omotta n da ‘thought’). At the same time, the use of the plain form in (24) can also be interpreted as the speaker’s attempt to avoid using addressee honorifics constantly at every possible place in order to hint at a certain degree of friendliness than always using sentence-final forms with addressee honorifics, but without going too far into narenareshisa ‘over-familiarity’. The lower-status participants in this data used this strategy quite frequently.

Although lower-status speakers’ use of plain forms in main clauses was more restricted than that of their higher-status interlocutors, there were some cases in which they used plain forms in main clauses with final verb forms. These seem to be related to strategies for discourse organization or particular speech acts. Examples where plain forms appeared included exclamations, soliloquy-like utterances, and emphasis.Footnote ²¹ We saw above that one of the effects of plain forms in dependent clauses, as in (23), is to avoid highlighting background information. Conversely, the avoidance of honorifics in main clauses may have the effect of focusing on the new information being conveyed. This is often seen when a speaker asks or answers a question focusing only on the information at issue in the exchange, as in line 8 in (24) or when the speaker is offering only a short supporting utterance or backchannel, as in line 3 in (25). In (24), for example, in lines 4 and 6, Y2 used addressee honorifics, but in line 8 she did not but simply provided an answer to O2’s question.Footnote ²² This efficiently meets O2’s current need for information, and addressee honorifics (Eigogaku-gairon deshita, Ano S-sensei deshita) may have been taken as too formal and stiff – or presented Y2 as less the friendly visitor she had been constructing herself to be.

In line 3 in (25), Y5 repeated the key phrase in O5’s utterance in line 2. She could have used an addressee honorific (pera pera desu), but here she is simply supporting what O5 said, and adding a verbal form, especially one with addressee honorification, would make her offering more intrusive. As it is, the focus remains on the most important part of the exchange in O5’s utterance. In general, plain forms seem more effective for performing a kind of exclamatory or emphatic speech act; and at the same time, they have the added benefit of supporting Y5’s overall construction of friendliness as instantiated in an office setting. In her examination of three dyadic conversations between professors and students, Cook (Reference Cook2006: 278–281) also found similar examples. For example, when one professor said that the proportion of Christians in Korea is 30% (sanwari), much higher (ōi) than the 1% of Christians in Japan, the student interlocutor only offers sanwari and ōi, which overlap with the professor’s utterance. Cook explains that the student’s use of these phrases helps co-construct an utterance with the professor. In examining interview discourses, Ikuta (Reference Ikuta, Jones and Ono2008) also observed many similar uses of plain forms by the interviewer who repeated or paraphrased (part of) what the interviewee said. Ikuta argues that such uses of plain forms can indicate that the speaker is playing a supportive role without taking the floor; they may function as backchannels. At the same time, depending on the content, they may function as indirect requests for clarification, elaboration, emphasis, or expressions of emotional reaction.

Speech acts (soliloquy, exclamation, complaint, joke) also seem to play an important role in speaker choice of honorific or plain forms. We found this in our data, when lower-status participants used plain forms in main clauses in soliloquy-like utterances, as illustrated in Example (26):

In Example (26), Y2 used a plain form in line 3 rather than ending with an addressee honorific (ja nai ka to omoimasu ‘I think it’s the case that …’). This may be because it is in a soliloquy-like utterance (see also Hudson Reference Hudson2011). If, instead, Y2 used the plain form omou ‘[I] think’ as in ja nai ka to omou ‘I think it’s the case that …’, directly addressing O2, that would transform the utterance into one of direct address to her interlocutor and she would have been making a more direct violation of the status-asymmetric norms for addressee honorifics. It is the overtone of self-address that allows a simultaneous encoding of normative alignment – use of a plain form when “addressing” oneself – and allowing Y2 to indicate a degree of informality or friendliness by producing a plain form in the presence of a status-superior interlocutor, thus avoiding the construction of excessive formality. In examining the discourse of faculty meetings at a junior high school which were primarily carried out using addressee honorifics, Geyer (Reference Geyer2008a, Reference Geyer, Jones and Onob) also found that plain forms were strategically used in monologue-like utterances in order to mitigate the force of face-threatening speech acts such as non-compliance (see also Makino Reference Makino, Akatsuka and Strauss2002 and Maynard Reference Maynard1993, Reference Maynard2004 for similar cases).

Lower-status speakers in our data also often used plain forms in main clauses in speech acts, such as exclamations, emphasis, and expressions of disbelief, as illustrated in (27):

Y3 used the addressee honorific desu in lines 1 and 3. But she simply said dare without an honorific in line 7, which can be considered a case of focusing on the key information as discussed above, while also expressing an informal attitude. Then in line 9, she uttered three sentences consecutively using plain forms. These utterances are exclamatory, expressing her surprise and strong disbelief about what O3 had said in line 8, which is more effectively conveyed using brief plain forms. And it seems to be the strategy that Y3 adopted here. This strategy was also used by some of the other lower-status speakers, particularly Y2 and Y5.

The nature of the speech event is another factor to consider. For example, Cook (Reference Cook, Pütz and Dirven1996, Reference Cook, Jones and Ono2008) found that students in elementary schools used addressee honorifics when engaging in happyō ‘presentations’, while they used plain forms when talking to their teachers individually. Setting is another factor. Okamoto (Reference Okamoto1998) observed the effect of the setting in the conversations she observed in sales encounters. That is, in department stores, sales clerks used mostly honorifics toward their customers, whereas vendors in casual marketplaces often used plain forms when addressing their customers. This also indicates that speakers make and react to assessments of the context in which an interaction takes place; in this case, department store clerks used honorifics to show deference, but also to create a formal atmosphere, indicate the class status of the store, etc., while the market vendors used plain forms to show friendliness toward the customers as well as to create a casual and lively atmosphere in the marketplace. Note, however, that again, while the setting may shape choices, it is not determinative; in both cases, there was also inter- and intra-speaker variation. Moreover, customers’ perceptions of the same use of forms by salespeople may differ; some customers who are addressed by plain forms in marketplaces may feel that the salespeople are rude or too familiar. The effect of setting may also be relevant to the six conversations we examine here, as mentioned above with regard to the difference in the use of honorifics between O2, who spoke in the office, and O4, who spoke at Y4’s home, but again, we cannot single out setting as more than a shaping factor in speaker choice.

The preceding discussion suggests that even if a lower-status speaker in a particular interaction tends to use addressee honorifics as normatively prescribed, she or he may “deviate” from that norm in order to achieve a variety of effects. These effects include the use of plain forms in dependent clauses to avoid highlighting background information and in structurally non-final forms to avoid constructing an excessively formal persona. In main clauses, we see plain forms used in soliloquy-like utterances and in performing speech acts such as exclamations and complaints. Plain forms were used in ways that highlight and support the main information offered by a higher-status interlocutor while playing a supportive role in the interaction. But this is not an exhaustive list, nor, we argue, is such a list possible. As our examples illustrate, speakers constantly negotiate what they consider normative for a particular context and make active and strategic choices of honorific or plain forms depending on the context.

Negotiation of what speakers believe to be the norm is, of course, not limited to lower-status persons. Higher-status speakers in our data also used plain forms in main clauses for particular speech acts and the like, as the lower-status speakers did.Footnote ²³ Moreover, as normatively prescribed, the high-status speakers seem to have more “freedom” to use plain forms in main clauses, and the data in Table 4.2 show that they indeed produced proportionally more plain forms than their lower-status interlocutors. This “freedom,” however, may actually create certain difficulties. As the metapragmatic comments seen in Section 4.1 indicate, a higher-status speaker’s use of plain forms can be interpreted minimally as a marker of vertical distance or one of egalitarian friendliness; conversely, their use of honorifics may be interpreted as an attempt to maintain horizontal distance (being unfriendly) or as an attempt to create an egalitarian relationship via reciprocal honorification (being egalitarian, without necessarily being friendly). These possible ambiguities may cause high-status speakers to negotiate between these various possible indexicalities by mixing both honorific and plain forms in interactions with the same addressee. Such uses of both forms were observed particularly in the case of Speakers O2 and O3, as illustrated in Example (28):

In line 1, O3 used a plain form in three main clauses (shiteta no, yū no kashira, and atta no ne). In line 3, she used the addressee honorific desu in two main clauses (two instances of desu tte ne). In line 5, she used a plain form (sō na no?) in the first main clause, and then an addressee honorific (arimashita yo) in the second main clause. Why did O3 switch back and forth between honorific and plain forms? We cannot, of course, know for sure, but if O3 had used only plain forms, it may have been interpreted by her younger interlocutor Y3 as marking her higher status in terms of her position and age, but that could well be taken as rude or as an excessive orientation toward social hierarchies, especially since the age difference is small and the seniority gap may also be small. And the possibility of using plain forms to mark social intimacy is ever-present. If O3 had used only addressee honorifics, on the other hand, that too may have been taken up as respectful, as an attempt to reciprocate the deference that her colleague was offering her, who did use relatively more addressee honorifics. But then, it could also be seen as too distant, because the two meet on a daily basis, and too formal for the kind of topic – baseball – being talked about. It is thus likely that O3’s mixing of honorific and plain forms is a result of her negotiation of these possible varying interpretations. These same sorts of negotiations can also be seen in the speech of the higher-status speakers in the other conversations.

And it may be, as Errington (Reference Errington1992) and Sankoff (Reference Sankoff and Sankoff1980) demonstrated with respect to the mixing of language in rural Indonesia and among the Headwaters Buang in Papua New Guinea, respectively, that it is the right mix of forms that is the key to some particular style (or register) and that it is the very mix that is “normative,” not in the sense of a dominant prescriptive discourse, but in the on-the-ground sense of people’s normal expectations of what a particular kind of conversation is “supposed” to sound like. Here, the dominant model fails not just because the inherent polyindexicality of the forms causes leakages of social meaning outside the scope of the dominant norms, or even because there are simply too many variant potentials (or potential variants) of all the non-linguistic variables that go into choice, but because the model fails absolutely to consider the probabilistic nature of verbal productions and hearer response to probabilistically “determined” productions.Footnote ²⁴

4.2.2 The use and interpretation of referent honorifics in relation to the use of addressee honorifics

The findings discussed above suggest that speakers negotiate what they believe to be the norm in a complex way and use addressee honorifics and plain forms strategically in order to construct a desired context. There are, of course, many other linguistic resources besides addressee honorifics and plain forms that speakers can employ as resources for the construction of a desired context. One such resource is referent honorifics. In what follows, using the same data, we examine how the speakers in the six conversations used referent honorifics and the corresponding plain forms in relation to their use of addressee honorifics. Since many of the topics of conversations concerned non-human matters, there were not a large number of places in which the participants used or could have used referent honorifics so our data here do not permit a systematic quantitative analysis. Yet, the data exhibited many interesting uses and non-uses of referent honorifics that can help us understand how they are related to the use of addressee honorifics and how the two kinds of honorifics together contribute to the construction of the context. In what follows, we first discuss the use of referent honorifics used for third persons and then those used for the addressees.Footnote ²⁵

Let us first look at Examples (29) and (30):

The speakers in these examples all talked about someone that they refer to as sensei ‘teacher/professor’, or their social superiors. But some of them used referent honorifics toward them, while others did not. In Example (29), O1 and Y1 were talking about a book on internet education. Both speakers used referent honorifics (o-kaki ni natta by O1 in line 1 and nasatteru by Y1 in line 2) when referring to Conrad, the author of the book. This segment had only two places that O1 and Y1 could have used referent honorifics. But later in the same conversation, when O1 was talking about Japanese professors and when Y1 was talking about American professors, both of them used referent honorifics (O1’s Kono sensei-gata ga iwarete iru koto wa ‘what these professors are saying is that’ and Y1’s Amerika no sensei-gata mo konpyūtā tsukatte, kitto komame ni shidō shite kudasaru deshō ‘I’m sure American professors will give careful instructions by computer’). In all these instances, O1 and Y1 also used addressee honorifics as well as the referent honorifics.

In Example (30), O2 and Y2 were talking about GH, a professor at the American university that both O2 and Y2 attended (at different times). In this segment, there were three places in which O2 could have used a referent honorific for GH, but he used plain forms in all three of them – iru instead of irassharu in line 1; yatte instead of nasatte or yatte (i)rasshatte in lines 3 and 5. Y2, on the other hand, first used a plain form (i-) in line 2 for GH, but immediately rephrased the same utterance using a referent honorific (irasshai-), laughing as she self-corrected as if embarrassed about the lack of deference in her previous utterance; Y2 then used a referent honorific for GH in the remaining two relevant places (yatte rasshai- in line 4 and irassharu in line 6). In each of the four cases, Y2 used addressee honorifics (imasu, irasshaimasu, yatterasshaimasu, and irrasharu n desu ka), while O2 used an addressee honorific in the second and third instances (yatteta n desu and yatte masu), but not in the first instance (iru).

The variation in the use of referent honorifics toward higher-status third persons in these examples demonstrates that speakers are not automatically following rules like “Use honorifics when speaking about a higher-status person.” A closer look at these examples and other similar examples in the data suggests that the use of referent honorifics is not simply triggered by the referent, but is even more closely related to how the speaker perceives, or wishes to relate to, the addressee and how he/she uses addressee honorifics. Table 4.2 shows that the speakers in Conversation 1 used the highest proportions of addressee honorifics. Even status-superior O1 used addressee honorifics 62% of the time, and Y1 72%. In other words, both speakers seem to be quite formal and reserved in their mutual markings of social distance. This seems also to be reflected in their use of referent honorifics toward higher-status third persons even in their absence, as illustrated in Example (29). In Example (30), on the other hand, O2 used fewer addressee honorifics (42%) than O1 did. In addition, unlike O1 in the first conversation, O2 used plain forms in main clauses frequently, as mentioned above. He also did not use referent honorifics to talk about his former (non-present) professor. In contrast, Y2 is seen to use addressee honorifics 68% of the time and, correspondingly, to use referent honorifics to refer to the same absent professor. Given the similarity in status differentials, age differentials, the settings where the conversations took place, and the like, it is easy to see that gaps in the prescriptions for keigo use exist and that more empirical study of action rather than attitude is urgently called for.

Let us now look at the use of referent honorifics for the addressees, including both the use of referent honorifics to refer to the addressee’s actions (other-elevating sonkeigo) and referent honorifics to refer to the speaker’s own actions that are related to the addressee (self-lowering kenjōgo I). Here, too, our analysis indicates a close relationship between the use of referent and addressee honorifics. To illustrate this, we return to the conversational pairs displayed in Table 4.2.

In Conversation 1, an excerpt from which is found in Example (29), both speakers used referent honorifics for the addressee quite frequently: O1 used them 63% of the time (14 out of 22 relevant tokens) and Y1 73% of the time (11 out of 15 relevant tokens). In line 1 in Example (29), O1 used the referent honorific motareta to refer to Y1’s action, although he is not consistent in this usage; later in the same line, he used the plain form moraitai rather than referent honorific itadakitai to refer to his own action as related to Y1. In line 2, Y1 used the referent honorific mōshi and another referent honorific kangaete rassharu.

In contrast, there were not many referent honorifics used for the addressee in the other conversations. O2 in Conversation 2, for example, used only one referent honorific for Y2 (the humble form o-hanashi-shita koto ‘what I told you’) in a total of thirty-four relevant tokens. As we saw earlier, O2 used addressee honorifics 42% of the time in Conversation 2. Yet, with a single exception, this did not extend to him using referent honorifics for her actions. Y2 also used plain verbals more than referent honorifics (two times out of six relevant tokens), even though she used addressee honorifics quite frequently (68%) in Conversation 2. We see, then, that even if a speaker marks deference (or “respect”) to an addressee via addressee honorifics, that is insufficient to predict that referent honorifics will also be used. Just as the mixed use of addressee-oriented plain and honorific forms seems to serve as a resource for achieving the right degree of deference and formality versus social solidarity and casualness, so does a mixture of referent honorifics and plain verbals index similar balances between the prescriptive and the socially constructed individual.

Other conversations show the same tendencies: in Conversation 4, O4 produced 19 relevant tokens where a referent honorific could have been used for Y4, but she used plain forms consistently (tsukaretā? ‘Are you tired?’; hanashi o kiite yatte ne ‘please listen to her story’). Recall, too, that her use of addressee honorifics was also very infrequent (3%) and limited to formulaic speech acts such as apologies. As mentioned earlier, the degree of intimacy between the participants in Conversation 4 is similar to that between the participants in Conversation 2, but O4 created a more informal speech style than O2 did in Conversation 2. Y4 had only six referent honorific relevant tokens and used referent honorifics for O4 in three of them. These examples hint at a richness of to-date underexplored relationships between the various components of the keigo typology and provide a bit of preliminary evidence that the speaker’s stance toward the addressee and, beyond that, toward the nature of friendly-but-respectful conversation, affects the use of referent honorifics in a non-trivial way.

We now discuss the third issue: the relation between honorifics and Standard Japanese, or that between plain forms and regional dialects. As noted in Chapter 3, there is a close relationship between Standard Japanese and modern honorifics, resulting from language policies developed by the Meiji state and continuing to the present. Honorifics as a core symbol of Standard Japanese “served as the social emblem of the modern Japanese nation-state” (Koyama Reference Koyama2004: 429) and through public education and the media, Standard Japanese honorific norms became available to most Japanese. Today, many Japanese seem to associate Standard Japanese with formal or polite speech, characterized especially by the use of honorifics, and regional dialects with informal speech, characterized especially by the use of plain forms.Footnote ²⁶ But these associations require more careful empirical analyses across many more regional dialect areas than they have received to date.

4.2.3 The relation of honorifics and Standard Japanese to that of plain forms and regional dialects

Based on the observations made above, one may conclude that there is a tendency for Standard Japanese phonological, lexical, and morpho-syntactic forms to co-occur with (Standard Japanese) honorifics, while regional dialect forms co-occur with plain forms. In this respect, we can see a clear effect of the dominant norm that dictates the use of honorifics along with – or as part of – Standard Japanese. However, a closer look at the data shows that such a characterization is too broad, as the distribution of these forms is not clear-cut and rather much more complex for a variety of reasons. Some regional dialects had honorific forms used prior to the modern era of language standardization; these may no longer be recognized as honorific systems, but simply as dialect forms (Miyake Reference Miyake1995).

Another complicating factor is the way in which – and the timing of which – Standard Japanese honorifics are introduced in schools. Although Japanese children do not get explicit school instruction about honorifics until the higher grades of elementary school, they are from the beginning guided to use ‘polite’ language (teinei na kotoba). And that, in the schools, means exposure to Standard Japanese and Standard Japanese honorifics. Thus, children acquire some ability to use Standard Japanese honorifics before they are at an age when the honorifics of their local regional dialect would be acquired. They are also absorbing the implicit message that “polite” language is Standard Japanese. Thus, especially in those Japanese dialects without a systematic, surface-segmentable set of “honorific” forms (especially predominant in eastern Japan), children’s attention is directed away from the more discursively constructed linguistic politeness of their dialects (K. Satō Reference Satō1996). How and whether these discursive politeness forms are absorbed later in language socialization or whether Standard Japanese honorific forms are appropriated into otherwise dialectal speech for displays of deference is a question that has yet to be fully addressed.

In order to begin to outline a process for examining the stereotypified association of honorific use exclusively with Standard Japanese, however, we offer examples of conversational data and written texts in the form of email messages from one area where regional honorifics are known to have survived – the Kansai area. Urban Kansai dialects (Kōbe, Kyōto, Nara, and Ōsaka) have their own honorific systems, and they may occur in relatively formal or even in informal speech. In other words, the disengagement of regional dialect speakers from regional honorifics may be exaggerated.

We return, then, to Speaker A, the 28-year-old woman from Ōsaka who worked in a Kyōto department store. In Chapter 2, we examined her use of Ōsaka dialect and Standard Japanese when talking with different people in different contexts. In Example (31), A is talking to a customer in the department store. In this and the next example, a solid-underlined item is an addressee or referent honorific and a double-underlined item is a plain form; a bolded item is a regional dialectal variant, and an item shaded with diagonal lines is a Standard Japanese variant. When an item is not marked as either regional or Standard Japanese, it is common to both varieties.

Speaker A’s speech style in (31) was very formal, as she was talking with a customer at the department store where she worked. She used addressee honorifics consistently in all three relevant places (imashi in line 1, mashi in line 3 and masu in line 4). In addition, she used four referent honorifics in her address to the customer (the humble form itada(ite) in line 1, the respectful form (omow)are in line 3, and the humble form itadak(imasu) in line 4 and the prefix go- for the noun sōdan in line 3). Speaker A also used the Standard Japanese form -nai in awanai ‘not fit’ instead of the Ōsaka form -hen, as in awahen. Speaker A’s consistent use of honorifics and frequent use of Standard Japanese morphology may be interpreted as an attempt to index speaker/referent-oriented deference, while at the same time indexing the class status of the department store.

However, Speaker A did not use Standard Japanese variants exclusively, and it is important to keep in mind that, throughout, the phonological features of her speech are uniformly Ōsaka dialect. In this respect, her speech indexes her regionality quite clearly, as a phonological matrix into which Standard Japanese morphological features can be inserted as needed. And they appear to be “needed” more in public contexts, as would be expected from at least a partial alignment with the dominant national discourse. But in the case of urban Kansai dialects, regional honorifics do not disappear even in the more private context. Example (32) offers another sample of Speaker A’s conversation, this time in a much less formal setting.

In contrast to (31), in (32) Speaker A was very informal. She used no addressee honorifics; she also used many Ōsakan lexical and morphological forms, as the bolded items show, as well as an Ōsaka phonological frame. Nor did she use any Standard Japanese referent honorific forms, as she did in her conversation with customers at work. This example can be said to illustrate the tendency for plain forms to co-occur with variants of a regional dialect. However, Speaker A did not use plain forms alone; she also used one referent honorific haru, which is an Ōsaka dialect form. Moreover, A used haru even in parts of the conversation with her husband. It has previously been observed that haru was used quite frequently by salespeople when talking with customers (Okamoto Reference Okamoto1998). In Okamoto’s study, this was particularly noticeable among market vendors, who often did not use addressee honorifics toward their customers but did sprinkle their speech with the other-elevating Ōsaka dialect referent honorific haru, as in okahattara kirei ya shi ne ‘it would be_PLAIN pretty if you put (this) on_HON’, where hat- is a variant form of haru. Note also that okahattara is followed by the plain form kirei ya, in which ya ‘is/would be’ is an Ōsaka dialect form, this time a plain (non-honorific) one. These observations suggest that the distribution of honorifics and plain forms is more complexly organized in relation to Standard Japanese and regional dialect use than the stereotypified dichotomous associations would have us believe.

A brief summary of another data set, consisting of email messages, will further illustrate this complexity. The messages we examined were written by IH,Footnote ²⁷ a man in his early sixties, to his former high school classmates. IH lives in Nara and is a native speaker of the Nara dialect, which is part of the Kansai dialect group. Our data included two kinds of messages written by IH: (a) approximately fifty messages sent specifically to one female recipient, MY, IH’s high school classmate, and (b) approximately thirty messages sent as mass mailings between 2010–2011 to the approximately one hundred classmates from his graduating class. Most of the messages were quite short, consisting of only a few sentences.

Both types of messages are in general written in Standard Japanese using addressee honorifics, but there were differences in the two sets of messages. First, the addressee honorifics used in the mass mailings were often higher-level keigo forms than those in personal emails to MY. In the group emails, IH occasionally used the very formal addressee honorifics gozaimasu, as in Chūshajō wa gozaimasen ‘There is no parking lot’, and de arimasu, as in X-shi was Nihon gaka de arimasu ‘Mr. X is a Japanese painter’. The individual emails contained no such forms. IH also used referent honorifics, both sonkeigo and kenjōgo (I and II), frequently in the group mailings, as in o-kaimotome kudasai ‘please buy them’,Footnote ²⁸ but only occasionally in his personal emails.

Furthermore, although IH generally used Standard Japanese and honorifics in all his emails, he did not use them all the time. Plain forms and Nara dialect forms were used not only for certain speech acts, such as joking or exclamations (for example, ore yori monoshiri yan ka! ‘you know better than I do!’), but also for dealing with the somewhat ambiguous relationship between him and his recipients. That is, IH and the recipients are not high school students any more, and he meets with most of them only infrequently. In this respect, Standard Japanese and Standard Japanese-based honorifics are likely to be expected. Yet, they had been classmates and friends in the past. IH’s use of plain forms and dialect variants may have been his attempt to construct a friendly or solidary relationship. Moreover, even in mass mailings, IH occasionally used dialect forms (for example, Doyōbi N-san ni noboru to erai hito deshita ‘On Saturday, when I went up Mt. N, there were lots of people’, in which erai is a dialectal form of Standard Japanese ōzei no. So, in these emails, we see that it is not simply a matter of using honorifics or plain forms, but that IH makes fine adjustments using Standard Japanese and dialect as well as honorifics and plain forms to convey manifold meanings, adding to our sense that what is needed in models for interpreting the social meanings of honorifics is a more thorough examination from the perspective of a probabilistic model of mixed production.

In this section, we have seen diversity in situated honorific use to match the diversity of opinions about such use as provided by attitude surveys and blog commentaries covered in the preceding section. We have demonstrated that dominant norms defined in terms of status difference and degree of intimacy definitely affect speakers’ decisions to use honorific or plain forms. At the same time, we have also seen wide enough inter- and intra-speaker variation to suggest that speakers do not automatically adopt and enact those norms but negotiate them in reference to multiple contextual features and deploy a variety of variant forms – involving (Standard and regional) honorific and plain forms as well as Standard and regional (morphological, lexical, and phonological) forms – in an intricate way to construct a particular persona, acknowledge particular interpersonal relationships, and display their understanding of other aspects of the context such as setting. Furthermore, we have shown that the social “meanings” of situated uses of honorifics and plain forms are potentially as multiple and ambiguous as the orientations, or ideologically based stances, toward normative usage that individual speakers articulated in the preceding section.

Through our analyses in the two chapters in Part II, we have considered the relationship between the meanings of honorifics at the macrosociological level and the meaning of honorifics at the micro-interactional level. As we saw in Chapter 3, the dominant ideology of Japanese honorifics has been tirelessly constructed and reconstructed in modern Japan through a variety of avenues, which has led to a common belief that honorifics are an essential element of the Japanese language and culture and that there are shared norms of honorific use that native speakers are expected to follow in order to be competent members of society.

The data presented in Chapter 4, however, have demonstrated that the notion of “shared” norms cannot be taken for granted. The linkage between honorifics and their stereotypical social meanings is in fact in constant negotiation with a wide variety of contingencies in actual practice, and this in turn requires us to recognize the polyindexical nature of honorific forms. The perpetual suppression (or erasure, see Irvine and Gal Reference Irvine, Gal and Kroskrity2000) of this polyindexicality in the dominant discourse is, we argue, one of the sources of speaker anxiety about the tricky art of evidencing their membership in the Japanese communitas in speech, since their real experiences with honorifics does not and in principle cannot match the standards of “correctness” set out in the dominant story, given the erasure of the multiply indexical nature of the empirical use of honorifics from the official narrative record. Our findings in Part II demonstrate that while the stereotypical meaning (Agha Reference Agha2007) of a certain linguistic form endorsed by the dominant discourse can be so powerful that it may appear natural and the only possibility, it is in fact only one of the many potential meanings in its indexical field (Eckert Reference Eckert2008). It is important to consider the diverse and complex processes by which ideologically dominant and alternative (or even oppositional) meanings are negotiated in interaction in the everyday social life of speakers in specific historical contexts.