The Qing empire ceded Taiwan to Japan as part of the settlement to end the Sino-Japanese war of 1894–95. As the Taiwan Government-General began setting up its capital in the face of armed resistance in June 1895, reports describing the curious folkways of the empire’s new subjects began to circulate in Japan. Especially prominent in the early wave of “first-encounter” documents were travel accounts of the hill tribes, collectively known as “banjin”, “seibanjin”, “yabanjin” or “banzoku”.^[10] Even before the Government-General could safely inhabit its capital, Takigawa Miyotarō published “Our New Territory: The Island of Taiwan” to popularize the quasi-ethnographic information contained in Ueno Sen’ichi’s famous military intelligence report on conditions among the Aborigines.^[11] Ueno’s report was an amalgam of first-hand accounts and information collected by British lighthouse-keeper George Taylor.^[12] Well into the 1900s, the occupation inspired popular ethnography for Japanese consumption, in the form of newspaper, magazine and scholarly accounts of life in “Darkest Taiwan’s” interior.^[13]

For the small coterie of anthropologists attached to Tokyo University, the ethnological bounty of the new colony proved irresistible. The intellectual backgrounds and institutional affiliations of the major players, Inō Kanori, Torii Ryūzō and Mori Ushinosuke, have been well documented elsewhere.^[14] For our purposes, it is enough to say that Inō Kanori (1867–1925), our major protagonist, set sail for Taiwan on November 3, 1895. At the time, Inō supported himself as an editor of an education journal while contributing notes on folklore to the Journal of the Tokyo Anthropological Society and attending the lectures of Japanese anthropology’s founding father, Tsuboi Shōgorō. Inō embarked under the auspices of the Japanese Army, thereafter working in the documents section of the Government-General and as an administrator of the Japanese Language schools, pursuing his interest in Taiwan anthropology between assignments.^[15]

Inō’s most remarked-upon contribution to Taiwan anthropology was precisely the kind of “epistemic operation” described by Nicholas Thomas as quintessentially modern: an ethnic map cum taxonomy of the Taiwan Aborigines. Inō sought to replace the casual observations of his amateur co-nationals and the pre-modern Qing descriptions of Taiwan Aborigines with a scientifically ascertained taxonomy based on the investigation of racial-cultural diversity in upland Taiwan. Inō Kanori succinctly stated these goals in mid-1895:

Inō’s manifesto (and subsequent writings) called for Japanese survey anthropologists to identify the unique features of each shuzoku (tribe/ethnos) on Taiwan in order to better understand the differences among the groups subsumed under the Qing terms shengfan (raw barbarian) and shufan (cooked barbarian). Inō also emphasized that anthropology should render faithful service to the state as a form of intelligence gathering. These two goals would come into conflict, I will argue, undermining Inō’s ability to construct a coherent account of Japanese relations with the uplanders, in effect forcing him to choose between loyalty to an emerging discipline or obedience to his bureaucratic superiors.

On May 26, 1897, Inō formed an expedition party to begin a 192-day ethnographic survey tour by order of the colony’s Bureau of Education. The Government-General ordered Inō and his partner, Awano Dennojō, to devise a portrait of Aboriginal society for the purpose of making recommendations on the subject of Aboriginal schooling. The results were sent to Gotō Shinpei in early 1899 as a report titled Taiwan Banjin jijō (Conditions among the Taiwan Aborigines).^[17] Considered Inō’s magnum opus, Banjin jijō is a rich, descriptive and internally conflicted document that speaks in multiple voices, reflecting Inō’s intermediary position in the colonial order of things. Relying on his own observations in the field (though never in any one spot for long), archival research in Chinese records and interviews with Pacification-Reclamation officers,^[18] Inō constructed a matrix of defining traits — physical features, everyday usages and implements (dozoku), cultural practices (kanshū), language and oral traditions — to classify the inhabitants of Taiwan’s interior into eight discrete ethnic groups.

The Janus-faced nature of this document, a testament to survey anthropology’s ambiguous legacy, is illustrated by Inō’s characterization of the Atayal peoples. On the penultimate page of his 283-page report, Inō warned Japanese officials against the temptation to caricature the Aborigines (banzoku) as savage headhunters.^[19] In Inō’s taxonomic grid, “headhunting” comprised a single item out of six elements called “customs”, while “customs” themselves stood beside other bundles of defining traits, such as “physical features”, “language”, “technology” and others. Inō emphasized that many Aboriginal groups had ceased headhunting, but even those who continued, like the Atayal, were also competent agriculturists and weavers. Moreover, continued Inō, the savage custom of headhunting was perpetuated as a form of defence against aggressive Han settlers. Inō finished by asserting that the tribes of Australia and Africa were much more primitive than Taiwan’s headhunting Atayal, thereby relativizing their backwardness by recourse to the accumulating world-wide database of “cultures” put into play by the armchair comparativists of yore.^[20]

Going completely against the grain of his conclusion, Inō began the substantive sections of Banjin jijō by fixing the Atayal peoples as Taiwan’s least-advanced tribe, describing them as preternaturally xenophobic, bloodthirsty headhunters responsible for over a hundred beheadings annually.^[21] Inō’s evolutionary ranking of the tribes, in what we might today call the “bullet-points of the report”, attributed the Atayal’s bottom position to environmental factors:

Thus, Inō’s evolutionary perspective reproduced elements of the old Qing “Sinocentric World Order” ethos that equated “civilization” with proximity to China’s sacral-political centre. Emma Teng, in her analysis of Qing nomenclature, discourse and travel writing vis-à-vis the Taiwan Aborigines, identifies a persistent strain in Qing documents which demonizes the Aborigines for their lack of civility. She terms such discourse the “rhetoric of privation”, in contrast to the more romanticized rhetoric of primitivism (the “Noble Savage”).^[23] Though Inō equated distance from Chinese influence with savagery in much of his ethnology, he severely criticized the rhetoric of privation as non-scientific in another venue, writing:

In this passage, the term “political compliance” is loaded. From the Sinocentric World Order perspective, the court of the Chinese emperor is the metonymous centre/apex of tradition, refinement, power and learning. The source of humanity-making benevolence is configured as a geographical node of virtue, which radiates outward and downward via the power of attraction, imitation and what we might today call acculturation. The boundaries of the realm of civilization are extended by bureaucracy, the repository of Confucian learning and instrument of Chinese statecraft. Thus, in the passage above “the absence or presence of political compliance” also denotes “cultural” submission to the Chinese centre.^[25] Thus, Inō’s critique prefigures Thomas’s characterization of pre-modern discourses about the Other; they are distinguished from modern scientific discourses by their fixation on lack or presence, their overbearing concern with the “values of the centre”.

There are, then, two major contradictions in Inō’s ethnology of Taiwan Aborigines. First, the relativizing rhetoric of the Taiwan Banjin jijō ’s conclusion contradicts the rhetoric of privation that permeates the body of the report; and Inō’s explicitly modern-pluralist approach to taxonomy is undermined by his ultimate recourse to the Sinocentric preoccupation with the Atayals’ physical and cultural distance from the Middle Kingdom. These glaring contradictions call for explanation, because Inō was, if anything, a deliberate scholar, a man obsessed with establishing himself as a member of the Meiji-period bureaucratic-literary elite.

As an ethnologist in 1899 colonial Taiwan, Inō was writing against a discourse that put the human status of the Atayal into question. Anthropologist David Scott uses the term vindicationism for such narratives.^[26] In other words, if the question is: “Are the Atayal beasts or human beings?” then Inō’s reply, in the vindicationist mode, is “They are human beings.” Inō’s contemporary Torii Ryūzō, fellow survey anthropologist and veteran of Tsuboi Shōgorō’s seminars, also laced his ethnological notes with vindicationist rhetoric. Moreover, Torii’s interpreter, Mori Ushinosuke, who would himself become a prominent government expert on Aboriginal languages, was an adamant vindicationist as well.^[27] Thus, it would be fair to characterize Taiwan survey anthropology of the Meiji period (1895–1912) more generally as a vindicationist enterprise.

Banjin jijō was, however, only partly an ethnological study. Primarily, it was edited and abbreviated for practical application as a report submitted to Gotō Shinpei, Taiwan’s Minister of Civil Affairs from 1898 to 1906. Analyzing the interplay between Gotō the powerful administrative superior and Inō the dutiful bureaucrat is as important as it is difficult.^[28] As a self-styled visionary and actor on the global stage, Gotō, an accomplished physician and public-health administrator, frequently invoked the scientific method as a rationale for his policy proclamations. Gotō’s avowed appetite for research on colonized populations, is matched by Inō Kanori’s reputation as a producer of such knowledge. Inō, the indefatigable, driven and scrupulous editor, compiler, analyst and fieldworker, is commonly regarded as the father of modern Taiwan Studies, and was certainly the government’s acknowledged expert on Aboriginal country around 1900. In addition, both men hailed from the area of northeastern Japan’s Iwate prefecture, giving them common cause as rising men from Japan’s rural periphery. Considering these factors, one would expect Inō’s ethnological labours to have had a large impact on Gotō’s view of Aborigines in Taiwan. Paradoxically, it appears that Gotō influenced Inō’s thinking instead; though Gotō of course had very little specific knowledge about the Aborigines themselves, and was ostensibly being informed by Inō’s work.

E. Patricia Tsurumi has aptly characterized Gotō’s rough-and-ready sociology of Taiwan as garden-variety Spencerian evolutionism.^[29] Tsurumi’s judgement finds evidence in a much-reproduced 1901 policy statement entitled “An Opinion on the Necessity of Conducting a Survey into Customary Law for the Governance of Taiwan”. Here, Gotō applied Spencerian logic to assert that Taiwan’s Chinese population was not ready for the sudden introduction of fully civilized Japanese legal codes, because it had become accustomed to a partially civilized legal regime during 200 years of Qing rule. In other words, the rights guaranteed to Japanese subjects under the 1889 constitution would not be granted to Taiwanese (though, of course, the obligations would) for fear that too-sudden a change would shock the “organism” of Taiwanese society. And as for the Aborigines, Gotō used the general marker for savagery, yaban, to degrade them and assert that they also could not be governed through modern law codes.^[30] Gotō referred to the “savages who dwell in the undeveloped lands” as living fossils from antiquity in a classic example of what Johannes Fabian has called “allochronic” discourse.^[31]

On one important point only, it appears that Gotō incorporated Inō’s ethnology into his own thinking. His declaration that the Aborigines and the Chinese were distinct populations was of a piece with Inō’s 1895 manifesto quoted above. We shall return to the significance of this agreement below. On the whole, however, it appears that Gotō was more hostile to than ignorant of Inō’s report of 1899. In what must have come as a stinging rebuke to Inō Kanori, Gotō applauded the efforts of government employees (like Inō?) to submit their hard-earned local knowledge to the government in the form of reports. Gotō rejected, however, the existing knowledge at hand as too unsystematic and non-specialist. Gotō wrote that Western nations had sufficiently developed scholarly communities to let specialists compete among themselves to study native customs, laws and economy; in these advanced nations, the government only had to convene these scholars and reap the harvest. For Gotō, Japan’s civil society (kokumin/“national people”) was still too immature for its government to take such a laissez-faire approach.^[32]

Matsuda Kyōko argues that Inō’s recourse to Social Darwinism stemmed from his visceral reaction to harsh research conditions. Poor infrastructure, lack of security and forbidding terrain combined to provoke Inō to project his “struggle to conduct a survey” onto the Atayal peoples as a “struggle for survival”. In short, Inō reasoned that the Atayal had been pushed to such extreme living conditions because they had been forced into the interior by superior forces (the Chinese).^[33] This analysis is attractive, for it shows the survey anthropologist responding to the local environment, yet in such a way that his own relationship to the culture-bearer is reified into an enduring characteristic of that society, exposing both the strengths and weaknesses of the genre. At the same time, Matsuda’s analysis does not explain the grave contradiction between Inō’s vindicationism and his rhetoric of privation.

My alternative explanation is admittedly speculative, but has the advantage of clarifying the contradictions within Inō’s corpus. I believe that Inō adopted elements of the rhetoric of privation and the language of Social Darwinism in the Banjin jijō to anticipate or answer to Gotō’s objections to his vindicationism. Inō began and finished his field survey before Gotō came to Taiwan, concluding on December 1, 1897. During a 13-month interval, Inō collated his data, read more deeply in Qing documents and drew his conclusions. He filed his report to Gotō on January 9, 1899, about eight months into Gotō’s tenure. During his write-up period, he was fired as part of Gotō’s and Governor-General Kodama Gentarō’s administrative house-cleaning of March 1898, only to be re-hired soon after. He then quit again in December 1898 to return to Tokyo for a year.^[34] Considering that Gotō’s fondness for evolutionary metaphors was well-known to Inō during a period of intermittent unemployment, it seems not unreasonable to expect that Inō would recast his survey ethnology to meet the expectations of powerful and sceptical readers. If Inō often portrayed himself as the centre of calculation vis-à-vis colonial policemen, military officers and amateur ethnographers,^[35] he in turn answered to an even more paramount centre of calculation in the person of Gotō Shinpei.

As we have seen, Gotō Shinpei did not think Inō’s survey worthy of the name “science” in 1901. Nonetheless, he sufficiently appreciated Inō’s skills as an editor and compliant underling to commission him for several more projects, making Inō, in effect, the Government-General’s in-house historian of indigenous Administration, for both the Qing and Japanese periods.^[36] In this new role, Inō would begin a second career in Taiwan as an historian,^[37] shifting his purview from the mapping of cultures in space to the identification of meaningful segments of linear time.