Divided Island

The Divided Island: Taiwan’s “Savage Border” Under Qing Governor Liu Ming-chuan and Japanese Governor-General Kodama Gentaro

FINAL PAPER

Name Student #

Timothy Hogan 97924012

Course Instructor

Taiwanese History Professor Chou Whei-ming

Date

January 22, 2009

International Masters in Taiwan Studies National Cheng-chi University

1

Introduction Although Taiwan came to the world stage relatively recently, its 400-year history has been the focus of much academic inquiry. Historians have explored Taiwan’s role in the European exploration and colonization of Asian and Pacific islands by the Portuguese, Spanish, and Dutch in the sixteenth and seventeenth centuries, the waves of Han immigrants that arrived from across the Taiwan Strait in the seventeenth and eighteenth centuries, relations between the Han settlers and the island’s indigenous inhabitants, Qing administrative efforts to govern the island and manage its opening to foreign trade in the latter half of the nineteenth century, and its colonization by Japan at the end of the nineteenth century. Taiwan also offers rich ground for social scientists, with its diverse inhabitants maintaining an amazing degree of heterogeneity in their religious beliefs, languages, and cultural practices, and for geographers and ecologists. The island’s Central Mountain Range, with over one hundred peaks above 3,000 meters in height and the highest peak in East Asia, Mt. Jade, and cleaved by deep river gorges that wind their way into the heart of the island, is home to a rich diversity of plant and animal species, many that are found nowhere else in the world. Finally, Taiwan’s experience under the rule of the Kuomintang (KMT) in the second half of the twentieth century, in which the island underwent serious political and cultural repression, rapid industrial development, and eventually democratic transition, has attracted considerable notice from scholars in economics and political science. Despite the breadth of the scholarship on Taiwan, there are areas where further investigation is justified, particularly when the topic of research straddles the lines of conventional academic divisions. One subject that draws together politics, geography, sociology, and economics into a historical narrative and analysis is the border that historically divided the island into two territories, one settled by the Taiwanese, a blending of people with Han and Sinicized indigenous ancestry, and one home to Taiwan’s un-Sinicized indigenous inhabitants.1 This border, known as the fan-jie (savage border) in Chinese, and bankai in Japanese, in its various forms and locations, has played an important role in Taiwan’s spatial, social, political, and economic development. In order to provide a greater understanding of the border in historical context, this paper will explore its history in the closing years of the nineteenth century and the dawn of the twentieth century, in the decades before and just after Japan claimed the island as its first 1

In this paper, Taiwan’s indigenous (or aboriginal) people, of whom there are currently thirteen “tribes” recognized by the central government, are referred to as Sincicized and un-Sinicized indigenous people. These terms were chosen because they reflect the Han conceptualization of the ethnic identity of the indigenous people living on the island during the periods discussed in this paper. The Mandarin Chinese terms used to describe these two groups of indigenous people were huafan (cooked rice) and shengfan (uncooked rice), and were meant to express the idea of assimilated/Sinicized and unassimilated/unSinicized indigenous people respectively. The Japanese initially adopted their own translations of the Chinese terms, kaban and seiban. In 1899, Japanese ethnologist Ino Kanori classified the indigenous peoples of Taiwan into eight tribes based on cultural and linguistic affinity (Barclay 2007, 71). This classification has largely remained in use, but since a reawakening of ethic consciousness in the 1990s, several more groups have been identified as a result of granting tribal status to groups previously listed as sub-tribes and also by recognizing Sinicized groups that were previously thought to have lost their indigenous cultural identity.

2

colony. It will examine the features of the border’s development by contrasting its administration during two separate periods of governance, 1884 to 1891, when Liu Mingchuan governed Taiwan for the Qing in Beijing, and 1898 to 1906, when GovernorGeneral Kodama Gentaro headed the Japanese colonial government on the island. The paper’s structure will include discussion of the border’s evolution prior to Liu’s arrival in Taiwan, the border policies established by both leaders, the development of a guard-line during their administrations, and an evaluation of the effectiveness of the border policy during the two time periods. Finally, the paper will offer a brief discussion of the significance of the historic border in contemporary Taiwan. In this paper, the term border will be used to denote that concept which Prescott (1965) refers to as a boundary, or an actual line of demarcation, whereas the term frontier will be used in accordance with his terminology, in which it refers to a zone adjacent to a boundary.2 Prescott’s discussion of the typography of boundaries is a useful starting point for this examination of Taiwan’s guard-line, given that he introduces historical examples of efforts by states to mark their borders by building walls to prevent foreign invasion, as in the case of China’s Great Wall, or Hadrian’s Wall, built in Britain by the Romans. These constructions are apt comparisons with Taiwan’s guard-line in that all three were defensive fortifications along which troops, both garrison soldiers and informal militia groups, were arrayed. One significant point of departure is the fact that Taiwan’s guardline was a fortified fence rather than an earthen or stone wall. Although an earthen barrier known as tu-niu was erected to demarcate the extent of territory open to Han settlement in Taiwan under the Qing government, the advancing frontier frequently made such a fixed border meaningless. As the need to open up new lands for settlement or to defend the camphor industry increased, so the border had to be adjusted to suit the changing circumstances. Border as Policy Tool From its inception in the latter half of the seventeenth century until the Japanese occupation of Taiwan in 1895, Taiwan’s border was a tool of Qing policy that functioned to divide the population of the island. This was done for several reasons. The first was to limit the expense of governing Taiwan. It wasn’t until 1887 that Taiwan became a province of the Qing Empire. Before that it was ruled as a prefecture of Fujian, across the Taiwan Strait. A small number of officials were posted on the island, due both to the Qing’s seeming lack of interest in developing Taiwan and the island’s low tax revenue. The civil administration of the island was minimal and centered only on those areas in the western coastal plains where Han farmers had settled. Most of the island was occupied by indigenous people. While those living in villages on the plains paid a nominal tax on their hunting lands as a show of acceptance of Qing authority, the mountain dwellers had little direct interchange with the Chinese and were generally hostile to all outsiders who ventured into the their territory. Prohibiting Han settlement in areas beyond the border was intended the minimize conflict and thereby limit the cost that the Qing would have to expend on administering the island. 2

Prescott, Geography of Frontiers, 30.

3

Another rationale for the border was to protect the land rights of the indigenes living on the plains, those who gradually became Sinicized and adopted Chinese language, dress, and life-style. Qing border policy was intended to maintain calm on the frontier where the two groups came into contact by recognizing the customary land rights of the indigenes and preventing Han settlers from taking their land. The indigenes did not conceptualize land ownership in the same way as the Chinese, and they could be tricked into giving over “ownership” of their land for the price of a few trinkets or a weapon. Unless the government had some means to prevent such practices, the indigenes would have become entirely landless, and subsequently a burden for the public coffers. Thus, the border was intended to maintain a status quo between the two populations while assimilation of the indigenous people into Han society could take place. The third reason for the demarcation of a border was to separate the growing population of Chinese settlers engaged in reclaiming land on the frontier, particularly on the fringes of mountainous northern Taiwan, from the un-Sinicized indigenous groups living beyond the extent of Qing authority in the mountains. Land pressure was greatest on the fringes of the forested mountains of northern Taiwan, particularly in modern-day Miaoli, Xinzhu, and Taoyuan Counties. The indigenes in these areas came into frequent contact with the Han only in the nineteenth century when the camphor industry developed. To protect their land from invasion, the indigenous groups would kill their victims and take their heads, an act that had a spiritual as well as practical significance. To reduce the risk of attack, the border was over the years of the nineteenth century developed into a guardline manned by armed men that offered protection to Han and Sinicized indigenes living and working on the frontier, especially in the zones where camphor was harvested. The development of the camphor industry in the mid-nineteenth century dramatically increased the number of attacks by the indigenes, and the guard-line became a method of insuring the stability of the camphor industry. Liu Ming-chuan, during his term of governor of Taiwan, developed the guard-line to become a more effective barrier to attack by the indigenes, especially the Atayal in northern Taiwan, in order to allow camphor exports to continue, and to increase the amount of Taiwanese territory that came under Qing authority. The Divided Island Throughout the eighteenth and nineteenth centuries, Taiwan was roughly divided into two territories based on population demographics and geography. The plains and rolling uplands along the western coast were populated by Chinese settlers and Sinicized indigenous peoples who largely supported themselves by subsistence farming of rice and other crops. The mountainous center of the island and its eastern coast, known as fandi or “savage territory” in Mandarin, was the domain of the un-Sinicized indigenous peoples, speakers of Austronesian languages with Malayo-Polynesian ancestry who engaged in shifting agriculture supplemented by hunting and the gathering of forest products. The two territories were divided by a fragmented border delineated by a series of blockades and guard-lines manned by militias made up of Sinicized indigenes and Chinese settlers from villages in the surrounding districts.

4

Administration of newly settled areas was more expensive for the Qing than prohibiting further development, so an east-west line was drawn, and Han settlement north of this boundary was not permitted.3 The land beyond the border was inhabited by groups of indigenous people, who engaged in swidden farming supplemented by hunting and fishing, They paid a nominal tax on their hunting fields where they killed deer for their meat and for their skins and horns that could be traded for the goods they could not produce themselves. As the Han population increased, the indigenes’ land was occupied by force, “purchased” for trade goods, or rented from the owners by agreeing to pay the tax on the land. The Qing government on the island, however, lacked the necessary troops and resources to prevent Han migration onto indigene lands, and new settlements sprang up in the lawless territory outside these borders. Rebellions against Qing rule, which had little context for settlers who had lived years and decades outside of Qing administrative districts, were commonplace. As settlement extended northward, the east-west border was advanced as well, and people living beyond the advancing lines were removed and resettled away from the border in order to minimize conflict between Han settlers and Sinicized indigenous people.

Figure 1. Approximate location of border in the eighteenth century.

In 1722, Fujian Governor-general Manpao proposed a significant alteration of the border, establishing a new line of control that ran north-south along the foothills of the Central Mountain Range.4 In a memorial to Beijing, he called for the construction of an earthen wall, known in Chinese as tu-niu (literally, clay ox), running alongside a deep trench, to serve as a permanent boundary, adding that all who crossed the border could be considered outlaws. The border was not a continuous wall, but instead existed as disconnected segments in areas such as strategic passes that gave access to mountain areas, and it was later redrawn in 1750 and 1760. The border was not formally established 3 4

Shepard, Island Frontier, 182. Shepherd, Island Frontier, 186.

5

in northern Taiwan until 1781.5 As the Qing policy of quarantining Han colonization of indigenous lands was relaxed and limits on settlement were moved forward, the system of internal boundaries to protect indigenous land rights evolved over the decades and can be seen as a “stopgap” measure intended to be used until Qing control could be extended to the all parts of Taiwan.6 Such borders and physical barriers had numerous historical precedents in China, most notably the construction of the Great Wall in the north, the Willow Palisade in Manchuria, and the wall along the southeastern coast to defend against the forces of the Ming-loyalist Cheng Chen-gung, who ruled Taiwan in the middle of the seventeenth century after defeating the Dutch colonists in Tainan.7 Attractive to Two Empires The administration of Taiwan under the Qing Governor Liu Ming-chuan and the Japanese Governor-general Kodama Gentaro was strongly influenced by the aims of the governments that sent them to act as their representatives there. Thus, a review of the issues that influenced the aims of the Qing and the Japanese is important for understanding the policies that the two governments applied to their rule of Taiwan, specifically those that relate to Taiwan’s internal border. Beginning in the 1870s, Beijing’s suspicion that other nations had ambitions for occupying Taiwan was aroused. In 1874, Japan made clear its interest in the island when it sent a punitive expedition against indigenous people in the southeast portion of the island whom Japan accused of murdering fishermen from Okinawa (an island that Japan would soon occupy but was at the time a kingdom that recognized Qing authority). The Qing denied responsibility for compensating the Japanese for the incident based on their claim that the eastern portion of the island was not under their control. The Qing did eventually offer to pay reparations to the Japanese for the losses they suffered in their expedition. While a practical response at the time, since it absolved the Chinese of responsibility for the actions of the indigenes living in their own lands, it indicated to observers that the Qing only claimed as their sovereign territory that part of Taiwan that lay on the Han side of the border. The Japanese were not the only power with a possible interest in Taiwan. The British and French were also attracted to the island, and the French ultimately occupied Keelung, Tamsui, and the Pescadores (Penghu) in 1884, the factor which ultimately resulted in Liu Ming-chuan to be sent to bolster Taiwan’s defenses and expel the French forces. No doubt recognizing the strategic and commercial value of an “unclaimed territory” so near the Chinese coast, the Qing determined to bring Taiwan more firmly into its empire. This decision was the impetus for efforts to strengthen the island’s military defense, increase the size of the civil administration, and promote the rural economy. The intention was to create an infrastructure on the island that would guide its development toward becoming a province and was influenced by ideas of Self-strengthening, a contemporary view that the adoption of western military methods and technology was necessary to aid 5

Knapp, Frontier Settlement, 58. Shepherd, Island Frontier, 191. 7 Shepherd, Island Frontier, 190. 6

6

China in its defense against foreign influence. Several reform-minded officials were named to head Taiwan’s administration during this period, including Shen Pao-chen in 1874, Ting Jih-chang in 1876, and Liu Ao in 1881. These first two were instrumental in developing Taiwan’s military capabilities and laying the groundwork for later developments, including improved harbors, roads, and even a railway. Liu’s appointment to Taiwan, while ostensibly granted in order to secure the island against an expected invasion by the French, can be seen as a continuation of the Self-strengthening efforts that were supported at the Qing Court by officials such as Li Hung-chang and Tso Tsungtang.8 There was not, however, complete support for the further development of northern Taiwan, particularly the harbor at Keelung, out of fear that such actions would make the area more attractive to foreign powers. In fact, the subsequent development of the coal mines in the Keelung area did play a role in attracting the French to occupy the northern coastal city in 1884. In the latter half of the nineteenth century, Japan was a rising power in Asia. Its government was reformed at the onset of the Meiji period, and over the succeeding decades various forms of modernization were introduced, not least of which to the military. In 1895, when Japan emerged victorious in the first Sino-Japanese War after defeating the Chinese navy at Dalian, it was ceded control of Taiwan by the Qing in the terms of surrender contained in the Treaty of Shimonoseki. The sudden addition of Taiwan to the Japanese Empire was unexpected, and there initially appeared to be no coherent policy for its administration.9 Japan’s army easily occupied the island’s capital, Taipei, which it renamed Taihoku, but found that inhabitants of other parts of the island were less welcoming. It took the army a further eighteen months to put down a rebellion in Tainan, and its forces throughout the island were often attacked by guerrilla forces who could blend into local communities or escape into the mountain when engaged. Japan was eager to begin developing the island’s agriculture production, particularly the rice and sugar industries on the plains and the lucrative camphor industry in the mountains, but first it had to bring stability to the society by eliminating opposition to its rule. In Japan, public discontent with the progress made in the new colony was expressed openly, and there were even calls to accept a French offer to buy Taiwan from the Japanese.10 From 1895 to 1898, three different Japanese officials headed the government, and though plans for Taiwan’s development were proposed, the turmoil throughout the island, particularly in the southwest, meant that little progress was made. The year 1898 was a turning point for the colonial government in Taiwan, for in that year Kodama Gentaro was named governor-general of Taiwan. During his eight year rule of the island, a clear vision of Taiwan’s future role in the Japanese Empire emerged, and projects would be undertaken that would ultimately have a major impact on Taiwan’s future development.

8

Speidel, Administrative and Fiscal Reforms, 441. Kublin, Evolution of Japanese Colonialism, 67. 10 Chang and Myers, Colonial Development, 435. 9

7

The Two Governors A comparison of Liu Ming-chuan and Kodama Gentaro reveals a surprising amount of similarity between the two military leaders, beyond the fact that both were chosen to lead government administrations of Taiwan. Both were in their late forties when posted to Taiwan, had been successful in military campaigns to put down anti-government rebellions early in their careers, and significantly, had been exposed to western military weapons and tactics. Liu had observed Gordon’s Ever Victorious Army, led by Englishman Charles Gordon, in the Suzhou campaign against the Taiping forces in the 1860s, while Kodama introduced German military organization to the Japanese army two decades later.11 Both governors can therefore be seen as reformers promoting the adoption of foreign military methods by their respective central governments, an effort that Kodama, with his access to the general staff, seems to have done this more successfully than Liu. Besides their shared credential as reformers, Liu and Kodama took very different approaches toward their military careers. Kodama remained active in the army, serving on the general staff and later as a major-general in the first Sino-Japanese War, the conflict that resulted in the cession of Taiwan to Japan. 12 Liu, in contrast, had retired to the countryside of Anhui after his military successes. With civil titles, extensive landholdings, and considerable income, Liu chose the relatively easy life of the gentry. Commanding a regional army, Liu remained a military asset for the Qing, and his abilities were not forgotten. In 1880, Beijing requested Liu come out of his retirement and lead Chinese forces in Xinjiang against Russian troops in the Ili River valley. Liu, avoided answering the call by delaying his eventual response until a diplomatic settlement negated the need for his involvement. In 1884, however, upon being summoned to Beijing, he was commanded to take up the role of Governor of Fujian and defend Taiwan from an invasion by the French. A more comprehensive review of the lives of these two generals is beyond the scope of this paper. Instead, the focus will now be on how both men, first Liu and then Kodama, incorporated the border into their policies for developing the island. An effective border was critical to both Liu and Kodama, and both men quickly took steps to secure the frontier zone once their initial military objectives were reached. To explain the development the border during the two administrations, this paper will now address the factors that influenced the border policies adopted by the two leaders, examining the role of the camphor industry, the guard-line system, land reclamation, and the indigenous people. The specific methods of managing the border that Liu and Kodama employed in support of their respective strategic goals can be seen as a result of the different values attached to the above-listed factors by their respective governments and to the extent of financial resources available to them. Finally, it should be noted that while Liu was fully occupied with Taiwan during this administration as governor, Kodama ruled Taiwan with the assistance of the head of the civil government, Goto Shinpei, to whom was given responsibility for many duties associated with day-to-day governance of the colony. In 1901, Kodama was named Minister of War and carried out this duty concurrently with his leadership of the colonial government in Taiwan. 11 12

Chu, Modernization of Taiwan, 37. Chang and Myers, Colonial Development, 435.

8

Consequently, he was frequently away from the island and abrogated much of his authority to Goto. Camphor and the Guard-line Historical records indicate that camphor has been used since at least the sixth century. In Taiwan, the camphor trade was made a monopoly of the Chinese government in the early eighteenth century. In 1855, American traders in Taiwan began buying camphor crystals for pennies on the pound, loading cakes of the dried camphor on their ships, and then sailing to Hong Kong and selling their cargo for a rich profit. The government’s camphor monopoly was abolished in 1868, and in the first six months after the trade was opened, the volume sold increased to 1,018,000 pounds from 41,700 pounds in the preceding six months. By the 1880s, camphor was Taiwan’s most valuable export, and while first used as an aromatic, it was later used in the emerging celluloid film industry, and was also a key component in the production of smokeless gunpowder. 13 In the nineteenth century, the method for obtaining camphor crystals was to distill wood chips from camphor trees, Cinnamomum camphora, by boiling them in water, capture the resulting vapor, allow it to condense into a liquid, and finally to dry into crystals. In Taiwan, camphor trees flourished at elevations of from 500 to 1,500 meters in elevation, in a zone stretching from central to northern Taiwan. Because the trees were so large and the terrain on which they grew so rugged, the distilling stoves were set up next to the felled trees, and the camphor crystals were produced on the spot. The camphor zones were in areas controlled by the un-Sinicized indigenous people, so businessmen in the camphor trade had to negotiate with the indigenes for access to the land and pay a fee per stove. This was necessary due to the indigenes’ custom of killing and taking the heads of anyone who invaded their land. In the nineteenth century, the Atayal group, spread across most of mountainous northern Taiwan, inhabited the lands on which the camphor trees flourished, and as they were known as the fiercest of the un-Sinicized groups and would punish any incursion on their land, the system of mountain use fees was thus in place when Liu arrived in 1884. Under Liu’s administration, the government earned revenue from the camphor trade in a number of ways. These included reinstituting a state-run monopoly, which was canceled due to complaints to the Qing by western governments, levying a tax on the transport of camphor to ports, also terminated at western insistence, and finally imposing a tax on each camphor stove in operation.14 To manage the taxation of the camphor trade, in 1886 Liu established the General Bureau of Camphor Affairs. Despite the mountain fee system, attacks on camphor workers took place with regularity, so it was necessary to maintain Han and Sinicized indigene guardsmen armed with rifles in the camphor zone to protect the laborers at the stoves. These guardsmen were stationed at guard houses along the border in the camphor zones, and they operated under two separate organizations, one composed of Han and funded by 13 14

Durham, Japanese Camphor Monopoly, 797-8. Speidel, Administrative and Fiscal Reforms, 456.

9

the state through the tax on stoves, and the other of Sinicized indigenous people, who were maintained privately by the businessmen involved in the camphor trade. Their duty was mainly to protect the workers at the stoves from attack by the un-Sinicized indigenous people. This required constant vigilance due to the attackers’ adeptness at using the natural cover of the forest from which to make their raids. The mountaindwelling indigenes were armed with spears, long knives, bows, and rifles. While not metal workers themselves, they traded forest products to Sinicized indigenes for modern weapons and ammunition. To stabilize the frontier and guarantee that camphor workers could harvest the crystal unmolested, Liu sought to improve the effectiveness of the guard-line system in much the same way that he carried out his reform of the military. Liu was not interested in fundamentally transforming the camphor industry; his main aim was to stabilize conditions in the camphor zone so that production could continue unhindered.15 To achieve this, he pursued a reform of the border guards that involved unifying the two forces under state control through the Bureau of Pacification and Reclamation (BPR) in order to increase their efficiency. To lead the BPR, Liu named Lin Wei-yuan, from a wealthy landowning family based in Pan-chiao, to head the agency’s main office in Daxi, a river transport town located near the zones where camphor was produced. For the BPR’s office in central Taiwan, Liu chose Lin Chao-tung, the head of another wealthy family with extensive landholdings in Wufeng. These men were responsible for overseeing the training and equipping of the guards assigned to defend the border. Liu’s reform of the guards had merged the ai-ting (Han guards), who numbered less than two thousand, and the tun-ting (Sinicized indigene guards), numbering approximately four thousand, to form the tu-yung (local braves) militia.16 The tu-yung forces were stationed at guardhouses along the border in the camphor zones, and their presence had the effect of reducing the number of attacks on camphor workers sufficiently to enable the number of camphor stoves in the forests to increase and full-scale production to resume. Although Liu centralized the administration of the guard-line, he did not seek to extend the line into areas where camphor was not being produced since such an ambitious program would have exceeded the state’s ability to fund it. The reform of the frontier guards was fundamentally successful in that it stabilized conditions in the camphor zone and guaranteed a steady supply of camphor for the export, and had it not been rolled back by Liu’s successor as governor, Shao Yu-lien, upon Liu’s departure in 1891, it might have continued functioning until 1895 when the Japanese assumed control of the island. Kodama Gentaro assumed the post of governor-general in Taiwan in 1898, and his main preoccupation at the time was the incidence of rebellion in the south of the island that had been ongoing since the Japanese landed on the island in 1895. He personally led many of the assaults on the rebel strongholds. The border played a role in these actions as well. The rebel groups would often escape into the mountains when pursued by the Japanese, so securing the border became a tactic for overcoming the resistance to colonial rule. The border’s development was primarily important, however for ensuring stability in the camphor zones. The revenue from the camphor trade was in important source of revenue, 15 16

Tavares, Japanese Colonial State, 366. Speidel, Administrative and Fiscal Reforms, 449.

10

so efforts to increase production were emphasized. To centralize their control over the trade, the Japanese made camphor a government monopoly in 1898. Violence in the camphor zone was frequent, and in that year alone, over three hundred attacks on camphor workers were reported, with 655 victims being killed or wounded.17 Rules governing the camphor trade were promulgated in 1895, and these specified that all camphor had to be sold to the state, the price paid for camphor oil and crystals would be set by the governor-general, the export of camphor through unauthorized ports would be forbidden, and stiff fines would be levied on anyone who adulterated camphor in order to increase its volume.18 While such prohibitions transformed the camphor trade and ultimately led to Japanese control of the production and sale of camphor, one clause of the law had an immeasurable effect on the land rights of the indigenous people. This stated that all forests and wastelands for which no Qing-era deed existed would be considered state land, and that on this land, no felling of trees nor production of camphor could be undertaken without a permit.19

Figure 2. Approximate location of guard-line in 1901.

In 1899, the Taiwan Camphor Bureau was established with offices towns in northern and central Taiwan. The terms spelled out in the camphor regulations were not applied immediately in the camphor zones, perhaps because the Japanese understood the potentially disruptive influence they could have on the indigenes and Han businessmen involved in the camphor trade. The mountain fee system did continue for a time, but over time these payments were phased out. Many indigenes involved in the camphor trade had sold their stoves to Japanese businessmen, and by 1900 no indigenous people held licenses to produce camphor. In comparison, in 1904, thirty-seven Japanese held licenses to produce camphor, and together they controlled 6,802 stoves, and an annual production

17

Takekoshi, Japanese Rule, 181. Takekoshi, Japanese Rule, 175. 19 Tavares, Japanese Colonial State, 372. 18

11

of over 3.5 million pounds of camphor crystals and 2.8 million pounds of camphor oil, far exceeding world demand.20 While initial improvements of the guard-line were designed to increase is efficiency and effectiveness for keeping the peace in the camphor zones that had been established during Liu’s rule, the guard-line was extended into the areas of Ilan to provide greater security from attack for settlers there. Under Kodama, a comprehensive strategy for defending the border was devised and it called for a radical restructuring of the guard-line, starting with making the border reach from the northern to the southern tip of the island, and secondly, building a system of guardhouses along the entire trail. Each guardhouse was located three to four hundred meters from the next one, and for every four or five guardhouses, there was one branch superintendent’s station, and for every four or five branch superintendent’s station, there was one superintendent’s station. The guardhouses were manned by two Han or Sinicized indigenes, the branch-superintendent’s by a Han or Japanese policeman and several guardsmen, and the superintendent’s station by a police inspector, a doctor, and extra troops. Reclamation and Pacification Liu’s efforts to wrest land from the indigenes followed the same principle that his predecessors Shen and Ting had adopted ten years earlier. Liu relied on the Bureau of Pacification and Reclamation to organize and conduct military operations into mountainous territory controlled by the indigenes. These forces would establish military bases and station armed guards there to farm the reclaimed land. 21 Roads were then built into the mountains to give the military and civilians access to the area. Farmers were encouraged to cultivate the arable land, growing rice in valley bottoms and tea on the hillsides. Forestry was also practiced, with trees being harvested for the camphor industry as well as for their timber. Areas in northern Taiwan that experienced such development include mountainous areas near Sanxia in Taipei Country, Daxi in Taoyuan County, Neiwan and Zhudong in Xinzhu County, and Nanzhuang in Miaoli County. Increased Han settlement in reclaimed areas close to major indigenous communities heightened the need for the maintenance of the guard-line, manned by Han and Sinicized indigenes, to prevent attacks on settlers and laborers. Liu’s main problem in reclaiming land from the indigenes was that they were not interested in giving it up without a fight. The officers of the Bureau of Pacification and Reclamation, aside from managing settlement projects on land wrested from the indigenes, were responsible for pacifying them as well. This was understood to mean bringing Chinese civilization to the indigenes living on the border of the empire’s territory. Similar efforts had been undertaken during the Sung Dynasty when Han settlement of the southern borderlands of Guangxi, Guandong, Yunnan and Sichuan had occurred.22 Under Shen Pao-chen, this had taken the form of encouraging the un-Sinicized indigenes to adopt Chinese dress, language, and customs, medicine and farming practices.23 The BPR under the leadership of Liu and the 20

Takekoshi, Japanese Rule, 177. Gardella, Treaty Ports, 191. 22 Shepherd, Island Frontier, 399. 23 Gardella, Treaty Ports, 185. 21

12

two Lins followed a similar approach, building schools for the indigenes in settlements next to Han communities. In 1887, Liu made an optimistic report to the Qing that tens of thousands of indigenes had been pacified, but the number of attacks and victims belied Liu’s ineffectiveness at managing the indigenes and he was informed repeatedly by officials in Beijing that his claims lacked credibility.24 An apt analogy for expressing the manner in which the indigenes were subdued and then later rebelled would be the notion that they might shave their hair to appear Sinicized, but they could always let it grow when it suited them. No doubt due to the difficulties that the BPR was having in convincing indigenes to adopt Chinese ways, Liu began to rely more heavily on military campaigns to subdue the unSinicized indigenes. Those groups living in the north, the Atayal, Sediq, and Saisiyat were frequently Liu’s targets, but the Amis and other indigenous groups living to the east of the Central Mountain Range were attacked as well. Despite the variety of weapons available to Liu in these campaigns, including older cannons, rockets, and land mines as well as newer technology such as machine guns, the Qing troops were often outmatched by the indigenes. Their warriors, fighting with spears, knives, bow and arrows, and muzzle-loading muskets, employed guerilla techniques against the well-equipped but ineffective Qing soldiers.25 In one battle near Ilan in 1889, Qing forces suffered a terrible defeat, losing 273 officers and troops, including Liu Ming-chuan’s own nephew.26 Some estimate the number of troops killed or wounded during Liu’s more than forty military campaigns against the un-Sinicized indigenes to total more than one-third of the 17,500 Qing soldiers under Liu’s command.27 Despite the high financial and human cost of Liu’s efforts to subdue the indigenes by force, the project can not be judged to have been very successful. Although advancements of the guard-line were undertaken, and new land was opened up to Han settlement that would increase the state’s tax revenue, very often the settlers cultivating these lands would be forced to retreat and abandon their fields when the indigenous people launched reclamation projects of their own. In common with Liu, Kodama ordered that a comprehensive survey of Taiwan be undertaken, with teams of surveyors mapping the coastline, the plains, and the mountains, giving the colonial government a firmer understanding of the nature of the island that they governed. In 1904, a 1:20,000 map was produced from the surveying data that gave the most detailed view of Taiwan’s landform that had been created. The map did not, however, include the island’s entire territory. Mountainous areas inhabited by the unsubdued indigenous groups living the north, east, central, and southeast portions of the island were left blank on the map, the surveying teams being unable to transport their scientific instruments into areas beyond the guard-line. Unlike Liu, who carried out policies of land reclamation of indigene territory to achieve political, social, and economic aims within the context of the Qing Empire, Kodama approached the problem from the perspective of a colonial ruler that aimed to exploit the island’s wealth without regard for traditional practices or economic models. The Japanese were not bound to customary arrangements about the use of mountain lands. If no Qing-era deed were 24

Speidel, Administrative and Fiscal Reforms, 458. Gardella, Treaty Ports, 191. 26 Speidel, Administrative and Fiscal Reforms, 458. 27 Gardella, Treaty Ports, 191. 25

13

available for land that one lived, farmed, or hunted on, it became the property of the Japanese government under one of the first laws that the government of Japan adopted after taking possession of Taiwan in 1895. Perhaps understanding that the sudden imposition of the new laws would cause disruption on the frontier, particularly for the non-Sinicized indigenes who collected substantial mountain fees from the camphor businessmen, these were phased in over several years.28 State ownership of forest land gave the government the power to lease or sell tracts of mountain land to Japanese corporations for camphor, timber, or even mining. The maintenance of the guard-line was essential to plans for the economic development of mountain areas, as workers could not safely venture into the forests unless they had armed guards protecting them. When the Japanese assumed control over the frontier zone in 1895, remnants of the guardsmen that Lin Wei-yuan had organized for Liu Ming-chuan were still in place, operating as a private force under Liu’s command, and securing the camphor workers from attack in areas near Nantou in central Taiwan and Sanxia in northern Taiwan. The Japanese began to pay these troops themselves, and they were later merged with governmental units of guardsmen on the border. The guard-line was extended in Xinzhu and Ilan Counties in 1897, and in 1899 and 1900, it was lengthened to stretch from Ilan, through mountainous parts of Taipei, Taoyuan, Xinzhu, Miaoli, and Taichung counties to Nantou in central Taiwan.29 In these early years of Kodama’s term in office, the guard-line was defensive in nature, being used to secure the frontier, not push forward into the indigenes’ territory. The rebellions that the Japanese were facing in southern and central Taiwan meant that actively reclaiming mountain land was not an immediate priority. In 1902, with major rebellions in the south suppressed, greater attention was paid to the development of the mountains. In the early years of the Japanese colonial period, the camphor economy on the frontier was integrated into a system of property ownership that was intended to encourage Japanese investment in the colony. The camphor permits that had been granted during the Qing era were not renewed, and instead Japanese business interests gradually established themselves as the major producers of camphor from 1897. The Japanese firms holding camphor permits began to ignore the payment of mountain fees to the indigenous leaders controlling the land where the camphor trees grew, and this led to an increase in hostilities along the guard-line, requiring additional troops to bring order. By 1899, when the camphor monopoly was established, the traditional camphor industry was well on its way toward transformation. This process involved numerous conflicts between the indigenous people who were displaced by the new system, and attacks on camphor stoves and workers caused production to be seriously reduced. Japanese troops were required to secure the camphor zones in order to reduce tensions enough for the trade to resume operation at comparable levels. In July 1902, an uprising known as the Nanzhuang Incident occurred in a camphor-producing district of Miaoli County, and it would have a significant impact on the later development of land reclamation and control of the indigenes in the years to follow. In this incident, Ri Ah-gui, a leader of a group of Saisiyat indigenes, orchestrated attacks on border guard stations and camphor stoves in protest 28 29

Tavares, Japanese Colonial State, 372. Bureau of Aboriginal Affairs, Report, 14.

14

over their exclusion from the transformed camphor trade.30 The Japanese response to the attacks was to pursue Ri and his followers into the mountains, attack their villages with artillery, use infantry forces to wipe out resistance and establish forward bases that were incorporated into the guard-line defenses. In November 1902, having been on the run since July, the last of Ri’s followers came down out of the mountains to submit to the Japanese, but they underestimated the reception that they received. First welcomed with drinks and an apparent lack of animosity, the indigenes let their guard down and expected that the Japanese intended to let bygones be bygones. Then, the Japanese turned their guns on the surrendering rebels and killed all that they could. Ri reportedly escaped the ambush but died a few months later, according to a police report from the time.31 Under Kodama, the Bureau of Aboriginal Affairs employed dual approaches to the control of the un-subdued indigenous groups. Against those groups of indigenes who were un-subdued but relatively restful, primarily groups in the south of the island, conciliatory measures were taken.32 With trade and communication restricted by the guard-line, the Japanese could win concessions from the indigenes by granting limited trade in goods such as salt and cloth. Toward the northern groups, the Japanese took an aggressive stance and used violent means to subdue indigenes who attacked the guards on the border. These included using artillery pieces to bombard the indigenes’ villages, but also to arm indigenous groups and incite them to attack groups that refused to submit. Between 1902 and 1906, more than seventy military campaigns were launched against the indigenes by the Japanese under the aim of advancing the border. The intention was gradually reduce the territory in which the indigenes had freedom of movement, with the long-term aim of bringing the entire island under Japanese control. Interchange between the indigenes and the Japanese was carried on by the policemen who were posted at the sub-branch superintendent and superintendent stations along the guard-line. These officials had responsibility for educating the indigenes on Japanese language and customs. In this respect, the Japanese efforts at pacification resembled those that Liu employed via the BPR, but in a much more thorough manner, owing to the greater resources that were made available for the project. In many cases, the Japanese were able to win the indigenes trust and convince them to move down from the mountains and settle in villages established to house them on the plains. The children living in villages of resettled indigenes were required to attend school and learn Japanese, and by 1907, the year after Kodama left Taiwan, there were thirteen such institutions, mainly for the subdued groups in the south.33 Such efforts had the effect of reducing the frequency of attacks by indigenes in the border districts but not eliminating such incidents altogether. That attacks continued was one of the main reasons for the use of increasingly violent measures taken by the Japanese in the years that followed Kodama’s departure. His successor, Governor-General Sakuma, showed much less patience for the slow progress in gaining domination over the indigenes that conciliatory methods produced. During his term of administration, from 1906 to 1915, major military campaigns were prosecuted against the unsubdued indigenes, especially the Amis, Taroko, Bunnun, Sediq, and Atayal groups. 30

Tavares, Japanese Colonial State, 361. Tavares, Japanese Colonial Policy, 380. 32 Bureau of Aboriginal Affairs, Report, 4. 33 Seiji Hishida, Japan’s First Colony, 277. 31

15

Evaluation Liu Ming-chuan’s reputation as a reformer is widely accepted, as is the notion that he did not achieve all that he set out to do. His policies for the guard-line can be seen in the same light. While he generally succeeded in stabilizing the camphor industry, his reform of the guard-line was not sufficiently institutionalized to outlast his administration, and under Shao Yu-lien the frontier reverted to the state that it had been in prior to Liu’s arrival. The indigenes collected mountain fees for access to the camphor zones, but raids on workers and stoves continued to take place. Further, Liu enjoyed little success at pacifying the un-Sinicized indigenes, let alone assimilating them, and was compelled to utilize military force to achieve his land reclamation aims in the face of fierce resistance. Liu’s efforts to “open the mountains, and subdue the savages” met with limited success and can be understand in the context of his other achievements in Taiwan. Liu was first of all a military man, with a character to match, and this may have contributed to the problems he encountered as Taiwan’s governor. He sometimes followed his own counsel too stubbornly and did not always take into account the possible consequences of his decisions. Therefore, his reach exceeded his grasp in that he was not able to achieve all of the goals that he set. These factors go some way toward explaining Liu’s eventual resignation from his office of governor and departure from Taiwan. In addition to the difficulties that his tax reform, land reclamation, and indigene pacification efforts experienced, his work to develop Taiwan’s coal mining industry and construct a trans-island railway encountered similar obstacles.34 His ideas were not always accepted by his superiors in Beijing, and he may have lacked the political skill required to convince others of the value of his tactics, even when they agreed on his basic strategy. Liu sailed away from Taiwan in 1891, leaving in a ship from Keelung harbor, and never returned. He returned to his estate in Anhui, and it was there that he died in 1894, just a year before Taiwan was ceded to the Japanese. Under Kodama, the guard-line along the border was thoroughly reformed, but this process was undertaken in stages according to need. While the anti-Japanese rebellion was occurring in the plains, the guard-line’s importance was due to its ability to block a means of escape by the rebel forces. Once the rebellion was largely put down, the guardline’s role was focused more directly on the need to maintain a defensive perimeter in the camphor zones. During this stage, the border was extended to entirely surround the unSinicized indigenes, from northern Taiwan all the way to the southern tip of island. This extension did not, however, give the Japanese adequate control of the frontier along the border. The mountain dwelling indigenes were still able to attack across the border when they chose and retreat to their own territory after their engagements. In 1902, the guardline’s defensive role was modified by the Japanese, and the strategy of advancing the guard-line was promoted. Troops and weapons were used to push forward into the mountains and subdue the indigenous inhabitants of the area. During this stage, the guard-line was no longer a simple perimeter boundary but was instead a network of guard-lines that took the form of a web, giving the Japanese a presence throughout the 34

Chu, Modernization of Taiwan, 40-42.

16

frontier regions and enabling government forces to exert more effective control. By the time of Kodama’s departure from Taiwan, the process of subduing the indigenous people was not complete. This can be explained by noting factors such as the rugged mountainous terrain, the fierce opposition of the indigenous people and the effectiveness of their guerrilla combat tactics, and the reliance in the early years of the Japanese occupation on conciliatory means of dealing with the indigenous people. The Japanese may have underestimated the need for the application of military force, but under Kodama, it became apparent that negotiation with the indigenous people would not result in a resolution that the Japanese would find acceptable. Kodama’s successor as governorgeneral, Sakuma, largely abandoned peaceful efforts to win over the mountain-dwelling indigenes and instead initiated large-scale military campaigns to destroy opposition. This policy did not bring immediate resolution, and attacks on the indigenes were carried out until the 1930s, when military aircraft dropped explosive and chemical weapons on indigenous villages following the Wushe Uprising in present-day Nantou County. Legacy of the Border The legacy of the border and the guard-line has had a significant impact on the sociopolitical development of Taiwan. Many of the administrative borders that exist in contemporary Taiwan, particularly those delineating the extent of townships that encompass the territory historically inhabited by the indigenous people, were established along the lines of the border and guard-lines that were developed during the Qing era, especially during the term of Liu’s administration, and the Japanese colonial era. Comparison of topographic maps from the Japanese period with those in use today reveal that the same mountain ridges, passes, and river crossings along which the guard-line formerly ran now serve as borders between administrative regions. It is possible to access these areas today and find relics of the guard-line’s infrastructure, including the remains of roads, survey markers, and even guard houses. In Guishan Village of Wulai Township, County, a guardhouse from the Japanese era, built of concrete and stone and having loopholes through which the guards could train their weapons on indigenes approaching the outpost, perches on a bluff overlooking the Nanshi River. It is overgrown with vines and bushes, but its location alongside an abandoned trail connecting Wulai, a township predominantly occupied by indigenous Atayal people, with Taipei City, just twenty kilometers to the north, testifies to its strategic importance. In Xinzhu County, along the borders between Neiwan, Jianshi, Hengshan, and Wufong Townships, telephone poles from the Japanese era with wires attached to their ceramic insulators, still stand along what was once the path connecting guard houses along the guard-line.35 In northern Taiwan, townships that encompass areas that were once located beyond the border and within indigenous territory include Wulai in Taipei County, Fuxing in Taoyuan County, and Jianshi and Wufong in Xinzhu County. Areas that lay outside of the border and were inhabited by Hoklo and Hakka speakers, the descendents of the Han settlers and the Sinicized indigenous groups, include the townships of Xindian, Sanxia, Daxi, Guanxi, Neiwan, Hengshan, and Zhudong. Travelers driving into the mountains and entering townships occupied by majority indigenous populations are in many places still required 35

In late 2008, the author explored the areas mentioned and documented the existence of these relics with photographs.

17

to stop and apply for mountain permits that allow passage beyond the border. Police officers at these remote stations are often members of indigenous groups themselves, and they occupy a position within the administration of the border and the guard-line that is closely linked to the historic guard-line system. The state still finds cause to regulate access into areas inhabited by indigenous people, and although these limitations have been considerably relaxed in the last decade, now existing as very nearly a formality, the linkage between the current practice and that implemented under previous governments, is undeniable. Conclusion To national leaders whose closest look at the extent of their territory comes from a map, a border is simply a demarcation of a political boundary. Its construction and maintenance are a matter for officials concerned with practical matters. Borders gain in importance when they are challenged and must be defended. In such cases, borders become military fronts, with forces stationed in the frontier zone to prevent invasion. Liu Ming-chuan was both a civil official and a military leader, and his development of Taiwan’s “savage border” was intended to assist him in achieving his strategic goals. This paper has examined the factors, political, economic, social, cultural, and military, that influenced Liu’s border policy, noting specifically the importance of camphor, the un-Sinicized indigenes, and land reclamation for settlement of Han farmers. Kodama Gentaro, and the three governor-generals who preceded him, inherited the border and guard-line that Liu Ming-chuan fashioned during his term in office. In the eight years that Kodama governed Taiwan for the Japanese, the border underwent major change, being first extended the length of the island and then advanced into the indigenes’ territory. While the motives for Kodama’s border policy were similar to those during Liu’s term, Kodama devoted more resources to the border’s development and created an infrastructure that allowed the Japanese to eventually overcome the indigenes’ resistance. This outcome did not occur quickly or easily. The violence which the Japanese relied on to subdue the indigenes was increased dramatically under Kodama’s successor Sakuma, reflecting the Japanese impatience with the intransigent indigenes. Both Liu and Kodama were acting in the interest of their respective empires, and the indigenes were obstacles to be uprooted and displaced. The guard-line was therefore a manifestation of political power on an imperial frontier. While the border is today a relic, one no longer indicated on contemporary maps of Taiwan, its existence for so many decades left an indelible mark on Taiwan society. When one crosses the historic but unseen border and enters the territory that formerly belonged to the mountain-dwelling indigenes, the change is palpable and apparent in the landscape, farming practices, language, religious worship, and attitude of the people. The conditions that once made the “savage border” and the guard-line practical no longer exist, but the border does have a potential purpose in contemporary Taiwan, that of education. From the troubled history of the border, we can trace the factors that contributed to Taiwan’s spatial, social, and cultural development.

18

Bibliography Barclay, Paul D. 2007. Contending Centres of Calculation in Colonial Taiwan: The Rhetorics of Vindicationism and Privation in Japan’s Aboriginal Policy. Humanities Research, Vol. XIV. No. 1: 67-84. Government of Formosa. Bureau of Aboriginal Affairs. 1911. Report on the Control of the Aborigines in Formosa. Taipei. Chu, Samuel C. 1963. Liu Ming-ch’uan and Modernization of Taiwan. The Journal for Asian Studies, Vol. 23, No. 1: 37-53. Chang, Han-yu and Myers, Ramon H. 1963. Japanese Colonial Development Policy in Taiwan, 1895-1906: A Case of Bureaucratic Entrepreneurship. The Journal of Asian Studies, Vol. 22, No. 4: 433-449. Davidson, James M. 1972 (1903). The Island of Formosa: Past and Present. Taipei: SMC Publishing. De Bunsen, E.H. 1927. Formosa. The Geographical Journal, Vol 70, No. 3: 266-285. Durham, Jr., Walter A. 1932. The Japanese Camphor Monopoly: Its History and Relation to the Future of Japan. Pacific Affairs, Vol. 5, No. 9: 797-801. Eskildsen, Robert. 2005. Taiwan: A Periphery in Search of a Narrative. Journal of Asian Studies, Vol. 64, No. 2 (May 2005): 281-294. Gardella, Robert. 1999. From Treaty Ports to Provincial Status, 1860-1894, in Taiwan: A Nnew History, edited by Rubenstein, Murray A. New York: M.E. Sharpe, Inc. Hishida, Seiji. 1907. Formosa: Japan’s First Colony. Political Science Quarterly, Vol. 22, No. 2: 267-281. Knapp, Ronald G. 1976. Chinese Frontier Settlement in Taiwan. Annals of the Association of American Geographers, Vol. 66, No. 1: 43-59. Kublin, Hyman. 1959. The Evolution of Japanese Colonialism. Comparative Studies in Society and History, Vol. 2, No. 1: 67-84. Lamley, Harry J. 1999. Taiwan Under Japanese Rule, 1895-1945: The Viscitudes of Colonialism, in Taiwan: A New History, edited by Rubenstein, Murray A. New York: M.E. Sharpe, Inc. Prescott, J.R.V. 1965. The Geography of Frontiers and Boundaries. Chicago: Aldine Publishing Company.

19

Shepard, John R. 1995. Statecraft and political economy on the Taiwan Frontier 16001800. Stanford University Press. Taipei: SMC Publishing. Shepard, John R. 1999. The Island Frontier of the Ch’ing, 1664-1780, in Taiwan: A new history, edited by Rubenstein, Murray A. New York. M.E. Sharpe, Inc. Speidel, William M. 1976. The Administrative and Fiscal Reforms of Liu Ming-ch’uan in Taiwan, 1884-1891: Foundation for Self-strengthening. The Journal for Asian Studies, Vol. 35, No. 3: 441-459. Takekoshi, Yosaburo. 1978 (1907). Japanese Rule in Formosa, translated by George Braithwaite, George. Taipei: SMC Publishing. Tavares, Antonio C. 2005. The Japanese Colonial State and the Dissolution of the Late Imperial Frontier Economy in Taiwan, 1886-1909. The Journal of Asian Studies, Vol. 64, No. 2: 361-385. Teng, Emma. 1999. Taiwan as a Living Museum: Tropes of Anachronism in LateImperial Chinese Travel Writing. Harvard Journal of Asiatic Studies, Vol. 59, No. 2: 445484.

20