JAPANESE DISCOURSE PARTICLES IN USE by EMMA C. PUCKETT A THESIS Presented to the Department of East Asian Languages and Literatures and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Master of Arts September 2014 ii THESIS APPROVAL PAGE Student: Emma C. Puckett Title: Japanese Discourse Particles in Use This thesis has been accepted and approved in partial fulfillment of the requirements for the Master of Arts degree in the Department of East Asian Languages and Literatures by: Kaori Idemaru Chairperson Lucien Brown Member and J. Andrew Berglund Dean of the Graduate School Original approval signatures are on file with the University of Oregon Graduate School. Degree awarded September 2014 iii © 2014 Emma C. Puckett iv THESIS ABSTRACT Emma C. Puckett Master of Arts Department of East Asian Languages and Literatures September 2014 Title: Japanese Discourse Particles in Use One of the distinctive features of Japanese is the presence of discourse particles. The only way to truly resolve what these particles mean and how they are used is to examine them in use and to study the entire system of many particles. In order to add to the attempt to do this by providing more data for the study of particles, this exploratory study examines a large corpus of naturally occurring telephone conversations in the online CallFriend Japanese corpus and conversations taken from the Japanese TV drama “HERO.” This analysis of both naturally occurring and scripted data will lead to further understanding both of actual patterns of particle use by real speakers and the language ideology that informs the usage of language created for a specific purpose. The results suggest that using this method of analysis shows a great deal not only about how particles are used in discourse strategies and in showing a speaker’s affective commitment to the conversation but about the semantics involved in their use. v CURRICULUM VITAE NAME OF AUTHOR: Emma C. Puckett GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene Willamette University, Salem Oregon Tokyo International University, Kawagoe, Japan DEGREES AWARDED: Master of Arts in East Asian Languages and Literatures, 2014, University of Oregon Bachelor of Arts in Japanese Studies, 2009, Willamette University AREAS OF SPECIAL INTEREST: Japanese Linguistics – Pragmatics Japanese Linguistics – Semantics Sociolinguistics Japanese History PROFESSIONAL EXPERIENCE: Graduate Teaching Fellow, University of Oregon, 2012-1014 Legal Secretary, Matthew Dubin, 2009-2010 GRANTS, AWARDS, AND HONORS: Graduate Teaching Fellow, East Asian Languages and Literatures, University of Oregon, 2012-present Senior Thesis Award, Department of Japanese Studies, Willamette University, 2009 Japan Student Services Organization Scholarship, Japan Student Services Organization Scholarship, 2007-2008 National Merit Scholarship, National Merit Scholarship Corporation, 2005-2009 vi Robert C. Byrd Scholarship, Federal Formula Grant, 2005-2009 The Phi Beta Kappa Society, 2009-present National Society of Collegiate Scholars, 2006-present Alpha Lambda Delta, 2005-2006 Cum Laude, Willamette University, 2009 vii TABLE OF CONTENTS Chapter Page I. INTRODUCTION .................................................................................................... 1 Forms of Discourse Particles ................................................................................. 2 Approaches Taken in the Study of Discourse Particles ......................................... 4 Modality Studies in Japan .............................................................................. 5 Information Territory Approaches ................................................................. 8 Approaches to Particles as Indexing Affect or Stance ................................... 10 Conversation Analysis Approaches ............................................................... 13 Studies on Yo ................................................................................................. 14 Studies on Other Particles .............................................................................. 16 Problems and Current Study .................................................................................. 19 II. ANALYSIS OF NATURALLY OCCURRING DATA ......................................... 24 Methodology .......................................................................................................... 24 Results .................................................................................................................... 25 Overall Particle Usage by Each Speaker and Particle Type ............................ 25 Particle Usage by Gender ................................................................................. 34 Particle Usage by Politeness ............................................................................ 36 Particle Usage by Utterance Position ............................................................... 38 Particle Usage by Conversation Makeup ......................................................... 40 Particle Usage by Response ............................................................................. 42 Particle Usage by Utterance Type .................................................................... 44 Response Type by Particle ............................................................................... 54 viii Chapter Page Discussion .............................................................................................................. 58 III. ANALYSIS OF SCRIPTED DATA ...................................................................... 65 Methodology .......................................................................................................... 66 Results .................................................................................................................... 67 Overall Particle Usage by Each Speaker and Particle Type ............................ 67 Particle Usage by Gender ................................................................................. 86 Particle Usage by Politeness ............................................................................ 87 Particle Usage by Utterance Position ............................................................... 90 Particle Usage by Response ............................................................................. 91 Particle Usage by Utterance Type .................................................................... 93 Response Type by Particle ............................................................................... 95 Discussion .............................................................................................................. 96 IV. GENERAL DISCUSSION .................................................................................... 101 V. CONCLUSION ...................................................................................................... 115 REFERENCES CITED ................................................................................................ 118 ix LIST OF FIGURES Figure Page 2.1. Overall particle usage by speaker ....................................................................... 26 2.2. Overall particle usage by speaker, percentage of total utterances ...................... 27 2.3. Overall raw frequency of each particle ............................................................... 34 2.4. Percentage occurrence of particles by gender ..................................................... 34 2.5. Politeness level of particle-containing utterances, percentages .......................... 37 2.6. Utterance position of particles ............................................................................ 38 2.7. Gender makeup of conversations ........................................................................ 41 2.8. Response to particle received.............................................................................. 43 2.9. Percentage of particles by types of utterances .................................................... 45 2.10. Responses to particles by type, percentage ......................................................... 56 3.1. Overall raw frequency of each particle ............................................................... 68 3.2. Overall particle usage by speaker ....................................................................... 69 3.3. Overall particle usage by speaker, percentage of total utterances ...................... 70 3.4. Percentage occurrence of particles by gender ..................................................... 86 3.5. Politeness level of particle-containing utterances, percentages .......................... 88 3.6. Utterance position of particles ............................................................................ 90 3.7. Response to particle received.............................................................................. 92 3.8. Percentage of particles by types of utterances .................................................... 94 3.9 Responses to particles by type, percentage ......................................................... 95 x LIST OF TABLES Table Page 2.1. CallFriend conversations analyzed ....................................................................... 25 2.2. Dominant vs. Non-Dominant Speakers ................................................................ 30 2.3. Different conversational pattern ........................................................................... 31 2.4. Number and type of particle by speaker ............................................................... 33 2.5. Percentage of types of utterances by particle ........................................................ 44 2.6. Types of responses to particles ............................................................................. 55 3.1. Episodes of HERO analyzed in this study ............................................................ 67 3.2. Main characters in HERO ..................................................................................... 67 3.3. Particles by speaker, percentages .......................................................................... 70 1 CHAPTER I INTRODUCTION One of the distinctive features of the Japanese language is the presence of particles, which have an important role in the formation of the Japanese phrase or utterance. Japanese particles have been defined in many different categories (Makino & Tsutsui, 1986; Maynard, 1993; Shibatani, 1990), but they can overall be split into two distinct types. The first of these are case particles, used to mark the syntactical function of various parts of a sentence, such as the particles ga, which marks a subject, or o, which marks a direct object. It has been argued that one of the effects of having this type of particle is to enable the sentence structure of Japanese to be relatively variable, as the particles mark what syntactic function the various words are serving in the sentence. The other type is that of discourse particles, such as ne, often translated as “isn’t it?” or yo, often described as a “verbal exclamation point.” The functions of these discourse particles are slightly more grammatically and semantically ambiguous and thus controversial in nature. In general they are thought to add emotional content to a sentence, specific to these particles, or to perform functions of sequential organization and conversation management, such as signaling opportunities for turn-taking or transition (Maynard, 1993; Saigo, 2006). However, what sort of emotional content they do carry, and their precise meanings, or if they even have any, has been a matter of debate. Though there have been a variety of explanations offered, there is as yet no firm consensus in the field, as the sheer number of new studies proposing alternate analyses attests. 2 Forms of Discourse Particles Discourse particles tend to be appended to the ends of phrases. They are, in fact, often called sentence-final particles, though many of them can in fact occur sentence- internally, initially, or singly. How often they tend to occur in each position seems to differ according to particle. The most common place for these particles to appear, however, is at the end of a phrase or an utterance. In Languages of Japan, Shibatani (1990) classifies discourse particles into two different types: - Interjunctive particles, which occur freely within a clause and whose presence or absence does not affect sentence formation - Final particles, which occur in sentence-final position. However, as Maynard (1993) points out, the same particles may occur both sentence finally and sentence internally, so they could easily be classified together as well as separately. In Japanese, the term shuujoshi (final particle) is used to refer to particles occurring sentence-finally, while particles that occur sentence internally are referred to as kantoojoshi (insertion particle). Some particles can appear in both positions, while some to be more restricted to the final position. For this reason, and the strongly interactional and discourse-central nature of the particles in question, I have chosen to refer to the particles in question as discourse particles, rather than final particles, and to look at all places in the phrase where these particles might appear, including, but not limited to, the ends of sentences. Discourse particles can also appear as combinations, rather than singly, such as the particles yo and ne appearing together as yo ne. Some of the particles can occur in combination with others, but not with all others—only some particular combinations seem to be possible. It also seems that the order of the combination is not free—when 3 two particles can appear in a combination, one must always precede the other, and the other order is not found. For example, yo ne is only found as yo and then ne, and does not ever appear as ne yo. This implies a certain order or hierarchy of various particles in phrases. A list of particles typically considered to be discourse particles in Tokyo Standard Japanese follows below (Shibatani, 1990; Makino & Tsutsui, 1986; Maynard, 1993; Saigo, 2006; Squires, 1994; Suzuki, 1990): ne, yo, na, sa, wa, zo, ze, yo ne, wa ne, yo na, wa yo Basic definitions of the particles above typically follow similar formats to those found in Makino and Tsutsui’s definitions of Japanese grammar, which follow below: ne—“a sentence-final particle that indicates the speaker’s request for confirmation or agreement from the hearer about some shared knowledge” (Makino & Tsutsui, 1986, p. 286) yo—“a sentence-final particle that indicates the speaker’s (fairly) strong conviction or assertion about something that is assumed to be known only to him” (Makino & Tsutsui, 1986, p. 543) na—“an exclamatory sentential particle which is used in informal male speech” (Makino & Tsutsui, 1995, p. 193), wa ne is described as the female equivalent of na sa—“a sentence-final particle used in highly informal speech by male speakers to express different degrees of assertion ranging from a light touch comment up to opposition or imposition” (Makino & Tsutsui, 1995, p. 358) wa—“a sentence particle used in weak assertive or volitional sentences by a female speaker” (Makino & Tsutsui, 1986, p. 520) 4 zo—“a sentence-final particle that emphasizes a male speaker’s emotion about something in his monologue or his strong desire to draw someone else’s attention” (Makino & Tsutsui, 1995, p. 609) The compound particles are generally not defined separately. It is generally stated in the literature (Saigo, 2006; Katagiri, 2007; Lee, 2007) that discourse particles only occur in spoken or written interaction where two parties are communicating, such as conversations, personal letters, and emails, but not in speeches, newspapers, essays, business letters, and so forth. Though Hasegawa (2010) showed that some discourse particles (namely ne), are commonly used in soliloquy, it is indeed generally the case that these particles are used either in interaction, or to give the impression of interactive speech. It is also worth noting (as Saigo does in the above 2006 study) that without the use of discourse particles, interactive speech of this nature in Japanese can sound dry, harsh, overly formal, rude, or unnatural. Thus, discourse particles are an important, even mandatory, aspect of natural speech in Japanese. Approaches Taken in the Study of Discourse Particles The approaches taken throughout the literature on particles in terms of analysis of the function and meaning can be organized into four broad categories. These can be separated into theories that see discourse particles as having some epistemic or modality- oriented function, those that deal with particles as indexing information agreement or territory, those that see particles as indexing affect or position, either toward the utterance, the information, or the other speaker, and those that take a discourse analysis perspective, or postulate that particles serve a communicative function as signals in turn- 5 taking or help to hold or relinquish ground in a conversation. This paper will take a position combining the third and fourth view. Modality Studies in Japan The use of the forms called te-ni-o-ha has been studied in Japan since Teniha Daigaishoo (ca. 1200) was written by Fujiwara no Teika, and began to be studied in depth in the Edo Period by scholars such as Fujitani Nariakira. Fujiwara categorized suffixes of verbals and adjectivals, particles, auxiliary verbs and conjunctions as te-ni-o- ha. Fujitani was concerned with understanding how an individual voice is represented in language, and thus found the categories he termed kazashi (pronouns, adverbs, conjunctions, exclamations, affixes) and ayui (auxiliary verbs, particles, suffixes) most important for his research. Similar concepts were used by Motoori Norinaga in his Te-ni- o-ha Himokagami (1771). Fujitani’s work influenced that of Suzuki Akira, who in his Gengyo Shishuron (1824), generally grouped words into two large categories, shi (referential words) and te-ni-o-ha (particles). He describes te-ni-o-ha as having no referential function, representing voice, “they are voices from the heart and attached to shi, like strings that connect precious beads, like hands that use or operate the containers, and that without shi they have nothing to be attached to” (Suzuki, 1979, p. 23-24). In many cases, in Japanese linguistics particles have been considered an aspect of modality. Modality studies, or chinjutsu-ron, in Japanese, gained in popularity in Japanese linguistics during the 1990s, especially in the work of Masuoka Takashi and Nitta Yoshio. Modality as defined by Masuoka deals with subjective sentence elements: “Proposition and modality are the two big elements that make up a sentence. I define them as the element that expresses objective facts, and the element that expresses 6 subjective judgments and attitudes, respectively” (Masuoka, 1999, p. 46). Nitta has a similar view, though he argues that modality does not necessarily modify the proposition, but can also modify the utterance, or something else entirely. He defined modality as “the linguistic expression of the speaker’s psychological attitude towards the verbalized state of affairs or towards the utterance the communication itself at the time of speech” (Nitta, 2000, p. 81). This view of modality is largely based on the work of Tokieda in the 1930s and after. Tokieda divided all morphemes into ‘objective’ (shi, or content words), and ‘subjective’ (ji, or function words), with the ji expressing the speaker’s judgments and feelings, expressing the “speaking self” in the sentence. As Narrog (2009) points out, “it is the illocutionary force-modulating sentence-final particles which appear to be grammatically more salient and to which definitions of Japanese modality apply best. Thus, there is a clear discrepancy between the focal points of the modariti concept in Japanese and modality in general linguistics” (p. 23). Despite this, some Western scholars have held similar views on modality, such as Lyons (1977), who defines modality as “the speaker’s opinion or attitude towards the proposition that the sentence expresses or the situation that the proposition describes” (p. 425) and Benveniste (1971), who describes modality as “devices suited to characterize the attitude of the speaker with respect to the statement he is making” (p. 229). Maynard (1993) works off of both Japanese and Western scholars when she creates her theory of Discourse Modality, which includes a theory of discourse particles and how they are used—she defines Discourse Modality as conveying “the speaker’s subjective emotional, mental or psychological attitude toward the message content, the speech act itself or toward his or her interlocutors in discourse” (p. 38). She goes on to state that “Discourse modality operates 7 to define and to foreground certain ways of interpreting the propositional content in discourse; it directly expresses the speaking self’s personal voice on the basis of which the utterance is intended to be meaningfully interpreted” (38-39). Maynard (1993) considers discourse particles “Discourse Modality Indicators” that express subjective and emotional aspects and deal with information status. However, there are major reasons to consider both modality in Japanese and the analysis of discourse particles from a different perspective rather than adopting this view, and to consider discourse particles entirely apart from concepts of modality. One reason is that Japanese does in fact possess markers of more traditionally defined modality from the Western perspective, such as markers of deontic and epistemic modality that deal with possibility, necessity, and permission such as may, might, should, would, and must, as in the forms beki and hazu (meaning should, or must). The question becomes that if Japanese does possess this category as defined generally in linguistics, why broaden or alter the concept of modality to deal with a different set of concepts and linguistic functions, rather than giving these their own specific classifications? This is especially important as using the chinjustsu-ron model of modality will make it more difficult to compare research with other work in linguistics, both on particles and with other topics. The idea of modality as described by Masuoka, Nitta, and Maynard conflates many categories that are treated entirely separately in general linguistics (such as modality, politeness, tense, information structure, and illocutionary force modulation, among others). While it may be difficult to analyze some of the forms in question without looking at multiple fields, looking at them as an intersection point of these multiple fields is more likely to prove more useful than considering them together. Also, as Narrog 8 (2005) points out, it is unlikely that a linguistic category or a set of specific forms actually can in fact be defined through a concept like subjectivity or speaker’s attitudes. Crosslinguistically, it appears that speaker’s attitude is expressed throughout the sentence rather than being confined to specific form classes or grammatical categories. It is also somewhat problematic in practice to actually divide linguistic forms into objective and subjective ones. There are many Japanese linguists who have pointed this out. Onoe (2001), for example, argues for a definition of modality based an irrealis vs. realis dichotomy. From this perspective, it is better to pursue the study of Japanese discourse particles outside the theoretical framework of “modality.” One of the drawbacks of the majority of studies on particles up until this point is that most of them have focused on ne and yo in comparison with one another, and largely ignored the other discourse particles in Japanese, such that the theories that have been formed have been based on the functioning of these two particles. For example, Maynard’s (1993) analysis of discourse particles as aspects of Discourse Modality disregards all except ne and yo based on frequency of use. In ignoring the other discourse particles, the possibility that the nature of choice of one particle over another (or over zero particle use) has been overlooked, thus materially influencing the arguments about the meaning of these particles. Information Territory Approaches Another, separate, major explanation of discourse particle usage in Japanese is that first advanced by Akio Kamio’s theory of the territory of information. Kamio (1995) states that the concept of territory is an important one in terms of understanding what takes place in discourse, and as part of his explanation claims that this concept can also 9 explain the usage of ne and yo in Japanese. According to him, this is because ne is used when the information being discussed is at least partly in the hearer’s territory (such as topics pertaining to that person’s personal life or experiences), and that yo is used when the information is in the speaker’s territory (such as that person’s personal life, experiences, and so on). For example, when talking about a friend’s sister’s birthday, if the speaker believes the information to be in the addressee’s territory rather than his or her own, he or she will use ne. (1) Kinako-san no tanjoubi wa futsuka desu ne. Kinako HON POSS birthday TOP the second COP ne Kinako’s birthday is the second, right? However, if it is the speaker’s own sister in question, or if the speaker for some other reason believes the information to be in his own territory and not the hearer’s, he will use yo. (2) Kinako no tanjoubi wa futsuka desu yo. Kinako POSS birthday TOP the second COP yo Kinako’s birthday is the second! However, there are many aspects of the usage of these particles that are left unexplained by this theory. For one thing, it does not explain why sometimes when these requirements are fulfilled, ne and yo are used, as in the examples above, and sometimes when these requirements are fulfilled, they are not utilized. Not every statement that is in one’s own information territory takes yo, for example—it would be quite acceptable to state the above sentence about one’s own sister as: 10 (3) Kinako no tanjoubi wa futsuka desu. Kinako POSS birthday TOP the second COP Kinako’s birthday is the second. Thus this theory does not explain why at times yo or ne is used, and not at other times. It also provides no explanation for the use of the myriad of other discourse particles, such as na, sa, or wa, or how they would fit into such a system. Thus it seems logical to conclude that there is something more than the concept of information territory that governs not only the use of these two particles, but the rest of the discourse particles as well, as they presumably form some sort of a system together, though information territory may indeed be a factor. Approaches to Particles as Indexing Affect or Stance Katagiri (2007) analyzes the use of the Japanese particles ne and yo based on the functions they serve in dialogue, arguing that they chiefly index the attitudes of a speaker toward the content of preceding utterances, and that the choice of sentence final particle along with intonational patterns makes up the choice of persuasion strategies for achieving shared informational and intentional states in joint activity with interlocutors. Katagiri claims that particles serve a wide variety of communicative functions, including assertion, question, confirmation, assent, inhibition, and exclamation, but focuses on their usage in terms of building collaboration through dialogues and joint construction of interaction, citing the lack of usage of such particles in writing or formal monologue. Through Katagiri’s analysis, yo presents the content of the preceding utterance as something the speaker has accepted, while ne presents the propositional content of the preceding utterance as something the speaker has not yet wholeheartedly accepted. This 11 analysis does not cover the usage of particles utterance medially or initially, or the use of certain particles on their own—or why only certain particles seem to be able to be utilized on their own and some do not. Katagiri argues that all discourse particles are used to persuade the hearer of something, which does not seem to cover all usages of such particles. It does seem that particles have an underlying purpose of indexing the attitudes of a speaker toward the content of their own utterances, as was also noted by those studying modality above, as a major part of their analysis of that concept. Katagiri is also right to point out that particles serve a wide variety of communicative functions, and by focusing on the underlying motivations for use rather than pure analysis of discourse seems to better be able to explain the appearances and use of multiple items in the discourse particle system. However, it is not entirely true that particles are not used in written Japanese, though they are used in written Japanese to give a sense of conversationality, as observed by Kataoka (1995) in a paper on the use of particles to convey affect in Japanese women’s letter writing. Similarly, Lee (2007) argues that the particles ne and yo can be analyzed using the theoretical framework of involvement. These particles would thus be seen as being interactive in nature, signaling the speaker’s attitude toward the utterance they are attached to in a way that is meant to invite the involvement of the conversation partner. However, Lee states that the difference between them is that they do so in different ways—that ne accomplishes this by aligning the speaker and the conversation partner with one another’s views (in a manner that the Lee terms “incorporative”), while yo does it by enhancing the speaker’s position as deliverer of the utterance, at the same time showing the speaker’s expectation of a response from the partner (“monopolistic”). Lee 12 argues that this analysis of the use of ne and yo is more explanatory than prior attempts using the models of information state, authority regarding the proposition, or affective common ground because it can not only explain the usage of both particles, but also yo’s use as both an intensifier and a softener depending on the utterance. In some ways, this recalls Tanaka’s (1996, 1997) arguments about the interactive nature of ne from the perspective of conversation analysis, discussed below. This argument also has the strong factors of being able to deal with the apparently contradictory nature of yo, both softening and insisting, and allowing for a variety of different types of involvement drawn upon by each separate particle. The problem with this view is that other studies have shown that while yo is not frequently used in self-directed speech, ne is (Hasegawa, 2010). Thus, it cannot solely be involvement with another speaker that determines the usage of discourse particles in Japanese. Hasegawa studied the use of particles in soliloquy, by inviting her subjects to her office, telling them to talk out loud as much as possible, and leaving them there for an hour while tape recording them. She collected a great deal of self-directed speech and found that the discourse particle ne occurred in 15% of the total self-directed utterances. She argued that this shows ne being used to monitor and control the speaker’s internal information processing, and that it cannot be entirely explained if the essential function of ne is characterized by the speaker’s assumption of shared knowledge with the addressee, the interlocutors’ information territories, affection common ground between them, or the coordination of dialogue. This suggests that a theory of usage for these particles based solely on their roles in terms of interaction in dialogue may not be sufficient to explain the ways in which they occur in the language overall. Of course, it could be argued that 13 the use of particles creates an impression of dialogue, or is prototypically used for interaction, but their use in self-directed speech is still an important finding that needs to be taken into account, especially as Hasegawa has the alternate interpretation of cognitive processing. Conversation Analysis Approaches Another major method used in studying discourse particles has been that of conversation analysis. An example of this can be seen in Hiroko Tanaka’s work (1996, 1997), which points out that discourse particles serve important roles in terms of organizing speech and sequential turn-taking in conversation. Her 1997 paper focuses chiefly on the usages of ne, arguing that it acts as a pivotal device used for turn and topic management in natural conversation, used to mark turn-entry points, acknowledgement- relevance places, possible transition-relevance places, and topic changes, depending on its position in the turn (initial, medial, or final). A major part of this argument is the importance of intersubjectivity in the ways in which ne is utilized. As Tanaka describes it, “the sequential positioning of ne in the interstices of parties’ talk potentially serves to link a current speaker’s action with a preceding or ensuing action by another speaker in multiple ways, and therefore plays a pivotal role in the mutual display of understanding” (p. 1172). This is another helpful look at the usage of ne by pointing out the links ne seems to have to the rest of the discourse. The use of ne thus addresses the understanding of the participants of a conversation as well as the coordinating of talk, either assuming understanding, appealing for it, or highlighting the lack of it, or, when used alone, as a final check for understanding before abandoning a topic for the next. Tanaka suggests that other discourse particles share some of these turn-management functions with ne, as 14 they often occur finally and can all be used to signal that the speaker’s turn is over, especially, Tanaka argues, ones such as yo or zo which she claims are more firmly fixed at the end of utterances than ne. Similarly, she claims that sa can be employed similarly to ne to solicit acknowledgements in turn-internal positions. However, as Tanaka’s research and analysis are focused on ne, there is no further exploration of the other particles. These previous studies have generally associated particles with the usage of the plain form rather the polite in Japanese, as in Cook (2002), Dunn’s examination of style shifting in a formal setting at a university club (1999), and Masuda’s article on particle usage in peer conversations (2007). Studies on Yo In general, studies on yo have been similar in tactics to studies on ne. Uyeno (1971) and Tsuchihashi (1983) take the view that particles encode speaker’s attitude toward the illocutionary force of the utterance, and that yo in particular expresses the speaker’s insistence when stating, ordering, asking or suggesting, thus providing extra impact. Uyeno observes that yo can have a softening effect as well, but does not explain how this can happen. Kendall (1985) describes particles as indexing the speaker’s attitude to the proposition expressed, and yo in particular as showing the degree of the speaker’s belief about the truth-conditional content or the strength or illocutionary force of the utterance, as well as a willingness to be held accountable to the content of the utterance. Part of the problem with this view is that yo as well as other particles can be used in irony, or when the truthfulness of the propositional content is not known, as well as with counterfactual statements, and that particles seem to be able to index more than the speakers’ responses to the simple facts of the utterance, such as general mood or 15 opinion toward the speaker. However, the description of yo as showing the degree of emotion or belief behind an utterance seems to fit with other descriptions of yo that have been used. Ooso (1986), Cheng (1987), and Masuoka (1991) argue that particles carry the speaker’s assessment of the hearer’s knowledge of the information communicated— i.e., that yo conveys the speaker’s judgment that the hearer has not yet had or been aware of the information conveyed. This is also Maynard’s view. Maynard (1993) argues that the particles ne and yo have a focusing/defocusing effect, such that yo directs the hearer to focus more on the information, while ne defocuses the informational content and leads the hearer to focus on communicative intent. This does not entirely explain how yo can be appended to some questions but not others (i.e., to less formal questions) and to indirect speech, or why it is sometimes taken as softening an utterance. However, the idea of yo focusing attention on truth value and ne focusing attention on communication could explain many of the usages of ne and yo in a way that does allow for their differences and preferences and may also be a useful lens. Matsui’s (2000) study on yo argues that it encodes the speaker’s desire that his or her informative intention be fully recognized by the hearer, which, in turn, functions as an explicit guarantee that the information in the statement is relevant, and that this is procedurally encoded, rather than contained conceptually in the particle itself. In short, the argument is that when the particle is appended to an utterance, it leads the hearer to seek greater contextual effects from the utterance than he or she would otherwise, that in return for putting in the effort to process the utterance, the hearer is guaranteed greater contextual effects and relevance—this study looks at the use of yo through the lens of relevance theory (Sperber & Wilson, 1986). However, Matsui agrees that the particle yo 16 communicates the speaker’s attitude to the information conveyed by an utterance and how it should be taken by the hearer (and implicitly argues this for all particles). In short, Matsui’s argument is that yo either indexes that the contextual effects of the utterance it is appended to are stronger than the hearer would have otherwise expected, or that there are additional contextual implications, and that in all cases the addition of a particle indicates the contextual effects of the utterance are greater than the hearer would have expected for the same utterance without the particle. This explanation has promise, as it offers a flexible explanation for particle use that allows for the multiple uses observed by researchers. It is extremely general, but seems as if it could be applied to any of the discourse particles in question and by doing so, further refined, especially by describing how differences in the particles are perceived and derived by listeners, as well as users. Studies on Other Particles There have been very few studies done on discourse particles other than ne and yo, despite the fact that there are a great many other particles in quite commonplace use. Suzuki (1990) examined the usage of na and sa in Japanese gossip discourse and concluded from qualitative analysis that na is a “contrastive” particle. The conclusion was that na has the effect of contrasting one’s statement without imposing it on the listener, while sa has the effect of “insisting,” in which the speaker attempts to convince the listener of the claim being made. For example, Suzuki provides an example of one speaker using sa repeatedly in the same utterance, trying to make a point about the topic the speakers are gossiping about. 17 (4) Sou iu rifujin n ate o tsukau tte iu no wa sa, That [the husband] did such an unreasonable thing sa Datte sonoo gakkoo o tooshite saa, Going through the school saa Koo atsuryoku o kakeru tte iu no wa sa . . . Putting pressure on sa (p. 316-317). Suzuki argues that sa, when used in this way, signals, “keep paying attention to the following,” or can be used to add a flavor of “obviousness” to statements. An example given for na is: (5) Ano hito ni wa sootoo sekinin ga atta to omou naa. [I] think that guy was pretty much to blame naa. (p. 318). Suzuki argues that this is contrasted with sa because it is used to offer a personal assessment that strongly emphasizes the speaker’s personal attitude toward the utterance. Squires (1994) also conducted an analysis of the particle sa, contrasting it with ne, using the framework of discourse analysis. His conclusion was as follows: “The Japanese particle sa effects a personal view on the information conveyed in the utterance. This use of sa contrasts with the use of ne which speakers use to create an empathic common ground with the hearer regardless of whether or not the information contained in the utterance is exclusively held by either participant” (Squires, 1994, p. 1). This argument in some ways recalls Kamio’s territory of information theory, mentioned above, though it does not in the end subscribe to it. Squires claims that ne is used to create empathy no matter what status the information holds or whose territory it is in, but that sa contrasts with that, carrying a personal view on whatever the information is, and thus marking it as strongly in the speaker’s territory alone. Nakamura (2013) points out that wa is used a great deal in the translation of foreign works into Japanese, when 18 translating the dialogue of the women, and sa seems to be used a great deal when translating the speech of young “cool” American men such as the main characters of the American television program 90210, seemingly to add a certain sense of “American coolness” to their dialogue. This may suggest a sense of coolness associated with the usage of sa in the minds of some, if it is used in this way in translations. Saigo’s dissertation (2006) examines the pragmatic properties and sequential functions of ne, yo, and yone from a conversational analysis perspective. A major part of Saigo’s argument is that zero, or the non-use of particles, must also be considered in terms of the analysis of the use of discourse particles, especially in terms of when discourse or discourse particles must be used in order to sound natural in Japanese, and the failure of non-native speakers to do so appropriately. However, the chief portion of his argument deals with the properties of particles in signaling whether the content of the argument should be treated as grounded, or be seen as to be grounded. He also fails to explain yo ne except that arguing that responding to ne is obligatory, but that a good conversationalist might also respond to the force of yo, while zero particle usage is when the speaker is giving no indication how the figure emerging in the discussion is to be grounded and thus has a topic closing function. Though it is important to take zero particle usage into account in any theory of particle usage, and it is also important to consider how the failure to use particles in interaction produces an unnatural effect, it seems from prior studies that there is a great deal more to the particles in question than that, and this analysis also fails to explain why yo ne should exist in any way that is not covered by his explanations of the two separate particles themselves. 19 Though there are a great many other studies on ne and yo, the studies mentioned above represent some of the only research available on na and sa. There are other particles also used in Japanese speech, but the literature on them is also rather thin on the ground. For example, in general when the particle wa is mentioned, it is described as occurring in women’s speech (Okamoto & Sato, 1992) or in an idealized and constructed image of Japanese women’s speech (Inoue, 2006). It has been argued to represent “gentleness” and/or “emotiveness” (Shibamoto, 1985). However, there is also some evidence that men also use this particle, which begs the question of how and when this occurs, and why it is typically seen as a woman-specific marker of speech. Problems and Current Study When taken together, analysis of the discourse particles in the previous studies seems to point to several major tendencies—that particles convey something concerning the speaker’s attitude toward the information in the utterance, his or her assessment of the hearer’s knowledge of the information, and his or her confidence level in the utterance. They thus depend on the speaker’s judgment of the proposition and relationship to the utterance. When this is looked at along with the argument that these particles are also or partly utilized as discourse cues, it is clear that the fundamental meaning of the particles is pragmatic in nature and must be interpreted in the current discourse context, at least to some extent. Thus their meaning is “procedural” in Matsui’s (2000) term. However, Hasegawa’s findings suggest that this usage may be possible even when a speaker is on his or her own. This may imply that the particles in question have some semantic values as well, or that their discourse functions are not solely reliant on an interlocutor. The relationship among the speaker (or his/her attitude), listener (or the speaker’s assessment 20 of her knowledge/attitude), and the information conveyed that define the choice of particle has been described using such concepts as territory of information, involvement, empathy, alignment, degree of belief, and relevance. Perhaps all of these have some part in the meaning and usage of the particles in question. The findings of the past studies are not necessarily contradictory with each other, and most of them seem to capture some facet of the meaning of the particles under examination. Rather, they shed light on the meaning of these particles from different vantage points, illuminating that the meaning and usage of discourse particles is complex and not easily defined under one single theoretical framework, at least so far. It is thus useful to look at numerous theories and frameworks when studying these particles. As far as yo and ne are concerned, the research thus far suggest that yo is used when the speaker is more committed to the utterance in question, or has a strong belief, and ne is used when the speaker wishes to obtain agreement with their utterance. However, it has become clear that the use of ne and yo does not ultimately hinge on the pure sense of “territory of information” as defined by Kamio (1993), but as many researchers pointed out it instead hinges rather on the speaker’s belief or feeling of ownership of the information or assessment of the listener’s knowledge and level of agreement with the speaker (Katagiri, 2007; Kendall, 1985; Lee, 2007; Tsuchihashi, 1983; Uyeno, 1971; Matsui, 2000), which may then lead to the role particles play as discourse markers that were noted by such researchers as Tanaka (1996, 1997) and Saigo (2006). Perhaps this, that the criteria are not the truth value or speakers’ possession of the information of the utterance but the speaker’s subjective judgment, gives room for some of the particles to be used creatively – in irony and in monologues. The wealth of analysis on ne and yo 21 should aid the analysis and understanding of the meanings and functions of other particles. The work on particles from the perspective of Conversation Analysis sheds still further light on the issues of particle meaning and use: these studies revealed particles are used as devices to manage conversation (e.g., Tanaka, 1996; Tanaka, 1997; Saigo, 2006; Suzuki, 1990). Broadly these studies seem to conclude that the particles as a class have the function of conversation management devices of various types and in various ways. While that might be the tendency, we can presume that different particles used in conversations are not always interchangeable. In fact, different particles may have different roles in managing conversations, due to their pragmatic/semantic differences, centering around the various theories discussed earlier. Thus, the question becomes how do the various particles differ from one another in terms of the conversation management functions they perform, and how do these relate to their pragmatic/semantic sense. My current work builds on the knowledge gained in these previous studies, partly by attempting to work them together, and attempts to contribute to the further understanding of the use of particles. One of the problems with previous studies has been the most obvious problem: there is next to no study that looks at more than three particles at a time. Even those that claim to provide an overall theory of particle usage tend to work solely with ne and yo, or perhaps one or two other particles. The vast majority of the studies above have also been based on intuitive judgments by the authors rather than working off of natural speech. Those that use natural speech have been limited, of course, by sample size, so further research in this area is always helpful. 22 It is my view that the only way to begin to truly resolve what these particles mean and how they are used is to look at exactly that—to examine them in use. The overall lack of examination of more of the particles used in Japanese in contrast with those most commonly analyzed is one of the major reasons why no scholars who have delved into this topic have been able to achieve a satisfactory consensus as to the basis of use for these particles. The other problem, of course, is the widespread use, until recently, of introspective rather than naturally occurring data upon which to base a theory. Thus, an exploratory study into the usage of the majority of Japanese discourse particles in use will add greatly to the understanding of these particles in the field. To this purpose and to further the study of Japanese discourse particles in general, I utilized corpus analysis to examine the usage of these particles in naturally occurring conversational data. I also examined the usage of these particles in the script of a Japanese TV drama. Because the particles under investigation occur most frequently in interaction, conversational data is the best source of data for analysis. However, it is also quite important to take into account speakers’ impressions and opinions regarding usage in terms of determining the true meaning of the words in question or formulating a theory of use. Thus, the idealized, yet not too heavily stylized, dialogue available in television dramas can also be a valuable source of data regarding how speakers see language and in what ways it is available to be used. The usage of naturally occurring conversational data is needed because this offers concrete data on how and in what situations these particles are truly used, so that some conclusions on their meaning can be drawn. However, data such as that offered by fictional materials such as TV dramas can still be useful, as this data comes from the minds of native speakers and thus represent an idealized version of 23 language, or how users of the language think the language is used. They can also be examined to see how language is utilized to create specific characterizations or influence audience perception of a character, as well as offering an opportunity to see language utilized in less frequently occurring and more dramatic contexts. Thus, the analysis of fiction offers another way to examine both the perceptions of native speakers and the use of the particles themselves. Due to these considerations, this thesis examined naturally occurring conversational data from the CallFriend corpus available on the website TalkBank (MacWhinney, 2007), as well as fictionalized conversational data from the 2001 Japanese television drama called “HERO.” There is currently no existing study that has examined the usage of more than four discourse particles in Japanese at the same time, or one that uses the combination of methods outlined above. 24 CHAPTER II ANALYSIS OF NATURALLY OCCURRING DATA Methodology In this study, naturally occurring data was analyzed through corpus analysis. This data was drawn from the Japanese language section of the CallFriend corpus housed on TalkBank1, which is made up of telephone conversations between native Japanese speaking friends residing in North America. The transcripts of these conversations are available online. The study analyzed six conversations (Table 2.1), two between one female friend and one male friend, two between two female friends, and two between two male friends. The speakers are all of roughly similar age, from late teens to mid-twenties. These conversations were examined for occurrences of the discourse particles ne, yo, sa, na, yo na, yo ne, zo, ze, wa, wa ne, and wa yo, and these instances recorded, while keeping track of (1) the gender of the speaker and other information available. Each occurrence of conversation particles was then coded in terms of (2) the topic of conversation and (3) the context and topic, (4) type of utterance (e.g. question, statement, etc.), (5) whether the utterance was responded to and (6) how (whether with agreement, disagreement, with a comment such as ii na (“it’s good, I think . . .”) or a response that introduces new information but orients to the past utterance, and so on), (7) the politeness level of utterance (plain or polite), (8), whether the particle was used utterance internally, finally, or alone, and (9) if the same speaker produced another utterance after the one with the particle or not. Both raw counts of particles and percentages with the overall number of utterances for each speaker and conversation are 1 http://talkbank.org/ 25 analyzed. Then the quantitative data described above is presented and analyzed, along with more in depth qualitative analysis of the most interesting features that emerge, and when it is needed. Descriptive statistics were used in the analysis of the results to reveal tends within the data itself, but not inferential statistics, as this study focuses on the trends that emerge in these pieces of data and compares the usage of each particle to the others within the same conversations. Table 2.1. CallFriend conversations analyzed. Conversation name Participants # of Overall Utterances 1773 Female and male 612 Hir Female and male 801 Mie Two female friends 700 Sum Two female friends 1171 Wat Two male friends 1480 4222 Two male friends 864 Results Overall Particle Usage by Each Speaker and Particle Type As stated above, there were twelve speakers, overall, two in each conversation. There were 1753 total particles uttered, out of 5628 utterances, so approximately 32% of utterances contained particles, though many contained more than one particle. The breakdown of overall particle usage by speaker can be seen in Figure 2.1 below. The first six speakers listed are female, the second six male. Figure 2.1. Overall particle usage by speaker. As is apparent from the above figure, particles overall than the others. speakers who are “heavy particle users,” and seven who are not, the heavy users being FE2, MIE, MA2, MA4, and WAT, the firs This suggests that heavy particle use is not a particularly feminine trait, nor particularly masculine. In the below figure appears time in terms of percentages of total utterances pattern emerges, though with slightly different speakers MA4, though with FE2, FE3, and MA3 not too far behind. There is still a noticeable pattern of heavy particle users vs. non 68 201 101 201 0 50 100 150 200 250 300 FE1 FE2 FE3 MIE SAC Particles 26 some of the speakers produced many more Simply looking at the raw counts, there appear to be five t two female speakers and the last three male. the same information as in the above figure, this (Figure 2.2). As can be seen, the same —these are MIE, MA2, and -heavy users visible, however. 87 136 97 118 187 106 274 186 SUM HIR MA1 MA2 MA3 MA4 WAT Figure 2.2. Overall particle usage by speaker, percentage of total utterances. As observed earlier, heavy particle use does not seem to be tied to gender, though of the heaviest particle users amongst to overall have been heavier particle users than females to a heavy particle user may cause a speaker to use more particles than they otherwise over the course of heavy particle user. For example, as can be seen in the conversation below, it is extremely common in the conversation between WAT and MA4 for end nearly every utterance with (6) WAT: amerika no hoo ga ie ga hiroi no ni American houses sure are wide though. MA4: soo da yo ne: That’s true, isn’t it? WAT: u::n Yeah. MA4: demo wakaru yoo na ki ga suru sore wa ore wa hora moo fakkusu: kenyoo de michatta kara 17 32 33 56 0 10 20 30 40 50 60 FE1 FE2 FE3 MIE 27 the speakers analyzed were male, and males seem . It may be the case that might have that particular conversation due to interacting with another the two of them to a particle of some kind, bolded below. na: sa:: 26 25 24 28 42 34 37 25 SAC SUM HIR MA1 MA2 MA3 MA4 WAT some speaking 28 But once I noticed this . . . because I saw that fax machines have more than one use, see . . . WAT: soo suru to: anoo coodo tsuki ni natchau kedo ne When you do that—but it’s attached to the cordless, right? MA4: u:n daitai soo da yo ne Mm, usually it’s like that, isn’t it? WAT: fakkusu kenyoo no coodoresu phones wa yahari nani ka nihon jin ga tsukura nai to deki nai no ja nai↑ → They don’t really use cordless phones with fax capability in Japan, you know? Japanese people can’t use them. MA4: a::↑::→ Hmm? WAT: nani ka yo Something (yo). First, WAT utilizes na when making a statement of his own thoughts and reflections (on the width of American houses), to which MA4 response with an agreement and yo ne. WAT agrees, and MA4 continues on with a statement of his own about the dual capabilities of American fax machines, using sa. WAT responds to this with a comment of his own, using ne, to which MA4 responds with agreement using yo ne again. WAT’s and MA4’s next utterances do not include particles, but when WAT starts talking again, moving toward making his next statement, he utilizes yo. This seems to show a pattern of using particles to respond to the other speaker, and to convey involvement in what the speaker is saying. The usage by one speaker may make it seem that if the other speaker does not use particles in response, that speaker would sound dry or uninvolved in comparison. Thus, one speaker’s frequent particle usage may prime frequent particle usage for a conversation partner, despite the fact that WAT does not in fact seem to be a heavy particle user. However, this did not hold true for all conversations that included heavy particle users—MIE, for example, was in a conversation with SAC, who produced far fewer particles. This may simply be because SAC, unlike, say, MA4, might be a speaker who 29 uses fewer particles in general, and so does not respond to this priming effect. Other factors might also be in play, however. MIE seems to be the more dominant speaker in the conversation—she dominates conversation topics and brings up most of the new ones, as well as speaking more frequently overall. SAC seems uninterested in usurping this general dominance. This tendency can be seen in the exchange below, where SAC simply responds to MIE’s continuing description of her current apartment. (7) MIE: apaato jitai wa ne: So, my apartment situation, you know? SAC: u::n Yes. MIE: hirokute ne: Big, you know? SAC: u:n Yes. MIE: anoo hot_water wa tada dashi ne: Uhh, the hot water is free, you know? SAC: hee:: Mmm. MIE: yasui shi ne: Cheap, too, you know? SAC: u:n u:n u:n u:n Yes, yes, yes, yes. In the conversation between MA4 and WAT, both of them share topics and bring up new ones frequently. Thus, while SAC may not be trying to provide an active conversational position and simply to seem involved in the conversation, MA4 and WAT may be doing just that, with WAT more interested in possibly becoming the driver of the conversation for some turns than SAC is. These two conversations seem to provide one example of how heavy particle use tends to appear in a conversation—when heavy particle use appears with the main, or dominant, speaker (MIE and MA4, in general, in the above conversations). However, 30 while some of these heavy particle users seemed to dominate their respective conversations, others did not, and these heavy particle users appear to fall into two separate types. One of these types (MIE, MA4, and MA2) is the sort that stands out in the percentage data above, leads and seems to dominate the conversation, providing most of the new information and clearly stating the majority of opinions (whether largely unchallenged, as in MIE’s case, or with challenge from the other speaker, as in MA4’s case), with the other speaker often tending to provide commentary on their statements in their utterances (as can be seen clearly in SAC’s responses above). The other type of heavy particle user (FE2 would be an example of this), provides a large number of utterances including particles, but most of these are comments on or spinning off of the other speaker’s utterances, as the other speaker drives the conversation. MIE, MA2, and MA4 can be seen to be the speakers driving their respective conversations through a variety of different analyses. For example, both produce both more utterances overall, and more utterances with particles, than their partners, as can be seen in Table 2.2 below. Table 2.2. Dominant vs. Non-Dominant Speakers Type of utterance (%) that included particle(s) Total utterance Number of utterances including particle(s) % age of utterances including particle(s) New information (%) Comments Response Opinion MIE 362 201 55.5 48 12 6 21 SAC 338 87 25.7 16 19 26 30 MA1 416 118 28.4 38 17 13 14 MA2 449 187 41.6 64 4 8 16 MA4 750 274 36.5 44 26 18 16 WAT 742 186 25.1 30 23 19 15 31 MIE produced 362 utterances while SAC produced 338; 201 of MIE’s utterances contained particles, and 87 of SAC’s, 48 percent of which shared new information in MIE’s case, and 16 percent in SAC’s case, while SAC’s utterances were mostly comments (19 percent), responses (26 percent), or statements of her opinion (30 percent). Similarly, MA4 produced 750 utterances to WAT’s 742, 274 of which contained particles compared to WAT’s 186. The pattern of these responses is not as clear as with MIE and SAC above, however, fitting the description of this conversation as one where WAT is more interested in sharing the floor than SAC is with MIE. MA4’s particle containing utterances involve sharing new information in 44 percent, while WAT shares new information in 30 percent. In the conversation between MA2 and MA1, this pattern is also observed, with MA2 producing 449 utterances to MA1’s 416, 187 of which contained particles to 118, respectively. For MA2, 64 percent of these share new information, as opposed to MA1’s 38 percent sharing new information. For presentation of a differing pattern, however, see Table 2.3. Table 2.3. Different conversational pattern. Type of utterance (%) that included particle(s) Total utterance Number of utterances including particle(s) % age of utterances including particle(s) New information (%) Comments SUM 540 136 55.5 45 9 FE2 631 201 25.7 37 13 Meanwhile, in the conversation between SUM and FE2, even though SUM produced 540 utterances to 631 of FE2’s, 136 and 201 of which contained particles, respectively, 45 percent of these involved sharing new information to 37 percent of 32 FE2’s, and 9 percent were comments on FE2’s utterances, while 13 percent of FE2’s were comments on SUM’s utterances. Moreover, SUM was the one to introduce and maintain the topic of conversation, her dislike of living in Wyoming, while FE2 responded to and commented on this topic without attempting to bring up a new one. In the case of FE2, it can be seen that this style of particle use also involves talking more than the other speaker, or, as in WAT’s case, close to the same amount, even if they do not drive the conversation as much in terms of providing new information or directing the topic under discussion. These observed patterns could imply two different methods of frequent particle usage—one being to use particles frequently to maintain the floor and control of a discussion (e.g. MIE, MA4), and the other to utilize them to show affective interest and participation in a conversation lead by another (e.g. FE2). It is worth noting that though SAC was not a heavy particle user, 45 percent of the utterances containing particles were simple comments or responses of the kawaisou da ne (ah, that’s sad, don’t you think?) or soo da ne (oh, is that right?) variety, in line with this style of particle use. A conflict between analyzing one or the other of these patterns of heavy particle usage could in fact be one of the major reasons for the lack of resolution of most of the prior literature on particles. If the results are viewed particle by particle (Table 2.4), it is apparent that the heavy particle users (i.e. MIE, MA2, MA4) tend to use all particles more heavily overall, rather than relying on extreme use of only one or two (though ne is by far the most frequent particle used), so the pattern is one that holds overall, rather than being focused on certain particles in particular, or changing in terms of the two discourse styles of 33 particle usage mentioned above. In other words, conversation initiators (e.g. MIE, MA4), use the same types of particles with similar frequency patterns as conversation followers (e.g. FE2, WAT, SAC). This may suggest that even different particles may have quite similar roles in conversation management or be available for similar operations. Table 2.4. Number and type of particle by speaker. ne wane yone na sa Wa Yo total FE1 28 0 0 11 7 1 21 68 FE2 135 7 3 23 1 10 22 201 FE3 36 2 3 8 22 2 28 101 MIE 92 5 14 24 33 11 22 201 SAC 42 0 5 8 12 1 19 87 SUM 63 4 20 19 1 4 25 136 HIR 51 0 11 6 7 1 21 97 MA1 54 3 11 17 11 3 19 118 MA2 83 2 13 16 36 1 36 187 MA3 29 1 1 10 17 2 46 106 MA4 119 3 15 34 71 5 27 274 WAT 55 9 12 39 9 6 56 186 Total 787 36 108 215 227 47 342 1762 The overall raw frequency of each particle in the data is shown in Figure 2.3 below. Yo na, na yo, wa yo, and zo were all used negligibly in the data, appearing 7, 1, 3, and 4 times, respectively. Ze did not appear at all. As can be seen, ne was overwhelmingly the most common particle used, followed by yo, and then na and sa somewhat distantly. From this it seems that the combined particles are much less common, but also that they reflect the frequency of their component elements—yo ne being by far the most common combined particle, and yo and ne being the most common particles overall, and so on. Figure 2.3. Overall raw frequency of each particle. Female speakers produced 794 particles, while male speakers produced 968. Each individual particle’s percentage (Figure 2.4) below. Figure 2.4. Percentage occurrence of particles by gender. 788 36 108 0 100 200 300 400 500 600 700 800 900 ne wane yone particles 15 1 2 13 1 2 0 2 4 6 8 10 12 14 16 ne wane yone 34 Particle Usage by Gender occurrence by gender can be seen in the figure 215 227 37 342 na sa wa yo 4 3 1 54 5 0.3 7 na sa wa yo f m 35 The occurrences of the majority of the particles by gender are nearly equal, though it can be observed that ne and wa are used more often by women and sa and yo by men. Because zo was so infrequent, it was not included in the figure above. However, in terms of the occurrences of zo, usually described as a hyper-masculine particle, two were produced by a woman (the same woman, MIE), and two by a man (the same man, MA3); however both productions by the female speaker were in quotations of some sort, as can be seen below. (8) こんな食べれないぞとか思っちゃった もの konna taberenai zo toka omottchatta mono this eat NEG zo or so think PAST so So I thought like, I really can’t eat this. In this section, MIE is discussing a possible reaction to certain song lyrics in a video that involve mitarashi dango, a type of snack. She offers a possible reaction, konna taberenai zo and then frames it as a quotation of a sort with “or something like that, toka, followed by omottchatta, “unexpectedly think.” Thus, this is somewhat imagined language. (9) それで「けすぞ」とかいって、大泣きして,で「消さないでくれだめ」と かいって、でね。この前は。 sore de “kesu zo:” toka itte oo nakishite de kesanai de kure that LOC erase zo or so say large cry do CONT LOC erase NEG OBJ give dame: toka itte de ne: kono mae wa no or so say OBJ ne this before TOP So I said I’ll erase it, and they cried a lot, saying no, don’t erase it. Before this. As can be seen above, the second usage of zo by MIE is again in a quote, and also relates to the same music video. Thus again, it is secondary, or somewhat imagined language, of a sort. 36 Male usages of wa include this utterance from MA3, about the location of a birthday occurring soon. (10) あれ近い わ are chikai wa there close wa Ah, it’s close. The rest of the male utterances including wa are very similar comments of this nature, and the rest of them occur between WAT and MA4, in the heavy particle use conversation mentioned above. Also, even though the particles sa and yo, described as assertive and masculine, were produced more frequently by male speakers and the “feminine” particle wa produced more frequently by female speakers, it is worth noting that male speakers did produce wa, and female speakers did produce sa and yo in great numbers. Thus, overall there is hardly any gender difference, and ne is by far the most frequent particle used. Particle Usage by Politeness The breakdown of individual particles based on plain versus polite form follows in Figure 2.5 below. The above figure presents percentages of occurrences in the plain and polite form out of the overall number of utterances in the plain or polite form. Particles appeared more often in the polite form, by percentage, than in the plain. Particles were considered to appear in the plain or polite form based on the appearance of the plain or polite endings in the utterance containing the particle. Figure 2.5. Politeness level of particle All occurrences of na the na-ending (of a verb or a noun) with the form appear more frequently in the quality that goes along with polite (though, of course, they clearly occur in the plain form as well). They may be less associated with a casual meaning that is strongest with sense of insistence or ownership of the utterance or a stronger sense of the affective position of the speaker. Ne may carry this sense of further distance in hence the equal percentages, though it is diffic numbers. However, it does seem possible that ne, are less associated with the casualness of the other particles, and are more polite speech. 14 1 2 20 1 10 0 5 10 15 20 25 ne wane yone 37 -containing utterances, percentages. in the polite form were in the form na no, a form combining no desu. Yo, ne, and yo ne polite form than other particles. Perhaps this implies a interactions better than that of the other particles sa and wa, such as a stronger wa ne ult to draw conclusions from these smaller yo, ne, and the combination of them 4 4 1 67 1 0 14 na sa wa yo plain polite seem to utterances, , yo needed in Particle Usage by Ut The breakdown of each particle analyzed by figure (Figure 2.6) below. Figure 2.6. Utterance position of particles. These results show that particles are indeed not solely final particles. In fact, are all used medially, almost as frequently in some cases. However, only certain particles, ne and sa, seem to be used initially, and only separate utterance, in this data. Some of the particles also seem more strongly final position than others such as seems to relate to usage as a mechanism for holding the floor in conversation particles appear frequently at the end of response in a continuing narrative, as can be seen below in the following passage of conversation between FE2 and SUM 431 15 74 299 21 34 0 50 100 150 200 250 300 350 400 450 500 ne wane yone 38 terance Position utterance position appears in the ne seems to appear alone, as a na. The usage of ne and sa medially, in particular, phrases when the speaker is pausing for a . 107 135 19 210 105 80 12 131 na sa wa yo final medial initial alone they tied to the , as these 39 (11) それでさむいはず だわ うんで子供もね FE2: sore de samui hazu da wa: un de kodomo mo ne That OBJ cold should COP wa yes child also ne ジャケットもたさず学校に 行かせた のね FE2: jaketto mo motasazu gakoo ni ikaseta no ne jacket also school LOC go causative past NOM ne あら SUM: ara: hhh Oh? (sound) FE2: hhh hhh hhh hhh hhh hhh (agreement sound) でも子供 は元気 でしょう SUM: demo kodomo wa genki deshoo: But children TOP energetic aren’t they うんうんうん FE2: u:n u:n u:n Yes yes yes だいたい体温 が熱いあのう高いし SUM: daitai taion ga atsui anoo takai shi Generally temperature SUBJ hot umm high そうそう で帰って来るころ にはね FE2: soo soo de kaette kuru koro ni wa ne oh yes, oh yes OBJ return come time LOC TOP ne ええ SUM: e: Yes 七十度 に 上がってしまうの FE2: nanajuu do ni agatte shimau no Seventy degrees at rise regrettably NOM ああ？ SUM: a::: Ahh? うんだから大丈夫 FE2: u:n dakara daijoobu Yes. because all right In this exchange, FE2 is relating information about the temperature and her experiences, with SUM responding with various modes of acknowledgement and backchannels. 40 FE2’s use of ne here seems to be one mechanism by which she signals she is not finished with her topic yet. Particles such as ne, na, sa, and yo also occur medially often when the utterance contains dislocation for emphasis or explanation, the dislocated portion following the particle. An example of this can be seen in the example below, from MA2. (12) 電話した よね 前 denwa shita yo ne mae phone do past yo ne before The phone call, you know? Before. There are 93 instances of dislocation involving particles used in this way in the data, 4 with na, 28 with ne, 2 with sa, 2 with wa ne, 6 with yo ne, 50 with yo, and 1 with zo. As yo makes up 54 percent of the particles in these utterances, clearly it has a specific interaction with this construction that the others lack. This may be related to the observations by Uyeno (1971) and Tsuchihashi (1983) that yo expresses the speaker’s insistence, thus providing extra impact, and thus after uttering it speakers often feel as if more information may be required to back up their point, providing words they might have otherwise ellipted at that point, or simply because yo and dislocation are both used to provide emphasis in an utterance and thus naturally co-occur more frequently. It is also worth noting that these instances of dislocation provide 38 percent of the medial occurrences of yo. Particle Usage by Conversation Makeup Below is a figure showing overall particle usage based on the gender of the conversation partners. As is clear in the following figure (Figure 2.7), there is a pattern of particle occurrence related to the gender makeup of the participants in the conversation—same-sex dyads consistently use more particles, both female male-male, than do conversations between a male friend and a female frie particles that do not show this precise pattern are of appearing less frequently in the female conversations, and also appearing less frequently in the female the male-male conversations. Figure 2.7. Gender makeup of conversations. This seems to suggest the male use of these particles is what is causing this result—as the conversations with one male participant still show more of the than the conversations without any male participants at all than that seen with the other particles, where both genders used more of the particles when speaking to other participants of the same gender, and suggest correlation sa and to some extent particles must make them more available in male 0 50 100 150 200 250 300 350 400 ne wane yone 41 -female and nd. The only sa and yo, both of which show a pattern -female conversations than the female -male conversations than —which is a different pattern s more of a yo with male speaker usage. Something about these -only speech than in other permutations na sa wa yo ff fm mm -male se particles , 42 while yo appears more frequently in both male and female-male conversations. These particles are often suggested to be casual, rough, or insistent (Makino & Tsutsui, 1986), and these qualities may be more valued or simply more socially acceptable in male-male speech, or they may be used to give a masculine “flavor” to the conversation, as it seems that any conversation involving men contains more of them than conversations involving only women. This pattern is particularly clear with yo, while sa has a large gap between conversations between men and women and solely women, which are roughly equal, and appearing far more frequently in the male-male conversations, which suggests a stronger effect of male-male speech being friendly to the higher usage of sa. While little difference was found in terms of individual particle use by the gender of the speaker, the gender makeup of the conversations seems to have a strong effect on particle use. This implies that speaking to someone of the same gender as opposed to a different gender is friendly to particle use—perhaps speakers feel freer to index a personal dimension in their utterances in these conversations that they do not in mixed- gender conversations, for example. One’s “best” friends are prototypically the same gender as oneself, which seems to suggest a societal expectation, at least, that a person is more comfortable amongst members of his or her own gender. This might also be one reason for the gendered impression some have of some particles, if speakers find themselves using a certain particle only with interlocutors of a certain gender. Particle Usage by Response Below is a figure (Figure 2.8) showing the type of response each particle- containing utterance received by percentage of that type of particle. “Response” means that the other respondent immediately replies verbally, “self” that the same speaker continues on, and "non-verbal responds with a sigh or some other non Figure 2.8. Response to particle received. The results here show that the majority of the time utterances containing particles are responded to. This seems to show that they do in fact have a pull on the other speaker to respond, or that in general the most approp particle is a response of some sort. occur before the same speaker continues on, percentage function of “holding the floor” during discourse continue speaking. However, they also occur fulfill this function, it is clearly not the only way they can be read in discourse differing patterns seen here do imply some differences in roles in discourse between the different particles. 72 61 81 25 33 17 3 6 2 0 10 20 30 40 50 60 70 80 90 ne wa ne yo ne 43 ” means that either there is a pause, or the other speaker -verbal sound. riate response to an utterance containing a Ne, sa, and wa ne are the particles that most often -wise. This may be a part of a , marking that the speaker is planning to before changes in speaker, so if they can 81 67 81 85 14 29 14 115 4 5 4 na sa wa yo response self non-verbal . The 44 Particle Usage by Utterance Type Below is a table (Table 2.5) that shows the types of utterances containing particles by percentage of the type of particle. Table 2.5. Percentage of types of utterances by particle. NE WA NE YO NE WA NA SA YO response 45 1 12 1 24 2 16 statement 21 1 7 4 3 19 45 reporting personal thought 38 0 5 0 30 11 16 sharing information 42 3 4 3 8 20 20 stating opinion 51 1 7 2 10 12 17 question 45 4 5 0 23 6 17 comment 71 2 6 1 2 4 14 The same information organized by percentage of type of utterance appears in the figure (Figure 2.9) below. “Response” refers to when the utterance is a response to another utterance, one with new information, agreement or disagreement, with some kind of original content, while not holding the floor. “Statement” is a simple statement, one that represents shared information, known facts, or some other statement that is not an opinion or novel information. “Reporting personal thought” refers to the speaker giving their own thoughts as a quotation or otherwise reporting their own thoughts in a way that provides some distance from them rather than simply stating them. “Sharing information” means the providing of novel information of some kind to the other participant while not responding to another utterance or holding the floor, or providing some other form of 45 original content, while stating opinion also refers to statements made while holding the floor, but ones that involve a statement of opinion rather than noncontroversial information. “Question” simply refers to when the utterance is a question. “Comment” refers to comments made about the other speaker’s utterance while not holding the floor or otherwise not in control of the conversation. Figure 2.9. Percentage of particles by types of utterances. First of all, some patterns of similarity are visible in Figure 2.9. There is a similarity between the usage of na and yo ne, for one, with both featuring responses and sharing information almost equally, followed by stating opinion at some distance, though yo ne has more noticeable instances of statements and comments, features it seems to share with ne (comments) and yo (statements) though yo ne occurs with these forms with less frequency, and na has more questions and instances of reporting personal thoughts. Na and yo ne also have lower occurrences of sharing information than other particles. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 NE WA NE YO NE WA NA SA YO comment question stating opinion sharing information reporting personal thought statement response 46 This may suggest a similarity in usage and thus meaning between yo ne and na in the data, and suggests the question of when speakers utilize yo ne and when they choose to utilize na. Na also shares with wa ne a tendency to be used in questions more than other forms. It seems that other particles plus ne maybe cause an effect that makes the resulting form more similar to na in use, as this does not seem to be an effect of wa. Also, as mentioned above, na is found noticeably more with reporting personal thoughts than the other forms. This seems to be a somewhat distinctive use of na, as it connects with personal thoughts in general, and many of the personal thoughts that are reported are found with na, even from speakers who do not frequently use the particle. It is also common to find na as part of the quotation, such as in the phrase ii na to omotta (it was good, I thought). Na seems to add a sense of personal reflection to the quote, and immediacy of a sort—it makes it seem more like it is a quote, something that came immediately straight from the speaker’s mind. An example of this is below. It was uttered by SUM, talking about her thoughts on the weather and an ideal place to live. (13) 住みたいな と 思って sumitai na: to omotte live want na QUOT think CONT “I want to live there,” I was thinking Na seems to function to make her quotation of her own thoughts seem more natural, with more affective content. Ne, meanwhile, in instances of reporting personal thoughts, seems to serve a similar purpose to the one it is usually described as having, seeking rapport or agreement, or acknowledgement that one would naturally think such a thing in the circumstance. 47 When wa, sa, and yo are observed, all three seem to feature more statements than the other forms. Wa and sa and wa ne are similar in featuring sharing information, while wa ne features less stating opinion than many other particles. Yo and yo ne seem to show the most variety between types of utterances they appear with overall. In terms of determining the usage and meaning of these particles, it is probably most useful to note that responses most often contained na followed by yo ne, and then ne at some distance, statements wa and yo followed by sa, personal thoughts mostly dominated by na, sharing information wa, wa ne, and sa, stating opinion similar throughout the usage of the particles but slightly more frequently with ne and yo ne, questions chiefly na followed by wa ne, and comments chiefly ne, followed by yo ne and wa ne. It seems a significant point that statements are so dominated by wa and yo. This seems to fit with the definition of yo as being assertive, or marking the speaker’s command over the utterance, and the descriptions of wa as having some sense of insistence or emotional content. An example follows below. It was uttered by MA3, talking about the American fondness for eating pizza. (14) ええ？アメリカ人みんな 喜んで 食べていたよ e:↑ amerika jin minna yorokonde tabete ita yo: eh? americans everyone happily eats PAST yo Eh? Americans all eat it happily (yo). It is interesting that yo does not tend to accompany the stating of one’s personal opinion in this way, instead, ne or yo ne being slightly preferred, as in the example below. (15) 多分 別 に 言われなくてもいいのだろう ね tabun betsu ni iware nakute mo ii no darou ne maybe not particularly say neg also good NOM perhaps ne 48 It might be all right if we stopped talking now, too (ne). This example also features many other hedges used to reduce the force of the opinion, such as tabun (maybe), betsu ni (whatever, not particularly), and darou (perhaps). The usages of ne in these contexts may be due to the desire to seek confirmation from the other speaker when one’s own opinion is being presented, reassurance and agreement so the conversation can proceed smoothly, while statements may not be obviously a statement of the speaker’s opinion, so yo is provided in order to give them more weight, or to make it more obviously the speaker’s opinion, or in the speaker’s territory, or make it more difficult to disagree. Comments, meanwhile, seem to be so heavily dominated by ne precisely because they are uncontroversial and generally agreed upon, or the speaker feels them to be, or feels that they should be. The below comment is uttered by FE2 about the frequency of hurricanes in Florida. (16) 信じられないね shinjirare nai ne: Unbelievable, isn’t it? Questions seem to often include ne to mitigate the sense of ignorance from the speaker, making it sound as if the speaker already has some knowledge. In a way, sharing information being dominated by ne seems to have a similar function, in the opposite sense—it carries an assumption that the speaker will agree with or already knows the information being presented, even if it is clear that they don’t, thereby bringing them into the speaker’s sphere of the conversation. Ne in responses seems to serve much the same function. 49 Na also seems to be used in some of the same ways ne is utilized as described above, as it appears as by far the most common particle used with responses, personal thoughts, and questions. The use of na when it comes to personal thoughts has been described above. It seems to give a somewhat more personal sense to a response or question when used rather than ne, more agreeing than questioning, as in the common response aa sou da na (ah, that’s right) rather than the similar response sou da ne (or that’s right, isn’t it). Ne has a more questioning sense than na, and thus seems to demand more of a response from the conversation partner, while na in its use in responses and reporting personal thoughts seems to have a more self-reflective quality. When na appears in questions, it might be as an attempt to soften the question in a different way, but utilizing these qualities. Sa appears most frequently in statements, in sharing information, and in stating opinions, and hardly at all in questions or reporting personal thoughts, or responses. The example below is MA1 sharing information about school. (17) 日本語 話せない のに さ nihongo hanasenai noni sa: Japanese speak NEG however sa They don’t speak Japanese but . . . . In many of these instances it seems that sa acts to give a sense of trailing off, or a sense of continuation (thus, if it is not followed up by the same speaker, it might give a sense of trailing off). It seems somewhat similar to the English “well,” in usage, in this way. Being used most often instating opinions, sharing information, and statements may imply that it has an insistent or strong character. 50 Referring back to Figure 2.9, it can be seen that the usage of ne is dominated by sharing information, stating opinion, and comments. As it is the most common particle used, it tends to dominate all the categories, except for that of statements. However, its use with sharing information (known by the speaker), stating opinion (the speaker’s territory), and comments (uncontroversial), seems to show that it is most frequently used when the speaker is somewhat in control of the information (though not exclusively). An example is below. FE2 is relating a story about her life, and talking about the temperature at the time. The information is not already known to SUM, her conversation partner, so it is an instance of sharing information. (18) 三十 いくら だね sanjuu ikura da ne thirty amount COP ne It was about thirty, you know? This seems to function to draw SUM into FE2’s relating of the story about herself. It may be used in order to try to draw the other speaker into these utterances or form an agreement, or because the speaker believes the statement to be agreed upon already. Wa ne’s usage is also dominated by sharing information, stating opinion, and comments, making it very similar to ne. Apart from appearing in questions more frequently than ne, percentage-wise, wa ne appears to be very similar to ne in use. This may be a function of the two particles sharing ne. However, use of questions and the more dramatic dominance of the sharing information type does suggest that there are some differences between these particles and their usages. The usage of yo ne is a bit different that that seen in ne and wa ne, implying that the combination of particles differentiates it more in some ways than the same factor did 51 for wa ne. The usage of yo ne is dominated by responses, sharing information, and stating opinions, as well as statements and comments. In this way, it is partly similar to ne, but does not feature as many comments, and instead has notably more instances of responses, and also more statements. The increase in statements may be explained by the influence of yo, but this does not seem to explain the number of responses, and responses are not particularly common usages of yo. Instead, this is particular to yo ne. This use in responses is often a case of strong agreement, as in the often repeated sou desu yo ne (yes, it is like that, isn’t it)? Thus, it seems to be an instant of yo serving to intensify the sense of rapport and agreement already conveyed by ne. When it comes to instances of sharing information, this can be seen to intensify the sense that the statement should be agreed with, makes sense, or could be seen as a sort of common knowledge, even though it is new to the listener. An example follows below, uttered by MA3 when describing how he met a female coworker he is interested in romantically for lunch. He is talking about the restaurant. (19) ご飯 食べた の だよね ランチ gohan tabeta no da yo ne ranchi: food eat PAST NOM COP yo ne lunch I ate there, yes, you know, lunch. This may also be noted as an instance of dislocation. Wa is clearly dominated by the sharing information category, more so than any of the other particles except wa ne (where this may be the influence of wa) and sa, followed by statements and sharing opinion. This seems to place it in the camp of similarity to yo and sa, which follow below and are dominated by the same categories. This could indicate a group of particles that have a more insistent or personal sense than the others. 52 Wa is often described as a feminizing or softening particle, but also as one of insistence, that indexes the affective position of the speaker rather strongly, though this appears to be mitigated by its perceived femininity. In this way, it might be seen as filling a role somewhere between na (personal, reflective) and yo (insistent, intensifying), but in a way that can be read as “soft” all the same (and could therefore be appropriate for a “feminine” woman). Of course, men use wa as well, from time to time, probably because this meaning niche is quite useful to both genders. Na is dominated by the response, sharing information, and stating opinion categories. It also shows more reporting personal thought than the other categories, as mentioned above, and some questions. Its usage in questions seems to largely be the composite form na no, which appears to have taken on its own, somewhat separate meaning (it is also the only source of na that appears in the polite rather than plain form). The usage of na in responses, personal thoughts, sharing information, and stating opinions seems to reflect the sense of it described above—portraying personal information and thought, and giving a strong personal tone to the utterance. It seems to give a more reflective, “talking to oneself” sense than most of the other particles, and is in fact often used in self-directed utterances, or utterances that are ambiguous in terms of direction. Sa, like wa, is chiefly dominated by the sharing information category, but it has notable instances of statements and sharing opinion as well, and little else, making it similar to wa in this way, though sa has noticeably fewer comments. This makes it seem likely that sa also conveys both an insistent sense and something about the speaker’s 53 personal affect as well as the sense of continuation described above. Sa may be used in a similar way to wa especially by men, for whom wa may not be as readily available. The occurrences of yo show a more even distribution, with notable frequencies of responses, statements, sharing information, stating opinion, and comments. Yo in fact has the most even distribution in terms of what types of utterances it appears in out of all the particles examined. This may in fact speak to a variety of different ways in which yo can be used or different senses it can have. Yo appears largely in statements, sharing information, and stating opinions, which seems to fit its generally agreed upon nature as an insisting particle or “verbal exclamation point.” However, it also appears in responses and comments in large numbers. This suggests that it is used in other ways as well. It is often used to intensify the sentiments of an utterance, and so may be used in responses and comments to make the statement more emphatic. This does not entirely explain its usage in questions, however. The questions in question in the data set are demands such as nani ka yo (what was it?) and hontou ka yo (really?), and thus it seems to give the sense of emphasis there as well. However, yo is not simply used to assert one’s opinion or views. It is often used to agree and to build rapport as well, sometimes by emphatically agreeing. It also can give a sense of engagement and interest in the conversation, a commitment to the topic at hand—more affect or lively speech can make a speaker seem friendlier, more casual, or warmer. In this way, despite yo’s nature of insistence and emphasis, it can have a variety of effects in discourse. It is also illuminating to note which particles do not appear with each utterance type. For example, wa doesn’t appear at all with questions, wa and wa ne don’t appear at all with personal thoughts, na, wa ne, and ne appear less frequently with statements than 54 yo, wa, and sa, with yo ne somewhere in between. Sa and wa also do not appear with responses much, while na and yo ne do, and yo and ne are somewhere in between. Na, sa, and wa occur far less frequently with comments than the other particles, especially sa and wa. This shows that wa and wa ne have a more limited range of types of occurrences they appear in frequently than many particles, wa appearing chiefly in statements, sharing information, and stating opinion, and wa ne in sharing information and stating opinion, with more questions and comments. Na, wa ne, and ne appear less frequently with statements, with yo ne in the middle, which could imply something about the ne component of these making them less likely to appear in statements, while the yo in yo ne makes them slightly more likely than they would be otherwise. This could be related to the nature of ne in terms of implying agreement, while na implies reflection on a personal thought or feeling more than stating something strongly, as well. Sa and wa also do not appear in responses often. This implies that they more frequently appear in statements that contain more novel content, rather than in responses to others. This could be a part of the insisting nature mentioned above. Response Type by Particle Another interesting category to consider is what sorts of responses are provoked by each particle. Below is a table (Table 2.6) showing the ways in which utterances containing particles were responded to, organized by type of response. Any variation of the response un, or “yes” in casual speech, was coded as “un,” though oftentimes it was repeated more than once. The category of aizuchi involves comments like ne, or soo na no, or aaa, any sound or short phrase made to reassure the speaker that the listener is continuing to listen (i.e. backchanneling). Response refers to a 55 response to the particle-containing utterance that contains new information, agreement or disagreement, with some kind of original content, while comment refers to comments made containing opinions or observations, some personal affective position to the utterance responded to. Continuation refers to when the other speaker simply continues with his or her own point or without obviously orienting to the utterance preceding it, while self-continuation refers to the same speaker continuing with another utterance. Question refers to questions, whether those confirming or asking about something else, and sound refers to a non-verbal response of some kind. Table 2.6. Types of responses to particles. NE WA NE YO NE WA NA SA YO un 201 11 35 14 54 95 74 aizuchi 75 6 14 13 29 17 30 response 165 3 26 7 40 20 76 comment 30 1 4 1 28 16 27 continuation 56 0 4 0 9 1 11 self- continuation 206 12 19 8 29 65 39 question 49 2 3 2 11 5 66 sound 21 0 1 2 11 8 13 Again, if we take the percentage, we get something like this, seen in Figure 2.10 below. What is most notable here is the prevalence of sa with the response type un and of wa ne, ne, and sa with self-continuations, as well as yo with questions, which are probably the most immediately striking results in the data. Ne, yo ne, and yo also seem to be responded to with responses slightly more frequently than the other particles do. Sound responses and continuations do not appear with wa ne at all, and continuations do not appear with wa, either. Na triggers the response of comments far more than any of the others, followed by sa and yo, while wa has noticeably more aizuchi response than the others (followed by wa ne). Ne appears most with the replies un, response, and self- continuation. Wa ne is responded to even more frequently by self but also by aizuchi. Yo ne, on the other hand, while still showing the tendency to be responded to by un or by the same speaker continuing, shows this second in a far lower percentage and also shows increased numbers of responses and different pattern. Wa seems to show the same pattern as far more aizuchi responses and the complete lack of the continuations shows responses in all of these, but also adds comments, far more than any other particle. Sa, like ne, is dominated by un respects than is the case for ne replied to with un, responses, and questions, a pattern not found amongst the other particles examined. Figure 2.10. Responses to particles by type, percentage. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% NE WA NE YO NE 56 -continuations and aizuchi, a slightly yo ne in many respects, but with yo ne shows. and self-continuations, but more dramatically in both . Yo, however, is noticeably different, instead mainly WA NA SA YO sound question self-continuation continuation comment response aizuchi un un, Na 57 Again, the response of un is chiefly dominated by sa, followed by yo ne, wa ne, and wa. The other particles trail them slightly. The response with aizuchi is predominantly to utterances containing wa, though wa ne slightly stands out as well, followed by yo ne and na. Responses mostly are motivated by utterances containing yo ne and yo, and ne. Comments are interesting split between those particles that appear to motivate this response frequently (na, sa, and yo), and those that motivate them hardly at all (wa ne, yo ne, ne, and wa). Continuations, however, most frequently happen with ne, not at all with wa ne and wa, and hardly at all with sa. They are not very common in the data and ne is the most common particle; however, their occurrence with ne is still somewhat striking. Yo ne occurring with a response more frequently while wa ne and wa hardly or don’t appear at all is similar to the situation in terms of comments, above. Self- continuations are dominated by being responses to utterances containing wa ne, followed by ne and sa. Sounds are dominated by na, wa, sa, and yo, and wa ne, yo ne, and ne are hardly ever responded to nonverbally or by a pause. Questions seems to follow yo more often than any other particle, followed by ne, wa ne, and na, with hardly any of the others provoking the reply of a question in any number. This seems to be related to the insisting or emphatic qualities of yo in some way, though this provokes the question of what causes questions frequently in terms of the others, as well. In some ways yo’s distribution of responses is similar to ne’s, except that yo has so many more questions and much fewer self-continuations. The most striking results from this analysis are the questions with yo, self- continuations with wa ne, ne, and sa, aizuchi with wa, and un with sa. It also seems significant that some of the particles provoke chiefly simple agreement (un) or are part of 58 a longer series of utterances produced by a single speaker, while others provoke a wider variety of replies such as responses and comments. Yo, na, ne, and yo ne seem to provoke this more widely varied response. Perhaps this is partly a function of yo, and it occurs with both yo and yo ne. However, this does not explain na or ne. Of course, these responses could also be affected by the type of utterance each particle tends to appear in, and this may in fact be a likely factor. Prior literature (Saigo, 2006) describes a strong constraint to reply to utterances that contain a particle, particularly utterances that end with one, and this can be observed in the data in the telephone conversations, even if the response is as simple as replying with a single short un. Discussion The small sample examined in this study suggests that there is substantial individual difference among Japanese speakers in the use of interactional particles. Some speakers seem to use them considerably more than others do. While further work with a larger sample size is needed to verify this observation and draw more definitive conclusions about what roles these speakers are playing in the conversations, it is possible to explore the role heavy particle use versus non-heavy particle use seems to play in the conversations at hand in this data. The usage of particles is clearly a tool to show involvement in the conversation, specifically affective involvement, as Lee (2007) claims, and can be seen in the particle-heavy exchange between WAT and MA4, as well as in FE2’s particle-using responses to SUM in their conversation. However, what function this involvement plays in discourse can take different forms. Specifically, it can take the form of directing and attempting to control the course of the conversation (as it seems to in the case of MIE, MA2, and MA4), or it can take the form of responding and 59 showing interest in the other speaker’s control of the conversation and topic (as in the case of FE2 and WAT, mentioned above). This implies that these speakers are using the affective resources of particles for a variety of purposes in conversation. Another interesting result is the overall lack of gender correlation. This is a novel finding, as it has been argued that at least some of the discourse particles examined here are gender markers (Inoue, 2006; Makino & Tsutsui, 1986; Okamoto & Sato, 1992; Shibamoto, 1985). It seems that in general, particles are a tool open to be utilized by both genders, though they are more commonly used overall in same gender conversation as opposed to different gender conversations. This seems to suggest that particles carry meaning that is widely useful in the speech of both genders, overall. Though some particles are found more often in the speech of certain genders rather than others, all the particles found were used by both genders. Thus, any association with gender seems to be one of degree rather than absolutes. Of course, this could be a function of change over time, with particles that at one time were more strictly associated with a specific gender losing that association as younger speakers become more creative or free with their usage of the particles in question across gender lines. All of the speakers in these telephone conversations were very young, under the age of thirty, and could be showing this tendency in their speech. This lack of gender difference could also be a function of the egalitarian nature of these conversations—the speakers are roughly the same age in each conversation and are already friends, perhaps leading them to use similar speaking styles to their interlocutors regardless of gender, with particles being one part of that. There were, however, some interesting effects of gender. These features of interest were visible in the fact that ne, wa ne, yo ne, na, and wa all appear much more 60 frequently in same gender conversations, while sa and yo don’t, instead appearing more frequently in male-male conversations than they do either in female-female conversations or male-female conversations. First of all, this suggests an association with masculine speech that sa and yo might carry. These are both insisting particles with a strong sense of vehemence. They do not exert the pull on the other speaker’s response that ne does (“and you should agree with me”), instead they give a sense of strength to the speaker’s sense of ownership of and belief in their own utterance. One could argue that they are more strongly based on the speaker, while ne draws more on the sense of the listener. This and a certain roughness often thought to be associated with male-male speech, wherein less care is shown for the feelings of the other speaker or plainer speaking is acceptable and not considered rude, may explain this result. Men use fewer of these particles when speaking to women. Meanwhile, this finding also raises the question of why ne, wa ne, yo ne, na, and wa all appear so much more frequently in same-gender conversations than they do in different-gender conversations. It is also worth noting that though wa is often described as a female particle, it occurs more frequently in same gender conversations between both women and men, and though na has been described as a somewhat “rough” or male marker, it also appears more frequently in same-gender conversations between both women and men. These particles might be seen as “softer” particles in some ways, based on Shibamoto’s (1985) classification of wa as a softener and the description of na by Suzuki as an indexer of one’s own personal thoughts (1996). Meanwhile, ne is frequently described as an attempt to seek affective common ground or agreement with one’s listener. Perhaps the use of these particles provide a sense of personal interest in the 61 conversation, or provide a way to give a sense of personal thoughts that is more useful in some way when speaking to a fellow member of one’s own gender. As mentioned above, ne also exerts a strong pull on the interlocutor’s response, so it may be that speakers are more comfortable making demands of this nature in same-gender conversation. The majority of particles and particle usage overall seems to be associated with the use of the polite form rather than the plain form in this data. Some of the particles show higher percentage frequency in the plain form than the others do, specifically wa and wa ne, which implies that wa is strongly associated with casual speech and the plain form, and yo ne seems to appear proportionally more in the polite form even than the others, but percentage-wise particles are more frequent in the polite than the plain. While one might expect that as a function of the emotive and affective sense contained in most particles the plain form would be more common, this is not the case here. This may be an attempt to give the slightly more distant-sounding speech of the more polite form a more emotive, involved tone to make it fit better in an involved discussion between friends. This might also have to do with different usages of the polite form. For instance, the usage of the polite form between friends would be quite different from usage of the polite form in a formal setting, such as the formal speeches at a university club examined in Dunn (1999) or a business meeting. Speech between friends, like speech between speakers of the same sex, may favor the use of particles in the way other usages of the polite form do not, which would explain why much prior literature has not found such a pattern of particles utilized alongside polite speech. It is also clear from these results that while some particles can appear utterance initially, medially, and finally, others are more restricted, appearing only finally and 62 medially. Ne and sa are the chief particles to occur initially, with ne appearing in the initial position far more frequently than sa. Ne is also the only particle to appear as an utterance on its own in this data. Ne is also the most heavily used particle. It seems likely these factors are related, though yo, the second-most common particle, seems far more restricted than ne, never occurring alone in any of the data, and chiefly tied to the final position, with 38 percent of its medial occurrences occurring as the result of sentence dislocation. Thus, it seems frequency alone can be ruled out as the reason for ne’s unusual behavior in this respect. Instead, it seems that ne has less semantic weight than the other particles, possibly as a result of semantic bleaching due to grammaticalization, the process of language change by which functions change from that of their original form to become grammatical markers, perhaps due to ne’s frequency (Hopper & Traugott, 2003). This would further broaden the usage of ne to allow it to cover multiple meanings and appear in multiple situations. This could be why it has been so difficult to pinpoint the core meaning of ne. The other particles are more dependent, bound to the rest of the sentence for their meaning, more of a comment on the rest of the utterance, which could mean they have undergone less grammaticalization and carry more inherent semantic content at this point. While ne still implies a meaning tied to other utterances rather than having completely separate semantic content, it seems that this meaning of ne can relate more broadly to discourse instead of being tied to the rest of a specific utterance in particular. An instance of ne uttered on its own still implies agreement with the utterances that come before it, and seeks agreement from the listener, but it is more general, tied to the discourse and not specifically to one utterance in particular. This may be because of 63 further progress down the path of grammaticalization toward becoming a function word than the other discourse particles, as Hopper and Traugott describe as a process of language change. This discourse application of ne while the other discourse particles are more tightly bound in terms of usage may be one reason for the common occurrence of ne compared to the other particles. It also simply appears to be more useful and commonly used in the data. This may be because of the nature of telephone conversations, with two speakers who are tightly focused on the discourse of the other. Ne may have greater utility than other particles in seeking out agreement in a situation of this time. However, it seems likely that ne is in fact the most common particle used in Japanese, for the combination of factors involving its greater productivity and greater utility. It is worth noting that the compound particles that appear in any number both involve ne. From looking at the types of utterances that contain particles and the types of responses they received, a variety of observations become possible. When looking at the types of utterances in which specific particles appear, it becomes apparent that na and yo ne are similar in many respects. However, na did differ from yo ne in terms of responses, being responded to with a great deal more comments than yo ne received. In this respect, yo differed the most from the other particles, especially in terms of being responded to with questions, while sa was responded to much more frequently with un than the other forms and the same was true of wa and aizuchi and ne, wa ne, and sa with self- continuations. Na is found more with reporting personal thoughts than the other forms. Perhaps this tendency hints at a connotation in na that is what leads na to be responded to more with comments, a more “comment-like” or reflective meaning in and of itself. It 64 also seems significant that statements appear most frequently with wa, yo, and to a lesser extent, sa, the most common instead being yo, seeming to support the arguments in past literature that yo has an assertive or confident sense. However, yo has a wide range of responses. Ne, however, seems to be preferred to be responded to with statements of personal opinion and comments, which seems to be at least part due to the desire to seek confirmation from the other speaker that they do in fact agree and mitigate the risk of presenting one’s own opinion. Ne also seems to be often used when the information is believed to be uncontroversial, or the speaker wishes it to be, or with questions, possibly to mitigate the feeling of ignorance from the speaker. Wa ne seems to be very similar to ne in use, but used more with questions and far more with sharing information, which wa is as well, as is sa. Sa appears most frequently in statements and sharing information, as well as stating opinions, while yo ne occurs in similar situations as ne, but not with as many comments and with more responses and statements, which is particular to yo ne. This is often in cases of strong agreement, and yo seems to be used in these cases to intensify the case of rapport involved with ne. Wa is dominated by sharing information, followed by statements and stating opinion, and seems to have a more insistent sense because of this. Questions, responses, and reporting personal thoughts don’t appear much, if it all, with wa. These observations begin to offer some insight into the nature of the meaning behind the usage of these particles, as well as allowing observation of some of their role in discourse, when it comes to the responses they provoke or encourage. 65 CHAPTER III ANALYSIS OF SCRIPTED DATA In this section, data from a Japanese television drama was examined. Obviously, the dialogue in this drama was the work of a scriptwriter. However, this still provides the researcher with opportunities to examine native speaker usage, especially in terms of information about what native speakers think about particle use, as this will come through in how the dialogue is written. Good dialogue must seem natural, so it will have features of naturalistic conversation, but it must not become so natural that nothing about the characters or storyline is conveyed, as narrative information or character development is not the purpose of natural conversation in reality. A novel opportunity for analysis here is an examination of the possible link between the methods of using conversational particles and the character’s role and depicted personality, and thus what particles “mean” in the mind of the native speakers who wrote the scenarios. The fact that discourse particles do appear in such dramas also supports the argument that such particles are necessary for conversation in Japanese to appear natural, and also that they convey important messages in these conversations, messages the scriptwriters of the drama make use of in writing their script. Dramas also offer more unusual dialogue and conditions for examination than are likely to appear to be recorded in the every day life of the researcher, and provide a wider pool of situations in which conversations occur. For these reasons, and most of all for the insight into the ideology behind the use of particles in the mind of native speakers, it is useful to examine scripted data such as that from a television drama. 66 Methodology Two separate episodes of a Japanese television drama were analyzed, both coming from the beginning of the series (Table 3.1). All utterances were transcribed, and all occurrences of discourse particles in the transcribed data were coded for the same factors used above in the analysis of the CallFriend telephone conversation data. The data used came from two forty-five minute episodes of the Japanese television drama “HERO,” which aired on Fuji TV in 2001. This drama centers around a metropolitan prosecutor’s office and the variety of characters who work out of it, including prosecutors, paralegals, and their immediate superior. In the episodes examined for this report, the major characters include the protagonist, Kuryuu Kohei, a highly unconventional prosecutor and a former juvenile delinquent who earned a high school equivalency degree, an ambitious female paralegal who aims to become a prosecutor, Amamiya Maiko, three other prosecutors in the office, Egami, Shibayama, and Nakamura Misuzu, and the two other paralegals, gossipy men by the names of Suetsugu and Endo. There is also their department head. Each episode features episode- specific characters involved with the case the prosecutors are handling—these include the characters of Shimano Saeko, a cooking instructor being investigated for marriage fraud, in the next episode, and Officer Danbara, a police officer who proves difficult for the prosecutors to work with. There are also other, incidental characters in each episode, though the particles these characters use have been counted in groups, separated by gender. All the major characters appear in each episode—however, the guest characters are different in each one and are thus specified by episode in the table below (Table 3.1). 67 Table 3.1. Episodes of HERO analyzed in this study. Episode name Featured Character Length Episode 3—The Crime Called Love Shimano Saeko 45 minutes Episode 4—What He Taught Me Policeman Danbara 45 minutes A brief description of the major characters in HERO follows below in Table 3.2. Table 3.2. Main characters in HERO. Character Name Sex Occupation Personality Kuryuu Kohei Male Prosecutor Unconventional, compassionate Amamiya Maiko Female Paralegal Conventional, ambitious Egami Male Prosecutor Conservative, interested in appearances, timid Shibayama Male Prosecutor Womanizer, manly Nakamura Misuzu Female Prosecutor Womanly, career- oriented Suetsugu Male Paralegal Gossipy, rule- oriented, older Endo Male Paralegal Gossipy, younger, flirtatious Shimano Saeko Female Cooking Instructor Cold, demure Policeman Danbara Male Police Detective Aggressive Bucho Male Head Prosecutor Anxious, worried about consequences Results Overall Particle Usage by Each Speaker and Particle Type The overall occurrence of particles over both television episodes is shown below, in Figure 3.1. Figure 3.1. Overall raw frequency of each particle Zo appeared 6 times, and ze once. It seems as if the particles can be separated into two groups—a group of frequently appearing particles ( infrequently appearing particles: common thread with the telephone data, but it is clear that frequency here, and sa less so. not appear in the television drama was counted as an instance of show, comparatively. This seems to be a function of it often appearing with polite speech (see below), which is much more common in the television show professional environment, than in the casual and friendly telephone conversations. When examined by episode, the particles are comparable in terms of instances, except that ne is by far more common in the first episode examined than in the Given this, the remaining analyses present data collapsed across the two episodes. 0 20 40 60 80 100 120 140 160 180 yo ne na 68 yo, ne, and na, and a group of more yo ne, sa, wa, zo, and ze). The overall rarity of na appears with greater Ze did not appear in the telephone data at all. except with the topic marker wa followed by ne alone. Yo ne is also more common in the television overall, set as it is in a sa yo ne wa zo ze wa is a Wa ne does ne, which second. Below in Figure 3.2 can be seen each character in the drama’s total particle use by percentage. No heavy particle users of the type observed in the telephone dat clearly, though it is noticeable that the category of “other female” is so dominant. This category refers to female extras with minor parts, most with only one or two lines, and this implies that these minor female characters use particles dispro number of lines they actually have when speaking. Figure 3.2. Overall particle usage by speaker. Though no obvious heavy particle users are immediately apparent, Kuryuu, Shibayama, Egami, Suetsugu, and Danbara all seem to use them frequently. Interestingly, these are all male characters. Misuzu and Saeko are the female speakers who use particles most, but not as much as these men. Amamiya, a female character, uses particles noticeably infrequently compared to the o male characters. The boss, or head of the office, might also be counted a slightly lower particle user. 0 10 20 30 40 50 60 Particle Percentage 69 portionally to the somewhat more thers, as do incidental a emerge Below in Figure 3.3 appears a presentation speaker, with the percentages of the characters’ utterances. Figure 3.3. Overall particle usage by speaker, percentage of total utterances. Below appears the same information in a table, in Table 3.3. Table 3.3. Particles by speaker, percentages. yo ne na Kuryuu 6 13 5 Amamiya 7 4 0.4 Shibayama 11 4 10 Misuzu 12 5 4 Egami 20 3 8 Endo 7 11 8 Suetsugu 7 10 11 Bucho 9 3 9 Saeko 10 11 1 Danbara 16 6 11 0 5 10 15 20 25 70 of the particles and their usage by usages of each particle out of their total sa yone wa zo ze total 2 3 0.4 0 0 29.4 0 3 0 0 0 14.4 6 3 1 2 0 37 4 1 5 0 0 31 3 4 0 2 0 40 0 2 0 0 1 29 0 4 0 0 1 33 0 0 2 1 0 24 0 1 6 0 0 29 0 0 0 0 0 33 yo ne na sa yo ne wa zo ze 71 Percentage-wise, it can be seen that it is Egami and Danbara’s use of yo that is most striking, followed by Kuryuu’s use of ne and Misuzu’s use of yo. The usage of yo from Danbara and Egami seems to be explicable because of the amount of time they spent arguing with other characters and insistently making their points, suggesting the insistent or emphatic usage of yo while Misuzu seems to use it often because she speaks forcefully more often than some of the other characters and it is often used in these situations. Most speakers tend to use only a few particles frequently, a tendency that comes through here, but in these cases the use of the rest of the particles is far overshadowed by the usage of one in particular. Danbara also uses na frequently, a tendency which is seen to a lesser degree in Egami as well. Both of these characters present a professional surface and seem to care about their job, but are also both at odds with the main characters and argue with and criticize them frequently. Both of these characters also seem to be attempting to present some form of “masculine” image, but without seeming too rough, over the top, or unprofessional. Na is used by most of the male characters quite frequently (except Kuryuu) and by the female characters hardly at all, except Misuzu. These exceptions seem to be due to characterization factors, as well. Misuzu presents a personality that had many feminine and “sexy” traits, but is also strong and independent, while Kuryuu uses more polite form and presents a less typically masculine persona than many of the other male characters. This may suggest that na can have a masculine sense, as suggested by some of the literature. As can be seen, Kuryuu uses ne more than any other character. The usage of ne seems to be a fundamental part of how his character is presented to the audience, and forms a major portion of the presentation of his casual and unconventional, but friendly 72 personality—showcasing the affective common ground and rapport he is especially shown building with the clients of the prosecutor’s office in particular. Endo and Suetsugu also use a great deal of ne, and they also attempt to build rapport, but for a different reason—most of their scenes show them gossiping with the other characters, or attempting to acquire gossip material, or, alternatively, acting in a sycophantic way. Kuryuu’s usage of ne was discussed above, and it certainly outnumbers his use of other particles by a great deal. He is also one of the only male characters to use the discourse particle wa. In general, his use of particles seems to be geared toward de- emphasizing a traditional or expected presentation of his social roles (as a public prosecutor, as a man), in favor of a more individualistic manner and attempts to build rapport with those he speaks to. He tends to use ne as a particle when stating his opinions instead of using yo or even na, and when arguing makes oblique statements instead of refuting his opponent’s points directly, as can be seen in the examples below. When arguing to the other prosecutors in the office that the police did in fact arrest the wrong man in Episode 4, he utilizes ne as he brings up his point: (20) ああ、このね。このジージャン．百万円 ぐらいするね。 Aa, kono ne. Kono jijan. Hyakumanen gurai suru ne. Aa this (ne) this jean jacket hundred ten thousand yen about do ne Ah, well, this, you know? This jean jacket. It would cost about 100,000 yen, wouldn’t it? This is followed by a further use of ne as he continues to make his point about the jacket: (21) それ千 九 百 三 十六年 ぐらいで来たね Sore senkyuuhyakusanjuurokunen gurai dekita ne. That 1936 about was made ne That was made in about 1936, wasn’t it? 73 The information he is giving here is not, in fact, shared information, and he is making statements in order to make his point, but he uses ne here, numerous times. This can also be seen in Episode 3, as when he is challenging Saeko about her tendency to take money from men by commenting on how expensive her class kitchen looks, talking about obviously apparent information to both parties. (22) ああなるほどね。んん誤解したわ 分かる なこれ な先生 先生 Aa, naruhodo ne. Mmm, gokaishita wa wakaru na kore. Na, sensei! Sensei Aa is that so ne misunderstanding wa understand na this na teacher teacher に は 今まで は付き合った男性いただいたのお金 何にばけて ni wa ima made wa tsukiaitta dansei itadaita no okane nan ni bakete LOC TOP now until TOP date PAST men was accept NOM money what turn か ね こう いうごうかな 食器です か ね ka ne. Kou iu goukana shokki desu ka ne. QUEST ne this way say expensive cutlery COP QUEST ne なシステムキッチンだ Na, shisutemu kitchin da. Na “system kitchin” COP Ah, now I get it. Mmm, now I understand how they misunderstand. Hey, Teacher! (What he calls Saeko) What did that money you accepted from those men turn into? This expensive cutlery? Or was it that full kitchen? These instances of ne usage show that while it may be a particle that indexes rapport or agreement in some ways, it can also be used in a way that seems more demanding because of the presumption of shared affective common ground, in a way that can come across as being almost passive-aggressive. This can show how ne might be used as a more assertive particle (and possibly how the more assertive particles, like yo, might seem “softer” in certain circumstances by comparison). Kuryuu’s particle usage differs from more argumentative characters such as Amamiya, Egami, and Danbara, as well as the more typical pattern found in the other male characters of favoring yo and na. He tends to use the polite form, at least in terms of the copula desu, but to also use other 74 markers of casual speech, such as verb contractions. However, despite this, his utterances often sound pushy and rude, and the other characters react to the inappropriate nature of his comments at times. It may be that his use of desu/masu polite form enables him to not cross the line completely, or is a nod in the direction of the professionalism he is supposed to be showing at work. This may also lower his usage of na and wa, as these are generally found in the plain form in the data. He also has far more dialogue than any of the other characters. Amamiya is the paralegal assigned to work with Kuryuu at the beginning of the series. She tends to be conventional, but she is an idealist and does believe in justice, so she is often swayed into going along with his more offbeat method of investigation. She is formal and a bit uptight, but rather direct. The fact that she uses far fewer particles than the other characters may be a result of these tendencies. The particles she does use are generally those found mainly with the polite form, no doubt because she almost exclusively uses the polite form. She does use yo more frequently, however, which may be a result of her tendency to state her opinion directly and argue with other characters. The below example is from a conversation with Kuryuu in Episode 3, where he states that it is next to impossible to prove marriage fraud, and she argues with him. (23) 実際 彼女が言った事 はうそなんですよ。 Jissai kanojo ga itta koto wa uso nan desu yo. Actually she SUBJ said things TOP lies COP yo The things she said were actually lies! In another example, she argues with Egami, who has been saying that their case is impossible to prove, using yo. 75 (24) むり って いうの よしましょうよ Muri tte iu no yoshimashou yo. impossible QUOT say NOM let’s not yo Let’s stop saying it’s impossible! She also uses yo when strongly agreeing with someone, as when, early on in the episode, she is shown to sympathize with Saeko as a fellow career women. When Saeko asks if her boyfriend was right to attack her just because she tried to break up with him, Amamiya responds with: (25) そうですよ Sou desu yo Like that COP yo Yes, exactly! It seems that Amamiya’s use of yo functions to present her as a person with strong ideas, and with a strong sense of what she thinks, which she is often vehement about, whether she agrees or disagrees. This is used to provide contrast to her generally somewhat clipped and formal speaking style, and perhaps to help give a sense of involvement and affect to her speech as well, and make her seem like a more emotional character than straightforward polite utterances would alone. This provides insight into one of the ways yo can be utilized to “soften” dialogues, as described, for instance, by Matsui (2000), by providing more intensity or a sense of emotional commitment to the utterance. The sense of emphasis yo provides can give a more personal character to an utterance and make it seem more emotional, thus offering a more personal feeling to the conversation. Thus, the particles appear in dialogue to illustrate a more emotional side of the character. This helps to illuminate one of their roles in discourse in general. 76 Kuryuu and Amamiya are the protagonists of the drama, and their characters are often contrasted. Kuryuu uses a great number of particles, twice as many as Amamiya (30 percent of his utterances contain particles, while 15 percent of Amamiya’s contain particles). This seems to be related to the presentation of Kuryuu’s demeanor as engaging and interested in the affairs of others despite his somewhat eccentric actions and comments, while Amamiya is presented as more abrupt, focused and goal-oriented, more aloof and less engaged with others, and more driven in terms of her job (she is very ambitious). This seems to suggest that social or personal engagement is reflected at least in part in the frequent use of particles, while lack of particle use may suggest a persona that is not interested in engaging with others. This contrast suggests the emotive nature of discourse particles in general. It is worth noting that when Amamiya does utilize particles, these seem to be used to help bring out her more emotional or the “human” side of her character. Amamiya has a tendency to be much more formal than Kuryuu does as well (though he often uses the copula desu associated with the polite form), so the correlation of particles with the casual, plain, and emotional comes up here as well. Amamiya’s general formality and professionalism also seem to be correlated with her lack of usage of any particles other than those that typically occur in polite form, or in professional contexts, yo, ne, and yo ne. This lack of use of the plain form may be one of the reasons she uses none of the more personally affective particles such as na or wa, or the more casual particle sa. This gives her a somewhat formal and not specifically feminine demeanor. Kuryuu, however, does not utilize the stereotypically masculine particles zo or ze, and his speech does not show the chiefly yo and na dominated pattern observed in the particle usage of the other male characters. It seems that both of these 77 characters present more about their personalities and approaches to life through their usage of these particles than they utilize them to mark their gender. By doing this, they also take a position regarding gender marking for themselves, as well—they do not play it up. Shibayama, on the other hand, another prosecutor in the office, has a strongly masculine presentation. Not only is he tall, with a deep voice, he speaks in a plain, casual form (using the personal pronoun ore, associated with roughness and extremely masculine men—most of the men in the series use the pronoun watashi, with Egami using boku, associated with younger or more boyish males) and casual endings such as changing –nai (the negative form) into –nee (a strong sign of casual speech). He also uses the stereotypically masculine pronoun zo several times in each episode, along with his speech showing the pattern of yo and na associated with most of the male characters and having a high incidence of the usage of sa. He is so masculine, in fact, that his masculinity comes off as almost parodic, and since he often fails to live up to the associations with his masculine persona, he seems to be “overperforming” his masculinity to some degree. For example, after getting a promotion and after apprehending a criminal, he is shown to shout in triumph (the second time, with the accompaniment of the particle zo), and this is not played entirely seriously. (26) よし、よしいいぞ！ yoshi yoshi ii zo Yes yes good zo Yes, yes that was great! This assertive mode of speaking seems to be one reason for his usage of yo and na more than any other particles, which seems to imply they can be somewhat strong in tone. Na 78 seems to be associated with stating one’s own thoughts or opinions, and thus can seem assertive, since it gives a personal tone to the statement, and injecting oneself into the conversation tends to be read as assertive in general. Shibayama uses this particle with almost every observation or statement of his opinion—and he also offers his opinions relatively frequently, as below in Episode 4 when he is offering his opinion about a lead Kuryuu is following. (27) それが 今度 の 事件となん の 関係 があんだろうな Sore ga kondo no jiken to nan no kankei ga ann darou na That SUBJ current POSS case and what POSS association SUBJ neg probably na That probably doesn’t have anything to do with the current case. Interestingly, he also seems to talk to himself more than any of the other characters do, and also frequently uses na in this context, as he does in the example below in Episode 4, after going into his own office. (28) ああ疲れた な Aa, tsukare na. Ah, I’m tired. These usages of na suggest some of its overall meanings, in the sense of having a strong or casual tone, but also being used to index or convey a sense of one’s own thoughts or reflectiveness, as stated before in the section relating to the telephone conversations. It is interesting that he also uses the discourse particle wa—it seems to be indexing the casualness of his speech, and the freedom with which he generally states his own opinions or lends his personal viewpoint to what he says. This also suggests that the usage of particles can be utilized to show a gender identity the person using them wishes to protect strongly, and that more than one particle is often used to accomplish this, 79 suggesting, perhaps, that the use of certain particles is associated with the use of certain others. Sa is mentioned above in the description of the telephone conversation data as often used in a series, or medially, as something of a marker that one is continuing a narrative or list. However, it seems to have another sense, that of an attempt at “coolness,” especially of a masculine nature, somewhat equivalent to using “well . . .” in English to start off an utterance, or to trail off. This may be related to the use of sa in translations as noted by Nakamura (2013). Both Shibayama and Egami tend to use sa in this way. However, these attempts at seeming “cool” are usually played for humor—both characters tend to use it when attempting to ask their female co-workers out on dates, only to be turned down in every case, such as when Shibayama tries to offer an invitation to Misuzu in front of the rest of the office, only to be turned down because he didn’t do a good enough job making it seem casual (as they don’t want their affair to become known). Nakamura Misuzu is another prosecutor in the office and the only female prosecutor (as Amamiya, the other woman, is a paralegal). She is contrasted with Amamiya in many ways—Amamiya is personally somewhat reserved despite being direct about her thoughts and opinions, while Misuzu is portrayed as being sexy and feminine and shows that overtly, Amamiya is quite formal and professional while Misuzu tends not to be. She is, however, shown to be extremely successful, and is also independent and straightforward, as well as feminine. Her speech seems to be coded rather feminine, but at the same time, she uses more “masculine” or assertive particles such as yo, na, and sa, and typically uses the plain form, though not as casually as some 80 of the male characters. At the same time, she uses the discourse particle wa in a way that is clearly meant to convey her femininity and be read as a stereotypical feminine speech marker, as in the utterance below, perhaps to contrast with her usage of “rougher” particles such as na and sa, or perhaps because she is shown as someone who states things like thoughts or opinions, and all of these can be used to express more personal thoughts and feelings, na, sa, yo, and wa: (29) 『マヒマヒ』なら付き合ってあげてもいいわよ。 “Mahi Mahi nara tsuki ate gete mo ii wa yo. Restaurant if you took me out if good (wa yo) It’s all right if you want to take me out to Mahi Mahi. In this statement, she is letting Shibayama know that she is willing to resume the affair they are having, in a somewhat oblique manner. In fact, every instance except one where she uses the discourse particle wa involves her affair with Shibayama in these episodes. In the other instance, she is making a comment to Amamiya about the case (“if anyone leaned over too far here and fell they would probably die”). Her usage of particles also shows a usage of particles to reflect a gender and personal identity, implying that choices of particles may be used consciously, such as the use of wa when speaking to a specific person or to present a particular sort of persona. Egami is another one of the prosecutors in the office. He was the prosecutor Amamiya worked with before Kuryuu was assigned to the office, causing her to split her time between the two of them, and harbors an infatuation with her. He is presented as the best at law in the office, but also something of a fool, obsessed with his dignity and not looking badly in front of others. He tries to appear manly, but is something of a weak character all the same. Despite often loudly expressing his opinions, he often fails to 81 back them up. His usage of yo is by far the most dramatic, percentage-wise, and this seems to be because he is often found arguing with others about the way various cases are progressing and should be handled. He is the other male character to use zo, and this tends to be at moments when he is trying to seem assertive or manly, as he does here, after arguing with Amamiya about a case: (30) もうあきらめたほうがいいぞ mou akirameta hou ga ii zo already give up would be good It would be better if you gave it up already. Here, he is attempting to seem assertive about a case he thinks Amamiya and Kuryuu are wasting their time on, after being insulted by Amamiya and turned down by her for a date earlier in the episode. After stating this, he leaves for the night rather than continue to help them. His usage of zo tends to come at such moments, after his dignity has been threatened. For instance, in the second episode, after he has spent most of the episode in trouble after arguing that a suspect is not the killer, only for him to escape, he uses zo when scolding the (newly exonerated) suspect for causing him so much trouble. His usage of na also seems to fit this protected persona of an attempt at “masculine coolness.” Likewise, he does not utilize sa often, but the times when he does seem to be as part of the above-mentioned usage of sa as an attempt to seem suave or “cool” in some way, especially vis-à-vis women, that is typically played for laughs, as below. (31) あまみや Amamiya! (she turns to see him, complete with leaning on something in a “suave” pose) フランス大使館の府連 パーティーに招待されちゃってさ Furansu taishikan no furen paatii ni shoutai sarechatte sa. 82 French embassy ‘s chapter party OBJ invi was CONT sa So I was invited to a party at the French embassy . . . . She then proceeds to turn down his invitation, citing that she “has to study” and asking him to “invite her next time,” and he is comically shocked. Endo is another one of the paralegals in the office, a young man prone to gossiping who has a somewhat less traditionally manly way of presenting himself. He is rather affected and over the top. He takes pride in knowing everyone else’s affairs and is often shown with Suetsugu, though he is something of the stronger personality despite being younger. This gossipy nature seems to be the reason why he uses ne the most frequently, as he is often found using it to convey that the information he is talking about is something everyone knows, or that he already knows the information someone else is talking about, a character trait commented on by the other characters. He is the only character to use ze, which he uses when urging Shibayama to hurry up in getting to the site of an investigation when dragging his feet. (32) 行こう ぜ しばやま。な？ Ikou ze, Shibayama. Na? Go VOL ze Shibayama na? Let’s go, Shibayama, right? Perhaps this is because he is not presented as manly enough to use zo, even when making a purposefully somewhat absurdly assertive statement. Suetsugu is the other male paralegal in the office, and most often is shown interacting with Endo, though he assists Kuryuu for Episode 4. He is an older man and portrayed as something as a sycophant, using mostly polite form. Most of his uses of na are comments or are self-directed. 83 The section head of the office (the boss, or bucho) is anxious, officious, also sycophantic to his higher ups, and self-absorbed. He spends a lot of time talking about his health problems. He also speaks in a very masculine style, mostly using plain form (to address his subordinates). He uses yo and na, and zo, as do many of the other male characters, though he also uses wa several times. Zo and wa seem to be associated with use in the plain form and in more casual and emotive situations, and yo is more insisting, while na seems to convey a personal feeling or reflection. These uses of particles seem to be part of his casual way of speaking. He also complains frequently about personal problems also using yo and na frequently in these instances, reflecting on the difficulty of a situation with na or using yo to give a statement more weight. This also shows particles being used in order to give a statement more affective content, or to make it seem more personal. Shimano Saeko is the victim of an assault case that the prosecutors’ office begins investigating for marriage fraud (inducing men to give her money thinking they would be married, only to dump them afterward) in Episode 3. She is a cooking instructor who seems to trade on her demure, ultra-feminine, “good wife” appearance, only to be in actuality cold, ambitious, and self-absorbed, with very little feeling. The presentation of her character in general is that she is shallow and self-serving. Despite her feminine way of presenting herself to others being a major plot point, she typically speaks in the plain style. She also uses the assertive particle yo quite a bit. It seems that this is also an example of how yo can be used to soften the tone of an utterance or interaction, as in the interaction below, talking about badly cut apples: 84 (33) だれが こんな下手なカットしたんですか Amamiya: Dare ga konna heta na katto shita n desu ka? Who SUBJ this badly did COP QUEST Who cut the apples this badly? しかたありませんよ 男性 は 普段 こんなことしないもの Saeko: Shikata arimasen yo. Dansei wa fudan konna koto shinai mono. Anything is not men TOP usually this thing can’t thing There’s nothing to be done, you know. Men usually don’t do this kind of thing. She recuts the apples, and everyone in the office is extremely impressed with her feminine style, saying things like “you’ll make a good wife someday,” and “what a woman.” Her usage of yo seems to be part of her overall warm, soft, feminine self- presentation, along with her usage of ne. It is notable that she uses wa the most frequently out of any of the characters, which is clearly to increase the stereotypical “softness” and femininity of the way she speaks. Her speech is marked by the usage of yo, ne, and wa, as well as by the plain form, which in this context seems to index familiarity or intimacy with those to whom she is speaking, and she does use the feminine form of the personal pronoun, atashi, as well as the question marker no rather than ka, all forms that have been noted as features of Japanese “feminine” speech in previous literature. Her speech could be described as self-consciously feminine, as she has purposefully crafted it as such. Officer Danbara is a police officer who plays a major role in Episode 4, which centers around a conflict about a case between the police and prosecutors, who believe different suspects should be focused on. He uses only three different types of particles, and most notable in his speech is his use of yo and na, as is shown below. 85 (34) 時間かけなくていいですよ。 Jikan kakenakute ii desu yo. Time take NEG good COP yo It’s fine if it doesn’t take much time. Here he is advising the prosecutors that they need not take much time on the case at hand. In the below example he is finally coming around to the idea that the prosecutors have helped apprehend the true criminal, and congratulates Shibayama on his role in capturing the perpetrator of the crime. (35) よく やった な 検事 さん Yoku yatta na, kenji -san. Good COP PAST na prosecutor HON You did well, Mr. Prosecutor. These seem to be sufficient to project his identity as a rather rough, manly sort of character despite an overall more frequent use of polite form (as he is speaking to his superiors on the job). He also uses ne in an assertive or pushy, argumentative way, as Kuryuu does, above, as can be seen in this exchange when he is criticizing the prosecutors’ work to their faces: (36) 今回 の 逃避で 検察 の まごつくかもしれませんだね Konkai no touhi de kensatsu no magotsuku kamoshiremasen da ne. recent POSS escape OBJ prosecutors POSS fluster perhaps POL COP ne こんな時だけで やってきってそうさ の砂子 の とされる Konna toki dake de yatte kite, sousa no manago no tosareru this time only OBJ do come CONT this NOM POSS lost child POSS いうの ふゆかい なんですよ iu no fuyukai nan desu yo. say NOM unpleasant what COP yo The prosecutors must be finding themselves flustered and confused by the suspect’s escape. He does not use wa, which may be a himself, and does not use the more stereotypically masculine particle the other characters, who seem more concerned with consciously presenting their masculinity. One might argue t and to some extent the section head, are attempting to overcompensating to a degree. Below in Figure 3.4 follows an analysis of the particles by the gend speakers. Figure 3.4. Percentage occurrence of particles by gender. Figure 3.4 presents the particles by gender out of each gender’s total utterances, thus compensating for the fact that men speak more frequently than women. Here, it becomes clear that men and women use dominate the usage of na and 0 1 2 3 4 5 6 7 8 9 10 yo ne na sa 86 reflection of the fact that he speaks very little about zo, unlike some of hat Danbara is masculine, while Shibayama and Egami seem masculine, and thus Particle Usage by Gender er of the yo and yo ne in equal numbers, while men sa, as well as make up all the users of zo, while women yo ne wa zo of male utterances of female utterances 87 dominate the use of wa. Men also dominate the use of ne, which seems to be at least partly a function of Kuryuu’s frequent usage of this particle, but also shows that ne is in fact frequently used by men in people’s minds (just as yo can be used frequently by women). The only particle in which women use more of the particle than men is wa, which does seem to fit with previous literature on it being a particle of feminine insistence or affective marking, though it is worth noting that men do use it, too, just as women do use sa and na. This is interesting because a created script might reflect more stereotype than real life, and yet the men use wa and the women na and sa at a rate somewhat comparable to the naturally occurring phone conversation data. Despite the language ideology associated with the feminine wa that is described in so much of the literature, and the way that ideology is used to present Misuzu’s and Saeko’s character, there is apparently enough recognition of the prevalence of wa in male speech or the usefulness in conveying an effect that it still appears. No women, however, use zo or ze whatsoever. This seems to suggest that the idea that these are male only particles and strong, almost outrageously, assertive is stronger. This may be because these particles occur so infrequently in real speech, with wa occurring much than they do, thus that the “ideal” form of them is what takes up residence in people’s minds, rather than a reality. Particle Usage by Politeness While the telephone conversations nearly all took place in the plain form, except one speaker who typically utilized the polite form, the polite form sees much more usage in the television show, no doubt because of its professional setting. In this data, in the cases of na, sa wa, and zo, particles appear more frequently in the plain form than in the polite, while for ne, yo, and especially 330 particles appeared in the plain form, while 164 appeared in the polite form, over the course of the two episodes, and these numbers are reflected in percentages (36 to 25, respectively). This supports the argument that engagement with a situation, as emotional content and personal opinion tend to be expressed more freely in casual or close conversations about particles. However, yo ne with yo and ne also appearing quite frequently in the polite form. form are close to being matched notably different pattern, with three times as many occurrences i the plain form. In the figure below (Figure 3.5) politeness level of particle utterances is shown by percentage. Figure 3.5. Politeness level of particle 0 2 4 6 8 10 12 ne yone 88 yo ne the opposite is true. Overall, in raw counts, some particles are a marker of affec , at least the way people think appears far more often in polite form than in the Yo and ne by utterances in the plain form, but yo ne presents a quite n the polite form as in -containing -containing utterances, percentages. na sa wa yo zo tive plain, in the polite plain polite 89 Yo ne is most often found in circumstances where the speaker is making a statement of some kind that they wish to mark with yo, but then either wish to soften it or make it seem somewhat less controversial by adding ne. It appears to retain the meaning of both of its components, but it does seem to give an effect of having slightly more of the sense of ne. Often the speaker extends the ne, or emphasizes it slightly, as well. The similarity in occurrence found with na in the telephone conversations comes to mind. The difference in usage between na and yo ne may be that yo ne is largely associated with the polite form, while na is largely associated with the plain. For example, Amamiya uses yo ne below in response to the section chief demanding if she and Kuryuu have thought out their plan to prosecute Saeko for marriage fraud. (37) 考えました よね Kangaemashita yo ne. Thought POL PAST yo ne We did think about it (yo ne) She says this in a rather polite and soft tone of voice, in general giving the impression that she is pleased that she was able to persuade Kuryuu to bring the case. This sort of indirect statement of opinion may go better with polite utterances than plain ones, which are quicker and more casual and thus fit with one particle or the other better. On the other hand, it may be because this sort of assumption is face threatening that it is used with the polite form in order to mitigate it and overall soften the effect, like the addition of ne might soften the effect of yo, or even vice versa, yo softening the effect of ne, which can seem insistent. Yo seems more personal and thus might cause the assumption of agreement inherent in the use of ne seem less presumptuous or demanding. Particle Usage Also coded for was the position of each particle in the utterance. These results a presented below in Figure 3.6 Figure 3.6. Utterance position of particles. It is clear that the final position is overwhelmingly the most used for all of the particles except sa. Again, similar to the results from the telephone conversation data, and sa both appear more frequently than other particles medially, while frequently medially than finally. “holding the floor,” or marking phrases as part of a narrative, list, or story of some sort as discussed previously. There including yo that it occur at the end of an utterance. The only utterances for which this is not true are examples of dislocation Ne, na, and sa are the only p all appear initially, though wa and ze, perhaps) have the strongest restriction in terms of location in the utterance. These 0 20 40 60 80 100 120 yo ne na 90 by Utterance Position . sa occur This seems to suggest that it has the function of also seems to be a strong restriction on any particle , similar to the telephone conversation data articles to appear on their own, and ne, na, sa, in a somewhat ambiguous case. It seems that yo sa yo ne wa zo ze final medial initial alone re na s more , . and wa (and zo 91 particles, notably, all have the function of insisting or asserting in common. Meanwhile, na is often used to mark personal thoughts, and sa as a filler or a marker that the utterance “belongs” to the speaker in some way, while ne is used to index some sort of agreement. All of these particles are freer in terms of utterance location, and can even be used as utterances on their own in this data. The meaning contained in each particle may relate to the difference in position in this case, or this may be a function of the type of role the particles are used to play in discourse (perhaps stemming from some inherent meaning). Particle Usage by Response The below figure (Figure 3.7) shows whether or not an utterance containing a particle was responded to, and in what manner. “Not applicable” is used when the scene ends on a cut, and so it is impossible to tell whether the utterance would have been responded to, or not. “Yes” refers to when there is a response, “self” to when the same speaker continues on with another utterance, and “no” when the speaker is not responded to in any way. The lack of responses was analyzed differently because in the television drama, as opposed to the telephone conversation, there were many much more obvious and dramatic instances of a complete lack of response, whereas in a phone conversation if one speaker stops participating it is the end of the conversation, thus providing a particular pressure to carry on. For example, in the drama many speakers turn away, or otherwise obviously withdraw from a conversation or make a clear choice not to respond to a given utterance, something not nearly as clear in the telephone conversations. Figure 3.7. Response to particle received. Immediately it becomes clear that the self prevalent in the data from the drama than they were in the telephone data. It seems likely that this is at least partly because of the difference between scripted dialogue in a television program and naturally occurring dialogue characters to finish their points and get across their thoughts and feelings to the viewer, things are not necessarily so neat in real life, and people are often interrupted by their interlocutors and lose or have to finish expressing their points or thoughts. In written dialogue, points are rarely left incompletely expressed except for character and story reasons. There are also more asides to oneself for dramatic effect in the situation telephone conversations present, with a strong requirement to respond to your partner and continue passes the discourse on and off, partly because you cannot see each other and the conversation is the only viable mode of interaction. might tend to switch off in discourse more frequently. 0 10 20 30 40 50 60 70 yo ne na 92 -continuations (the “self” category) are far more —while the drama is written to allow negotiate for control of the conversation before they can scripted dialogue, and a wider variety of situations than Thus, people Ne and yo ne have the feature of sa yo ne wa zo y self no n/a 93 being immediately followed by another utterance from the same speaker slightly more often than they are responded to by a conversation partner, while yo, wa, and zo are all noticeably more often responded to than followed by the same speaker. All three of these particles have some sense of insistence and personal emotion. This may lead them to be more often placed at the end of a speaker’s turn, in order to show that speaker’s strongest or final thoughts on a matter, or they may have a stronger draw in terms of provoking a necessary response than the other particles. Particle Usage by Utterance Type The figure below (Figure 3.8) indicates the types of utterances containing particles in terms of percentages. These were coded as statement, reporting personal thought, sharing information, stating opinion, question, and comment. It is worth noting that the sharing information category is less common overall in the data from the television drama, which might be a result of the differing situations that occur in a phone conversation and the situations in a workplace in a television drama, where the characters see each other every day and have some shared knowledge and information about the case at hand. Statements also seem to appear in great numbers, which may also be a result of the nature of the interactions in the television drama. The single utterance containing ze is a statement. This information shows some interesting differences between the particles. For instance, wa is dominated by statements, as is zo, which helps to show that these particles have a strong insisting element, or at least strongly code a personal emotion or opinion in an utterance. Na is overwhelmingly dominated by comments, and also features more in direct reports of personal thoughts than any other particle, which fits with it giving a personal or self- directed sense. Sharing new information is common with much so with the others. Zo Thus, wa may have a sense of insistence, but it is more affective in nature and not directly tied to an opinion directly expressed in the utterance itself. by statements and comments, which seems to fit with the description of it provided above, as it is a combination of the tendency found with found for yo (the statements), seemingly dropping most of the sharing information function of both yo and ne. Ne information, which gives a picture of it being used to either comment on a situation and seek agreement or index what is believed to be a common belief, or to share new information that the speaker wants to have accepted, or wishes to soften in some way, as was the dominant tendency in the telephone conversation data. share information, and, as stated above, often to hold the floor or present narrativ present new information in a way that is an attempt to seem detached or suave. Figure 3.8. Percentage of particles by types of utterances. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% NE YO NE WA NA 94 yo, ne, sa, and wa, but not as and yo are most often used to state opinions, but Yo ne is domin ne (the comments) with that is also dominated by comments, along with sharing Sa is chiefly utilized to SA YO ZO statement reporting personal thought sharing information stating opinion question comment wa is not. ated es, or to The below figure (Figure 3.9) particles were responded to. Figure 3.9. Responses to particles by type, percentage. Of course, there is only one utterance containing self-continuation. Self-continuation is the chief way that many of the particles are responded to, percentage-wise, except for response equally, while wa is dominated by responses. more instances in which an utter the others. This may indicate that they carry less pressure for a conversation partner to respond. This emphasis on self tendency for speakers to speak for some time and then finish as part of the 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% NE YO NE WA 95 Response Type by Particle presents the manner in which utterances containing ze, and it is responded to with a zo and wa—zo has self-continuation and Na, yo ne, and zo seem to have ance containing a particle is not responded to than any of -continuation may show, as discussed earlier, a greater scripted NA SA YO ZO n/a response comment continuation self-continuation question no nature 96 of the dialogue. Yo ne is not responded to at all by any comments or questions, which may imply something specific to the use of this particle and its role in discourse. This may relate to the usages of yo ne described above, that it is used in polite situations to soften a strong point, one that does not usually provoke either direct questions (further indirectness) or comments. Notably, yo is not as dominated by responses that are questions here, though they still appear more frequently than with most of the other particles, and instead sa features responses that are questions notably more than any other particle, as well as the most comments. This may imply some differences between sa and yo in the natural data and the scripted data, or some differences between phone conversations and face to face interaction, such as the ones in the drama. However, it is worth noting that these are both particles that seem to be of the insistent type, as described previously. Na, sa, and yo all show more comments and questions than the other particles (though ne also has some responses that are questions and a few comments, but not nearly as dramatically as these others). Wa and zo are also somewhat similar in terms of response type, featuring no questions or comments. Discussion These results allow a further refining of the understanding of particle usage, not simply how they are actually used, but how they are believed to be used and what they mean to the people who write television dramas (and, ostensibly, to the people who watch those dramas). This allows an interesting look at the underlying meaning of these forms in terms of how people actually think of them, the ideas they have about them, and their language ideology. It also allows insight into how particles are used in order to present a given identity to others, as these particle usages were chosen for the characters in order to 97 give them certain characteristics in the eyes of viewers and to encourage certain responses toward the characters. When observing the particle usage in these episodes, it becomes clear that particles form a major part of the way the personalities of the characters are presented to the viewer through each character’s dialogue—which particles each character uses, for example, or does not use. The breakdown is very clear between them, with some characters using some particles frequently that others never utilize at all. The vast majority of the instances of some particles (such as wa or zo) come from one or two characters. This seems to suggest that the particles used by each character form a major part of the presentation of their chosen characterization to the audience, and thus that a person’s particles use is something that can index personality traits or a mode of presenting oneself to others. Each of these characters has a certain strong impression in terms of how they are presented to the viewer, and their usage of particles adds to that impression. Thus, particles can be used to influence the way a person appears to others, and they can be associated with the effects they have by writers and provided intentionally in order to produce these effects. It is also clear that some particles occur far more frequently than others. The frequency in general reflects the same distribution found in the natural telephone conversation data, with one noticeable difference—in the television drama, yo appears more frequently than ne, while in the telephone conversations, ne outnumbers all other particles by a sizable margin. This could be because of the nature of phone conversations versus in-person speech, or because the situations presented in the television drama are of a far more dramatic nature than a casual conversation with one’s friend, involving 98 arguments, revelations, and insistent statements, or a combination of both reasons. However, yo and ne still clearly outnumber the other particles, though na is also noticeably more frequent than the rest. This implies a much wider gap between na and sa usage than was present in the telephone conversation data, where they were virtually equal in terms of frequency. Ne appears throughout the television drama data in a variety of ways. The most striking is probably Kuryuu’s use of ne. Not only does he use ne to present information that is not common knowledge (and thus is in his territory of information), he uses ne in a way that could be seen as pushy or passive-aggressive, presuming affective common ground where there is none, or that might in fact be something his interlocutor wishes to resist, as can be seen in the conversations with Saeko related above. Meanwhile, the use of ne to mark information already known or understood can be seen in Endo’s use of it, where it seems to be mainly used to show that he knows the information being spoken already and agrees as part of his personality as an inveterate gossip. One might even go so far as to classify ne into types, one of which is agreement ne, and the other is “pushy” or passive-aggressive ne, where the particle’s assumption of shared attitudes can be used to undercut an interlocutor’s argument and include them in the action of agreeing with you even if they might not wish to. Yo ne seems to be associated in the data with the polite form more than any other particle. It seems to be found in situations where the speaker is making a fairly assertive statement that may not be entirely fitting in this polite context, so that ne is added to give a sense of shared agreement. In general it has a sense of emphasis, but the ne and yo elements seem to work together in different ways. In all these cases, yo ne seems to be 99 used when there is shared information, or information the speaker feels should be generally agreed upon, that they want to bring special emphasis to. In this way, yo might be used to show that the speaker does feel something strongly, but the ne used to give a sense that the speaker believes that this is either a shared feeling or should be. Wa is used by both women and by men in the data. Some of the cases of women seem to be instances of using it to play up its role as a stereotypically feminine particle, especially its usage by Misuzu and Saeko. This usage of wa does seem to index personal affect and feeling, especially in Misuzu’s case, as she uses it in situations where she might be expected to be more emotional (with her lover, or when annoyed), but it is portrayed as a conscious, almost manipulative strategy to play up femininity on the part of these characters. Wa is also utilized by male characters, such as Kuryuu and Shibayama, and this seems to be chiefly as an emotive particle that conveys a strong emotional tone in the utterance. Na is typically used by male speakers in this data, which is a somewhat different result than that from the naturally occurring telephone conversation data. It may be utilized in this manner in accordance with the roughness literature such as the dictionary by Makino and Tsutsui describe it as reflecting—this may be an incidence of language ideology concerning na, or perhaps a stereotype associated with its use. It still seems to be used in order to reflect personal or self-directed thoughts and to add a reflective tone to the utterance in which it is used. Sa is found more frequently than other particle in the medial position, even more than it is found in the final position. This seems to be related to a holding the floor function, as previously discussed. It can also be seen to be used to start off utterances, 100 especially those where men in particular might be trying to seem cool, or may sound somewhat awkward. This seems to be related to sa’s use as a hedge overall, and it may be associated with male usage, particularly in this way, as another form of stereotyping. This may be one reason it is responded to more often by questions in the television data than that from the telephone conversations. Yo was the most common particle in the television drama data, which seems likely to be a result of the situations presented in the drama as compared to the phone conversations. It seems to be used chiefly as an insisting or emphatic particle, but we also see another use of yo, namely to show a more enthusiastic, emotional, or human side of characters who may not use many particles, such as Amamiya. This could be one way the softening sense of yo described by authors such as Matsui (2000) arises. Zo is rare in the data, though certainly not as rare as it was in the telephone conversations, and in the drama it is used solely by males. It seems to be used to index a stereotypical form of masculinity, even one that is over-the-top in some ways and played up almost absurdly. This might suggest the linkage of zo with a fictional or stereotyped form of masculinity, as well as its rarity in actual speech, such that it is available to be used this way, or might even be associated with an exaggerated presentation of masculinity of the type one is far more likely to find in fiction. It may sound “too” masculine to use in everyday conversation, or be seen as such. It also might be the male equivalent of the linkage of wa with highly stereotyped feminine speech, and thus be avoided by those who do not wish to fall into a stereotype. 101 CHAPTER IV GENERAL DISCUSSION First of all, when the two sets of data are compared, obvious differences appear. To begin with, yo is the most commonly occurring discourse particle in the television drama, while ne is by far the most common in the telephone conversations. Yo ne is also much more common in the television drama than in the telephone data. As discussed above, this may be a function of there being far more utterances in the television drama in the polite form than there are in the telephone conversations. Similarly, sa is comparatively less common in the television drama, which may be a function of the same factor, as sa seems to be associated with the plain form, at least somewhat. Wa and zo appear more frequently in the television drama, as is perhaps to be expected when it comes to created speech that can reflect the language ideology associated with these forms, and use that for its own purposes, rather than natural language. Similarly, ze appears in the drama (once) while it does not at all in the phone conversations. This may suggest some of the differences from the different mediums, the different types of interactions, and the differences between natural and scripted data. One factor commonly brought up when it comes to particle usage is that of gender marking through language. This factor has been widely described in prior studies and in descriptions of sentence particles, such as the ones offered by Makino and Tsutsui (1986, 1995) and replicated previously (pages 13-14) and in Shibamoto’s (1985) description of Japanese women’s language, Inoue’s (2006) work regarding the same topic, and in Kataoka’s (1995), Nakamura’s (2013), and Okamoto and Sato’s (1992) studies. One question is whether or not this was observed in the study. Were any of these particles 102 found to be used as gender markers, or were any of them used predominantly by one gender? In fact, the breakdown by gender in this study is relatively the same in both datasets, in that wa is used by both genders but more so by women, and sa and na are used by both genders but more so by men. There is a difference between the two datasets here, however, first of all in that the more dramatic difference in the phone conversations is between the male and female usage of sa, while in the drama the more dramatic discrepancy is between the male and female use of na. Yo ne and yo usage is equal between the genders in the television drama, while in the phone conversations yo was utilized more by males than by females. Also, while women used ne more in the phone conversations, men used ne more frequently in the television drama. This effect is most likely due mainly to the central character, who uses ne a great deal. However, there is no such obvious effect of character on the other differences. Perhaps this reflects a discrepancy between the idea of language usage held by the individuals who wrote the scripts versus the reality of natural conversation, or perhaps it is a factor of the very different situations between the drama and the set and focused interactions of a phone conversation. This seems to imply that overall, men and women are most likely use yo, ne, and yo ne in roughly equal amounts, men use na and sa more often, though women use it too, and vice versa for wa. This suggests that broadly speaking, male and female speakers are likely to use particles in generally equal proportions. This would be, overall, a quite different pattern of particle use as relating to gender than that typically described in previous studies. In a related point, particle usage was found in the telephone data to occur much more often in situation where the two speakers were of the same gender rather than 103 different genders, a result that has not been previously reported in studies on particle usage. This may be due to an attempt to provide a sense of personal interest in the conversation, providing a way to give a sense of one’s emotions and personal thoughts and opinions that is either perceived as more acceptable in same-gender conversations or, perhaps, less threatening. Another question explored in this study was whether or not particles appear more frequently in the plain or polite form. The most obvious difference in politeness level is that the polite form barely appears in the telephone conversation, but it is quite common in the drama. In both sources of data, particles appear frequently in the polite form, and in the telephone conversations appear more frequently in the polite form than in the plain, percentage-wise. This runs counter to prior descriptions in the literature, which may be because of the friendly nature of the conversations in the telephone data, even those involving usage of the polite form. However, in the drama particles appeared more frequently in the plain form than the polite form overall. This may also have to do with the more formal situations that appear in the drama, in a workplace, in legal and law enforcement situations, and so on, more similar to the situations described previously in the literature as provoking use of the polite form. Some of the characters in the drama, however, do seem to have moments of speaking more casually or in a friendly manner despite their usage of the polite form. This may suggest that when analyzing the form of speech in which particles appear, it is important to take the situation into account and what the speaker may be attempting to accomplish with the usage of particles or not, as well as with the plain or polite form or not. A speaker may use the polite form to seem refined or project an image of themselves, or be polite in general, but still wish to seem 104 engaged and interested rather than distant, especially in a friendly conversation, and thus use particles and the polite form, while in a more formal situation also utilizing the polite form, one’s personal feelings may be inappropriate and a distant image may be more desirable to maintain, and thus particles would not be used. Also notable is which particles appear in polite form in the drama—yo, ne, and yo ne, chiefly, with some na appearing as part of the construction na no. This is very similar to which forms appear in the polite form in the phone conversation (ne, yo ne, na, and yo), though in far greater numbers. Wa, wa ne, zo, and ze do not appear in the polite form at all, and sa only very occasionally in both data sets, which may imply something about the nature of these particles as compared to the others. In both situations, particles seem to occur quite frequently in the polite form, as well as in the plain form. However, some particles seem to be associated more with casual speech than others, while others, such as yo ne, seem to be dramatically more common in the polite form. Discourse particles are often described as final particles in much of the earlier literature (Hasegawa, 2010; Katagiri, 2007; Makino & Tsutsui 1986; Matsui, 2000; Saigo, 2006). In terms of utterance position, the major difference between the phone conversations and the drama is how much more common it is for particles to occur medially in the phone conversations than the drama, except in the case of sa, which occurs more frequently medially in the drama and slightly more frequently finally in the phone conversations. It seems likely that this medial use of sa is something of a prototypical usage, one that comes readily to people’s minds—thus, it is more available for use in fiction than it necessarily occurs in natural data (though it does also occur in natural data). Yo hardly ever occurs in any position but final in the drama, but does occur 105 more often in medial position in the phone conversation data, though these occurrences seem to be chiefly a result of dislocation, suggesting that the constraint that yo appear finally is quite strong there as well. However, it does seem that certain particles—na, wa, and yo ne, for example, have a stronger restraint that they not appear in positions other than final than others do. The initial and alone positions do not occur with wa ne, yo ne, na, wa, or yo in the phone conversations, and do not occur with yo, yo ne, zo, or ze in the drama, and their occurrences with wa are marginal at best (it is unclear what the speaker means to say, entirely). Na, however, does occur medially and alone in the drama. This does seem to suggest, though, that these particles do, in general, have a stronger constraint than ne or sa in this area. The usage of ne and sa in places other than the sentence final position has been described previously by Tanaka (1999, 2000), who describes ne as serving a variety of discourse functions depending on where it appears in the utterance, and both Suzuki (1990) and Squires (1994), who describe the appearance of sa in various positions in an utterance. However, the fixedness of yo, wa, and zo seem to be somewhat novel findings in this regard. In terms of responses to utterances containing particles, there seems to be far more pressure to reply to particle-containing utterances in the phone conversation data than in the drama. Part of this is no doubt for dramatic effect in the drama, as well as the effect of cuts away from the action or the end of a scene, and part of it is likely the function of the differences between a phone conversation and face-to-face interaction— the very nature of a phone conversation in which verbal means of communication are the only ones available, and you have communicated with a person specifically to talk to them, produces greater pressure to reply. The drama also features more self-response 106 than the phone conversations did, perhaps because of its scripted nature versus the naturalistic data of the phone conversation, complete with interruptions, wasted time, tangents, and everything that goes along with natural conversation. These results support Saigo’s (2006) claim that there is discourse pressure to respond to most utterances containing particles. Comparing the naturally occurring data to the created dialogue for the drama allows us some insight into the nature of what particles might mean in people’s minds and their ideas of them and the differences between that and the usage of particles in reality. The drama, for example, plays up the feminine use of wa and the masculine use of zo for largely comedic effect, as in Shibayama’s exclamation after apprehending a criminal (p. 78). It also, more subtly, utilizes the underlying meanings inherent in the particles and their usage to provide details and produce an overall effect in terms of characterization and the way the characters interact. For example, over the course of a scene, Egami starts off by saying, muri da ne (It’s impossible, you know?) to muri da yo, later on (it’s impossible!). Kuryuu and Amamiya, who are working on the case on which he is dubious, ignore him. Later, he repeats muri da yo, and later, mou muri da yo (it’s already impossible!), and then, right before he leaves for home in a huff, he states: (38) もうあきらめたほうがいいぞ mou akirameta hou ga ii zo already give up would be good It would be better if you gave it up already (zo). This escalation of the emphatic nature of his particles is used to show his increasing impatience and frustration with the other two, which leads him to clash with Amamiya, and then leave, progressing from ne to yo to the even more emphatic and over-the-top zo. 107 This method of utilizing particles is applied in a broader sense, as well, characterizing Kuryuu through his use of ne and other particles (yo, na, despite his use of the polite marker desu), Amamiya through her overall lack of particle use, Misuzu through her use of sa and wa, and so on. In a sense, the participants in the phone conversations can also be thought to be doing this, on their own. By choosing which particles they use and when, they are conveying something about their overall personality, as well as their stance on a given issue and their attitude toward the content of the utterance itself. Thus particle choice means a great deal in terms of personal expression and conveying attitude, stance, and personal affect. A person not using particles at all in Japanese, as Saigo (2006) points out, sounds oddly robotic. This is an important point—particles convey emotion and forms of response that speakers expect. The choice of particle and whether or not to use a particle at all is an important part of an utterance, and a consistent choice, one that is always facing a speaker when he or she begins to speak. Are particles best understood through analysis from a discourse analysis standpoint, and do they serve roles as discourse management markers, or are they better understood from the perspective of providing affective weight to an utterance or indexing speaker stance? After analyzing the data from both the telephone conversations and the television drama, I propose that the best way to understand the meaning of Japanese discourse particles is by analyzing them on two axes—the personal emotional response to the utterance encoded in the particle and the stance the particle projects toward the other speaker. The affect contained in an utterance can, in fact, be quite a different thing than the speaker’s overall stance toward that same utterance. For example, a speaker may be using a particle to show that he or she is engaged in a conversation (a discourse usage), 108 that he or she agrees with the conversation partner (stance), and that he or she is confident of the information being presented (affect). Both of stance and affect combine in order to convey the full meaning of the particle and fulfill the role the listener expects, that of an affective response or a stance taken on the part of the speaker. Thus, particles both provide affective weight to an utterance and index speaker stance, as two separate things. By choosing a particle, the speaker chooses a stance, as well as a method by which to reflect other personal factors such as gender presentation, professionalism, warmth versus coolness, or closeness versus distance. Particles can be used for pragmatic and discourse purposes as well, to escalate or de-escalate an argument, to soften a face-threatening act, or to hold or give up the floor. It seems that both of these methods of analyzing particle usage have some validity. Thus, a model that attempts to explain particle usage and provide a sense of what meanings particles carry, as well as how they are used, must include a way to account for all of these usages—indexing affect, stance, and serving as discourse markers. Thus, a combined model is the most effective manner in which to examine the usage of Japanese discourse particles. It is made clearer by the data examined that Kamio’s (1994) territory of information theory does not fully account for the usage of discourse particles. Specifically, it does not explain why ne is also commonly used with information that is fully in the speaker’s territory and not the hearer’s (an utterance such as, “tada ne, watashi ima sunde iru tokoro wa,” (“only, you know, the place where I’m living now”) from the data collected for this paper, for example, which refers to the speaker’s own experience, firmly in one’s own information territory, and is followed by the speaker continuing to describe her apartment, sharing new information that only she is in 109 possession of)—and simply this dismisses this widespread phenomenon, calling it “optional ne” (Kamio, 1995, p. 240) without explaining why it might occur. A vague sense of the meanings and usages of each of the particles examined has been given earlier in the analysis by examining each factor in turn. A more concise definition of the use and meaning of each particle, based on the data examined, follows. With ne, there is some expectation of shared understanding. This can be used with information that is not shared, but it has an affective connotation of suggested understanding all the same, “you agree with me, don’t you?” Thus it can be used to build rapport, but it can also sound demanding, assuming, bumptious, or insisting, depending on how reasonable it is to project this shared stance of understanding toward the conversation partner. We have seen examples of ne being used to build rapport, such as the conversations between WAT and MA4 on page 38 and between SUM and FE2 on page 48 in Chapter II, and another example where it is used to demand an agreement (such as Kuryuu’s discussion with Saeko presented on page 82 and 83). It can be used to hold or take the floor in discourse. The same speaker often continues to talk after using ne, as can also be seen in the conversation between SUM and FE2, as well as in the example from the drama mentioned above. Ne alone or initially is also often used to interrupt, or to provide oneself with an opportunity to take the floor, as described in some of the prior literature (Saigo, 2006; Tanaka, 1997). Tanaka (1997), for example, describes ne as a device used to mark turn entry points, acknowledgement-relevance places, possible transition-relevance places, and topic changes, while Saigo suggests that ne is used when a speaker wishes that “the figure emerging in the talk should be treated as a ground for the next proposition without further ado . . . and thus directs the 110 addressee’s acceptance” (2006, p. 35). Meanwhile, Kataoka describes the use of ne in letter-writing as reflecting both “realities such as age difference, politeness, and the degree of solidarity on one hand, and affective/emotional strength and other metalinguistic intentions on the other” (1995, p. 449-450). Kendall (1985) describes the use of ne as showing that the speaker would like the hearer to confirm or agree with his or her statement (or that a speaker might use it just to seem as if he or she wants confirmation in order to be polite, or for conversational implicature). This seems to reflect the idea that ne is used to imply that sense of agreement or confirmation, though the possibly more negative senses in how this could be used and the passive-aggressive nature of some utterances have not been much examined in prior literature. The description of an “empathetic common ground” (Lee, 2007) offered by prior studies does not seem to be inaccurate either when it comes to the description of ne, but it is important to note that this common ground can be created whether or not the participants agree, and can seem forced, unwanted, or pushy, as well. Ne can also be used to argue or make statements, despite the meaning it carries being one of assumed agreement rather than direct emphasis, as can be seen in Danbara’s dialogue as well from page 94, as well as the Kuryuu and Saeko example mentioned above. An example of this is Kuryuu’s use of ne when giving the information about the jacket, on page 82. This assumption of agreement is what seems to allow ne to be used both to project rapport and the affective common ground mentioned in the literature, and the more assertive and pushy, passive-aggressive usage as well, when it projects a common ground that isn’t accepted or desired by the other speaker. 111 Na, which was often used in a more reflective/thinking context, seems to have an effect of softening utterances or making them seem more personal. It seems to be irrelevant to inviting or demanding an agreement with the propositional content of the utterance or to requesting an acknowledgment of the content from the listener. Because of this, na seems to give a feeling that it is less insistent than ne or yo. The personal nature of this particle can make it seem more emotive or casual. The overall impression projected with na is that of a personal thought that is happening right now, or a reflection, not a strong claim. It is often used in comments and to report personal thoughts, as well as in the personal thoughts themselves. Na is quite casual in tone, as we saw used more in the plain form than the polite form, and not often by female speakers, so it seems to have slightly masculine connotations. However, female speakers do use the form na no, and this also appears in polite form. Sa also has a personal tone and gives a personal, subjective quality to the utterance, one with a connotation of personal experience, but it is more insistent. It is described as “insisting” by Suzuki (1990) and giving a flavor of obviousness to certain statements, and Squires describes utterance-internal and final use of both sa and ne (1994, p. 25), and states that though sa can be insisting or assertive, it can also serve to make a statement of opinion seem lighter in tone, and serves to establish ownership over the information presented. In the data examined in this study, sa appeared frequently in both data sets, more by males by females, more in male-male conversations than any others in the telephone data, in the final and medial positions in utterances, and more in the medial position in the television data. It most often appears in the contexts of sharing information, stating opinions, and statements, and it is most often associated with 112 responses of un and self continuations. In this sense, it could be seen to be in line with Squires’ description of implying ownership over a piece of information. It also gets more continuations from the same speaker than many of the other particles. It is often used to hold the floor or when continuing a conversation, or to take the floor, and offers a sense of continuation, as can be seen in the example on page 58, whether the utterance is in fact continued or not; sa seems to imply that there is more to follow, as can be seen in the instances of relating a narrative using sa from the telephone data. As mentioned above, it is clearly more freely used medially than most of the other particles except ne. Wa can be seen as softening, but can also be insistent or assertive in that it encodes a sense of emotional connection to the utterance. It has previously been described (Inoue, 2006; Makino & Tsutsui, 1986; Nakamura, 2013; Shibamoto, 1985) as being a typical indicator of Japanese’s woman’s speech, and as applying weak assertiveness (Makino & Tsutsui, 1986, p. 520). In the data from this study it does indeed seem to be less strong or “plain” than na, but still casual and carries a strong affective connotation. It appeared in similar situations as the other more insistent particles such as sa and yo—sharing information, statements, and stating opinions. In both the drama and the telephone conversations it was used by both men and women, but the usage by women in the drama seemed to be to play up its stereotypical feminine sense, as can be seen in the speech of Misuzu and Saeko. It seems to appear in more emotional utterances, such as Misuzu using it mainly when she seems annoyed, and in the telephone data as well, such as in the example on page 44. As has been observed in previous studies such as Matsui (2000) and Katagiri (2007), yo does give a sense of insistence, emphasis, or importance to the utterance. This 113 can either index disagreement toward or agreement with the other speaker depending on context. This can be seen in its use with more assertive types of utterances such as statements, stating opinions, and sharing information. It can also show engagement in or interest in the discourse, as was seen in the case of Amamiya’s use of yo in the drama, for example. It was also found in a variety of different types of utterances, including usage in questions, adding credence to this sort of usage, as well. This gives yo a broader usage than has often previously been described, and explains how it can be used in such apparently contradictory situations, both as assertive and as a “softener” of a sort, such as in the usage described by Uyeno (1971), as it can be used to soften by showing interest, commitment in the conversation, emotion, or affective engagement. Matsui’s analysis of yo as guaranteeing relevance seems relevant to the results, and more useful in explaining the usage of yo both to agree and disagree and to show interest in the discourse at hand than the presenting new information that the interlocutor does not yet have or the strong belief in truth value described in much of the other literature (Cheng, 1987; Kendall, 1985; Masuoka, 1991; Maynard, 1993; Ooso, 1986; Tsuchihashi, 1983; Uyeno, 1971). The combination particles such as wa ne and yo ne do seem to carry the meaning of both component particles, but they also seem to have some composite meaning and to be used in more specific circumstances. These particles have not been frequently described, though yo ne was mentioned by Saigo (2006) and Kendall (1985). It seems as if in general wa ne mostly follows the usages of ne in many situations, though it does show some differences in others, such as appearing more often in questions and with slightly different response pattern, while yo ne seems more unique from its component parts, being used frequently with the polite form, and to provide a sense of commentary 114 similar in some ways to na. Wa ne was provided by Makino and Tsutsui (1995) as a feminine alternative to na, however, yo ne seemed more likely to perform some of those functions in the data examined here. Yo ne seems to be used in cases of strong agreement or insistence, as in sou desu yo ne (yes, it is like that, isn’t it)? This seems to be an instant of yo serving to intensify the sense of rapport and agreement already conveyed by ne, or yo being softened by ne’s indexing of shared agreement. A major factor in the choice of whether or not to use na or yo ne may be politeness level, as yo ne appears more in the polite form. Zo, which has not frequently been studied in past examinations of particles, and which is usually described as a strongly masculine particle (Makino & Tsutsui, 1986; Kendall, 1985; Nakamura, 2013) did not appear frequently in either the telephone data or the television drama, though it did appear more in the television drama, all in the usage of male speakers who were presented in situations where manliness might be called for. It appears to be used more in scripted data, as a marker of an idea of manliness, than in actual speech. Two of the usages of zo in the telephone conversations were from a female speaker and were quotations, while the other two were from a male speaker. In all uses it seems to have a very strong sense of emphasis and assertiveness. This may imply that zo may be or may be becoming more of a fictional or stereotypical, idealized particle that portrays masculine in an extreme or over the top fashion than something people may use in typical conversations in everyday life. 115 CHAPTER V CONCLUSION Clearly, particle usage varies from situation to situation and from speaker to speaker. The widest field possible will make any conclusion drawn that much more valid. It is hoped that this study will add to that field, with both the scripted data of a TV drama and natural data from telephone conversations being analyzed. The difference between the phone conversations and the face-to-face communication in the drama alone suggests a wide and productive area of study for the variety of ways particles are used to take a stance and reflect personal views, positions, and identities toward the situation at hand. It is also clear that particles are indeed used in the polite form in some situations, if not even more often than in the plain form at times, and that this usage of particles may also be an interesting area of specific study, how and when particles do indeed appear in the polite form, not simply the plain. Certain particles, such as ne or sa, appear in various sentence positions, while others are more fixed, such as yo. Particles also frequently serve as a pivot for dislocation. It is clear that particles help to convey much of the tone and affective content of an utterance in Japanese. The lack of a particle has been observed to sound terse or overly direct, which may be one reason so many particles appear in phone conversation, and is borne out by the characterization of Amamiya in the television drama. However, overuse of a particle may sound unnatural or forced, or like one is “trying too hard,” as can the usage of a particle that does not seem to fit one’s personality, as can be seen in the case of Egami. These might prove to be interesting factors to keep in mind when examining in more detail particle use in naturalistic conversation, as well—how speakers 116 may respond to particle usage they perceive to be “too much” or out of place, or insufficient. Particles are an essential feature of discourse, and also of personal presentation. While actual usage is more fluid, much of what is found in the television drama is backed up by the natural data gathered, including the use of wa by both men and women. Speakers are more likely to use particles, it seems, when speaking to a listener of the same sex as themselves, perhaps because the atmosphere is inherently more casual, or there is less pressure to speak gender-appropriately. At the same time, certain speakers use particles more than others—both in a given conversation, and overall, across conversations, as revealed by the drama. When a speaker heavily uses particles in a conversation, it seems as if this can take two forms—either as the dominant speaker, or as the more passive, commenting speaker. It can be seen from the data that there is a strong pressure to reply to utterances containing particles, though at times and with certain particles it is stronger than at others. Comparing and contrasting the particles to one another enables a deeper analysis of their meaning and use. It becomes clear where various particles fit vis-à-vis the other particles in the system, in what ways they are used more and less than the others, and in what situations the same speaker chooses to use one over the other. More in depth analysis of these processes is required, but it is hoped that this can serve as a general overview of the discourse particles of Japanese and how they compare to one another in use in more than one discourse situation. Particles seem to be able to be classified broadly into “insisting and emotive” particles (wa, sa, yo, zo, ze) and “questioning and thoughtful particles” (ne, wa ne, yo ne, na), though they each have clearly specific connotations and uses, as described above. Ne, yo, and na are by far the most common 117 particles, across both data sets. This may be because they are more productive, as a result of grammaticalization, or because the situations in which they are used are broader and come up more frequently in discourse. The use of corpora for the study of pragmatic elements such as particles offers many possibilities, as it offers a wide range of data usage, and it is hoped that others will utilize this area in studying Japanese discourse forms such as particles, along with others such as evidentials and modals. The comparison of natural data to scripted data such as television also seems to be a productive area that may illuminate such concepts as identity construction and the presence of language ideology and how it is reflected in media. As media goes on to influence those who consume it, this in turn creates the language ideology that goes into it again. Studying media is a good way to ascertain what people think of language, and this in turn influences their actual language use. This can then also be compared to natural language use, and the differences and what they might mean examined. As it is impossible to use Japanese in a natural fashion without using discourse particles, it is hoped that an in-depth study of these particles will prove useful to those attempting to teach natural-sounding Japanese to others. The particles themselves have broad senses of meaning, but they are used in fairly specific ways, and often to mean fairly specific things—thus they do seem to carry some meanings. Hopefully, this study, despite rather broad definitions, will aid in determining what exactly it is these particles mean and accomplish in Japanese utterances, by helping to show which prior explanations prove the most convincing and to provide new insight as well, as well as aid in the explanation of that to students of the language. 118 REFERENCES CITED Cheng, C. (1987). Syuuzyoshi: Hanashite to kikite no ninshiki no gyappu o umeru tame no bunsetsuji. Nihongogaku, 6 (10), 98-109. Cook, H. M. (2002). The social meanings of the Japanese plain form. In N. Akatsuka, S. Strauss, & B. Comrie (Eds.), Japanese/Korean Linguistics, 10, (p. 150-166). Stanford, CA: Center for the Study of Language and Information Publications. Dunn, C. (1999). Public and private voices: Japanese style shifting and the display of affective intensity. In G. Palmer, & D. Occhi (Eds.), The Languages of sentiment (p. 107- 127). Hasegawa, Y. (2010). The sentence-final particles ne and yo in soliloquial Japanese. Pragmatics, 20 (1), 71-89. Hasegawa, Y. (2003). Soliloquy for linguistic investigation. Studies in Language, 35 (1), 1-40. Hopper, P. J. & Traugott, E. (2003). Grammaticalization. Cambridge: Cambridge University Press. Inoue, M. (2006). Vicarious language: Gender and linguistic modernity in Japan. Berkeley: University of California Press. Kamio, A. (1994). The theory of territory of information: The case of Japanese. Journal of Pragmatics, 21, 67-100. Kamio, A. (1995). Territory of information in English and Japanese and psychological utterances. Journal of Pragmatics, 24, 235-264. Katagiri, Y. (2007). Dialogue functions of Japanese sentence-final particles ‘Yo’ and ‘Ne’. Journal of Pragmatics, 39, 1313–1323. Kataoka, K. (1995). Affect in Japanese women's letter writing: Use of sentence-final particles ne and yo and orthographic conventions. Pragmatics, 5(4), 427-453. Kendall, S. A. (1985). Japanese sentence-final particles as commitment markers. Proceedings of the 11th Annual Meeting of the Berkeley Linguistic Society, 166-174. Lee, D. Y. (2007). Involvement and the Japanese interactive particles ne and yo. Journal of Pragmatics, 39, 363-388. 119 MacWhinney, B. (2007). The Talkbank project. In J. Beal, K. Corrigan, & H. Moisl (Eds.), Creating and digitizing language corpora, vol. I: Synchronic databases. New York/Houndmills, Basingstoke, Hampshire: Palgrave-Macmillan. Makino, S. & Tsutsui, M. (1986). A Dictionary of basic Japanese grammar. Tokyo: Japan Times. Masuda, K. (2007, March 26-28). Use of Japanese sentence-final particles in peer conversation. Paper presented at the 17th International conference on Pragmatics and Language Learning, Honolulu, Hawaii. Masuoka, T. (1991). Modariti no bunpō [The Grammar of modality]. Tōkyō: Kurosio. Masuoka, T., & Takubo, Y. (1989). Kiso Nihongo bunpō [Basic Japanese Grammar]. Kuroshio Shuppan, Tokyo. Matsui, T. (2000). Linguistic encoding of the guarantee of relevance: Japanese sentence- final particle YO. In G. Andersen & T. Fretheim (Eds.), Pragmatic markers and propositional attitude (p. 145 – 173). Amsterdam: John Benjamins Publishing Co. Maynard, S. K. (1986). On back-channel behavior in Japanese and English casual conversation. Linguistics, 24 (6), 1079–1108. Maynard, S. K. (1993). Discourse modality: Subjectivity, emotion, and voice in the Japanese language. Amsterdam: John Benjamins. Nakamura, M. (2013). Constructing the Japanese language through translation. Tokyo: Gendai Shokan. Narrog, H. (2005a). On defining modality again. Language Sciences, 27 (2), 165-192. Narrog, H. (2005b). Modality, mood, and change of modal meanings: A new perspective. Cognitive Linguistics, 16, (4), 677-731. Narrog, H. (2009). Modality in Japanese: The layered structure of the clause and hierarchies of functional categories. Amsterdam: John Benjamins. Nitta, Y. (2000). Ninsiki no modaritii to ninsyoo. Tokyo: Hituzi. Okamoto, S. & Sato, S., (1992). Less feminine speech among young Japanese females. In K. Hall, M. Bucholtz, & B. Moonwomon (Eds.), Proceedings of the Second Women and Language Conference (p. 478–488). University of California Berkeley. Onoe, K. (2001)[1990]. Bunpo-ron – chinjutsuron no tanjoo to shuuen. Bunpo to Imi, (p. 278-300). Tokyo: Kurohio Shuppan. 120 Ooso, M. (1986). Goyoo bunseki: ‘Kyou wa ii tenki desu ne’ – ‘Hai sou desu.’ Nihongogaku 5 (9), 91-94. Saigo, H. (2006). The pragmatic properties and sequential functions of the Japanese sentence-final particles ne, yo and yone. (Doctoral dissertation). University of Durham, Durham. Shibamoto, J. S. (1985). Japanese women’s language. New York: Academic Press. Shibatani, M. (1990). The languages of Japan. Cambridge: Cambridge University Press. Sperber, D. & Wilson, D. (1986). Relevance: Communication and cognition. Oxford: Blackwell. Squires, T. (1994). Discourse analysis of the Japanese particle SA. Pragmatics, 4 (1), 1- 29. Suzuki, H. (1976). Gendai Nihongo ni okeru shuujoshi no hataraki to sono soogo shoosetsu ni tsuite [On the function of sentence-final particles in modern Japanese and their correlations]. Kokugo to Kokubungaku [Japanese Language and Literature], 11, 58–70. Suzuki, R. (1990). The role of particles in Japanese gossip. Proceedings of the Sixteenth Annual Meeting of the Berkeley Linguistics Society (p. 315-324). University of California Berkeley. Tanaka, H. (1999). Turn-taking in Japanese conversation: A study in grammar and interaction, Amsterdam: John Benjamins. Tanaka, H. (2000). The particle ne as a turn-management device in Japanese conversation. Journal of Pragmatics, 32, 1135-1176. Tsuchihashi, M. (1983). The speech act continuum: An investigation of Japanese sentence-final particles. Journal of Pragmatics, 7, 361-387. Uyeno, T. (1971). A study of Japanese modality: A performance analysis of sentence particles. (Doctoral dissertation). University of Michigan, Ann Arbor, Michigan.