Katai and Homei: "Surface" Versus "Single-Dimensional" Description

Futon differs in one important respect from shishosetsu to follow: Katai's narrator accommodates several points of view. Although focusing primarily on Tokio, the narrator presents freely if briefly the thoughts of Yoshiko, Yoshiko's father, Tokio's wife, and even a curious neighbor. Katai never felt compelled to restrict narrative perspective to that of the protagonist. Rather than strive for a unified point of view in his later works, he moved in the other direction, with the stated goal of adopting no point of view at all. This was the substance of his famous "surface description" (heimen byosha ); and Futon , although by far his best-known work, might profitably be characterized as no more than a dry run for works like Sei and Inaka kyoshi (The country teacher, 1909), which more successfully exemplified his technique. In "Sei ni okeru kokoromi" (1908), Katai claims to follow in the footsteps of the Goncourt brothers by championing a style that allows the narrator to treat all events and characters with the same degree of studied aloofness. Katai argues that his purpose in writing Sei was to describe accurately and dispassionately what he saw and heard and to avoid all interpretation of events and characters. "By refraining from petty, subjective interpretation and from analyses of phenomena about which I had no direct understanding, and by presenting my material in unaltered [ari no mama ] form," he concludes, "I believed that my descriptions would actually come closer to the truth and would of themselves suggest the inner significance of things."^[38]

[38] KBHT 3:450. Katai's alleged debt to the Goncourt brothers should be viewed as skeptically as that to Zola. Katai was no doubt familiar with the Goncourts' various pronouncements on the novelist's role as historian, including the following: "We passed through history to arrive at the novel.... On what basis does one write history? On the basis of documents. And the documents of the novel are life" (quoted in Richard B. Grant, The Goncourt Brothers , 30-31). But Katai apparently took no note of the Goncourts' concern with style and their awareness of the mediative powers of their form, which they argued could miraculously attract the reader's interest in "a human story that we know never took place" (ibid., 114). These self-styled clinicians of truth frequently acknowledged their aesthetic motivations: "Art for Art's sake, art which proves nothing, the music of ideas, the harmony of a sentence, that is our faith and our conscience" (ibid., 16). The Goncourts, then, never aspired to the "unaltered" transcription of life any more than Zola did.

In his effort to legitimize this narrative technique, Katai had important company. Shimazaki Toson's first autobiographical text, Haru (1908), was being serialized in another newspaper at the same time as Sei . Toson, whom Edwin McClellan argues was "as responsible as any single writer could be for the prevalence of the idea in Japan that the novel need not or should not be a creation of the dramatic imagination,"^[39] also utilized in this and later writings an aloof narrator who makes little effort to probe his characters' minds. Katai was quick to recognize the similarity of their efforts and took note of Toson's achievement in “Sei ni okeru kokoromi." The same dispassionate style, carried if anything to an even greater extreme in the writings of Tokuda Shusei, prompted Natsume Soseki, for one, to describe it disparagingly as an endless stream of words with no direction, no informing idea, no life.^[40]

Katai, too, had his critics, most notably Iwano Homei, who argued that Katai's failure to invest his works with a specific point of view deprived them of meaning and interest.^[41] Homei championed what he called "single-dimensional description" (ichigen byosha ), arguing that it avoided the epistemological pitfalls of Katai's technique. "There is no life outside the self," he asserts in one of his first major critical statements, "and literature is the product of the

[39] "Toson and the Autobiographical Novel," 348.

[40] "Bundan no kono goro," in Soseki zenshu 16:723-25.

[41] Homei is an anomaly in Japanese letters. Although usually grouped with the naturalists, he was something of an outsider who relished his role as gadfly to the bundan . Like Mishima Yukio, he took as much, or more, pride in being a man of action as he did in being a man of letters. Confronted with failure at every turn in private life, from his business ventures to his relations with women, he celebrated his defeats in his writing, creating a hero who presided bellicosely over his own demise. After establishing his reputation as a fiction writer with Tandeki , which describes his liaison with a syphilitic geisha, he went on to produce a monumental, five-part account of his pursuit of wealth and love in Tokyo, Sakhalin, and Hokkaido (usually referred to as Homei gobusaku , 1910-18).

author's subjective tone or attitude.... A writer cannot hope to portray life conscientiously if he cuts himself off from his own subjectivity."^[42] It is clear from the ironic dedication to Katai of his first anthology that Homei was conscious from the beginning of his polar position vis-à-vis the author of Futon , despite (by Homei's own admission) the latter's influence.^[43] Nor was Homei alone in sensing this polarity. Major naturalist critics like Shimamura Hogetsu and Hasegawa Tenkei saw Homei and Katai as embodying two divergent trends in the naturalist movement. Tenkei, for example, contrasted Homei's passionate, "self-revelatory" (jiko kokuhaku ) style with Katai's aloof, "self-contemplative" (jiko seikan ) style and argued that Homei's was a throwback to the romantic period while Katai's represented naturalism's true path.^[44]

While his critics were dwelling on degrees of emotional distance between author and hero, Homei was concerning himself with the mechanics of representation. In an essay entitled, with typical immodesty, "Gendai shorai no shosetsuteki hasso o isshin subeki boku no byosha ron" (My theory of writing, which will revolutionize thinking on the shosetsu , present and future, 1918), he articulates his position on narrative·

This is not a world in which we know everything about everyone else. All people and all things are actually reflections of our own minds We reign as sovereigns over our private worlds and we allow no rights to others.... This said, let us consider what it means to be a writer. The author does not understand others as he understands himself.... It is impossible for him to be purely objective ... and by trying to be so he irresponsibly takes a position only God can take.... I will use my subjectivity, not narrow-mindedly but to its fullest potential, to enter into the feelings of a single character, whether A or B or C. Just as we cannot know what another person thinks in real life, the author cannot know what B or C thinks if he has identified with the feelings of A. There is almost no one here or abroad who realizes this fact, except for myself and those who have fully digested my theory on this subject.^[45]

[42] "Gendai shosetsu no byosha ho" (1911), in KBHT 3:366.

[43] See the preface to Homei's Tandeki (the anthology), in Homei zenshu 18: 78-81.

[44] "Jiko bunretsu to seikan" (1910), 207-9.

[45] KBHT 5:88-90. For discussions of this essay and its place in Homei's critical project, see Yoshida Seiichi, Shizenshugi no kenkyu 2:449-60; Wada Kingo, Byosha no jidai , esp. the chapters on Katai's "surface" and Homei's "single-dimensional" description, 103-72; and Noguchi Takehiko, Shosetsu no Nihongo , 197-210.

In this and other writings, Homei never insists that the author restrict his point of view to a character modeled after himself, only that he be consistent. It was the Taisho literary establishment that would go on to interpret Homei's definition of "single-dimensional description" in an even stricter sense than Homei himself had ever advocated: namely, that the author could do no better than to narrate lived experience from his own point of view.^[46]

Yet Homei presents here something rather close to what would emerge as the mainstream junbungaku worldview (even though he portrays himself as the lone voice of reason in a critical wilder-ness-encouraged, to be sure, by a literary establishment that took him only half-seriously).^[47] That is surely why he aroused the ire of junbungaku opponents. Ikuta Choko, for example, whose critique of the shishosetsu we noted in Chapter 3, argued that however much Homei's thesis held true for works of a confessional nature (diaries, travel essays, autobiographies, etc.), in which a single point of view was most convincing, it had no bearing on fictional works, which appropriately might have several protagonists and several points of view. The "best" method of representation, Ikuta concluded, was simply the one that best suited a particular writer.^[48] The philosopher and critic Tsuchida Kyoson (1891-1934), continuing the attack, insisted that Homei was on extremely shaky epistemological ground when he argued that the author's perceptual world overlapped the protagonist's. Tsuchida's Kantian system divided the world into two categories of perception—the things, people, and other "objects" (taisho ) in nature that exist independent of each in-

[46] Homei stressed in many of his essays that the point of view need not be the author's own. Among his own writings, the group of stories published near the end of his life focusing on a character named Osei are examples of this. (See, for instance, Osei no shippai [1920], which chronicles the heroine's loss of her boardinghouse to a wily carpenter on whom she has naively relied to make improvements. [lwano Homei zenshu 8:387-488]) Homei also insists in the face of repeated attacks that the hero, even when modeled after the author, should not be equated with the author himself. See Wada, Byosha no jidai , 132 and 136, for extracts of his protestations, of which Wada himself seems skeptical. Noguchi Takehiko, meanwhile, argues that Homei was more successful than Katai or Toson at transforming his personae into autonomous characters. See shosetsu no Nihongo , 204-9.

[47] Ino Kenji, in his study of Meiji writers, notes that Chikamatsu Shuko aligned himself solidly with Homei against the likes of Hogetsu and Katai on the issue of narrative perspective and suggests that it was Shuko and not Katai who, as heir to Homei's narrative theory, fixed the shishosetsu's course. See Meiji no sakka , 401.

[48] "Iwano Homei shi no byosha ron" (1918), in KBHT 5:97-100.

dividual's apprehension of them as images, and "subjects" (naiyo ) in need of an "I" to see them—and he claimed that Homei made no distinction between the two.^[49] Both reached the same conclusion: to argue as Homei did that "this is not a world in which we know everything about everyone else" and that "all people and all things are actually reflections of our own minds" was, in a literary context at least, the height of naïveté.^[50]

Homei's "naïveté," however, was characteristic of a great many Taisho writers who, while making no pronouncements of their own, adhered to the tenets Homei set forth. Ikuta and Tsuchida's condemnations notwithstanding, Homei had an unerring sense for the mode of narrative presentation that would triumph in the Taisho era. In essence, he was championing what we identified in Chapter 2 as the written reportive style. Homei's critics correctly pointed out its constraints, but they underestimated the enormous power it held over the Taisho literary community. Homei's theory helped legitimize the shishosetsu's mode of presentation to the extent that it came to be regarded as "natural," and finally transparent.

A few years after Homei's death, Katai conceded in his memoir Kindai no shosetsu (shosetsu in the modern era) that Homei's "single-dimensional description" was, in principle, the ideal to which literature should aspire and that the truly great writers utilized just such a technique.^[51] Such a concession on Katai's part was probably less a shift in his own position than an acknowledgment of the shishosetsu's dominance in the early 1920s. Given the popularity by this time of first-person narration and the critical insistence on the equation of author and hero, the Jamesian-style narrative that focused on a single character distinct from both author and narrator never became the dominant mode in Taisho letters. Even though Homei had some success (as did Tokuda Shusei and of course Natsume Soseki) in building a narrative around a character not modeled after himself, the Taisho literary establishment took far less interest in works to which the author-hero formula could not be applied.

[49] "Iwayuru ichigenteki byosha o ronzu" (1918), in ibid., 101-6.

[50] Ibid., 200, 101.

[51] In Tayama Katai zenshu 17:357.