The 2000 ATJ Seminar took place in San Diego on Thursday, March 9. Several hundred members and others attended nine sessions throughout the day, hearing presentations on linguistics, literature, language pedagogy, and second language acquisition. On the following pages are reprinted abstracts of the individual presentations, ordered alphabetically by the names of the presenters.

"The Usage of Kanji in Modern Japanese and the Selection of Kanji in Instruction"
Nobuko Chikamatsu (DePaul University), Hironari Nozaki (Aichi University of Education), Shoichi Yokoyama (National Language Research Institute of Japan), Steve W. Ryner (DePaul University)

In Japanese language instruction, many teachers introduce kanji characters according to the selection and ordering of characters found in their selected textbooks. However, it is not clear that kanji characters selected in this manner are the most useful to learners of Japanese. For many teachers and students, kanji present a great obstacle to reading authentic materials such as newspaper articles at the advanced levels. This study focuses on the question, "How many and which kanji should be introduced for advanced Japanese reading?" In developing a systematic approach to teaching kanji, one of the important criteria for selecting "useful" kanji is kanji character frequency. A recent Japanese corpus linguistic study showed that the top 500 most-frequent kanji characters used in a major newspaper accounted for approximately 80% of total kanji usage (Chikamatsu et. al., in press). In this study, we focus on these 500 most-frequent characters to examine how useful it would be to teach these 500 characters for the purpose of reading authentic materials in Japanese. The newly developed kanji word frequency list is closely examined and analyzed to answer the following questions: (1) How are these top 500 characters used in kanji compounds in newspapers? (2) What are the highest-frequency kanji compounds in newspapers? and (3) How are these top 500 characters introduced in the most commonly used basic/intermediate textbooks of Japanese (e.g., Nakama, Yookoso, etc.)?

Finally, as the pedagogical application of the current study, a new Web-enabled software application is presented, JAVA KANJI FLASHCARD 500, in order to introduce the top 500 newspaper kanji characters. The program contains 500 character cards with the frequency ranking, on/kun reading, English gloss, stroke number, stroke animation, five high-frequency two-character compounds, and sample sentences. It has three functions: Browsing, Searching, and Drilling. The program is available at www.nuthatch.com/java/kanjicards with no requirement of any Japanese operating system (Japanese Language Kit, Kanji Kit, etc.).

"The Importance of 'Tsukiage' "
Yoko Collier-Sanuki (University of British Columbia)

In rating the language proficiency level of a speaker, the ACTFL OPI (Oral Proficiency Interview) employs a "proving" technique, "tsukiage." Its purpose is "to discover the ceiling or limitations of the interviewee's proficiency, i.e., the patterns of weakness, by raising the level of the interview to the next-highest major level" (Buck 1989, pp. 4-6). This definition of "tsukiage" unfortunately gives a negative impression to many teachers. In fact, "tsukiage" may mean in daily Japanese an attack on company administrators by union members, which underscores the negative connotations. It is, however, important to understand that the use and implications of "tsukiage" are different in tests such as OPI and in classrooms. For example, letting students take the initiative in the classroom and encouraging them to speak more in conversation is a simple but hard-to-accomplish "tsukiage" task, especially for sympathetic and/or talkative teachers. The goal of this paper, therefore, is to clarify the misapprehension, discuss the benefits of "tsukiage" in raising students' awareness and motivation for improvement, and encourage teachers to apply the concept in the classroom. "Tsukiage" is an important and effective technique to effectively develop a student's ability.

"Functions of Pitch Contours in Japanese Clauses: A Quantitative Comparison of Native and Non-native Perception"
Sanae Eda (The Ohio State University)

In this talk, I present the results of an experiment that examined the process of the acquisition of intonation by comparing the ways in which native and non-native speakers of Japanese perceive the differences of pitch contours in Japanese clauses and arrive at the intended interpretations of the utterances. The subjects were first asked to listen to a pair of utterances that differed minimally in some aspect of their prosodic structures, and judged whether the two utterances conveyed the same intention on the part of the speaker (discrimination task). They were then asked to "interpret" each utterance by performing forced choice paraphrasing task (interpretation task). The research findings can be summarized in the following three points: 1) among the NS of Japanese, speakers of the Tokyo dialect performed differently from speakers of other dialects; 2) NS group performed significantly better than NNS group in both discrimination and interpretation tasks; 3) while the proficiency test score turned out to be a significant predictor of NNS performance in the interpretation task, the same was not the case with their performance in the discrimination task. The pedagogical implications of these findings are discussed.

"What Does the Recalling Past Construction Tell Us about the Extended Predicate 'no'?"
Naoya Fujita (Pacific University)

I argue that the recalling past construction exemplified as "Kyoo-wa kinyoobi datta ne" reveals the pragmatic mechanism of the extended predicate ~n(o) desu. The recalling past construction presents an intertesting syntactic restriction: if a non-stative predicate appears in this construction, then the predicate must be in the form of extended predicate, as in "Kimi-wa ashita nihon-ni iku-ndatta ne," as opposed to "*Kimi-wa ashita nihon-ni itta ne." Based on the fact stative predicates do not have such a restriction in the recalling past construction, I argue that a function of the extended predicate is to "stativize" a predicate. Stative predicates are characterized by [-controllable], meaning that Event X is beyond the speaker's control. For instance, "Dooshite nihongo-o naratteimasu ka?" sounds as if the speaker might not be interested in the listener's reaction, because "naratteiru" is [+controllable] and remains within the speaker's discourse territory. On the other hand, "Dooshite nihongo-o naratteiru-ndesu ka?" sounds natural because the extended predicate neutralizes the controllability and makes the discourse information accessible to the listener. This analysis further explains the psychology of recalling. Recalling involves one person who acts as the speaker and listener simultaneously. In order for the speaker to talk to himself, he must make the discourse information available to himself by stativizing the predicate and then neutralizing its discourse territory so that it becomes accessible to himself as the listener.

"Tanizaki Genji and Manga (Yamato) Genji"
Mayuko Fujiwara (San Francisco State University)

Tanizaki's "Genji monogatari" (1979) is the third translation of Genji that he worked on. His aim in the third version was to popularize "Genji monogatari" to as many people as he could by using easier words than his first two versions. His faithful yet sophisticated translation has been attracting readers for years.

Yamato's "Asakiyumemisi" is well known as a comic (manga) version of "Genji monogatari." There have been a lot of arguments both for and against considering it as a literary work. Yet, in both good and bad ways, "Asakiyumemisi" has a powerful influence on readers, and it has great literary value even though it is a comic.

Basically, "Asakiyumemisi" is faithful to the original "Genji monogatari." However, under elaborate study and a lot of research, Yamato cut some parts and added her own ideas effectively to make the story more dramatic and passionate. Her great sensitivity and insight into each character's feeling deepened the story as well. These efforts to make it "fun and easy" to read have encouraged more people to become interested in the original text and other translations. Even though "Asakiyumemisi" is a valuable work, it is not complete by itself as a literary translation of "Genji monogatari." Still, "Asakiyumemisi" has something that the original text does not have, and it has played a major role in introducing "Genji monogatari" to many people as enjoyable reading.

"Analysis of L2 Writing Strategies in Japanese"
Yukiko A. Hatasa (University of Iowa)

Previous research on writing has shown a similarity between L1 and L2 writing processes, suggesting L1 transfer. Both L1 and L2 writers produce text through three sub-processes--planning, actual writing, and revising--which interact in a non-linear fashion. Skilled writers tend to pay attention to organization and pragmatics, evaluate global aspects, and do all levels of editing. The L2 writing process is constrained in that L2 writers focus more on linguistic level and code-switch between L1 and L2 (Whalen and Jard, 1995; Sasaki and Hirose, 1996). However, the target language in most of these studies is English, and little is known about other languages, especially a non-cognate language with a radically different writing system. Therefore, the present study examined the L2 writing processes of seven advanced JFL writers with English backgrounds. The subjects were asked to write two argumentative texts, one in English and the other in Japanese, and to verbalize their thoughts during composition. The data were analyzed qualitatively and quantitatively in terms of the time spent on and the frequency of occurrence of each writing strategy. The results confirmed the overall similarity between L1 and L2 writing processes at the pre-writing stage, but also contradicted the previous findings in that the subjects of the current study spend more time on pre-writing than those in other studies. The pre-writing time, however, was negatively correlated with the writing quality. Likewise, the writing fluency was positively correlated in the quality of L1 composition but was negatively correlated with that of L2 composition. In addition, the reliance on L1 during pre-writing was not necessarily characteristic of poor writing. Furthermore, extensive evaluation of surface-level features, especially spelling, was performed even at this advanced level, reflecting the complexity of the writing system.

"Development of the Oral Simulated Interview System (OASIS)"
Kazumi Hatasa and Atsushi Fukada (Department of Foreign Languages and Literatures, Purdue University)

The most common method to evaluate learners' ability in oral production is an individual interview. The problem, however, is that giving oral interviews and evaluating each student is a time-consuming and logistically difficult task for teaching staff, especially when the enrollment becomes large. The proposed system, named the Oral Simulated Interview System (OASIS), is designed to reduce the burden on instructors without making interviews too unnatural. OASIS makes use of digital video clips, photographs, and drawings to elicit students' oral production. Students' utterances are recorded as digital sound files for easy retrieval. Instructors can later listen to the sound files for evaluation. Digital files can be listened to in no particular order for more efficient evaluation. It is also possible to listen to the interviews with students for feedback. OASIS, however, is not intended to recreate the smooth interview that a human interviewer can perform. Twenty-seven students in a second-semester Japanese course volunteered to evaluate the system. They thought positively of the system as a tool to practice for oral interviews. The majority still preferred a real interview test over the computerized test, although none of them found the computerized test to be unacceptable. The current system appears to be a useful tool for practice, but requires improvement as an alternative to real interviews for testing.

"Diachronic Variation in Honorific Forms"
Junko Hibiya (Keio University)

There are two types of honorific form which are used to show the speaker's respect for an individual or the individual's action or state. Type One: prefix 'o' + verb root + 'ni naru' or special honorific verb (lexical) Type Two: verb 'reru/rareru' form (regular) According to Makino & Tsutsui (1986) and Miura & McGloin (1994), the differences between Types One and Two are: 1. Type Two is less polite than Type One. 2. Generally, Type One is used in women's speech whereas Type Two is preferred in men's speech. 3. Type Two is more common in formal writings, e.g. newspapers.

My preliminary observations suggest that in recent years the use of Type One honorific form has become more frequent in formal writings. In order to explore whether a change is in progress, 24 essays which were written in memory of the deceased between 1950 and 1999 for the journal "Gengo Kenkyu" were examined. Analysis of the data revealed the following three points: 1. Before 1980, those who were born between 1890 and 1919 used only Type Two honorific forms. 2. Since 1980, the use of Type One honorific form has increased (5% -30%) regardless of writer's age. 3. After 1990, an innovative double honorific form, e.g. "ossharareru, o-yomi-ni-narareru" is sporadically observed in essays written by those who were born after 1925.

"Textbook Talk versus Real Talk: What We Think We Do Is Not What We Do"
Kimberly Jones and Tsuyoshi Ono (University of Arizona)

While dialogues in commonly used Japanese textbooks are far more natural than in the past, they are still clearly distinguishable from real talk. Textbook dialogues: 1) often lack contextual notes to help explain the contexts in which they would sound natural; 2) are typically short, representing very brief encounters or seeming as if they are decontextualized segments from longer interactions; 3) are primarily for exchanging information; 4) are composed of utterances neatly matched into pairs of complete sentences, such as "question-answer" pairs; 5) contain a high amount of new information per utterance; 6) contain little negotiation of meaning between interactants, and little use of linguistic devices used for negotiation (e.g., repetition, repair, postposing, "aizuchi," sentence-internal particles such as "ne," and "fillers" such as "ano"); and 7) contain few examples of multiparty talk.

Unlike textbook dialogues, naturally occurring Japanese conversations, both informal and formal ones, are highly interactive. Utterances are typically very short, are not complete sentences, and contain relatively little new information at a time. Meaning is constantly negotiated by devices such as those listed in item 6 above. Some features of actual talk would be easy to incorporate in dialogues used for teaching. We could create dialogues with more fragments and postposing, with fewer neat pairs of sentences, and with utterances that do not present too much new information at a time. Longer sentences could be broken up into short phrases with pauses between them. Repetitions, pauses, fillers, sentence-internal final particles, and back-channels are all easy to incorporate.

Teachers and textbook writers should examine real talk so they can continue to create ever more realistic dialogues. In addition, Japanese language students and students of Japanese language pedagogy should be given opportunities to hear and analyze real talk so they can learn how to effectively contextualize and use existing pedagogical materials. This will help students learn to participate in Japanese conversations outside the language classroom in a myriad of interactional contexts.

" 'Otoko no Genji monogatari' no choosen: Hashimoto Osamu. 'Yoohen Genji monogatari' ga mezashita mono"
Shiho Kanazawa (San Francisco State University)

Hashimoto Osamu's modern translation of "Genji monogatari," "Yoohen Genji monogatari" (Tokyo: Chuokoronsha, 1991), is quite unique compared with other modern translations because of its monologue style of Hikaru Genji for the first forty-one chapters, from Kiritsubo to Maboroshi. He focuses on two viewpoints: men's viewpoint (although the original which is written by Lady Murasaki has a women's viewpoint) and the modern viewpoint. By using monologue style from these viewpoints, Hashimoto tries to fill in the meanings between the lines which are not expressed by words in the original--for example, the description of the workings of Hikaru Genji's mind told by his monologue, the political situation of Hikaru Genji which cannot be expressed from the women's viewpoint, and historical background information about the culture, society, and common sense of the Heian era. Since he adds many sentences and avoids the ambiguity of the original, readers can easily read the story. On this point, his translation is successful. But his translation gives an impression which moves a bit far from the original, although it includes all the stories and structures of the original. The identity of Hikaru Genji, which derives from the viewpoint of modern men, sets Hashimoto's translation apart from the original, and it produces the unexpected transformation that he intended. The expression "Rokujo Miyasudokoro" indicates a conflict between the original and Hashimoto's translation. From the standpoint of accuracy, it is quite difficult to evaluate the "Yohen" (which means transformed) "Genji monogatari." But it is highly admirable that Hashimoto suggests the depth of the story of "Genji monogatari" through his translation.

"Setouchi Jakucho yaku 'Genji mongatari' ni tsuite"
Yoko Katagiri (San Francisco State University)

"Genji monogatari (Tale of Genji)" translated by Jakucho Setouchi into modern Japanese has sold nearly 2.1 million copies since its publication in 1996. This huge commercial success of Setouchi's translation is quite recognizable, and this phenomenon also raises the question, "Why 'Genji' now?"

There are at least four major factors to answer this question: (1) The reputation of Setouchi: she is one of the most popular writers in Japan with her interesting stories; she is also known for having become a Buddhist nun in 1973. (2) The relevance of "Genji monogatari" to today's people and society: Setouchi claims that although Japanese society has modernized rapidly, the tale still teaches us the basics of love. (3) Gorgeous yet well-figured book design: It is published in a boxed cloth edition with colorful drawings, which easily remind consumers of the dynastic magnificence of the Heian era. (4) Setouchi's literal interpretation: In order to make "Genji monogatari" readable, she inserted many grammatical subjects, avoided the use of too much "keigo," and broke up long sentences.

Beyond those factors, the popularity and the fact that today's readers largely welcomed Setouchi's close translation reflect the tale's universality and the conservatism of Japanese society, in spite of Setouchi's strong feminist voice arguing that Japanese women have made much progress in the last 10-20 years. To make Setouchi's characteristics clear, I compare part of her translation with Akiko Yosano's translation, which was published in the Meiji era. Both Se touchi and Yosano are famous as modern female writers and feminists among Japanese, ye t their translations have articulate distinctions that display the possibility of various interpretations at different periods of modern Japanese society.

"An Analysis of Allo-repetition for Agreement and Disagreement in Japanese Conversation"
Chisato Koike (University of California at Los Angeles)

In this study I analyze the uses of immediate allo-repetition, i.e., repetition of other speaker's utterances, for agreement and disagreement in videotaped conversations between Japanese native speakers. I examined the uses of allo-repetition in terms of syntax, prosody, non-vocal movements and sequential organization, and incorporated participants' comments into my analysis. I found that allo-repetition for agreement and disagreement manifested different characteristics. When speakers used allo-repetition to indicate their agreement with the prior utterance, they repeated with falling intonation while looking at their interlocutors and nodding. When speakers used allo-repetition to indicate disagreement with the prior utterance, they mitigated disagreement by using syntactic devices, i.e., the use of "X tte iu ka" ("to say X or"), "maa" ("well ..."), and "ka naa" ("I wonder ...") as monologue. Speakers also used prosodic devices, i.e., lengthening the last mora of an utterance, and non-vocal devices, i.e., tilting the head and diverting the gaze. In addition, while allo-repetition for agreement occurred early in a sequence, often overlapping with the prior turn, allo-repetition for disagreement was sequentially delayed. My analysis suggests the importance of analyzing conversational data from multiple perspectives, and it also has pedagogical implications for the teaching of repetition as a conversational strategy.

"Typological Differences between Japanese aru and English do"
Yoshimi Maeno (Wellesley College) and Takehiro Kanaya (University of Montreal)

This paper compares, from a typological viewpoint, two verbs: Japanese aru and English do. Those two play the most characteristic role in their respective languages, not only as a verb but, even more importantly, as a verb auxiliary. Whereas the use of the auxiliary do is as generalized as in negation, question, emphatic and proverbial, that of aru remains without being fully accounted for. Our search for "recycled aru" extends in such surprisingly vast areas as: intransitive verbs, passive, nominal sentences, perfective aspect, so-called Na-adjectives, conditionals, among many others.

The second part of this paper describes how often aru and do verbs are used in oral narrative stories by native speakers of Japanese and native speakers of English. In this study oral narrative stories were collected from fifteen native Japanese-speaking bilinguals and fifteen native English-speaking bilinguals, both in English and in Japanese. Each narrative clause was scored as one of the following categories: attention, orientation, action, evaluation, outcome, and coda. The results of this study indicated that the native Japanese-speaking bilinguals used more do or "action" in their English narrative stories than in their Japanese narrative stories. Likewise, the native Japanese-speaking bilinguals used more aru or "outcome" in their Japanese narrative stories than in their English stories. The native English-speaking bilinguals showed the similar results.

"Efficient Reading of Japanese Texts: Elements and Implications"
Sachiko Matsunaga (California State University, Los Angeles)

This paper reports a study which identified knowledge of Japanese pronunciation and oral proficiency as important elements for developing reading fluency of Japanese texts. The data from the study indicated that even those who can obtain the meanings of many of the kanji words in the text are not efficient readers if they do not know the Japanese pronunciations. The data from the same study also showed that the higher the learners' Japanese oral proficiency (as measured using a simulated Oral Proficiency Interview) was, the more kanji words they knew in the Japanese pronunciation, regardless of the learner's background. Moreover, their oral proficiency was highly correlated with both their reading comprehension and speed. Based on these results, the paper argues that the development of the ability to decode kanji words in Japanese pronunciation, as well as the development of oral proficiency, appear to be crucial to achieving fluency in reading Japanese as a foreign language. Pedagogical implications from these findings were discussed with the audience.

"Learning through Talking? Analysis of Small Group Activities in Upper-level Japanese Language Courses"
Junko Mori (University of Wisconsin-Madison)

Using the methodological framework of conversation analysis as a central tool of analysis, this study examines the sequential development of talk-in-interaction observed in a small group activity in an upper-level Japanese language classroom. Given similar instruction of a task, each individual or each group approaches the task in a different way and develops the interaction accordingly. On the one hand, the ways in which the students develop their talk represent their understanding of the activity and their existing skills for accomplishing various interactional routines. On the other hand, the fact that each group ends up having different kinds of exchanges also means that the students will experience different kinds of learning opportunities. This study explores how the institutionalized nature of the task becomes realized in the talk and how challenging it could be for the students to go beyond a series of questions and answers, the pattern recurrently reinforced in language classrooms. It emphasizes the significance of students' ability to produce an assessment of the co-participant's responses for creating a discussion-like, or conversation-like interaction.

"Learning New Kanji Words through Reading"
Yoshiko Mori (Georgetown University)

In my presentation, I discuss the advantage of using multiple sources of information in interpreting new kanji words while reading. In so doing, I focus on three findings of my research investigating how learners of Japan infer the meanings of unfamiliar kanji words usiing information from component characters as well as contextual information, and discuss their pedagogical implications for reading instruction. First, overreliance on a single source of information increases the chances of generating incorrect inferences, which justifies the recommendation that students always be encouraged to check their guesses against surrounding context. Second, the lack of correlation between the ability to use kanji clues and the ability to use contextual clues suggests that good context users are not necessarily good at semantic analysis of internal word structures. Consequently, in any attempt to influence students' learning, teachers should consider individual differences in the ability to use a particular type of information. Lastly, proficiency in Japanese correlates with students' ability to use contextual clues, but not with the ability to use kanji clues. This suggests that success in using contextual information can partially be explained by proficiency, but successful word analysis must be explained by abilities beyond linguistic knowledge of Japanese. Thus, both novice and advanced learners might need some form of metacognitive instruction aimed at raising their awareness of the complexity of the semantic structures of kanji words.

"Effects of Computer Feedback in Second Language Acquisition: How Explicit Should It Be?"
Mariko Moroishi (Purdue University)

Providing corrective feedback to learners' errors is considered to be one possible technique to draw learners' attention to errors they have made and to help them to improve their target language skills. However, the optimal degree of explicitness of feedback judged to be most beneficial for language learning is still controversial and needs to be further investigated by more empirical studies. This study compares two types of corrective feedback provided in a timely, systematic manner in computer-assisted language learning (CALL) tasks. The software system the authors designed provides learners with: 1) metalinguistic feedback that presents learners with grammar rules when they made an error, and 2) corrective recasts that request clarification by repeating a learner's error and highlighting it on the screen. In this presentation, we discuss the design of the software system and the instructional philosophy behind it, focusing on its feedback features. We also discuss the design of an upcoming experiment designed to examine the issue of which kind of feedback facilitates language learning better.

"Fissure, Blindness and the Gaze in Natsume Sseki's Meian: Reading the Diseased Body"
William Ridgeway (University of Hawaii)

In Natsume Sseki's final, never-completed novel Meian (1916, translated Light and Darkness, 1971), the story begins in an examination room with the protagonist Tsuda Yoshio undergoing an anal probe. The doctor tells him that a "basic operation" must be performed. Before leaving the room, Tsuda recalls the microscope he had just peered into and asks the doctor if he is tubercular. "No, you're not," the doctor says. "I know just by looking at you." Within these brief two pages of Section One, Sseki has set the tone for the entire novel and established the basic tropes of fissure, blindness and the gaze. In a poetic leap of association I link Tsuda's physical condition (fissure) and the fact that he cannot observe his own disease with the metaphor of blindness (cecity, from cecum); the microscope and the doctor's diagnosis with the need for ocular proof, with ocularcentrism; and medical scrutiny with the gaze, the ability to know and "read" people by looking at them.

Organizing fissure, blindness and the gaze as metaphors or metonyms for understanding the text, I attempt to show how all interpersonal relationships in the novel are characterizable by this rhetoric. Fissures, for example, erupt on many levels: physiologically, there is Tsuda's anal fistula or fissure and Kiyoko's miscarriage; psychologically and socially, there are fissures between husband and wife (Tsuda, O-Nobu), brother and sister (Tsuda, O-Hide), superior-subordinate (Mrs. Yoshikawa, Tsuda), friend and "foe" (Tsuda, Kobayashi), and ultimately between "lovers" (Tsuda and the woman to whom he was once engaged to be married, Kiyoko).

Just as Tsuda must go under the knife, we as readers, too, must "mesu wo ireru" (or make a searching inquiry, apply a scalpel of sharp criticism to the text. In doing so, we can expose the blindness of the characters and the many blind spots, including the author's own. O-Nobu is said to possess senrigan or clairvoyance, but her powers ultimately fail her. Tsuda is famously blind to the concerns of others. "The blind egotism of the characters and the power each, in his blindness, desires over the other," in Janet Walker's phrase, point to a symptom of a disease which is larger and more serious than "mere" self-centeredness. The character of Kiyoko remains a blind spot since she appears only in the concluding (though unfinished) chapters of the novel. The author's own blindness to his country's imperialism and racism, as pointed out by James Fujii in his reading of Man-Kan Tokorodokoro, is overlain on Kobayashi, an ideologue who nevertheless has nothing to say about the propriety or significance of his trip to Korea.

Finally, the gaze, or the "terrorism of the gaze," to use Michael Moon's phrase, is to be found everywhere: O-Nobu seeks to exert control with the "streak of lightning" and "hate beams" of her gaze; Madame Yoshikawa seeks to control all within her domain by means of her normalizing gaze or panoptic gaze. Ultimately, though, fissure, blindness and the gaze can only describe symptoms without naming the disease. The disease itself is found in the unutterable ano koto (money matters) which plague and rupture most of the relationships in the novel. The disease, also represented by Uncle Fujii's myyaku (the cure-all) in the form of a cheque, remains unnamed in the end, but we can see it or its symptoms somewhere amid the paradoxes of modernity, the cash necessity, the family pressures, personal tyrannies, the impinging of one personality on another which, taken together, constitute the major concerns of Sseki's text.

"On the Use of Japanese Demonstrative a-Series"
Mitsuyo Sato (George Washington University)

This paper examines emotional deixis and cataphoric deixis in the Japanese demonstrative a-series. The English demonstrative that seems to have certain parallels with the Japanese a-series. Lakoff (1974) defines that in English that has the function 1) to establish emotional closeness between the speaker and the hearer and 2) to establish solidarity between them. Kitagawa (1979) claims that the Japanese demonstrative a-series also can function as emotional deixis, as seen in one of her examples, "Tanaka ni zuibun atte nai naa. Ano hito mo toshi daroo naa." This use of a-series represents a sense that both the speaker and the hearer share the same perspective based on their background knowledge. A distance between the speaker and the referent is in her words "heart-felt." In the data I collected, what the a-series points out lies in a certain "psychological distance" away from both the speaker and the hearer. In "Jibun de iu nomo chotto are nan desu kedo, ma uta niwa jishin ga arun desu yo," the speaker expresses some hesitation in saying something boastful about himself by using are in the utterance. Are can be replaced by such words as jiman shiteiru and hazukashii etc. Also, in "Kon'na koto iu no are desu kedo, anata kimochi waruin desu," are can be replaced by shitsureina and warui etc. Comparing Kitagawa's example and my data, emotional deixis a-series in Kitagawa's example refers to a tangible entity as a referent (i.e. a person or a thing). Emotional deixis in my data corresponds to a certain "semantic" area that is not physical, but is psychologically distanced from both the speaker and the hearer. English demonstratives have a cataphoric function in such sentences as "I met this guy yesterday. He and I went to the same high school." The indefinite this in colloquial English cataphorically refers to something which is mentioned in the next sentence. Masuoka and Takubo (1992) assert that only ko-series in Japanese demonstratives have a cataphoric function. However, my data suggest that a-series have a cataphoric function as well. For example, in the sentences "Kazumi wa are ka. Oya kara moratta karada ni fuman demo aruno ka," are refers to the following sentence: Oya kara moratta karada ni fuman demo aruno. There are several reasons why the speaker would use the cataphoric are: 1) as a filler in the conversation, 2) to soften negative and/or hesitated utterances and 3) to get hearers. There are significant linguistic and teaching implications for emotional and cataphoric deixis. With more research, it will be possible to redefine and clarify the Japanese demonstrative system. Also, most Japanese textbooks do not introduce extended and emotional uses of demonstratives. For advanced students, a more comprehensive understanding of the demonstrative a-series can increase their awareness of politeness strategy in a variety of delicate situations.

"The Attitudes of Japanese Language Educators Toward the Teaching of Kanji"
Hideko Shimizu (Vanderbilt University)

The purpose of the current study was: (1) to identify the underlying attitudes of Japanese language educators toward kanji; (2) to identify which teaching strategies are employed by educators; and (3) to investigate the relationship between (1) and (2). The current study developed a survey to which 251 educators in the US and Canada responded. Using principal component analysis, six statistically reliable domains (cultural tradition, difficulty of kanji, positive affective orientation, aptitude, usefulness of kanji, and the future of kanji ) and three strategies (context-based, cognitive memory, and rote-learning teaching strategies) were identified. Descriptive statistics showed that most people saw kanji as useful and that the most popular instructional strategy was the "rote learning strategy." Canonical correlation analysis revealed a significant correlation between a set of attitudes (i.e., affective orientation, usefulness of kanji, and cultural tradition) and a set of strategies (i.e., cognitive memory and context-based strategies). One of the interpretations of the results is that attitudes toward teaching kanji can be linked to motivational domains and a specific set of instructional strategies which is expected to promote meaningful learning.

"Regarding Death in Ooka Shohei's Musashino Fujin and Kaei"
David Stahl (Middlebury College)

Ooka Shohei was a battlefield survivor, and the key to fully appreciating his most important literary works is to approach them as survivor's literature. For many years, impaired mourning for his dead comrades frustrated his literary efforts to come to terms with his traumatic battlefield experience. Ooka sought to resolve his death-related issues in fictional works such as Musashino Fujin (1950) and Kaei (1959). Focusing on narrative perspective, I argue that while in the former he was unable to identify and empathize with Michiko as she despaired and killed herself, in the latter he did so for Yoko almost to perfection. I account for this important change in emotional capacity by highlighting early 1958 diary entries in his "Sakka no Nikki." These clearly show that after thirteen years spent repressing the memory of his dead comrades, Ooka finally recovered the ability to mourn. Recovering his ability to embrace, empathize with and grieve for the war dead enabled him to do the same for Yoko when he wrote her story some six months later. Moreover, by the time he began serial publication of Reite Senki in 1967, he had discovered the capacity to do the same-and more-for more than 90,000 of his fallen countrymen.

"Descriptive and Comparative Analyses of Motivation Factors for the Study of Japanese"
Junko Ueno (Indiana University)

In the field of second language (L2) learning, the research of learners' motivation is considered essential, and numerous studies have discussed the issue (Buonomo; 1990, Day; 1987, Ralph; 1989). This area calls for even more exploration in teaching non-cognate languages such as Japanese, because non-cognate linguistic aspects have unique influence on American student affective variables including motivation, as Samimy (1994) pointed out.

In fact, Tanaka (1998) discovered that the motivation of American high school students studying Japanese was more varied and complex than the integrative-instrumental distinction which has been prevalent in L2 motivation research. While Tanaka's study is beneficial to help teachers of Japanese to grasp their students' learning backgrounds, her findings appear to be limited because of the following reasons: 1) the study was conducted solely in a high school setting, 2) it does not describe the motivation factors of learners of Japanese in comparison to other cognate languages, and 3) it lacks pedagogical implications for teachers to apply to their classroom practices.

The primary purpose of my study is to examine student motivation in JFL (Japanese as a Foreign Language) classrooms compared to that of cognate language classrooms and explore the pedagogical applications specific to JFL instruction. This study utilized a survey research approach, based on Tanaka's questionnaire, in which 48 second year Japanese students and 17 second year Spanish students participated at a Midwestern state university. Participants responded to the following four areas in the survey: 1) previous Japanese (or Spanish) and other foreign language study, 2) previous exposure to Japanese (Spanish) culture, 3) motivation orientations for taking Japanese (Spanish), and 4) importance of Japanese (Spanish) language study.

The data analysis involves two kinds of comparative approaches. The first analysis compares the motivation factors of students of Japanese and those of Spanish students and discusses influence of Japanese as a non-cognate language on student motivation. The second analysis focuses on different motivation factors between high school students and university students based on Tanaka's findings and explores why the differences occur between high school and university levels and how teachers should cope with them. Lastly, some pedagogical implications are presented along with classroom practice for teachers of Japanese in the U.S.

"The Role of Illness as Seen in Shiga Naoya's Short Stories"
Guohe Zheng, Ball State University

More than two thousand years ago, Ssu-ma Ch'ien (145 BC?-86 BC?), the Grand Historian of China, made an insightful statement on the relationship between the adverse condition a writer found himself in and the literary achievement he made in Chinese history. A pattern in which writers achieve their artistic goals throughout history seems to have been captured in Ssu-ma Ch'ien's remarks, for amazingly similar ideas have been expressed again and again two thousand years later by many modern Japanese writers. Adverse conditions also have a role to play in many works of Shiga Naoya (1883-1971), a modern Japanese writer who has often been acclaimed as shsetsu no kamisama, or "the god of fiction." This paper presents an analysis of the underlying role illness plays in Shiga Naoya's short stories. The paper begins with a consideration of the role of the author's physical illness by comparing two of his short stories: "Dekigoto" (An Incident, 1913) and "Kinosaki nite" (At Kinosaki, 1917). I argue that, as the author was not directly involved in the accident described in the former, it is no more than a minor piece commending the good nature of the ordinary folks; in contrast, the latter is a true masterpiece with a serious theme crystallized from the author's personal experience of some life-threatening mishap. Then the paper goes on to explore the role of the author's spiritual illness in his short stories. Citing examples from such short stories as "Seibei to hytan" (Seibei and Gourds, 1912) and "Aru toko, sono ane no shi" (A Certain Man, His Sister's Death, 1920), I argue that the animosity between writer Shiga and his father prompted him to write these pieces. In conclusion, even though Shiga Naoya may not be a "god of fiction," his short stories, inspired by his physical or spiritual illness and written in his own style, make him a self-reliant writer, unique and great. On the other hand, the role of illness in his short stories illustrates the operation of one of the most consistent dynamics in the literary world.