Writer, text, context

Anyone involved in the study of modern Japanese literature soon becomes familiar with Itō's name, which appears with great frequency in connection with so many major writers of the twentieth century. As a central player in the Japanese literary world, Itō is best known as the archetypal man of letters, member of the bundan, and scholar and editor engaged with a wide range of roundtable discussions, journals and other publications related to literary criticism. In fact, he was no less prolific in his production of fictional works throughout his writing career. Itō reached the peak of his popularity as a writer of both critical and fictional literature during the 1950s, but it was his earlier upbringing and the literary trends he encountered when he was coming of age as a young man that led him to articulate a colonial mentality through “Yūki no machi”.

Itō's early years were spent in or near the town of Otaru, Hokkaido, but in 1928 when he was twenty-three years old Itō escaped what he felt to be its oppressively provincial atmosphere and moved to Tokyo with the ambition of making his mark in the literary world. Like other aspiring writers, Itō threw himself enthusiastically into the various literary and artistic currents that had emerged in the wake of the 1923 Great Kantō earthquake. Proletarian writing did not particularly excite him, and in 1930 he even wrote several articles critical of left-wing ideology. On the other hand, he was thrilled to come across transition [sic], a literary journal founded in 1927 that featured surrealist, expressionist and Dada art under the editorship of Paris-based Eugene Jolas (1894–1952). Given Itō's interest in such experimental writing, it is not surprising that, among contemporary Japanese writers, he had a penchant for the works of the neo-sensationalist (Shinkankaku-ha) writer Yokomitsu Riichi (1898–1947), whose ground-breaking modernist novel Shanghai (1928–1931) was being serialized just as Itō moved to the capital (Kockum Reference Kockum1994: 66).

In his enthusiasm for such works, there is little to distinguish Itō from other equally ambitious young Japanese writers of the time. However, his literary development took a distinctive turn when he fell decisively under the influence of James Joyce (1882–1941) and Sigmund Freud (1856–1939), whose writings he first encountered during the autumn of 1929. Indeed, Itō was so excited by these iconoclastic thinkers that he characterized himself as staggering under the burden of three great intellectual loads: Freud, the French language, and Joyce (Kockum Reference Kockum1994: 78). The depth of his indebtedness to Joyce can be surmised from the fact that, together with two other writers, in 1934 he produced the first full translation into Japanese of Ulysses (1922), Joyce's shockingly new modernist text that incorporated, among other things, a stream of consciousness prose style. At the same time, Freud's writings, which first became widely available in Japanese translation in 1929, had an equally profound effect. This becomes evident when reading Itō's essay “Bungaku ryōiki idō” (A shift in the literary realm, 1930), in which he enthused over the potential role of psychology in future literature. The essay might be viewed as the first step towards Itō's participation as a founding member in the New Psychology School (Shin Shinrigaku-ha) of writing, which sought to give shape to Freudian themes in literary form (Kockum Reference Kockum1994: 96).

“Yūki no machi” has been described not only as Itō's most successful example of a stream of consciousness writing style but also as the best example of New Psychology fiction. (See, for instance, Kockum Reference Kockum1994: 159; Keene Reference Keene1984: 673.) However, it took the author several years of refining his writing skills before he reached the point where he was able to produce literature of this calibre. During the first part of the decade, Itō made various attempts to incorporate the ideas of Joyce and Freud into his fictional pieces, but these early efforts were received with a less than enthusiastic critical response. For example, in a review of Itō's story “Kanjō saibō no dannen” (Cross-section of the emotional cells, 1930), Kawabata Yasunari (1899–1972) declared himself unimpressed by what he felt to be the author's clumsy efforts to employ Freudian psychoanalytical descriptions. In Kawabata's opinion, the modernist experimentations of the poet Hori Tatsuo (1904–1953) worked far more successfully as literature (Senuma Shigeki 1955: 73–4).

The lukewarm critical response led Itō to acknowledge that he was yet to find his own literary voice, and in 1931 he announced his intention to revisit his northern roots both figuratively and literally with the aim of becoming a more accomplished and self-possessed writer. The first of these retrospective literary self-analyses took the form of a short story entitled “Umi no shōzō” (A portrait of the sea, 1931), in which the narrator recalls his coming-of-age experiences through a series of complex entanglements with women (Kockum Reference Kockum1994: 135). Itō also began to use literature as an opportunity to reflect upon and try to make sense of his relationship with his father, who died in Hokkaido in 1928, which had been problematic since childhood. For example, “Seibutsu-sai” (The feast of living creatures, 1932) centres on the death of a father, and his son's lingering sense of guilt about their unsatisfactory relationship (Itō Sei shū, Itō Sei 1970: 651–8). Works such as this indicate that the author saw the period when he was growing up in Hokkaido as a rich source of inspiration for developing deeper insights and more gravitas in his own literature.

In “Yūki no machi”, the narrator's anxious tone undoubtedly continues to convey the impression of personal insecurity, but the actual style of the piece hints at a writer who has reached a level of far greater confidence in his own voice. Certainly, the story draws on the author's actual visits back to his native home, and this contributes to the vividness of the literary representation of Otaru. Indeed, Itō felt certain that this work marked an important stage in his literary development, to the extent that he singled it out as his favourite and most successful piece of fiction.

There are so many twists and turns in this story that it is impossible to do it justice with a short summary: briefly, it depicts the experiences of a narrator called Utō who has returned to Otaru from Tokyo after an absence of about ten years. Several factors strongly hint that the story might be read in that mode of reading known as the shishōsetsu (I-novel), first identified by Japanese literary critics during the mid 1920s, which assumed close parallels between the lives of narrator and author.Footnote ² For instance, the names of author and narrator are virtually identical, they share the same native place to which they both return from Tokyo after a similar period of time away, and distinctions between author and narrator are blurred further through the literary device of narrating the story in the first person (watakushi) voice.

A consideration of the text's overall style of writing, however, indicates that it cannot be neatly pigeonholed as a straightforward I-novel. As the story opens, Utō emerges from Otaru railway station and takes in the sight of the surrounding buildings and streets in a way that promises to convey very ordinary, everyday experiences in a matter-of-fact descriptive style. However, such mundane realism proves short-lived, and the story soon takes on a far more experimental and fantastical tone. As the narrator walks through the town, he encounters a number of people from his past – some of them already dead – through a series of hellish and hallucinatory experiences. The majority of these people are young women who were deeply traumatized by Utō's mistreatment of them during his carefree younger days. But he also bumps into the ghosts of Proletarian literature writer Kobayashi Takiji (1903–33, given the name Ōbayashi Takiji in the story) who was murdered in Tokyo while in police custody, as well as the renowned author Akutagawa Ryūnosuke (1892–1927, Chirigawa Ryūnosuke in the text), whose suicide had shocked an entire generation of writers a decade earlier. The story's ending can hardly be described as positive: so many vengeful ghosts have pursued Utō that he has almost lost his mind. However, the very last sentence offers at least a glimmer of hope when he expresses a determination that, come what may, he will find the means to survive.

A central characteristic of stories that employ the fantastical mode is that they often resist easy interpretation. And yet, as Rosemary Jackson (Reference Jackson1995: 3) has pointed out, even fantasy writing might be understood as being rooted ultimately in specific historical conditions. One way to interpret the despairing tone of “Yūki no machi” may be as a reflection of the traumatic political and social vicissitudes to which Japanese writers had been subjected during the previous decade. When Itō first arrived in Tokyo he was intellectually invigorated by modernist experimentation, but that moment was now well and truly over. In the very year that Itō was writing his own story, Yokomitsu had already discarded his earlier innovations in favour of a far more conservative and introspective approach. His novel Ryoshū (Melancholy Journey, 1937–46), which grew out of his recent travels around Europe, sought consolation from a threatening and divisive world through a literary meditation on what it meant to be essentially Japanese. Meanwhile, the Proletarian literary movement had collapsed under internal ideological disagreements as well as governmental crackdowns. In 1933 Kobayashi Takiji was murdered by the authorities, but also that year imprisoned communist leaders Sano Manabu (1892–1953) and Nabeyama Sadachika (1901–79) renounced all ties with the Party, which they no longer viewed as a legitimate organ of the working class. As a consequence, large numbers of left-leaning writers came to recant their dearly held political convictions through the medium of confessional, so-called tenkō literature (Lippit Reference Lippit2002: 204).

Okuno Takeo spells out how “Yūki no machi” might be linked to these major social and political changes. He identifies Itō with a generation of writers – including Takami Jun (1907–65), Hori Tatsuo (1904–53), Sakaguchi Ango (1906–55), Dazai Osamu (1909–48) and Ishikawa Jun (1899–1987) – who participated in an artistic revival (bungei fukkō) during the mid 1930s. This group was very conscious of the fact that many of their number had come to renounce, often under duress, their most fundamental beliefs. Though Itō himself may not have aligned himself with the left-wing literary approach that now found itself under attack, he was still sensitive to the increasingly oppressive nature of the times. Against this background, the nightmarish demons in “Yūki no machi” may be understood as embodying an irrepressible sense of guilt, self-loathing and hypocrisy that many writers felt after having betrayed their own core identities (Okuno Reference Okuno1980: 118–20).

I take it as axiomatic that it is possible to assume some kind of link between any literary text and its wider social and historical environment, and this paper pursues the approach further by associating those demons in Itō's story with the broad colonial mentality of the late 1930s. My own research takes inspiration from earlier writings that have explored Japanese literary articulations of colonialism. Recent work by Michele Mason on the nature of Japanese colonialism in Hokkaido, and the writing of Komori Yōichi that reflects upon links between Japan and postcolonialism immediately come to mind (see in particular Mason Reference Mason2012b: 65–), and also Komori Yōichi (2001). In this article I turn to a series of close textual analyses in order to flesh out my argument, but let me acknowledge from the start that I am highlighting structures of colonialism in ”Yūki no machi” that are almost never expressed directly. In fact, there is only one clear reference to the theme – the term used is colonial culture (shokuminchi no bunka) – in the whole story (Yūki no machi, p. 829).

Furthermore, some critics have argued that Itō's story is unsuccessful precisely because it failed to take into account any wider social context. For example, when the story first came out, Communist literary critic Miyamoto Yuriko (1899–1951) complained that it painted a reality falsely based on a bourgeois subjective perspective and without links to the world at large. Itō himself compounded the impression that he was not the sort of writer with an interest in broader social matters through the waspish tone he used to counter Miyamoto's criticism: his response was that, having been poisoned by left-wing literature, she was simply out of touch with a new generation of intellectuals (Kockum Reference Kockum1994: 168).

Notwithstanding these exchanges, I would suggest that any relationship between fictional works and the wider social context is more complicated than either Miyamoto or Itō appear willing to contemplate. Rather, it pays to read a little more carefully between the lines of any literary text. In this connection, Alan Tansman has offered some interesting comments about Shiga Naoya's (1883–1971) writing during the 1930s that also throw useful light on Itō's story. Post-war Japanese thinkers have asked themselves how far an author as influential as Shiga should be considered complicit in the pre-war social environment when Japanese fascism was on the rise. In short, to what extent did Shiga's literary efforts contribute to the general political climate of the age? According to Tansman, the fact that, compared to the work of overt political ideologues:

No less than Shiga, Itō was a literary man with a real sensitivity to the evocative power of words. And, as Tansman puts it so aptly, writers of high repute respond to their environments indirectly, almost as magicians, casting spells and creating their own distinctive atmospheres. On the other hand, if even a magician is unable to exist in a vacuum, writers cannot help but contribute in one way or another to the ideological formations of the wider world. The fact is that, in the Japan of 1937, colonialism was in the air, so it should come as no surprise that it also found expression in the pages of Itō's text.

The controlling shape of ideas

Structures of colonialism, which speak of uneven power relationships between a dominant centre and a distant weaker locality, are observable in a variety of forms in “Yūki no machi”. Most obviously, such a relationship can be identified between the culture of mainland Japan centred round Tokyo and the peripheral outpost of Hokkaido. And from a wider perspective, a global configuration of power between Japan and its colonies on the Asian mainland was entering an increasingly aggressive phase during the late 1930s and was on the point of leading to full-scale invasion and war. In the pages below, I will address the significance of both of these paradigms of colonial power in “Yūki no machi”, but firstly it is worth giving consideration to a form of colonialist mentality that appears within the text more at the level of the individual; a mentality, moreover, shaped less by the larger matters of geopolitical relations than by currents of thought informing Itō's personal interpretation of the world. Put another way, the author's intellectual frame of mind had a determining influence on the topographical features that he came to highlight in the story.

In certain ways “Yūki no machi” stands out as a spatial network of streets and neighbourhoods through a painstakingly detailed sketch of Otaru, with the result that the reader is left with an extremely strong impression of realism. The western-style Hokuyō hotel is described as being located “on the left-hand corner of First Street at Inaho-chō, where it intersects with Number Two Fire Lane in front of the station” (p. 799). What is more, references are made to real-life places such as the Otaru Commercial Higher School that Itō attended as a student, and the Tsukimi Bridge near the waterfront, described as spanning “the canal between the reclaimed land and Minami Hama-chō. It leads to the Number One Fire Lane in Ironai-chō on the other side of the canal” (p. 805). This degree of precision implies that the text constitutes a faithful reproduction of the real town, an impression of verisimilitude consolidated further by the fact that a map of Otaru was inserted into the first edition of the story (Kockum Reference Kockum1994: 160).

However, there is more going on in the text than meets the eye. It cannot be denied that the concise layout of streets and districts, and the highlighting of various architectural landmarks – hotels, banks, post offices and so on – convey the impression that the text serves as a kind of guidebook for the visitor to Otaru. Indeed, many of the structures named within the text – for example the Nippon Yūsen Shipping Line building down by the waterfront (p. 802) – correspond to buildings located in the real town. But while buildings have their own concrete reality and undeniable presence, they cannot be reduced merely to structures made up from brick or wood or plaster. Buildings always signify something more.

The nature of this additional dimension becomes clearer through a consideration of the ways in which architecture functions in the text. As Utō walks down the long avenue that stretches from the railway station to Otaru harbour, the town passing before his eyes is pieced together in terms of a series of prominent buildings. One of these landmarks is a “three-storey brick-built shipping company with a nineteenth-century Russian style metal column and gallery” (p. 802). Attention to this degree of detail suggests a concentration on the “art or practice of designing and building structures” that is the primary definition of architecture. However, the narrator goes on to expand upon a secondary meaning when he reflects how “all memories of the past are caught up” in the various architectural forms of the landscape (p. 799). To be sure, Utō is making the obvious point that physical sites dotted around his native town embody memories relating to his own past. But he is also hinting at a more metaphorical (though no less significant) definition of architecture, namely as “a unifying or coherent form or structure”, an interpretation that makes it possible, I would suggest, to conceive of architectural forms not only as physical objects but also as structures of thought and feeling.Footnote ³

This secondary meaning can be fleshed out in “Yūki no machi” through an examination of the various ideas that, having fired Itō's imagination after his shift to Tokyo, went on fundamentally to shape his literary representation of Otaru. Consider, for example, the Western-style Hokuyō hotel that is depicted more in terms of a Freudian architecture of the unconscious than as a realistic mapping out of a physical structure. The hotel takes centre-stage at the beginning of the story when Hisae, a former lover who now looks shockingly aged, appears from nowhere and waylays Utō. Ignoring his protestations, she drags him into this hotel where ten years earlier they used to meet. An almost anthropomorphic bond between narrator and building is implied through Utō's observation that some grim fate, which surely awaits them within the hotel, “has tied a cord around us, and it is pulling Hisae and me deep into its dark corridors” (p. 800).

Once inside, a truly Freudian nightmare unfolds. Hisae pushes Utō along a dark corridor lined with countless identical rooms, until they come to the room where they used to have their rendez-vous. She points out a dirty white enamel basin and a bar of soap sharpened with a knife, and Utō is shocked to recall not only that this is the very place where he persuaded Hisae to abort her foetus, but even that he deserted her afterwards when she contracted an infection and was forced to enter hospital. The horrific nature of these sordid past events takes a disturbingly graphic form when the couple continue along the corridor towards the kitchen area and catch sight of two hairy-armed men preparing food amidst the rattling of cutlery, surrounded by “deep red tomatoes mashed together in a large, white bucket” (p. 801). Of course these labyrinthine corridors have some correlation to the realistic layout of a hotel, but only to a point; after all, it is hardly believable that the same bar of soap could have remained in the room for ten years. More important, however, is the way in which the corridors function as pathways that lead the narrator back into long-repressed memories.

Freud's influence can also be identified in the specific nature of the locations that the author chooses to depict in his story. For example, although the text represents Utō in a range of situations mostly designed to highlight disreputable aspects of his past, there is a distinctly scatological turn of events when he enters a filthy pubic toilet where “white urine stains are splashed everywhere and water has collected in cracks in the concrete corridor” (p. 810). As he stands at the urinal, the tormented ghost of another woman from his past appears. She informs him that, even though she is dead, she is still driven to return repeatedly to this toilet where she disposed of her own aborted foetus. Not surprisingly, Utō flees in terror and disgust the moment he feels her hand reaching out to touch him. The author once denied any autobiographical foundations to his story, but the confessional style with which it mercilessly uncovers sordid aspects of the protagonist's personal life carries some strong echoes of the I-novel. Nevertheless, at least as compelling is the way the text articulates a New Psychology School viewpoint that values Freudian-inspired themes as a tool for analysing the darker side of the self through literature.

The other intellectual framework that had a powerful impact on Itō's literary representation of Otaru relates to James Joyce, whose long-term influence cannot be overemphasized. Indeed, Michael Ainge (Reference Ainge1993: 330) has pointed out that, when Itō discovered Joyce while still a university student, he recognized a kindred spirit, since they had both spurned their hometowns for literary careers in a cosmopolitan centre of modernist culture. Moreover, 1936 was the year in which Itō re-read Ulysses and produced a new, detailed, synopsis of the novel; and the following year (just as he was writing his story) he declared his continuing devotion to Joyce in an article entitled “Yurishiizu yodan” (Ulysses, A Digression). In other words, Itō's loyalty to the Irish writer continued long after most other modernist-inspired writers in Japan had turned to more conservative literary forms (Ibid., 341; Kockum Reference Kockum1994: 158). The result is that, notwithstanding its realistic representation of Otaru as a specific town in the northern part of Japan, “Yūki no machi” depicts a locality that may equally be interpreted as a textual construction modelled along the same lines as the Dublin portrayed in Ulysses.

One point at which the ambiguous significance of Otaru as both a real and imaginary site comes to the fore is when Utō wanders into Yamada-chō, a district that now specializes in second-hand clothes stores but which, as the red-light area until a few years earlier, contains all too many memories connected to Utō's disreputable younger days. In keeping with the fantastical thread that runs through the story, the garments arrayed in these shops transform themselves into accusatory ghosts that rise up and fly after Utō, reminding him of his terrible past deeds: in the end, they engulf him in wave after wave until he feels himself drowning “at the bottom of the sea” (p. 820). Purely as an example of fantastical writing, the passage makes very effective use of a real district of Otaru in order to convey a disturbing and surrealistic impression of claustrophobic intensity. However, the scene has wider resonances. Kockum (Reference Kockum1994: 161) argues convincingly that this literary evocation of a sense of inescapable horror has close parallels to the grotesque and hallucinatory Circe section of Ulysses where Stephen rushes out of Bella Cohen's bordello, full of bitter remorse and self-recrimination. In this way, even events pertaining to a particular Japanese town become subsumed into a fantastical portrayal that goes beyond the local and hints at a broader sense of crisis and unease common to much modernist fiction.

Another way in which the apparently realistic depiction of an actual locality proves to resemble more closely a Joycean modernist expression of reality is through formal properties of the text such as the presentation of long lists of people and places. For example, Kobayashi Takiji was a young left-wing firebrand who attended the same high school as Itō in Otaru. The two boys did not get on. In Itō's story, the narrator encounters Ōbayashi's ghost who carries him off on a breath-taking flight through the air. At one point, they catch sight of hordes of people who, having committed the sin of failing to be “saved” by the Marxist theory that Ōbayashi espoused during his lifetime, now cry out for help as they plummet headlong into hell. These people are depicted as representing a wide array of more than twenty-five professions: a university professor, an engineer, a stockbroker, prostitutes, steel workers, chimney sweeps, and so on (p. 807).

The literary effect of reducing individual human beings into long lists is twofold. First – and this may be inspired by a proletarian literary approach – a kind of hyper-reality emerges whereby each person is identified according to precise aspects of their social background, thus confirming the complexity of interactions between different types of people through which any society is forged. On the other hand, such lists also evoke strong parallels with Joyce's literary techniques in the way they replicate what Ainge (Reference Ainge1993: 345) describes as “musical rhythm and momentum as sounds, apart from their denotations”. In “Yūki no machi”, these lists help transform specific people and places into streams of language that run through the text, producing the same sort of stream-of-consciousness effect that is found in Joyce's modernist work.

In Itō's text, there is even a moment where a stream-of-consciousness writing style actually takes the shape of flowing water, and the author explicitly spells out his literary indebtedness to Joyce. After fleeing from the toilet, Utō decides to walk along the Myōken River, where he notices triangular wavelets on its surface that gradually transform themselves into little speaking tongues. The mouths begin to taunt him, and even go so far as to liken him to Leopold Bloom, the fictional protagonist of Ulysses, as he “strolled along the Liffey River”. A succession of faces rises to the surface and reminds Utō of his misdeeds. For instance, Yuriko, whom he played along for years before deserting her and going off to Tokyo, emerges fleetingly to lament that she ended up marrying a bully who beats her so terribly that she is now close to death (p. 811). By appearing as an integral element of the river flowing through the town, fragmentary characters such as these serve to confirm how Utō's personal history is intimately woven into Otaru's topography. At the same time the fact that the Japanese author specifically references the landscape of Joyce's Dublin amounts to a conscious reconfiguration of a local and undeniably lived experience into a more general modernist structure of thought that translates both Irish and Japanese sites into abstract, textual constructions.

There are other ways to explore the relationship between Itō's encounter with Freud and Joyce and his literary depiction of Otaru, but enough has been said already to raise some pertinent questions centred around the colonial frame of mind that pervades the story. To begin with, if Itō's aim in “Yūki no machi” was to rediscover his roots in the specific locality where he grew up, why did he choose to articulate his personal history through the controlling lens of those Western ways of seeing that he had only recently acquired following his move to Tokyo? And in any case, since this textual representation of the author's provincial hometown unfolds through what is in effect a Tokyo-centred point of view, perhaps it would be more correct to argue that the text is not so much a case of Western colonialism per se than of a Japanese metropolis-centred colonial attitude towards its peripheral regions?

There is a lot to unpack here. After all, any easy separation of “Western” viewpoints from the diverse ways of seeing common to many Japanese writers and intellectuals since the earliest days of Meiji had been problematic. Moreover, the discourses of cultural power surrounding the relationship between artistic and literary movements based in Tokyo and the provinces are more complex than any simple bifurcation would suggest (see, for instance, Long Reference Long2012: 17–34). What is certain is that these big questions informed much more than a single story by Itō Sei. Indeed, they concerned a whole generation of writers during the 1930s, and would culminate in the July 1942 wartime symposium entitled “Overcoming Modernity” (kindai no chōkoku), when a group of Japanese intellectuals – including major figures such as literary critic Kobayashi Hideo (1902–83) and Kyoto-school philosopher Nishida Kitarō (1870–1945) – debated the nature of modern Western culture and its reception in modern Japan. They sought, albeit with limited success, to answer the question of what it meant to be modern, and whether it was possible to conceptualize a Japanese form of modernity that remained outside the orbit of Western influence.

The present paper touches only tangentially on these broader issues as it seeks to sketch out structures of colonialism in Itō's “Yūki no machi”. But any attempt to take up even this narrower concern needs to consider the author's relationship with Otaru and Hokkaido in a wider context. In other words, it is necessary to identify the text in terms of a history of place that includes, but also extends beyond, the experience of a single writer.

The bigger picture

July 1937 was a pivotal moment in Japanese history. A clash between Japanese and Chinese troops at the Marco Polo Bridge near Beijing escalated into a full-scale Japanese invasion of China and the tragedy of the second Sino-Japanese war (1937–45). At the same time, this incident might be identified as just one example in a long chain of events related to Japanese colonialism since the beginning of Meiji. In 1879, the tributary Ryukyu Kingdom was unilaterally reconfigured into the Japanese state proper in the form of Okinawa Prefecture. Japan's modern empire is usually dated from 1895 when Taiwan was ceded to Japan after the first Sino-Japanese war (1894–95). Other signposts of empire building include the annexation of Korea in 1910, and the invasion of Manchuria in 1931, followed by the establishment of the puppet state Manshūkoku in 1932. However, it could be argued that Japan's empire began as early as 1869, when Japan formally incorporated the island of Hokkaido into the Japanese state. If this is true, then Itō's story can be usefully explored not merely because the date of its publication coincides with the fateful incident that led to further colonialist expansion into China, but also because a closer look at Itō's literary representation of Otaru throws into sharper focus the wider history of Hokkaido that is fundamentally rooted in structures of colonialism.

During the early years of Meiji when the northern territory of Hokkaido had only recently been opened up for large-scale development, it was sometimes likened to Japan's very own Wild West. This phrase gained currency, partly from certain shared attitudes between the newly arrived mainland Japanese (Wajin) and the American settlers as they moved westwards across the United States. For example, both groups believed they had access to virgin territory they could claim as their own, and both had very few scruples about wresting territory from the grasp of the native inhabitants who were already there. But there are even clearer parallels that make the term appropriate. Kuroda Kiyotaka (1840–1900), the director of the Hokkaido Development Agency (Kaitakushi), invited the United States commissioner of the Department of Agriculture, Horace Capron (1804–85), to Japan to advise on Ainu affairs and the development of Hokkaido. Arriving in 1871, Capron carried out surveys for farming, logging and mining locations that were based on colonization programmes first employed against American native peoples after the opening of the transcontinental railroad (Komori Yōichi Reference Mason and Lee2012: 60–75).

There is another way of identifying the island of Hokkaido, known as Ezo until 1869, with a history of colonialism. During the late sixteenth century, the Matsumae warrior clan had crossed over into Ezo from Honshū and settled in the southern part of the Oshima peninsula, from which they pushed further inland as they gradually began to extend control. At the same time, the trade in seal pelts drew Russians into contact with the Kuril Islands to the north so that, by 1770, they had visited virtually every part of the island chain. In 1792–93, Catherine the Great (1729–96) sent an expedition further southwards into Ezo. Though the Matsumae clan treated their Russian visitors with due courtesy, they informed them in no uncertain terms that by right Westerners should not have entered the island, thus clarifying their understanding that Ezo already lay within the Japanese sphere of influence. This Russian advance pushed the Japanese Bakufu into spelling out their formal claim to Ezo in 1798, when it placed the southern half of the island under its own rule. In 1806 it went further by declaring the rest of the island as well as Sakhalin to be part of its own territory (Jansen Reference Jansen1996: 95). From a historical perspective, then, the northern island was generally assumed to be firmly under Japanese jurisdiction from the early nineteenth century, even if it was not formally renamed until the beginning of Meiji.

Given Hokkaido's location at the edge of Japan's sphere of influence and its consequent association with contested histories, it is not surprising that “Yūki no machi” should depict a native space that still carried traces of a Russian presence. For instance, on a soulless piece of reclaimed land by the waterfront, Utō comes across a small temporary hut that neatly encapsulates a story of cultural and political struggle between Japan and Russia. Hisae introduces Utō to his former Russian friend Vladimir, who has now become Hisae's lover. Flags adorning the hut advertise the nature of the event taking place inside: “Judo vs. Boxing: Grand International Match”. It turns out that this is a travelling show in which Vladimir's daily task is to perform as a Russian boxer, slugging it out against his Japanese judo opponent. The nationalistic and competitive nature of this ritualized performance is spelt out through another hyperbolic sign: “Spirit of the Martial Arts East and West” (p. 802).

However, this is no portrayal of a battle between equals. On the contrary, Vladimir's role confirms that colonialism is always a story of unequal power and domination. After expressing his delight at meeting Utō after so many years, the Russian immediately launches into a complaint that the Japanese audience “are never satisfied unless the Russian gets hammered by some flashy move from the Judo man” (pp. 802–3). In the broader context of the story, this is more than one individual's lament about daily personal humiliation. In various ways, Vladimir symbolizes all those emasculated Russians whose ability to claim a space for themselves in Otaru, and more generally in Japan, was always limited. He may be likened, for instance, to the Russian visitors from the eighteenth century whose presence was tolerated but not welcomed; he is the defeated side during the Russo-Japanese war of 1904–05; and he represents the many Russians refugees who fled the turmoil of the Russian Revolution to find themselves defenceless and adrift in Japan during the 1920s. In short, Vladimir's sorry circumstances in “Yūki no machi” serve to reconfigure the long struggle between two cultures into a theatrical spectacle in which the foreign threat now appears to have been domesticated and reduced to the level of entertainment.

But if Hokkaido reveals itself as a liminal space with a less than stable identity, a similar degree of ambiguity can be related even to the native place in which Itō grew up. Shioya now exists as a fully incorporated suburb of western Otaru, but originally it stood as an independent village on the coast facing Ishikari Bay. With no railway extant in the region in the early years of Meiji, the village was founded by herring fishermen who gradually migrated northwards along the coast bringing with them the culture, language and traditions of the Matsumae clan. And Shioya also became home to other immigrants, including Kansai farmers and merchants from the Niigata region. As a youngster, Itō experienced these demographic movements in terms of the village's social structure. The fishermen's children, with their Tōhoku accents, were known as the “beach kids” (hamakko), while the farmers' children, or “mountain kids” (yamakko), continued to imitate the Kansai dialect of their parents (Hayakawa Masayuki Reference Hayakawa1975: 1–5).

At a time when immigration into Hokkaido from mainland Japan was leading to an enormous increase in the Wajin population – from 60,000 in 1850 to 1,800,000 by 1913 (Totman Reference Totman2001: 320) – the mixture of different identities and traditions of the sort found in Shioya cannot have been unusual. However, family circumstances meant that Itō experienced an especially difficult relationship with his native place. The fact that his mother came from local Matsumae stock ought to have placed him firmly in the camp of the “beach kids”, but his father's itinerant background made this impossible. His father was a self-made man, originally from the Hiroshima region, who had escaped village life by enlisting in the army. After being posted to China during the Sino-Japanese war of 1894–95, he moved to Taiwan, thence to Hokkaido in 1901, at which point he left the army to become a schoolteacher (Kockum Reference Kockum1994: 3). The young Itō's literary bent endowed him with a strong sensitivity to language, and the result was that seemingly trivial matters, such as accent, mattered. Itō even suggested that he failed to develop a close relationship with his father (and his Hiroshima accent) because they lacked a common means of expression (Hayakawa Masayuki Reference Hayakawa1975: 2). In terms of his relationship with his native place, Itō's lack of any clear affinity towards either the children of fishermen or of farmers symbolizes the way in which he felt an even more ambiguous connection to place compared with other settler families.

Itō's tenuous sense of belonging emerged once more in his literary representation of Otaru, and it may be possible to trace its genealogy through the lens of a colonial mentality. If Shioya owed its existence to disparate groups of colonizing immigrants, Otaru likewise developed into a town of considerable stature thanks to a raft of ideas and physical objects – architectural structures in both senses, as it were – that had their origins elsewhere. For example, in “Yūki no machi”, when Utō notices souvenir shops selling maps of “the eight scenic views of Otaru” (Otaru meisho hakkei) (p. 799) near the railway station, the town seems to fit seamlessly into a mainland Japanese cultural lexicon of traditional landscapes. Throughout Japan, it was common practice to enumerate touristic sites in this way. However, it turns out that this list of Otaru's scenic sites was first put together during the 1880s, apparently by Saigō Tsugumichi (1840–1902), head of the Hokkaido Development Project (Kaitaku chōkan) (Hidaka Shōji Reference Hidaka1985: 85). Furthermore, colonial influences left their physical imprint on the shape of Otaru through the buildings dotted throughout the story. Satachi Shichijirō (1856–1922) designed the Nippon Yūsen Shipping Line building, and Sone Tatsuzō (1852–1937) created the Otaru branch of the Mitsui Bank. These leading architects had both studied at the Imperial College of Engineering (Kōbu dai gakkō) in Tokyo under the English architect Josiah Conder (1852–1920), designer of the Rokumeikan pavilion in 1883. The Mitsui Bank stood out as a prime example of English Renaissance style, while all the landmark buildings in Otaru revealed a typical Meiji-period mixture of Renaissance, Baroque, brick and stone styles (Hidaka Shōji Reference Hidaka1985: 93). In other words, Japanese metropolitan architectural standards, which were already drawn from European models, stamped the provincial town of Otaru with signs of central power and authority.

Another distinctive marker of colonial ties between mainland culture and Hokkaido was the railway. After abandoning his walk along the river, Utō enters the main shopping district of Hanazono-chō, where the commodification of modern life is on full display. He notices a variety of shops, including “toyshops, post card stalls, billiard halls, bookshops, Janome Sushi shops, teashops and camera shops” that satisfy an equally diverse array of modern consumers; “students, shop assistants, surveyors, farmers, mothers”, and so on (p. 813). The central role of the extensive railway system in facilitating the circulation of these goods becomes clear when the railway safety gate that crosses the main street goes down, and Utō watches a freight train rumble by. This turns out to be the Hakodate Main Line, a track that not only serves as the principal means of transportation linking Otaru to Sapporo, but also represents one small link in a nationwide transportation network tying the peripheral regions into an economic system centred on Tokyo.

In fact, the railway has a central role in fleshing out colonial structures of power from a variety of angles. For example, Utō reaches a point where the prospect of staying in Otaru a moment longer becomes unbearable, and he tries to buy his return ticket to Tokyo at a ticket booth. It soon becomes clear that the official in the booth, dressed as a representative of state authority in his smart uniform with its “stand-up collar and decorated with gold buttons” (p. 816), has considerable control over the lives of others. This first becomes evident through his condescending treatment of Utō: he does not even bother to look his desperate customer in the eye when talking. However, the official has a more specific role as guardian of colonial power. If the rail network can be understood as a system of economic and cultural capital emanating from the metropolitan centre to the provinces, then the official's refusal to sell Utō a ticket to Tokyo amounts to his prevention of what he identifies as an undesirable element from infiltrating the very heart of the empire. His specific reason for not providing Utō with a ticket is that his name appears on a list of Ainu names.

Reference to race-based discrimination is not at all surprising, since this way of categorizing hierarchies of people was another central aspect of colonialism prevalent in Japan and elsewhere at the time. We might recall, for example, that the United States had already passed the National Origins Act in 1924 with the aim not only of limiting European immigration, but also essentially of excluding any further Japanese immigration. Itō's story draws from race-based discourses in a variety of guises. For instance, negative race-specific physical attributes are used to confirm Vladimir's pathetic and alienated situation. He is depicted as voluble and clumsy, as having “thick (atsubottai) warm hands, and it felt a bit creepy touching his skin” (p. 803). At another point in the story Karl Marx's ghost, appearing in the sky over Otaru, is defined in terms of stereotypical physical attributes (hair, eyebrows, nose) that identify him as a Jew. To compound matters, Marx is introduced as the author of “The New Jewish Das Kapital” (Shin Judaya shihon ron) (p. 806). In effect, Itō's literary text reflects the culture and times in which he lived. Lest we forget, his story was written in 1937, just a year after Japan had signed the Anti-Comintern Pact with its closest ally Nazi Germany.

On the other hand, Itō should be credited with the fact that “Yūki no machi” offers no support for the crudest racialist ideologies which were circulating at the time. As Utō waits his turn at the ticket booth, a rabid xenophobe standing in the queue proclaims the need to defend the “pure Japanese spirit” against “people with impure or mixed blood”. This man is particularly suspicious of Utō's crinkly hair, considered a distinguishing characteristic of the Ainu. But while Itō clearly places the xenophobe in a negative light by presenting him as a loud and blustering fool, the author is still trapped within the racial discourses of his age. When Utō casts doubt on the Japanese-ness of the xenophobe, with his “earthy coloured face and hands like those of aborigines from the South Seas (nanyō no dojin)” (p. 816), he is ridiculing the xenophobe for his hypocrisy rather than for his racist attitudes. Moreover, Itō's acceptance of this colonial perspective is confirmed when an angry Utō walks away ticketless from the booth. He is mortified, not because he feels a sense of outrage against discrimination in general, but because he has been mistaken for an Ainu.

This was not, of course, the first time that a work of literature had displayed negative attitudes towards the Ainu. Indeed, a history of discrimination goes back to the first arrival of mainland Japanese on Ezo during the fifteenth century. From the Wajin perspective, the Ainu were generally considered childish and backward barbarians in need of protection from their more civilized neighbours. One Wajin legend even asserted that the Ainu were descended from dogs: it was this kind of extreme viewpoint that facilitated their demonization and the guilt-free exploitation of their labour (Siddle Reference Siddle1996: 40–44). Certainly, harsh treatment and the influx of diseases against which the Ainu had little resistance had a devastating effect on their population. It became so entrenched that by the early twentieth century the Ainu were commonly described as a “dying race” (metsubō naru jinshu) (Ibid.: 77).

On the other hand, “Yūki no machi” highlights ways of placing the Ainu within race-based ideologies during the Meiji period which may be specifically related to the development of a colonialist mentality. Michele Mason has suggested that the colonization of Hokkaido from Meiji went hand-in-hand with the construction of a modern nation state. Her argument is that Japan was only able to define itself as an advanced, civilized nation through the creation of binary relationships with apparently less developed colonial territories (Mason Reference Mason, Mason and Lee2012a: 6). Certainly, a link between the Ainu and other colonized groups emerges in “Yūki no machi” where the xenophobe is depicted as an aboriginal (dojin) from the South Seas. In fact, this derogatory term, with its implication of primitive backwardness, was first used during Meiji in connection with the Ainu, in the writing of an 1879 law relating to the legal position of the Ainu, who were renamed the “former natives” (kyū-dojin) of Hokkaido (Tierney Reference Tierney2010: 8). The term reappears in the story when Utō attends a lecture given by Chirigawa at the Inaho Elementary School. In a direct reference to Akutagawa's famous satire Kappa (1927), Chirigawa miraculously walks into a film being projected onto the stage, where he turns into the mythical Kappa creature and “climbs from branch to branch up a huge crape myrtle (saru-suberi) tree like a South Sea aboriginal (nanyō no dojin)” (p. 826). The Japanese name for the tree makes a less than subtle link between primitive peoples and monkeys (saru) in a way that echoes the association between the Ainu and dogs. This passage also reveals a more general point that, by the time Itō had come to write this story, the word aboriginal had become a generic term to describe colonized racial groups within the Japanese structure of power.

Itō, Otaru, the world

Between 1900 and 1940, the populations of Hakodate, Sapporo and Otaru were all roughly equal. Hakodate had a long history dating back to the Tokugawa period, but Otaru's rapid development from the Meiji period was remarkable. It began as a humble village in 1865, but by the time of Itō's birth, it was developing into an important commercial centre:

Otaru's good fortune arose from its two main industries – herring fishing and banking. The town was blossoming as a port for trade with Russia and the Asian mainland. On the financial side, the Bank of Tokyo chose Otaru as the location for its Hokkaido branch during the Meiji period, and this contributed to the city becoming well known as the Wall Street of Northern Japan until the economic crisis of 1927.

By the time Itō came to write his story it is true that Otaru had lost much of its earlier promise. However, it still bore memories of earlier glories and its links to the wider world. In “Yūki no machi” there is a point where Itō spells out precisely such a wider economic and historical framework of power and influence. In a rare moment of respite from his nightmarish experiences, Utō and an old friend escape to sit quietly for a moment in the Suitengū Shrine located on top of a hill within the city. He enjoys the surrounding view:

The fascinating detail, in which Otaru is identified through its world map reference, offers evidence that the town is perceived as a significant landmark within a global frame of reference that extends far beyond the narrow confines of Japan. It has, in other words, its place in the world. The nature of that place finds an echo through the British consulate, an indication of the ubiquitous presence of British Imperial power at the time. By association, Japan's own powerful impulse towards empire building is being spelled out. Moreover, the motives for colonial expansion also emerge through the billboard's seductive flickering. The sign speaks of battles over commodities, people and raw resources that were being played out on a far broader scale on the Asian mainland. “Yūki no machi” presents a town inhabited by ghosts; the irrepressible characters and incidents of an individual's past life, but also the disturbing spectre of Japan's engagement with the wider world in the late 1930s.

In this paper, I have attempted to expose some of the tentacles of Japanese colonialism as articulated in Itō's short story from a variety of angles. I have suggested how “Yūki no machi” sheds useful light on the real fears, tensions and expectations surrounding colonialism that had become an integral part of Japanese culture and society. My study has also touched on physical and geographical structures of colonialism that speak of uneven power relationships between a dominant centre and a distant, weaker, locality; from the very shape of Otaru's buildings to the national railway network that controlled the flow of goods between periphery and centre. More generally, I have drawn attention to Itō's story because I believe it effectively highlights a global power configuration between Japan and its colonies that was entering an increasingly aggressive and bellicose phase during the late 1930s.