¹ The so-called ‘aboriginal inhabitants’ here are of Austronesian or Malayo-Polynesian descent. In terms of languages used, indigenous tribes in Taiwan are of Austronesian origin. Austronesians are widely distributed over a large part of the world—east from Easter Island, west as far as Madagascar on the eastern coast of Africa, south from New Zealand and Papua New Guinea, with Taiwan the northernmost part. The most ancient Austronesian site in Taiwan are the ruins at Dapenkeng in Bali District of New Taipei City. The Dapenkeng culture dates back to the Neolithic Age between 4,000 and 3,000 bce. See Peter Bellwood, ‘The Austronesian Dispersal and the Origin of Languages’, Scientific American 1991, pp. 88–93; Liu, Yi-chang, Cunzai de weizhi—Taichung diqu de kaogu yizhi yu shiqian wuhua (The Existing Unknown—Archaeological Sites and Prehistoric Culture of Taichung Area), Cultural Center of Taichung County, Taichung, 1999Google Scholar.

² Hung, Li-wan, Shufan shehui wangluo yu jiti yishi: Taiwan zhongbu Pingpu zuqun lishi bianqian, 1700–1900 (Relationship between Social Network and Ethnic Identity of Shufan—Historical Transformation of Plains Aborigines in Central Taiwan under Qing Rule, 1700–1900), Lian-jing Books, Taipei, 2009, pp. 41–89Google Scholar. For a discussion of the long-term relationship between the Han and the aboriginal people, see Brown, Melissa J., Is Taiwan Chinese? The Impact of Culture, Power and Migration on Changing Identities, University of California Press, Berkeley, Los Angeles and London, 2004, pp. 43–50Google Scholar.

³ The officially recognized Indigenous Peoples of Taiwan today were shengfan tribes during the Qing era. Most of the shufan or plains aborigines had lost their tribal status and were still lobbying for recognition. In this article, the terms ‘aborigines’ and ‘aboriginal peoples’ are used to denote the native inhabitants of Taiwan.

⁴ Yi-gui Zhu of Luohanmen (present-day Neimen Township in Kaohsiung County) led a rebellion against the Qing government.

⁵ Lan, Ding-yuan, Dong zheng ji (Expedition to the East). Taiwan Wenxian Congkan, Vol. 12, Tai-wan yinhang, Taipei, 1957 (first published in 1721), p. 40Google Scholar; Huang, Shu-jing, Taihai shicha lu (Record of a Tour of Duty in the Taiwan Strait). Taiwan Wenxian Congkan, Vol. 4, Tai-wan yinhang, Taipei, 1957 (first published in 1724), pp. 167–168Google Scholar.

⁶ Prohibited access to aboriginal territory was neither successful nor effective, as evidenced by the constant immigrant-shengfan conflicts and frequent redrawing of fanjie. See Hong, An-quan (ed.), Qinggong gong zhong dang zouzhe Taiwan shiliao (Historical Materials of Taiwan in Qing Court Memorial Archive), Vols 1–12, The National Palace Museum, Taipei, 2001Google Scholar.

⁷ During the reign of Qianlong, the aboriginal boundary was redrawn many times. See Shi, Tian-fu, ‘Hongxian yu lanxian: Qing Qianlong zhongye tawain fanjie tu’ (Red Line and Blue Line: Aboriginal Boundary Maps of Qing Taiwan during Qianlong Reign), Newsletter of Taiwan Field Research 19, 1990, pp. 46–50Google Scholar; Shi, Tian-fu, Qingdai tai-wan de diyu shehui (Regional Society of Taiwan in Qing: Research on the Historical Geography of Zhuqian Area), Cultural Affairs Bureau, Hsinchu County, Xinzhu, 2001, pp. 65–116, 229–232Google Scholar; Lin, Yu-ju and Wei, Dong, ‘Lin Shang-wen shijian qian de Taiwan bianqu tuxiang: yi Qianlong sishijiu nian Taiwan zixian tu wu zhongxin’ (Image of Taiwan Frontier before Lin Shuangwen Incident by Taiwan Purple-Line Aboriginal Boundary Map of 1784), Taiwan Historical Research 19:3, 2012, pp. 47–94Google Scholar.

⁸ Hung, Shufan shehui wangluo yu jiti yishi, p. 76.

⁹ The Mudanshe Incident of 1874 triggered Japan into dispatching a punitive expedition to avenge shipwrecked soldiers who had been slaughtered/headhunted by Mudanshe inhabitants. For the removal of the aboriginal boundary, see Da-chun Luo, ‘Shang zhifu qing jingli taiwan houshan fandi’ (Governor, Please Manage the Aboriginal Territory in Taiwan Mountain), 19 April Tongzhi 6, Taiwan History Digital Library (THDL), file name: ntu-1130713-0006500068.txt.

¹⁰ Inō, Kanori and Awano, Dennojou, Taiwan fanren shiqing (Taiwan Aboriginal Affairs), Taiwan Governor-General's Office, Taipei, 2000 (first published in 1900)Google Scholar; Mori, Ushinosuk, Taiwan fanzu zhi (Gazetteer of Taiwan's Savage Tribes), Southern Materials Centre Publishing Inc., Taipei, 2008 (first published in 1917)Google Scholar. The academically informed model of human diversity in survey anthropology did not enter Japanese colonial discourse uncontested, as seen in the relationship between Kanori Inō and the government anthropologist in Barclay, Taiwan. Paul D., ‘Contending Centres of Calculation in Colonial Taiwan: The Rhetorics of Vindicationism and Privation in Japan's “Aboriginal Policy”’, Humanities Research 14:1, 2007, pp. 67–79Google Scholar, has made a thorough analysis.

¹¹ Liao, Shou-chen, Taiya zu de shehui zuzhi (Social Structure of the Atayal), Tzu Chi College of Medicine and Humanities, Hualien, 1998Google Scholar; Pei-wen Liao, ‘Taiwan yuanzhu min lieshou xisu yu chuanshuo’ (The Headhunting Custom and Legend of Taiwanese Aboriginals), Master's thesis, Department of History, National Chung Hsing University, Taichung, 2010; Guo, Ming-zheng, You jian zhenxiang: saideke yu wushe shijian (Another Truth: Seediq and the Wushe Incident), Yuan-Liou Publishing Co. Ltd., Taipei, 2012Google Scholar; Yamada, Hitoshi, 首狩の宗教民族学 (Research of Religious Ethnology on Headhunting), Zhumo shufang, Tokyo, 2015, pp. 207–377Google Scholar; Yen-hsiang Hsiao, ‘Jiaoyi, lie, yu shehui zhixu, 1874–1898’ (Trade, Head-hunting and Social Order in Borderland Taiwan: Study of Zhiwu She at Upper Wu River, 1874–18), Master's thesis, Department of History, National Chi Nan University, Nantou, 2015, pp. 45–77.

¹² Simon, Scott, ‘Politics and Headhunting among the Formosan Sejiq: Ethnohistorical Perspectives’, Oceania 82, 2012, pp. 164–185CrossRefGoogle Scholar.

¹³ Kojima, Yoshimichi, Fanzu xiguan diaocha baogao shu (Report on the Customs of Savage Tribes. Vol. 1, Atayal Tribe), Huang, Zhi-hui (ed.), Institute of Ethnology, Academia Sinica, Taipei, 1996 (first published in 1915), p. 24Google Scholar; Sayama, Yukichi, Fanzu diaocha baoguan shu—di si ce (Aboriginal Research Report—Part IV), Liu, Bi-zhen (ed.), Institute of Ethnology, Academia Sinica, Taipei, 2011 (first published in 1917), pp. 37–42Google Scholar.

¹⁴ The informants interviewed were mainly descendants of the Atayal tribe with branches, including the Seediq and Tstole who used to live in the Puli basin and its peripheral areas, a key region for the pacification of aborigines from late Qing rule until the early Japanese colonial era. Some came from Wugonglun and Fangliao where headhunting was prevalent and folk religions against such practice were popular. Prior to the foreign colonization of the island, the tribes had no written record of their history. Hence, the oral histories related by the informants became their collective memories. The informants who were interviewed were chosen after actual field visits.

¹⁵ Yang, Mu, Wu Feng, Hong Fan Publisher, Taipei, 1979, p. 1Google Scholar; Ye, Shi-tao, Yu Shou (Headhunting), Paise wenhua chuban she, Kaohsiung, 1991, p. 1Google Scholar; Pu, Zhong-cheng, Bei Yiwang de Shengyu: Yuanzhu Min Shenhua, Lishi, yu Wenxue de Zhuisu (Forgotten Sanctuary: A Retrospect of Aboriginal Mythology, History and Literature), Wunan Book Inc., Taipei, 2007Google Scholar; Sun, Da-chuan (ed.), Taiwan Yuanzhu Minzu Hanyu Wenxue Xuanji (Selected Chinese Literary Work of Indigenous Peoples of Taiwan—Poetry), Inknet, Taipei, 2003Google Scholar.

¹⁶ According to Hitoshi Yamada, the Japanese had a tradition called ‘shoushou’ which means cutting off the victim's head to execute them or as a military exploit or sacrificial offering to the gods of war. This tradition became obsolete in 1874. See Yamada, 首狩の宗教民族学, pp. 11–22. However, the nature and meanings of Japanese headhunting were different from those of aboriginal peoples of Austronesian or Malayo-Polynesian descent.

¹⁷ Torii, Ryuzo, Tanxian Taiwan: Torii Ryuzo de Taiwan Renlei Xue Zhi Lu (Explore Taiwan: Torii Ryuzo's Anthropological Tour in Taiwan), Yang, Nan-jun (trans.), Yuan-Liou Publishing Co. Ltd., Taipei, 1996, p. 206Google Scholar.

¹⁸ Wan, Bima'Daxi Wula, Taiwan de yuanzhu min: bunong zu (Taiwan Aborigines: Bunun), Taiyuan wuhua, Taipei, 2003, p. 57Google Scholar.

¹⁹ Yamada, 首狩の宗教民族学, pp. 238–242; Shepherd, John R., Statecraft and Political Economy on the Taiwan Frontier, 1600–1800, Stanford University Press, Stanford, California, 1993, pp. 54, 57, 59, 81Google Scholar; Campbell, William, Formosa Under the Dutch, Described from Contemporary Records, Kegan Paul, London, 1903, p. 318Google Scholar.

²⁰ Apart from oral history on the headhunting custom of plains aborigines, the Kaxabu Museum in Shouchangfan (present-day Puli Town) also has on display bags used by warriors on headhunting expeditions.

²¹ In 1903, Taiwan was home to 2,852,461 Han Chinese and 102,979 aborigines. The aboriginal tribes were distributed in northern, central, and southern Taiwan. Among them, the largest tribe was Wangzi fan with a population of 28,073, which accounted for 27.3 per cent of the shengfan population. See Mori, Taiwan fanzu zhi, p. 2; Taiwan zongdufu zongdu guanfang wenshuke (ed.), Taiwan zongdufu di qitongji shu, 1903 (Taiwan Sotokufu Seventh Statistical Yearbook, 1903), Taiwan zongdufu zongdu guanfang wenshuke, Taipei, 1905, pp. 229–230, 261–272.

²² The Atayal firmly believed that after death, they would drop to an abyss where a fierce sinister monster hid in a deep pool. Above the pool was a beautiful hongu utux or spirit bridge made up of rainbow colours. The ‘rainbow bridge’ was guarded by a deity who would strictly assess whether in life the deceased had contributed to the tribe and whether they abided by the gaya (the tribe's moral and ethical standards). Only those who passed the appraisal could cross the bridge to join their ancestral spirits. See Wu, Hui-lian, ‘You ma—ni hai hui jixu zhibu ma’ (You ma—Will You Continue Weaving?), Xin Guxiang 1, 1999, p. 40Google Scholar.

²³ Chen, Shu-mei, ‘Xiaoshi zhong de jingmian wenhua’ (Vanishing Tattooed Culture), in Yu Lu Gong Wu—Taiwan yuanzhu min wenhua san (Dancing with Deer—Taiwan's Indigenous Culture, III), Taiwan Panorama, Taipei, 1997, p. 172Google Scholar.

²⁴ Huang, Taihai shicha lu, p. 167; Sayama, Yukichi, Fanzu diaocha baoguan shu—di wu ce taiya zu qianpian (Investigation Report on Savage Tribes. Vol. 5, Atayal Tribe, Part I), Lu, Xin-chun (ed.), Institute of Ethnology, Academia Sinica, Taipei, 2012 (first published in 1918), pp. 120–122Google Scholar; Sayama, Yukichi, Fanzu diaocha baoguan shu—di wu ce taiya zu qianpian (Investigation Report on Savage Tribes. Vol. 6, Bunun Tribe, Part I), Yang, Shu-yuan (ed.), Institute of Ethnology, Academia Sinica, Taipei, 2008 (first published in 1919), pp. 95–98Google Scholar; Torii, Tanxian Taiwan, p. 206.

²⁵ Simon, ‘Politics and Headhunting’, pp. 164–185.

²⁶ Kiyoto, Furuno, Taiwan yuanzhu min de jiyi shenghuo (Ritual Life of Aborigines in Taiwan), Ye, Wan-qi (trans.), Yuan min wenhua, Taipei, 2000 (first published in 1945), p. 17Google Scholar.

²⁷ Interview with Pihu Obing, A-dou Li, Huzhu Village, Renai Township, Nantou County, 14 May 2015.

²⁸ Mori, Taiwan fanzu zhi, p. 315.

²⁹ Ibid. Fix, Douglas L., Kanjian shijiu shiji Taiwa (Curious Investigations: 19th-century American and European Impressions of Taiwan), Lo, Charlotte (comp.), Ru guo chuban she, Taipei, 2006, p. 38Google Scholar.

³⁰ Torii, Tanxian Taiwan, p. 207.

³¹ Wei, Hui-lin, Yu, Jin-quan and Lin, Heng-li (eds), Taiwan sheng tongzhi (Annals of Taiwan Province), Vol. 8, Historical Research Committee of Taiwan Province, Taipei, 1952, p. 87Google Scholar.

³² In the 1900s, Ushinosuke Mori conducted an anthropology survey with Torii Ryuzo in Taiwan. He concluded that headhunting was a superstitious act rather than a form of resistance against the ruling regime. Nevertheless, the Wushe Incident, led by Mona Rudao, was indeed a revolt against long-term suppression by the Japanese colonial authorities. See Mori, Ushinosuke, Shengfan Xingjiao: Mori Ushinosuke de Taiwan tanxian (Exploring Aborigines—Mori Ushinosuke's Adventures in Taiwan), Yang, Nan-jun (trans.), Yuan liu, Taipei, 2000, pp. 589–590Google Scholar. For a thorough analysis of the Wushe Incident of 1930, see Ching, Leo T. S., Becoming ‘Japanese’: Colonial Taiwan and the Politics of Identity Formation, University of California Press, Berkeley and Los Angeles, California and London, England, 1962, pp. 133–151Google Scholar.

³³ Ushinosuke Mori, a Japanese-era anthropologist; Shou-chen Liao, a post-war Truku; and Ming-zheng Guo, a retired Seediq school teacher, presented different perspectives on headhunting in different time periods (in the nineteenth and twentieth centuries). See Mori, Taiwan fanzu zhi pp. 325, 329; Liao, Taiya zu de shehui zuzhi, pp. 219–221; Guo, You jian zhenxiang, pp. 48–49, 59, 65–68.

³⁴ Apart from the head, the weapons and personal belongings of the hunted were also brought back to the tribe as spoils.

³⁵ Liao, Taiya zu de shehui zuzhi, p. 222.

³⁶ Mori, p. 325.

³⁷ Joseph Beal Steere explained that the special bags were for carrying the hunted head of the victim. See Steere, Joseph Beal, Formosa and its Inhabitants, Jen-kuei, Li Paul (ed.), Institute of Taiwan History, Preparatory Office, Academia Sinica, Taipei, 2002 (first published in 1873–1874), pp. 52–53Google Scholar.

³⁸ Neither the Atayal nor the Seediq had toumu. Instead, the Atayal had pasapung, which in their language means the arbiter or policymaker, while the Seediq had qbsuran (elder sibling), which in their language denotes the leader or decision-maker. It was pointed out by one of the reviewers that ‘qbsuran’ means elder sibling. The fact that ‘qbsuran’ was used by the tribe indicates that there were no other expressions for ‘leader’ in their language.

³⁹ Sayama, Fanzu diaocha baoguan shu, p. 37.

⁴⁰ Yamada, 首狩の宗教民族学, p. 312, highlights the close relationship between collective animal hunting and headhunting.

⁴¹ Guo, You jian zhenxiang, pp. 81–85.

⁴² Walis Iyung (Ren-he Jiang), Nanfeng Village, Renai Township, Nantou County, interviewed on 12 June 2015. Walis Awi (Fu-sho Bao), Huzhu Village, Renai Township, Nantou County, interviewed on 14 May 2015. Both mentioned that the taboos of lbuwy halung would also vary according to the individual character of the hunters.

⁴³ Interview with Walis Awi.

⁴⁴ Mori, Shengfan Xingjiao, pp. 589–590. One of the reviewers pointed out that in Simon's article he mentions that elders used headhunting to sanction or even shame youth. Needless to say, there were different incentives and disincentives to engage in headhunting, and these varied within social transformations. For example, some tribes, although not many, did collaborate with outsiders, including Han Chinese and the Japanese. There are different ways to interpret gaya, which merits further study. See Simon, ‘Politics and Headhunting’, pp. 168–178.

⁴⁵ Hsiao, ‘Jiaoyi, lie, yu shehui zhixu, 1874–1898’, pp. 54–69.

⁴⁶ Huang, Taihai shicha lu, p. 167.

⁴⁷ Jiang, Yu-ying, Taiwan fuzhi (Taiwan Prefecture Gazetteers), Xiamen University Press, Xiamen, 1895 (first published in 1685), p. 61Google Scholar, described aborigines as bloodthirsty, merciless murderers who displayed the hunted heads at home; and the greater the number of heads hunted and displayed, the mightier the hunter was.

⁴⁸ Shuishalian liao she, made up of six tribes, including Mei she, Pu she, and Shuishalian she, which comprised four tribes, all collected rent kangwuzu from new settlers who used their land for cultivation. See Hung, Li-wan, ‘Qingdai taiwan bianqu shehui zhixu zhi kaocha: yi zhuoshui xi and wu xi zhongyiu “kang wu zu” wei zhongxin’ (Reconsidering Social Order in Border Region of Qing Taiwan: Study of ‘Kang Wu Zu’ at Midsteam Zhuo Shui and Wu Rivers), Taiwan Historical Research 20:4, 2013, pp. 29–40Google Scholar.

⁴⁹ These headhunting tribes of Wangzi fan feared that renting and selling land to new settlers would invite the immigration of alien ethnic groups, which would jeopardize tribal peace and prosperity. The exchange of private land among fellow tribesmen was allowed only upon approval from the tribal community. Trespassing into the land of fellow tribesmen, although prohibited, was not deemed punitive and encroaching upon the territory of enemy or alien tribes was encouraged and considered a legitimate way to claim new living space. Headhunting was a means to settle disputes arising from the fight for land. Hence, the importance of land to headhunting tribes would determine how they would deal with immigrants infringing on their living space. See ‘Puli she difang zhimin di diaocha hengshan jishi wai er ming’ (Survey Report of Colonial Government on Puli Area), in Taiwan zongdufu dangan (Taiwan Sotokufu Archives), Vol. 302-2-12-18, 9 April 1898, pp. 18–19; Kojima, Fanzu xiguan diaocha baogao shu, pp. 201–204.

⁵⁰ Feng, Yong (ed.), Liu Ming-chuan fu tai qian hou danan (Archives Before and After Liu Ming-chuan as Governor of Taiwan), Taiwan lishi wenxian congkan, 1997Google Scholar, reprinted from Taiwan wenxian congkan, Vol. 276, 1969, pp. 14–15.

⁵¹ See ‘Puli she difang zhimin di diaocha hengshan jishi wai er ming’, pp. 18–19.

⁵² See Li-wan Hung, ‘Jianan pingyuan yanshan diqu zhi zuqin guanxi: yi “alishan fanzu” wei li’ (Relationships between Ethnic Groups in the Hillside Areas of Jia Nan Plain, 1700–1900: Study of ‘Alishan Fan Zu’), Taiwan Historical Research 18:1, 2011, pp. 64–82.

⁵³ Ibid., pp. 82–83. On the emergence of the Feng Wu cult, see Zheng, Ying-yi, ‘Tongshi zhidu, xinyang yu yanshan bianqu shehui— Taiwan Feng Wu xinyang de xingcheng’ (The Tongshi (interpreter) System, Cults and Frontier Society: The Formation of the Cult of Feng Wu in Qing Taiwan), Journal of History and Anthropology 12:2, 2014, pp. 51–84Google Scholar.

⁵⁴ Leo T. S. Ching analysed the Japanese eulogized versions of Feng Wu's death as told in ‘The Story of Feng Wu’ and ‘The Bell of Sayon’ and pointed out that between 1910–1930, the colonial administrators had the intention of turning the aborigines into loyal subjects who would fight for the imperial empire. However, Ching did not explore economic factors behind the ethnic conflicts. See Ching, Becoming ‘Japanese’, pp. 151–160; Ching, Leo T. S., ‘Savage Construction and Civility Making: Japanese Colonialism and Taiwanese Aboriginal Representation’, Position: East Asia Culture Critique 8:3, 2000, pp. 795–881CrossRefGoogle Scholar.

⁵⁵ See Hung, ‘Jianan pingyuan yanshan diqu zhi zuqin guanxi’, pp. 65–67, 82–92.

⁵⁶ This was pointed out by one of the reviewers.

⁵⁷ Simon, ‘Politics and Headhunting’, p. 175.

⁵⁸ Xu, Jin-sheng (ed.), Xu tai yu shi xi zu pu (Lineage Genealogy of Xu Tai-yue), Jisi gongye Xu tai yuguanli weiyuan hui, Taipei, 1983, p. 20Google Scholar. For the occupation history of Xu Ming-shi, see hui, Taiwan shiliao jicheng weiyuan (ed.), Taiwan zongdufu dangan chaolu qiyue wenshu— di er ji (Taiwan Sotokufu Archives—Copies of Contract Documents—Series II), Ministry of Culture and Yuan Liu Publishing Co. Ltd., Taipei, 2007, Vol. 34, p. 29Google Scholar. It contains a record of Ming-shi Xu paying a Han Chinese, En-ze Lin, money for land use, indicating that Lin was an ‘early bird’ who had established his foothold in the aboriginal territory before Xu's arrival.

⁵⁹ Xu (ed.), Xu tai yu shi xi zu pu, pp. 27–28.

⁶⁰ See Zhao, Shen-zhen et al. , ‘Dao Xian Tong Guang Si Chao Zoyi Xuanji (Selected Memorials of Four Reigns from Daoguang to Guangxu)’, Taiwan wenxian congkan, Vol. 288, 1984, pp. 88–89Google Scholar; Inō, Kanori, Taiwan wenhua zhi (Gazetteer of Taiwan's Culture), Southern Materials Centre Publishing Inc., Taipei, 1994 (first published in 1965), p. 525Google Scholar.

⁶¹ See Zhao et al., ‘Dao Xian Tong Guang Si Chao Zoyi Xuanji’, pp. 88–89.

⁶² Feng (ed.), Liu Ming-chuan fu tai qian hou danan, p. 101.

⁶³ Qing court statistics of 1877 show the presence of 6,000 shufan immigrants in the Puli basin, while Japanese colonial government statistics of 1902 reveal that out of a total population of 8,264, 4,659 (56.4 per cent) were Han Chinese, 3,589 (43.4 per cent) were shufan, and 28 (0.3 per cent) were shengfan. There is an obvious discrepancy in the number of shufan settlers. By 1905, their population rose to 11,377, indicating an exponential increase in immigrants settling in the Puli basin. See Feng (ed.), Liu Ming-chuan fu tai qian hou danan, p. 14; bu, Linshi Taiwan hukou diaocha (ed.), Mingzhi san shi ba nian rishi shiqi Taiwan hukou diaocha jijiyuan biao( shang) defang zhi bu (Raw Data of Household Survey in Colonial Taiwan of 1905, Part I), Wensheng shuyuan, Tokyo, 2000Google Scholar; ‘Puli paichusuo puli she bao ji wucheng bao jie Zhuang ming diaocha’ (Survey on Street and Village Names Conducted by Police Station of Puli Community), in Taiwan zongdufu gongwen leizuan 4246.52, 1902.

⁶⁴ Feng (ed.), Liu Ming-chuan fu tai qian hou danan, p. 101.

⁶⁵ The Japanese colonial government continued the Qing practice of issuing permits for land use. Although the inhabitants might have occupied or used the land for years without the official permit, the land would have been considered state-owned. To gain access to the land and forests, the Japanese built roads, again infringing on the living space of shengfan, thus stirring their resentment and resistance. See ‘Puli she difang zhimin di diaocha hengshan jishi wai er ming’, pp. 84–85; Taiwan Church News, Vol. 255, June 1906; Taiwan Church News, Vol. 256, July 1906.

⁶⁶ Hong (ed.), Qinggong gong zhong dang zouzhe Taiwan shiliao, Vols 1–12.

⁶⁷ He, Meng-jiao, Lin, Zheng-hui and Wu, Jun-ying (eds), Wufeng lin jia wenshu ji—dongjun deng xiang guan xinhan (Selections of the Wufeng Lin's Archives in Qing Taiwan: the Correspondence of the Tung Army), Academia Historica, Taipei, 2014, pp. 57, 69–79Google Scholar.

⁶⁸ Taiwan Church News, Vol. 184, July 1900.

⁶⁹ Ching, Becoming ‘Japanese’, pp. 133–173.

⁷⁰ Campbell, William, Fuermosha sumiao—gan wei lin mushi Taiwan riji (Sketches from Formosa—Diaries of Rev. Dr. William Campbell), Xu, Ya-qi (trans.), Avanguard, Taipei, 2005 (first published in 1915), p. 51Google Scholar. According to Steere, Formosa and its Inhabitants, p. x, he and Campbell conducted the survey in shengfan territory in 1873–1874; hence, their observations are of occurrences that took place in the late nineteenth century.

⁷¹ These shufan were the Sizhuang fan who had exchanged land with the original inhabitants of Puli basin, the Pushe fan. They were headhunted by the Wangzi fan. See Steere, Formosa and its Inhabitants, p. 58; Hung, ‘Qingdai taiwan bianqu shehui zhixu zhi kaocha’, pp. 24–26.

⁷² Taiwan Church News, Vol. 218, May 1903, recorded that in February, the Japanese confiscated all the guns of Pushe fan. Knowing that Pushe fan were without guns, the headhunting tribes launched attacks every day, killing more than ten Pushe fan within a week. The latter pleaded with the government to return their guns to them and the Japanese took pity on them and gave them back their weapons.

⁷³ Interview with Kaxabu senior Hui-zhen Wei, 9 October 2015.

⁷⁴ State intervention in eighteenth-century Qing Taiwan was different from that which took place in twentieth-century Japanese colonial Taiwan. For example, Brown states that ‘The Qing regime, like the Dutch before them, exploited ethnic tensions in Taiwan to their advantage.’ She also points out that ‘Japanese rule focused on making Taiwan a model colony, both to economically support Japan's then-growing imperialist ambitions and to demonstrate to Western nation-states that Japan was as capable an imperialist power as they were.’ See Brown, Is Taiwan Chinese?, pp. 43, 53. In my opinion, Qing rule prevented interaction among ethnic groups through boundary demarcation, while the Japanese established military communities for defence against shengfan and to control transactions between tribes as well as their economic life. While the Qing empire failed to institutionalize border trade or control weapon deals and the salt trade, the Japanese used business exchanges as a means to control tribal life and thus force them to comply with their governance. In addition, the Japanese wanted to seize control of mountain resources. See Shepherd, Statecraft and Political Economy, pp. 105–380; Brown, Is Taiwan Chinese?; Tavares, Antonio C., ‘The Japanese Colonial State and the Dissolution of the Late Imperial Frontier Economy in Taiwan,1886–1909’, The Journal of Asian Studies 64:2, May 2005, pp. 361–385CrossRefGoogle Scholar; Barclay, ‘Contending Centres of Calculation’, pp. 79–82; Yen-hsiang Hsiao and Li-wan Hung, ‘Jiaoyi yu fufa: yi pulili she zuowei biabqu zhili zhongxin, 1874–1898’ (Trade Exchange and Pacifying Aborigines: Study on Puli she as Strategic Center for Borderland Management, 1874–1898), in Proceedings of Yilan yanjiu di shi yi jie xueshu yantao hui— yilan Atayal (11th Academic Conference of Research on Ilan—Atayal in Ilan), Part III, pp. 532–539.

⁷⁵ Feng (ed.), Liu Ming-chuan fu tai qian hou danan, pp. 14–15.

⁷⁶ See ‘Puli she difang zhimin di diaocha hengshan jishi wai er ming’, pp. 18–19; Barclay, ‘Contending Centres of Calculation’, p. 78.

⁷⁷ At that time, the Japanese colonial government had launched an attack on the Alang she in the north of Puli basin. Fear spread and the chieftains of four she (Man, Wu, Touza, and Duluguin) to the east of Puli took hundreds of tribesmen with them to seek assistance from Gantaman fan and Du fan in the south. These two southern tribes pretended to enter into alliance with the four eastern she, then ambushed them on their way back, killing 122 tribesmen. See Taiwan Church News, Vol. 225, 1903.

⁷⁸ Ibid.

⁷⁹ These headhunting tribes were the Shuishalian tribe and Alishan fan, as abovementioned, as well as the six tribes of Neiyou, including the present-day Thao, Tsou, Kanakanavu, Hla'alua, Bunun, and Rukai in southern Taiwan. See Li-wan Hung, ‘Qingdai nanzixuan xi and laonong xi zhongyou zhi shengfan and shufan zuqin guanxi, 1760–1888: yi “fufanzu” wei zhongxin’ (Ethnic Relationships between the Plain Tribes and the Mountain Peoples in the Nan-tzu and Lao-nung Valleys during the Ch'ing Period, 1760–1888: A Study of ‘Fu Fan Tsu’), Taiwan Historical Research 14:3, 2007, pp. 1–71.

⁸⁰ Research studies on land occupation through the establishment of ai include Xue-ming Wu, Jinguangfu kenai yu xinzhu dongnan shanqu de kaifa (Establishment of Defense Fort Ai and Development of Southeastern Suburb of Hsinchu,1834–1895), Institute of History, National Taiwan Normal University, Taipei, 1986; Wu, Xue-ming, ‘Qingdai yige wushi tuoken jiazu de yanjiu: yi xinzhu Jiang Chaofeng jiazu wei li’ (A Practical Land-Reclaiming Family in Ch'ing Taiwan: The Case Study of the Chiang Family of Hsin-chu), Taiwan Historical Research 22, 1995, pp. 5–52Google Scholar; Shi, Tian-fu, ‘Qingdai Taiwan beibu neishan de diyu shehui ji qi diyu hua: yi Miaoli neishan de jilong xi liuyu wei li’ (Regional Societies of Inner Mountains in Northern Qing Taiwan: Case Study of Jilong River Valley in Miaoli), Taiwan Historica 56:3, 2005, pp. 181–242Google Scholar; Zeng, Sian-wei and Hung, Li-wan, ‘Qing Qianlong nianjian Wulixue xi and Xiulang xi shangyou zuqin guanxi’ (Study of Land Relations and Shan shui yin around Wulixue River and Upstream of Xiulang River during Qing Qianlong Era), Taiwan Historical Research 22:2, 2015, pp. 111–150Google Scholar.

⁸¹ See Wen-rong Lin, ‘Jiekai “chankui zushi” zhi mi’ (Unveiling the Mystery of ‘Chankui Zushi’), in Lin, Wen-rong (ed.), Taiwanshiji congbian, Part I, Guozhang chuban she, Taizhong, 1987, pp. 1–15Google Scholar; Wang, Zhi-yu, ‘Zhushan diqu de gongmiao’ (Local Temples in Zhushan Area: Xyuantian Shangdi and Chankui Zush), Feng Chia Journal of Humanities and Social Sciences 4, May 2002, pp. 183–210Google Scholar; Xie, Chong-guang, ‘Tongzhi xing, chengchuan, bianqianyu yanjiu quxiang—min yue taiminjian xinyang yanjiuyinlun’ (Homogeneity, Inheritance, Transition and Research Orientation—An Introduction to the Research of Folk Belief in Fujian, Guangdong and Taiwan), Journal of Fuijan Normal University (Philosophy and Social Sciences) 4, 2011, pp. 29–35Google Scholar; Zhang, Zhixiang, ‘Min yue zhishu suo jian chankui zushi simiao yu xinyang tankao’ (Discussing the Temple and Belief of Master Can Kui on Min Yue Gazetteers), Feng Chia Journal of Humanities and Social Sciences 18, 2009, pp. 119–148Google Scholar.

⁸² Findings from fieldwork, 14 October 2014. gong, Xindian rixing (ed.), Jiawu nian nongmin li (Farmers’ Calendar of Jiawu Nian), Rixing gong, Taipei, 2014Google Scholar; gong, Runji (ed.), Nongmin li (Farmers’ Calendar), Runji gong, Taipei, 2014Google Scholar. Inscriptions at the Taiping Temple in Xindian (Xindian taiping gong yange beiwu), year unknown.

⁸³ Whoever touched the bamboo fences under the spell would find themselves stuck to the fence, unable to escape. Interview with Kaxabu elder, Ying-yu Pan, Winter 2011 and Spring 2012. Interview with Zhongyuan senior, Walis Awi (Fu-shou Bao), 10 April 2015.

⁸⁴ Interview with Kaxabu elder, Hui-zhen Wei, 9 October 2015.

⁸⁵ ‘Mingzhi sanshi nian qi yue zhong puli she fujen shu shiwu baogao’ (Report from Land Reclamation Office of Puli she, mid-July 1897), in Taiwan zongdufu gongwen leizuan (Taiwan Sotokufu Documents), Vol. 19, 13 September 1897, pp. 158–159. Moreover, Taiwan ri ri xin bao, the daily newspaper, also reported frequent headhunting incidents in Puli she. Taiwan ri ri xin bao, 2 November 1898, p. 4; Taiwan ri ri xin bao, 14 October 1899, p. 4, reported inhabitants who were taking a nap at a local shrine and collecting firewood in the forest being headhunted.

⁸⁶ Taiwan Church News, Vol. 218, August 1903.

⁸⁷ Ibid.

⁸⁸ Campbell, Fuermosha sumiao, p. 99.

⁸⁹ Karankawas, a native American tribe from Corpus Christi, is said to have the practice of consuming the flesh of their dead enemies, not for food, but to acquire their power and courage. However, more evidence is needed to prove that such a practice existed. See Murphy Givens and Jim Moloney, Corpus Christi: A History, Nueces Press, Corpus Christi, Texas, p. 2.

⁹⁰ Campbell, Fuermosha sumiao, p. 101.

⁹¹ Chuan Hu was a Qing official responsible for military and defence, and travelled widely in Taiwan. His book, first published in 1893, recorded the aboriginal customs he witnessed in the Puli basin. See Hu, Chuan, ‘Tai wan ri ji yu bing qi’ (Diaries and Reports on Taiwan), Taiwan wenxian congkan, Vol. 71, 1960, pp. 31Google Scholar.

⁹² More details on such interaction can be found in my other articles. See Hsiao and Hung, pp. 519, 543; Li-wan Hung, ‘Cong zixi shouchu yang guanxi kan bianqu zuqun jiaoliu xian xiang, 1834–1946’ (Adoption Relations and Ethnic Interaction: A Study of Danan Village in Qing Taiwan Borderland, 1834–1946), in Li-wan Hung (ed.), Archaeology, History and Indigenous Peoples: New Perspective on Ethnic Relationships of Taiwan, Shung Ye Museum of Formosan Aborigines and Institute of Ethnology, Academia Sinica, Taipei, pp. 303–363; Li-wan Hung, ‘Taiwan bianqu zuqun jiaoliu tuxiang: ti wan wu xi liuyu danan juluo shufan hunyin guanxi wei li, 1834–1946’ (Ethnic Interaction along Wan Wu River of Taiwan Borderland: Study of Marriage Network of Shufan in Danan Village, 1834–1946), Tunghai University Journal of History, 3 September 2017, pp. 109–174.

⁹³ Kuilei fan are the present-day Paiwan with some Rukai. They were found in south Taiwan, from the Nanzixian River to the Hengchun Peninsula. Hence, geographically speaking, Kuilei fan were often considered the present-day Rukai. The tribe got their name ‘Kuilei’, which means marionettes used in puppet shows, from their colourful tribal costume, as explained by one of the reviewers.

⁹⁴ Hung, ‘Qingdai nanzixuan xi and laonong xi zhongyou zhi shengfan and shufan zuqin guanxi, 1760–1888’, pp. 43–58.

⁹⁵ Interview with elders Xiu-ying Pan, Qing-wen Pan, and Yong-kun Yang, 6 March 2015, 22 April 2015, 1 May 2015, and 23 June 2015.

⁹⁶ Wei, Yu and Lin (eds), Taiwan sheng tongzhi, pp. 86–90.