Like the 1900 Taiwan Banjin jijō , Inō Kanori’s 1905 Ryō Tai jūnen shi (10-Year History of the Occupation of Taiwan) was compiled for the Government-General’s second-in-command, Gotō Shinpei.^[38] Gotō’s preface stressed that the history of Taiwan was testament to Japan’s achievements as a modern colonial power. Inō of course inscribed Gotō’s progressive view of history into this historical digest, though he stumbled in his short chapter on indigenous Administration, the topic he knew best. As a government scribe, Inō imposed a linear, progressive narrative structure upon the confused history of Japanese-indigenous relations by making Japan’s “punitive policy” into the dynamic element of the narrative. Inō thereby de-emphasized the record of conflict within the administration and the complex story of frontier diplomacy in the earlier period. This simplification, in turn, erased the pluralistic view of Taiwan’s internally differentiated Aboriginal population from the official narrative, while maintaining the major distinction between Chinese and Austronesian races.

In its chapter on indigenous Administration, the “10-Year History” ignores the first eight months of martial law on Taiwan (August 1895–March 1896) to open with the Government-General’s declaration of civilian rule on April 1, 1896. This opening gambit is important, for it establishes the Confucian subtext of Inō’s preferred and intended narrative structure: civil government is normal, ideal and laudable, while martial law is a last resort, an expedient for failed policies. Again we can detect Gotō’s hidden hand here, recalling that the Minister of Civil Affairs insisted, upon taking the portfolio in 1898, that he be paramount to all military men in Taiwan, except for the Governor-General, Kodama Gentarō. To dramatize his much-publicized belief that military rule was ruining the colony, Gotō actually struck a naval officer in front of a military audience to defend his own honour. Kodama, ever Gotō’s protector, approved of Gotō’s brash action.^[39]

The first event of Inō’s history, then, is the establishment of the Pacification-Reclamation Office (bukonsho). The Bukonsho, wrote Inō, was chartered “solely to enact ‘moral suasion’ (kyōka) among the Aborigines”. Inō then added, contradictorily as it turned out, that the Bukonsho was also charged with overseeing the “economic development” (kaihatsu) of the “Aboriginal territory” and “finding useful employments for the Aborigines (banjin no jusan)”. The tension between kyōka and kaihatsu becomes clear if we comprehend “moral suasion” as a spatial metaphor rooted in the “Sinocentric” topographical political imagination,^[40] and conceive of kaihatsu as a temporal metaphor more appropriate to Enlightenment theories of progress. In the former model, the centre of calculation is the Imperial Centre itself, eternal, patient and inevitably triumphant. In the latter model, the centre of calculation is the state’s political leadership, which resolves conflicts and defines efficiency in the context of national interest in a world of competing nation-states.

Perhaps anticipating Gotō’s views on the subject, Inō posited headhunting as the defining trait for “certain tribes in the northern half of the island” to introduce the Aborigines in his “Ten-Year history”. This stereotype, based on a single trait of the population in question, was precisely the kind of demonizing Inō decried in his vindicationist mode five years earlier. Adding force to the “trope of the savage headhunter”, Inō used the contrasting term ryōmin to describe their victims. In Qing-period usage, ryōmin (Chinese: liangmin) referred to tax-paying artisans, merchants and agriculturists: literally, the “good people”.^[41] Inō attributed ryōmin victimhood to the atavistic Aboriginal “custom” of headhunting, initially constructing a culturalist explanation redolent with the rhetoric of privation. In the discursive field of “moral suasion” (kyōka), then, such “evil customs” would ideally be reformed by the civilizing, edifying influences of the centre, as transmitted by civil (Chinese: wen, Japanese: bun) institutions like the Bukonsho.

In the first turning-point in his narrative, Inō recounted that the Bukonsho could not stop Aboriginal attacks on ryōmin through moral suasion alone. Therefore, the Government-General formulated a system of punishments (chōbatsu) directed at Aboriginal headhunters in late 1897. Despite this concession to expedience (force), Inō assured readers that Japanese policy remained organized on the principle of “reassurance through acts of kindness (suibu)”, to argue that the “civil” impulse was still ascendant around 1898.

Quite abruptly, Inō then changes tack to describe headhunting incidents as “acts of murder and assault” (kyōkō ) to explain the government’s expansion of police forces (keisatsu) along the Aboriginal border in 1898. The new intolerance of “assault and murder” can be read as headhunting’s redefinition from “custom” to “crime”. This reconsideration was warranted, according to Inō, by a “fear that headhunting would stop plans for Aboriginal-territory development dead in their tracks”. Curiously, Inō neglects to mention any specific commodities or economic activities that might have been connected to headhunting at the time (though he surely knew, as we shall see below).

As violence became unmanageable on the Aboriginal frontier in 1898, the Government-General lacked a unified plan. The argument over whether to consider headhunting as a custom in need of reform or as crime in need of punishment fomented a “clash of opinions” within the bureaucracy. Still maintaining the Confucian perspective, Inō called the “moral suasion” emphasis of Bukonsho civil administration the “positive policy” and referred to “punishments” as the “negative policy”.

Then, in what Inō called a “Great Revolution”, in June 1898 the Government-General dissolved the Bukonsho, for allegedly leaning too far in the direction of leniency/attraction to the neglect of force/punishment. Henceforth, Aboriginal Affairs was put under the rubric “severity tempered with leniency” (on’i narabi okonawaru), a dignified locution for “carrot and stick”. Subsequent narratives characterized the dissolution of the Bukonsho as a necessary response to Aboriginal savagery. Inō, however, intimated that perhaps it might have been made effective if given more time. Such a hypothesis would explain why Inō switched back to the vindicationist mode in this narrative, now casting the northern Aborigines as history’s victims. Halfway through Inō’s account, the Han are transformed from “ryōmin” (good people) into “Chinamen”. Temporarily abandoning the rhetoric of privation that explained headhunting in terms of “savagery”, Inō implied that both Han and Aborigines were to blame for the mayhem that was impeding Japanese development in the highlands. Recalling an early staple of Japanese official rhetoric before the Kodama-Gotō era, Inō now claimed that Taiwan’s Chinese settlers provoked Aboriginal bloodlust by taking advantage of their ignorance and stupidity (gumō ) to perpetrate land swindles.

Ostensibly acting as an impartial broker to stop the revenge cycle, the Government-General decreed that all non-Aborigines (Chinese) obtain permits to reside in “Aboriginal territory” in February 1900. At this point in the narrative, Inō’s confusion and discomfort become palpable. Even after situating Aboriginal-initiated violence within a context of Han perfidy, Inō back-tracked to describe Aboriginal “assault and murder” as irrational behaviour. In a tortured locution that I read as a concession to the preferences of the administrators above him, Inō wrote: “As the number of people entering the Aboriginal districts increased, the ignorant Aborigines harboured suspicions of invasion, prompting the Aborigines to perpetrate outrages on an immeasurable/disproportionate scale.”

Accordingly, from 1902 onward, heavily armed and staffed guardlines (Japanese: aiyū; Chinese: aiyong) were extended to physically separate Aboriginal territory from the rest of Taiwan. In January 1903, all of northern Aboriginal country was placed under police jurisdiction, in a complete concession to the “expedient” of police rule. Bringing his narrative up to the present (1905), Inō wrote that the Aboriginal population north of Puli in Nantou prefecture was now governed solely under the rubric of “force and intimidation” (iatsu kyōsei), while the southern Aborigines would be governed under the banner of “education and largesse” (keihatsu suibu). In this perplexing document, Inō has the northern Aborigines commit atrocities because they are provoked by cunning Chinese invaders who take advantage of their ignorance. And yet the Japanese government responds by ruling these historical victims under the banner of “force and intimidation”. At the same time, readers are to believe that by 1905 the “southern Aborigines” were willing objects of non-coercive policies of tutelage and guidance. Inō concluded this report with an unconvincing assurance that the Government-General’s two-pronged approach would likely produce good results sometime in the future.

For students of the Taiwan Banjin jijō , as well as for Japanese administrators at the time, there could be little doubt that the choice of Puli, Nantou prefecture, as a dividing line in Inō’s narrative meant that Inō was defining the “northern tribes” as the Atayal. In Inō’s taxonomy of 1899, all tribes south of Puli, which is the geographic centre of Taiwan, do not practise the art of facial tattooing (although limb tattoos were found throughout the island). Other than this single distinguishing feature, there was no ethnologic basis for bifurcating Taiwan’s Indigenous Peoples into southern and northern halves. This division was, instead, political. While Inō’s survey anthropology of the late 1890s demonstrated internal diversity and a welter of varied political conditions in rural Taiwan, his narratology, like the old Qing paradigm he once tried to overturn, employed a typology that sorted Aborigines into “good” (southern) and “bad” (northern) imperial subjects. Inō did not invent this Manichean nomenclature; he merely reappropriated it from the “amateurs” he once ridiculed as unfit to perform ethnological analysis.

From as early as 1896, Japanese officials, notably Nagano Yoshitora, found that Bunun warriors in the environs of Puli could be persuaded to bear arms for the government in its skirmishes against villages north of Puli, who fell under the broad rubric “Atayal”.^[42] From this time forward, the “southern tribes” became symbols of compliance in official documents, while the Atayal became stock villains in Japanese rhetoric. It is important to note here that Bunun were also known as headhunters and, in fact, could be induced to take heads at the behest of the Japanese state. We can thus conclude that Inō’s 1905 narrative located savagery not in the act of murder, but in the act of non-compliance with government demands.

According to several contemporary sources, the Atayal tribes clustered around Wushe (“Musha” to the Japanese) made themselves odious to the Japanese when a few villagers murdered 14 Japanese road surveyors in February 1897, the ill-fated Fukahori expedition. The long saga of investigation, recovery of remains, and economic blockades to punish those responsible is indeed a central thread in the area’s history under Japanese colonial rule.^[43] Probably more important, though, is the fact that many of the settlements just north of Puli abutted rich stands of camphor. Several “northern tribes” were active in the lumbering trade, charging outsiders like the Japanese fees for access to the forests in the early years of colonial rule. Disputes with a particularly powerful Saisyatt entrepreneur-chief named Ri Agui erupted into a large-scale war in late 1902. Like the Atayal, the Japanese subsumed the Saisyatt under the term “northern tribes” in later discourse, again suggesting that “northern tribes” was really shorthand for “non-compliant” villages. Mochiji Rokusaburō is considered by many to have been the brains behind Gotō’s Aboriginal policy. In his famous position paper of late 1902, he proposed that the Aboriginal Territory be divided into northern and southern sectors, with the former earmarked for military conquest, the southern for “moral suasion”.^[44] And thus, the rough-and-ready demarcation of operational zones in the war to increase the empire’s wealth obliterated fine distinctions made by survey anthropologists like Inō Kanori, who wrote in 1899 that “the disparities in cultural attainment within tribal divisions is often just as varied as the disparities between tribes themselves”.^[45]

The image of the north/south division among Aborigines was cemented in a rather grisly affair on October 5, 1903. Then, Japanese officials induced their Bunun allies from Kantaban to entrap, ambush and slaughter more than 100 Atayal men (from Palaan and Hogo) in a single morning near Puli. These so-called southern tribesmen actually redeemed the heads at a Japanese outpost, and were photographed doing so. Thus, by 1903, Japanese administrators began to define all Aborigines north of Puli, who were distinctive because of their facial tattoos, as a problem population as a result of economic disputes over rights to camphor stands and local resistance to Japanese attempts to survey remote areas. The empirical poverty of this conceptual apparatus is laid bare when we observe that Japanese army and police forces fought a number of pitched battles with villages of “southern tribesmen” (Paiwan and Bunun especially) well into the 1930s.^[46]