EPILOGUE: THE SHISHOSETSU TODAY

No one can reveal oneself completely. Yet writing becomes impossible if one discloses nothing of oneself. Rousseau was enamored of self-exposé; yet he does not reveal himself fully even in his Confessions. Mérimée despised confession; yet does not Columba reveal something very profound about him? The line between "confessional" and "non-confessional" literature is a hard one indeed to draw.
Akutagawa Ryunosuke, "Shuju no kotoba"

This book is not a book of "confessions"; not that it is insincere, but because we have a different knowledge today than yesterday; such knowledge can be summarized as follows: What I write about myself is never the last word: the more "sincere" I am, the more interpretable I am ... my texts are disjointed, no one of them caps any other; the latter is nothing but a further text, the last of the series, not the ultimate in meaning: text upon text, which never illuminates anything.
Roland Barthes by Roland Barthes

Why end a study of shishosetsu with Kasai Zenzo? It is because, as one writer has remarked, pure literature in Japan met its demise with Kasai's passing.^[1] Hyperbole notwithstanding, this opinion expresses a certain truth: Kasai's death did coincide with the end of the shishosetsu's heyday. After the Taisho period, the bundan would never again be as receptive to writers who wrote only about themselves and only in their own voices. By the mid-1920s, the central place of junbungaku was being challenged by the first wave of shi-

[1] The writer is anonymous, but the eulogy is frequently cited. Tanizaki Seiji quotes it in "Kasai Zenzo hyoden," 495.

shosetsu criticism, as we noted in Chapter 3, and by the emergence into the limelight of more aesthetically inclined neo-perceptionist and more socially conscious proletarian literature.

The aftermath of the Pacific War brought with it self-examination on the literary as well as other fronts; postwar critics typically charged the shishosetsu with being a vestige of what was "feudal" and unprogressive. Since the early 1960s, with the rise to prominence of writers like Abe Kobo (1924-), Oe Kenzaburo (1935-), and Kurahashi Yumiko (1935-), all of whom are steeped in the western intellectual tradition, Japanese literature has entered into an age of internationalization. Junbungaku has lost its strict autobiographical connotation, and both the form and content of "serious" literature have regularly strayed from the shishosetsu mode.

Yet there is no question that the shishosetsu is still deeply entrenched in contemporary literature. Despite the western novel's continued impact on Japanese fiction, the traditional poetics of authorial "presence" has remained largely intact. One of the "attractions" of the shishosetsu , Masamune Hakucho wrote soon after the war, was its apparent longevity, and this of course is all the more true today. The shishosetsu , like haiku and waka , would survive as long as Japanese culture did, he predicted, because it was similarly bound to the native sensibility and observed the same crucial literary convention: the depiction of a particularized, personal, narrowly defined world that was still the most "real" to a Japanese.^[2] Literary realism, of course, is "plausible not because it reflects the world but because it is constructed out of what is (discursively) familiar."^[3] Our awareness of this fact, however, far from undermining the shishosetsu's reality, actually helps us better appreciate its persuasiveness, realizing as we do that its sensibility and conventionalism are bolstered by a powerfully supportive epistemology that valorizes lived experience.

This essentially antifictional (which, as we have seen, is not the same as "unimaginative") view of literature has prompted the shishosetsu writer Kanbayashi Akatsuki (1902-80), for example, to remark on numerous occasions that he regarded each of his stories as if it were a final document, that he recorded his life as if leaving

[2] "Watakushi shosetsu no miryoku" (1947), in Masamune Hakucho zenshu 7:380.

[3] Catherine Belsey, Critical Practice , 47.

a testament. "I write in the shishosetsu mode not because I believe it to be the greatest literary form," he reflected early in his career, "but because I feel that I have no choice.... I acknowledge the worth of the more 'objective' kind of fiction ... but I cannot write unless I have a great personal stake in the material."^[4] We can easily recognize in Kanbayashi's words, written just before the beginning of the Pacific War, the same underlying view of literature that we noted in Yasuoka Shotaro's remarks three decades later: "Truly creative writing involves ... discovering something in yourself, not dreaming up some formally satisfying story."

Few critics would deny the shishosetsu's continued prominence, but most argue that it has undergone a transformation, in the decades following the period we have studied, that has been variously described as a "fictionalizing" or "distorting" of the form. Nishida Masayoshi's (1931-) observations typify this view. The shishosetsu entered into a period of decline soon after the early years of Showa, Nishida observes. With the exception of a handful of "dispassionate," "life-affirming" authors like Kanbayashi and Ozaki Kazuo, most later writers are prone to exaggeration or idle reverie. Thus, Makino Shin'ichi's (1896-1936) writing is characterized as "hallucinatory," Kajii Motojiro's (1901-32) as "undisciplined," Kawasaki Chotaro's (1901-85) as "decadent," Dazai Osamu's as "nihilistic," and Shimao Toshio's (1917-86) as "fantastic."^[5]

What Nishida and many others see as a decline, however, can more profitably be regarded as a sign of the form's diversification, something that, moreover, has been apparent from the beginning. If our study of three shishosetsu authors has taught us anything, it is that their common struggles with "authentic" narrative representation were resolved in various ways that, for all their supposed fidelity to lived experience, reveal considerable editorial license and imagination. Hiraoka Tokuyoshi (1929-) notes that the shishosetsu's seemingly most logical and comfortable definition—that it is a first-person (or third-person) narration about the author's own experience, with no fictional embellishment—quickly reaches an impasse over the meaning of "experience." One might of course

[4] "Watakushi shosetsu shikan" (1941), quoted in Matsubara Shin'ichi, “Gusha" no bungaku , 227-28. Kanbayashi expresses similar sentiments in "Watakushi shosetsu no unmei" (1947), in Kanbayashi Akatsuki zenshu 15:285-91.

[5] Watakushi shosetsu saihakken , 243-77.

argue that it includes such everyday activities as eating, sleeping, and walking; but a case could easily be made as well for including what the author sees or hears or reads. And in fact an author like Shiga—not to mention other writers like Takii Kosaku (1894-1984), Nagai Tatsuo (1904-), and Shono Junzo (1921-)—typically writes about the latter as much as the former. The very recording, however "faithful," of these acts, surely requires no little creativity, Hiraoka suggests. But if what the writer has seen or heard or read is part of his personal experience, Hiraoka argues further, then is not also what he himself has imagined? In that case, is not an ostensibly "fictional" story like "Kamisori" (or like "Kozo no kamisama," we might add), in which the author faithfully records the experience of his imagination, just as much a shishosetsu as any of Shiga's other works? And if so, then are not all shosetsu actually shishosetsu ?^[6]

Hiraoka's rather facetious complication nevertheless exposes the fragility of conventional wisdom about the shishosetsu: that it depicts, with greater accuracy and authenticity than any other form, life as it is really lived. Still, the myth of sincerity dies hard. Kanbayashi Akatsuki, the self-styled writer of testaments, for example, tells fellow participants at a roundtable discussion in amazement that he received a letter inquiring about the health of a character in one of his stories. Such an inquirer obviously does not know how to read fiction, he concludes.^[7] Yet he then goes on to defend his habit of neglecting to describe characters and narrative situations because, he claims, they are bothersome and redundant, as readers would be familiar with them anyway—clearly a move by the author to link his writing unequivocably to the referential world.^[8]

This example is hardly unusual. In 1981 Matsubara Toshio (1905-86) published a story in which the narrator reminisces about his visit as a boy to Kinosaki, the spa made famous by Shiga Naoya's story. He describes a chance meeting with a man whom he later deduces to have been Shiga himself at a Kinosaki bathhouse and

[6] Bungaku no doki, 44-45.

[7] Kanbayashi Akatsuki et al., "'Watakushi shosetsu' ron," 42.

[8] Ibid., 50. We have noted, of course, that the author's previous texts themselves provide the necessary "background" information (thus suggesting a tradition or convention of reading, as opposed to a "simply" referential reading) but this is certainly not what Kanbayashi means here.

recalls the scene in Shiga's story, which he witnessed, that depicts a rat struggling in a stream running through town to escape a crowd bent on killing it.^[9] We learn in a sequel published two years later, however, that this experience is not only untrue but also that its untruth, which he finally confesses to a Shiga disciple out of good conscience, greatly annoys the disciple and other Shiga admirers, who have read the first story as nonfiction and invited the author to a reception commemorating the tenth anniversary of Shiga's death. The author is told by one acquaintance that never again will he be able to believe what the author has written.^[10]

The very massiveness of Matsubara's attack on referentiality in his second story would seem to privilege it as a "truthful" account. But might not this, too, be a trap? We cannot help wondering whether we are really being told the whole truth even here, in a story that rewrites (after the manner of Kasai Zenzo) the earlier story in a way that alters forever our reading of it. Just as we now reject the first story as a true account, might not the appearance of a third story later on make us reject the second? Matsubara is able to "deceive" his readers not because he is the first shishosetsu writer to try it (we have learned in this study that deception—in the guise of "sincerity"—is in the nature of the form) but because of the powerfully persuasive rhetorical apparatus that still binds the shishosetsu narrator "inside" a text to the author "outside" it.

The attraction of the shishosetsu's rhetoric is so great, in fact, that it continues to win converts from among its opponents. Akutagawa Ryunosuke, the frequent critic, as we noted in Chapter 3, of the junbungaku chroniclers of personal experience, took to writing shishosetsu himself late in life. Of his posthumous "Haguruma" (Cogwheel, 1927), Kasai Zenzo is reported to have said, "At last Akutagawa has written a true shosetsu ."^[11] For Kasai, being "true" to the word shosetsu of course meant rejecting the transcendent grammar of fabrication. More recently, Oe Kenzaburo, once the literary world's enfant terrible whose distinctly alien prose style shocked his elders into stern remonstrances about the limits of the language's resiliency, has become a shishosetsu writer of note, al-

[9] "Hanasanakatta hanashi."

[10] "Kutsu."

[11] Yamamoto Kenkichi, Watakushi shosetsu sakka ron , 28.

though he himself would no doubt recoil at the characterization. Oe expressed his defiance of the form as early as 1961, when he set forth a manifesto of his goals as a writer: "Kobayashi Hideo wrote, in the very year I was born, 'The shishosetsu is dead, but have people vanquished the self? The shishosetsu will doubtlessly reappear in new forms, so long as Flaubert's celebrated formula, "Madame Bovary—c'est moi," lives on.' But I believe otherwise. The shishosetsu is not dead. People have not vanquished the self. And the shishosetsu has not appeared in new forms."^[12] Indeed, the shishosetsu seems a poor vehicle, he concludes, in which to embark on his own admittedly personal yet at the same time highly imaginative and metaphoric journey of self-exploration.^[13]

Beginning perhaps with his "Rain Tree" series, however, which appeared in the early 1980s, Oe presents a new kind of narrating voice. The first-person boku , modeled dearly as never before after the author himself, serves as a thoroughly accessible and concrete locus of perception, guiding his readers through a world that is both linguistically and epistemologically "authentic." To be sure, this seeming authenticity, which pays its respects repeatedly to the referential world, actually provides a springboard to an often densely allegorical realm. But Oe has succeeded in putting his new wine in an old bottle and showing them both off to good advantage. One of the shishosetsu tradition's most vociferous rebels has come nearly full circle, using Japanese literature's most realistic discursive mode to push realism to its limits. In an interview following his receipt of the Yomiuri Prize for his “Rein tsurii" o kiku onnatachi (Women who listen to the "rain tree," 1982), Oe denies that his work is a shishosetsu and emphasizes the process of fictive reconstruction to which he subjects personal experience, while adding demurely that he has given more direct expression to his thoughts than ever before.^[14] Yet it is not his "thoughts," somehow more directly expressed, that lead us to label this and much of his recent writing as shishosetsu ; rather, it is the specifically located narrating voice, which fairly exudes authenticity by its apparent denial of fictive imagination. Oe presents a narrator who relates only what he can

[12] "Watakushi shosetsu ni tsuite," 192.

[13] Ibid., 197.

[14] Yorniuri Shinbun (evening edition), 2 Feb. 1983, p. 5.

see or hear himself. Whether or not this nontranscendent narrator tells us the "truth" about Oe's lived experience is not our concern; he presents his story as if to chronicle that experience and nothing else. Thus, Oe, like Kasai and other writers before him, has found a way to transcend lived experience by paradoxically forfeiting all pretensions to transcendence.

Oe is not alone in pushing the "realism" generated by Japanese literature's most trusted mode of narrative representation to its limits. Fujieda Shizuo (1908-), a disciple of Shiga Naoya, adamantly maintains a nontranscendent stance even as he explores manifestly transcendent realms. At the conclusion of Gongu jodo (In search of the Pure Land, 1970), for example, the narrating "I" dies and rejoins his deceased relatives in the family grave. Confronted with so blatant a transgression of "reality," even as it is told in the quintessentially authentic shishosetsu voice, the reader is forced to question the truth of all details in this story, which has seemed to correspond so faithfully to the referential world.^[15]

To say that the shishosetsu is being exploited to suit the contemporary writer's idiosyncratic needs and that "fantasy" cannot be separated from "truth," however, is merely to say that it is business as usual with this deceptively straightforward form. Hiraoka Tokuyoshi, one of the more sensitive shishosetsu readers, notes in a suggestive essay that the shishosetsu , having now become so well established, has broadened its discursive perimeters. Today's shishosetsu no longer needs a first-person narrator, he argues; a third-person narrator will do just as well. Nor does it need to be based strictly on the author's own experience; it need only give the impression that it does.^[16] The argument is true as far as it goes, of course, but we have learned from our survey of the Taisho shishosetsu that it has in fact always been true. Indeed the difference from author to author or from generation to generation in the amount of narrative "deception" is of little significance, as the potential for it is always there. We have seen that a wide range of

[15] Ironically, Fujieda is the model for the character (identified by an initial) in Matsubara's story, cited above, who is offended by the author's "untruth." This lack of sympathy for another writer's "fabrications" has informed the consumption (as opposed to the production) of shishosetsu throughout the history of the form—a sign, no doubt, of the myth of sincerity's unwaning power.

[16] "Watakushi shosetsu no suitai to tensei," 154.

morally and artistically disparate writers have all found a common ground of expression in the shishosetsu —not because they were equally motivated to tell the truth but because they found in it a readily available medium for marketing their candor.

The shishosetsu seems quite capable, therefore, of thriving in an age of "internationalization." For what marks this form as tradition-bound is not its thematic content, as is often suggested, but its mode of discourse, which is by no means intrinsically ill-suited to the contemporary scene. That is why the critic Isoda Koichi's (1931-87) argument that the shishosetsu is an expression of an era that now belongs to the past, is problematic. "In my own view," he writes, "the style of the I-novel is closely associated with the traditional Japanese way of life; the world portrayed by Shiga Naoya and Ozaki Kazuo belonged to the Japanese-style dwelling with its tatami, its shoji, and its back garden. As a literary form, it was ill suited to life in high-rise, reinforced-concrete apartment houses."^[17] On the contrary: the shishosetsu is very much at home with "modern" life in Japan and will continue to be as long as its epistemological base remains intact. The presence of so many major shishosetsu writers of both sexes active today—Yasuoka Shotaro, Irokawa Budai (1929-), Miura Tetsuro (1931-), Abe Akira (1934-), Setouchi Harumi (1922-), Hayashi Kyoko (1930-), and Tsushima Yuko (1947-), to name just a few—moreover, belies the form's alleged affinity with a bygone world.

The history of narrative in Japan that produced the shishosetsu is a long one, but the relatively brief history of the modern shishosetsu itself is still unfolding. Now that the first generation of shishosetsu writers are in their graves and their works have achieved the status of classics, they will surely be reread by those with far less to invest in the writers' private lives. Shiga Naoya's wishful comparison of the writer's anonymity with that of the Kannon goddess's sculptor may yet become a reality. The evolution in language necessary for a clean break between narrating subject and narrated object, however, will not take place overnight. Readings of the narrated object in shishosetsu will no doubt continue to be constrained by the subject—the particularized voice that refuses to assume a presence independent of the author-hero, regardless of the narrative person. We confuse

[17] "The Historical Con text of Postwar Japanese Literature," 8. See also Isoda's discussion in Sengoshi no kukan (from which the article cited here is adapted), 197-99.

watakushi —which Kobayashi Hideo erroneously equates with Flaubert's je in his famous essay—with first-person pronouns in any western language, at our peril. As long as that difference remains, the shishosetsu's nontranscendent, communal narrating voice will overlap its principal actor and even you, the reader. And as long as narrator continues to merge with actor—watakushi/boku/ore , etc., with kare or kanojo or even anata —shishosetsu will always occupy the heartland of language and literature in Japan.