The Territorial Dispute and Japanese Intervention

Historians have carried out extensive studies related to the territorial conflict over the north bank of the Tumen River between Qing China, Korea, and Japan, especially the Qing-Japanese treaty negotiations that took place between 1907 and 1909.Footnote ² Although much has been said about historical origins, conflicts, and negotiations of the dispute, few have paid sufficient attention to the arguments of a Japanese bureaucrat who played a critical role in the Qing-Japanese negotiation. This man was Shinoda Jisaku, an expert in international law who, acting as the director of general affairs in a local colonial institution, was in charge of an investigation of the disputed borderland. From 1909 to the 1930s, Shinoda published extensively on the territorial issue on the Sino-Korean border. Over this period, he developed and repeated his basic judgment of the territory: that the land north of the Tumen and Yalu rivers was a “neutralized zone” and a “no man's land” during most of the Qing dynasty. This definition remains influential in contemporary academia. Since the 1960s, numerous Korean scholars have published studies on the Sino-Korean border issue. Many of them adopted the theory of “no man's land” as a standard description.Footnote ³

Before analyzing Shinoda's discourse in detail, we need to review the historical background of the Qing-Chosŏn territorial dispute and the Japanese intervention. Starting in the 1870s, tens of thousands of Korean peasants risked their lives crossing the Tumen (Tuman in Korean) River, fleeing natural disasters and extreme poverty in their homeland in Hamgyŏng Province and seeking refuge in southern Manchuria and the Primorsky region of Russia.Footnote ⁴ As refugee communities began to emerge on the northern side of the river, conflicts arose between the Qing and the Chosŏn in relation to two interrelated problems. The first concerned the question of which country had the right to govern the Korean refugees. The Qing, in order to recruit human laborers to secure the Jilin frontier against Russian infiltration, proposed to register these Korean squatters as Qing subjects. The Chosŏn, however, was against this policy for fear of losing control of its population. Instead, it urged the Qing to repatriate these trespassers.

Soon another challenge occurred before the two governments could settle the first problem. This time, the issue was raised not by the states, but by the Korean refugees occupying the north bank of the border river. Refusing to be repatriated to their home country, nor to be registered as Qing subjects, they asserted that according to a demarcation stele installed by a Qing official in 1712, the “Division River” was not the Tuman but a river north of it. For hundreds of years, they said, the land between the “Division” and the Tuman rivers had been an “unused wildland” (K. hanhwang chiji; C. xianhuang zhidi; 閒荒之地) on which no one was permitted to live. Tring to arouse the Qing's sympathy, they employed a typical zongfan (tributary) rhetoric and Confucian doctrine that “all land under Heaven are king's land” (C. putian zhixia mofei wangtu 普天之下莫非王土) (Hansǒng T'onggambu Reference Tʾonggambu1907). Since the Chosŏn was a subordinate state of the Qing, the refugees begged the Qing Empire's mercy to allow them to cultivate and settle on this wildland. Their theory, that there was a “true” border other than the Tuman, was later adopted by the Korean officials and king, who argued that the lands north of the Tuman River should be Korean territory and those poor peasants did nothing wrong. Was the Tuman/Tumen the boundary? Which river was the Tumen? Which stream constituted its source? Debate surrounding such questions led to territorial disputes that were to last for decades.Footnote ⁵

From 1885 to 1887, the Qing and the Chosŏn conducted two joint demarcations in order to clarify the “old boundary” between them. Both governments dispatched local officials to not only carefully observe the terrain, but also search for evidence from various ancient records and documents to prove their territorial claims. After two years of repeated bargaining, the joint projects ended in a stalemate and failed to reach a final resolution. Meanwhile, more Koreans were attracted to cross the river and resettle on the north side of the Tumen, which they called “Kando.”

It is beyond this article's scope to investigate the Qing-Korean demarcations in detail. But we need to understand that such an effort was not made in the spirit of international law but in accordance with traditional zongfan rhetoric. Sure, international law was not exactly a new thing in East Asia in the 1880s, at least not for the Qing (Tian Reference Tao2001). Yet neither the Qing nor the Chosŏn attempted to frame their boundary dispute in international law terms. Rather than seeing the bilateral relations as one between two equal “nations,” both countries viewed it in the imperial hierarchical order, in which the Qing was the superior empire and Korea its most loyal tributary kingdom. The main task for demarcation, therefore, was not to negotiate a new division line between them, but to prove the historical legitimacy of an old border, the border determined on the order of the Qing emperor in 1712. Both sides heavily employed the zongfan discourses and principles to argue for their best interests. As Andre Schmid (Reference Schmid2002, 209) suggests, the demarcation was “a fascinating example of how spatial issues could be contested within the tributary relationship, without explicit reference to Western discourses on space.” As a matter of fact, it was not until the end of the Sino-Japanese War (1894–95) and the “independence” of Korea that a few Korean local officials started to reframe the territorial conflict in line with European international law (Kim No-gyu Reference No-gyu1903).

The Qing-Korean debates began to take place as Manchuria was being opened up for development, spurring fierce competition among all of the surrounding countries, each vying for a share of the enormous economic and geopolitical benefits to be gained from control of its virgin lands (Lattimore Reference Lattimore1932). The first Sino-Japanese War (1894–95) and the Russo-Japanese War (1904–5) led to a radical reconfiguration of regional power dynamics. Defeating China and Russia, Japan gained full control of the Korean peninsula. Following the model of European colonialism, it turned Korea into a “protectorate” in 1905 and immediately moved to expand its influence in Manchuria. The China-Korea territorial dispute provided Japan with a convenient excuse for political, economic, and military interventions. As East Asia was drawn into the global capitalist and imperialist system, the resource-rich Tumen River area, a remote frontier that had been maintained in a largely wild and undisturbed state for hundreds of years, was suddenly transformed into a geopolitical hotspot, a contested space in which regional, even global, ambitions began to be played out (Inoue Reference Manabu1973; Tsurushima Reference Setsurei2000).

Japan's intervention in the northern Tumen River region (called “Kantō” by the Japanese) initially served its grand strategy of competing against Russia, which had invaded Manchuria in 1900 and threatened Japan's dominance of the Korean peninsula. Even prior to the Russo-Japanese War, Japanese strategists and pan-Asianist activists had already noted the crucial economic and geopolitical value of the region lying to the north of the Tumen River. Several investigation teams were dispatched to gather local information (Li Reference Huazi2015). After expelling the Russians from Korea and southern Manchuria, Japan took advantage of the Sino-Korean dispute to encroach upon the Kantō area, which was now home to over 100,000 Korean immigrants.

In 1907, in the name of “protecting the Koreans from being bullied by local bandits and Qing officials,” the Japanese Resident-General in Korea (Tōkanfu) established the Temporary Kantō Branch of the Tōkanfu (hereafter referred to as the Kantō Branch) as a local colonial administrative institution (Ch'oe Reference Chang-gǔn1996). Japanese military forces intruded to the north bank of the Tumen River and set up the Kantō Branch in Longjing (now Longjing city in Yanbian Prefect, Jilin Province, China). Japan announced that the sovereignty of Kantō was “undecided,” and that since Korea was now a Japanese “protectorate,” Japan had the right to govern local Koreans in the disputed border area (Shinoda Reference Jisaku1984). The Chinese, of course, strongly rejected this claim. The conflict continued for two years until a series of bilateral treaties were signed in October 1909, in which Japan forced a fragile Chinese government to yield mining and railway privileges in Manchuria in exchange for Japanese recognition of the Sino-Korean Tumen boundary.Footnote ⁶

Shinoda Jisaku's “No Man's Land”

A key figure in the Kantō Branch, Shinoda Jisaku is mostly remembered today as a high-ranking colonial bureaucrat in Korea during the period of Japanese rule (1910–45). In the 1920s and 1930s, he served as the Minister of the Korean Royal Household (J. Riōshiki; K. Iwangchik), supervising the everyday lives of members of the (former) Chosŏn royal family. While in this position, he presided over the compilation of the last two Veritable Records of The Annals of the Chosŏn Dynasty (Chosŏnwangjo-sillok): that of Kojong (Kojong Sillok) and Sunjong (Sunjong Sillok). Contemporary Korean scholars generally criticize these two volumes as being heavily tarnished by Japanese censorship and distortion (Chang Reference Yong-suk2014; Kang Jae-eun Reference Jae-eun and Lee2005, 218). From 1940 to 1944, Shinoda was appointed as the ninth principal of Keijō Imperial University, the predecessor of today's Seoul National University and one of only two imperial universities established in the Japanese colonies.Footnote ⁷ During this period, he was awarded the Grand Cordon of the Order of the Rising Sun for his service in Korea (Arima, Matsubara, and Takano Reference Arima, Takatoshi and Shinji2008). When all is said and done, Shinoda Jisaku's career as a colonial official began with the Tumen border negotiation.

When plans to establish the Kantō Branch were being formulated, Shinoda, then an acting lawyer in Tokyo, was among the first to be recruited by the director of the Branch, Saitō Tsuejiro (1867–1921). Shinoda's main duty was to study the territorial dispute and prepare arguments in favor of Japan, providing evidence from both historical and legal perspectives. As Shinoda was to later recall, the Tōkanfu had determined from the very outset that the northern bank of the Tumen River, adjacent to Manchuria, Korea, Russia, and the Sea of Japan, should be developed as a Northeast Asian transportation corridor. Railways should be built to connect the Korean harbor at Ch’ŏngjin with Kantō to facilitate the movement of Japanese commodities and Manchurian resources (Shinoda Reference Jisaku1999, 477, 478). The most convenient way in which to realize this plan was to prove that Kantō belonged to Korea (and hence to Japan). However, the Japanese government made no definitive statement that Kantō was Korean territory (a difficult argument to make, as many Japanese investigators pointed out at the time). As a jurist, Shinoda needed to find a way of justifying Japan's claim with respect to this territory.

The lawyer was picked for this job because of his excellent background in both legal and military services. Born to a peasant family in Shizuoka Prefecture, Shinoda graduated from Tokyo Imperial University in 1899, majoring in law. A member of the newly established Japanese Association of International Law, he was a close friend of Akiyama Masanosuke (1866–1937), a pioneering legal scholar in Japan. Following Akiyama's encouragement and recommendation, he joined the Imperial Army during the Russo-Japanese War as a jurist (Shinoda Reference Jisaku and Hensankai1941, 403, 404). For Japan, fighting a war against a major world power such as Russia served as proof that Japanese civilization and the Japanese race could not be considered inferior to Western civilization and the white race. Since all European colonial powers were advocating international law as the “universal norm,” Japan also needed it to fashion itself as a “civilized” nation—more civilized than Russia. Allocating at least one international law expert to each of its armies, then, was a crucial strategy in ensuring Japan victory not only on the battlefield, but also in the context of internationally recognized moral and legal standards.

Shinoda served in the Third Army, commanded by General Nogi Maresuke (1849–1912). Most of the battles in which it engaged took place on the Liaodong Peninsula in Manchuria. Involving by far the greatest degree of mechanized warfare, these were arguably the most horrific conflicts in the whole war, with both Russia and Japan suffering massive human casualties. Even General Nogi himself lost two sons, in the Battle of Nanshan and the Siege of Port Arthur. When the Russian troops in Port Arthur finally surrendered in January 1905, Shinoda took part in designing the legal procedure of capitulation. Negotiations mainly drew upon precedents from the Franco-Prussian War, as well as the Laws and Customs of War on Land created at the first Peace Conference at The Hague in 1899. To a lesser degree, according to Shinoda, reference was also made to the ceremony that took place in 1636 when the Chosŏn dynasty surrendered to invading Manchu forces (Shinoda Reference Jisaku1930). The whole handover process was conducted smoothly and peacefully, which Shinoda later boasted about as “establishing a new example for international war in the Far East” (Shinoda Reference Jisaku and Kinenkai1934, 360). What Shinoda emphasized, however, was the way in which this example marked an extraordinary Japanese contribution to the laws of war: “foreign observers at the time, such as military attachés and journalists etc., highly praised the capitulation, to the extent that it established a model for the future” (Shinoda Reference Jisaku and Kinenkai1934, 360, 363).

Indeed, Japan wanted to ensure that every step of its expansion looked “civilized.” International law provided the appropriate trappings, though often at the cost of those who were deemed to be “uncivilized.” Designating Korea as a “protectorate,” for example, showed Japan's mastery of the colonial discourse embraced by the “international community.” It publicly denied Korea's autonomy and deprived the Korean people of their own state (Dudden Reference Dudden2005). In the case of the Kantō dispute, Japan portrayed local Koreans as the victims of a humanitarian crisis, so that “protecting Koreans” became its noble excuse for encroaching on the territorial sovereignty of another state. However, underneath the surface, what Japan was really worried about was the potential for Kantō, a region adjoining the Japanese-controlled peninsula with only a loosely controlled border, and already home to the world's largest overseas Korean community, to become a hotbed for anti-Japanese activism. Independent activists and Righteous Army (K. Ŭibyŏng) militias, referred to by Japan as “indocile Koreans” (J. futei senjin), moved back and forth across the Tumen River, attacking Japanese garrisons and gaining support among local populations, causing major headaches for the Japanese colonial rulers. Therefore, aside from its economic benefits, gaining and maintaining control of Kantō was a key objective for the Japanese in their attempt to eliminate Korean resistance (Shinoda Reference Jisaku1984, Reference Jisaku1999).

Shinoda's strategy of defining Kantō as a “no man's land” developed over years. In 1910, he wrote and published a paper in The Journal of International Law (Kokusaih ō zasshi). In order to demonstrate the “fair” and “scientific” basis for his argument, he denounced Qing and Korean territorial claims, asserting that Kantō was an unused wildland with unclear sovereignty (Shinoda Reference Jisaku1910). Later, in an essay titled “Retrospective on the Kantō Issue,” he wrote: “regarding the Kantō Issue, the conclusion of my research is that: ‘Kantō belongs neither to the Qing nor to Korea. It is a naturally formed, uninhabited, neutralized zone’” (Shinoda Reference Jisaku1999, 485). But this statement needs to be read in context. Shinoda's “Kantō” encompassed a much larger region than the area for which the term was normally used, including not only the lands north of the Tumen River, but also those north of the Yalu River. That is to say, his “Kantō” constituted a huge area including almost all of southern Manchuria. This move deliberately confused the north bank of the Tumen, which was under dispute, with the north bank of the Yalu, which was not. “Since the land north of the Yalu was recognized as Qing territory,” he argued, “it was fair to regard the land north of the Tumen as Korean territory” (Shinoda Reference Jisaku1999, 486).

What is more, although Shinoda was not the first to try to neutralize the territory under dispute, he did it in a new way. Recall that the early Korean squatters also argued that the land in the Tuman north bank was “unused wildland.” Certain Japanese researchers before him had already viewed Kantō as a “neutralized area” or an “area beyond civilization” (JACAR, n.d.; Morita Reference Toshitō1906). However, Shinoda's definition did not simply echo these theories but had a much subtler implication when taken within the framework of international law. Shinoda explained that the “neutrality” of Kantō was a result of the Manchu-Chosŏn agreement reached in the early seventeenth century following the two Manchu invasions, in which the two regimes swore to maintain their original division. But, the original text of the agreement did not indicate where the border was. Taking advantage of this ambiguity, Shinoda cited Description de la Chine (generally known as The General History of China in the English-speaking world), an influential encyclopedic record of the history, geography, and society of the Qing published in 1735 by French Sinologist Jean-Baptiste Du Halde (1674–1743). Based on one document in this book (discussed below), Shinoda determined that Korean territory extended to the north of the Tumen and Yalu rivers and that a vast area of southern Manchuria was a “no man's land”:

Further, in his 1938 monograph, The demarcation stele of Mount Paektu (Hakutōsan teikaihi), Shinoda clarified and consolidated this definition, devoting an entire chapter to the process of “the formation of a no man's land” (mujinchitai no seiritsu) (Shinoda Reference Jisaku1938, 20–31). According to him, since the Qing government had severely limited civilian access to the borderland area, such a policy presented itself as a form of “territorial abandonment.” And since no facilities had been established after the Qing “abandoned” the land, he argued that “this place did not belong to the Manchu.” This was how the land north of the Yalu and Tumen rivers “became a complete no man's land” (Shinoda Reference Jisaku1938, 26–27). Despite the lengthier historical narrative contained in the monograph, the logic remained essentially the same as that contained in his previous essays. The most important evidential support, again, was Du Halde's Description de la Chine.

But Shinoda deliberately confused two concepts here: a land with few inhabitants and a so-called “no man's land.” It was true that after invading Korea and conquering China, the Qing made a significant part of Manchuria (including the Qing-Korean borderland) a royal reserve and forbade Han civilians to enter. Yet the Qing set up banner troops to defend the Jilin frontier and regularly patrolled the border with Korea. Although the area was indeed depopulated, at least in theory,Footnote ⁸ it became so precisely because the Qing court executed its governorship relatively successfully. Furthermore, the Qing claim to southern Manchuria was well documented by both the Qing and Chosŏn governments. There should be no confusion between a lightly populated place and an ungoverned space. The key question here is not so much the “sovereign” nature of the land, but how the Qing-Chosǒn borderline became problematic. A brief examination of this question links the imperialist expansion of Japan in the early twentieth century with another empire-building project in eighteenth-century Qing China.

The Problematic Qing-Korean Border

The current Sino-Korean boundary—that formed by the Yalu and Tumen rivers—is arguably one of the oldest state borders in the world that is still effective. The western section of the boundary, following the course of the Yalu (K. Amnok) River, was formed in the late fourteenth century, while the eastern part, following that of the Tumen River, dates to the fifteenth century. While the lands to the south of these two rivers were governed by the Chosŏn dynasty of Korea (1392–1897) for more than four hundred years,Footnote ⁹ the same period witnessed a significant change in governance with respect to the lands on the north. Before the mid-seventeenth century, the Jurchen people had inhabited these lands under the nominal rule (jimi) of Ming dynasty China (1368–1644) for generations. After succeeding in unifying the Jurchen population, Nurhaci, the leader of the Jianzhou tribe, established the “Later Jin” regime (1616–36), leading a rebellion against the Ming. His son, Hong Taiji, renamed the Jurchen “Manchu” and the regime “the Great Qing” (1636–1912) in 1636. Before conquering the rest of China and establishing a new and long-lasting dynasty, the Jurchen regime invaded Korea twice (in 1627 and 1636), with the objective of eliminating the possible threat of a Ming-Chosŏn alliance. The two wars forced Korea into acknowledging the superior status of the Qing in the bilateral relationship and confirmed the traditional geographic division between the two countries. Both states agreed to forbid their subjects from settling in the lands adjacent to the border. The ban was once rigorously enforced, leading to the public beheading of illegal border crossers, whether for poaching or squatting, and the demolition of houses and farms situated near the rivers (Li Reference Huazi2006). Moreover, to maintain the distinguished identity of the ruling group, the Qing Empire kept a major part of its mythical Manchurian homeland (including the Tumen River region) as a royal reserve (Elliott Reference Elliott2000). Several willow palisades (liutiaobian) were established to segregate the nomad zone, the agrarian zone, and the forest zone in Manchuria (Zhang Reference Cunwu1987). During a period of almost two hundred years, the Qing-Korean border area was deliberately depopulated, and the Qing regime prohibited access to the rich soil of Manchuria to all but a limited number of Banner troops and officials.

A closer examination of disputes related to the Tumen boundary requires us to trace back to the first attempts at demarcation made by the Qing in the early eighteenth century. In 1710, a criminal case was brought to the attention of both the Qing and Chosǒn courts. Nine Koreans trespassed the Yalu River in order to poach ginseng. Running into five Qing, they murdered them and took their belongings. In recent years, trespassing incidents like this one had occurred repeatedly. Qing Emperor Kangxi (r. 1661–1722) decided that the time had come for a comprehensive and conclusive clarification of the geographic division between his empire and the Chosŏn.

The background of this decision was that Emperor Kangxi, obsessed with the new geographic surveying and cartographic techniques, was conducting a grand mapping of his entire empire, including not only inner China, Mongolia, and Tibet but also Korea. In seventeenth-century China as well as Europe, cutting-edge cartography was a critical tool for imperial powers (Akerman Reference Akerman and Akerman2009). Introduced by Jesuits, the early modern surveying and mapping technologies proved advantageous over indigenous technologies, especially in a military sense. This had been well proven in a recent armed confrontation between the Qing and Russia in the Amur River basin (1685–89). Kangxi's geographic project served two practical purposes: allowing his regime to acquire updated geographical data on Manchuria that would assist in countering Russian expansionism, while also permitting the creation of a highly accurate imperial atlas that incorporated Chinese as well as inner Asian territories (Hostetler Reference Hostetler and Akerman2009; Perdue Reference Perdue1998). Prior to 1710, Kangxi had already sent several survey teams to Manchuria. Through these surveys, Kangxi promoted Mount Changbai (K. Paektu) as the royal ancestral mountain, further strengthening the foundation myth of the monarch and the sacredness of Manchuria (Elliott Reference Elliott2000). Naturally, the project included the border region between the Qing empire and Chosŏn Korea. Although no one disputed the boundary formed by the Yalu and Tumen rivers as the Qing-Korean boundary, a small section between the two rivers’ respective headwaters had yet to be clearly defined. Surrounding the summit of Mt. Changbai, this area was characterized by an extremely harsh climate and a complicated drainage system. The 1710 homicide case provided Kangxi with a good excuse to demarcate this ambiguous section of the border once and for all (Qingshilu 1986, 246:Kangxi 50–5).

The emperor entrusted a local Manchu official, Mukedeng (1644–1735), with this task. Since Mukedeng would be supervising the homicide trial in a Korean border town, Kangxi ordered that he take the chance to extensively investigate the Qing-Korean boundary in the area between the two headwaters. The first attempt in 1711 failed because of a lack of cooperation from the weather and the Korean officials alike. Aware of the situation, Kangxi, for a second attempt in 1712, asked the Chosŏn regime “to lend a bit of assistance” (Schmid Reference Schmid and Lary2007; Yang Zhaoquan and Sun Yumei Reference Zhaoquan and Yumei1994).

The Chosŏn received this request with great reluctance and caution. Having suffered two invasions by the Manchu “barbarians” and forced to submit to the Qing, the Korean kings and literati were hostile toward the Manchu regime. Through its intelligence network in China, the Chosŏn court assumed that the Manchu regime could not fully control China for an extended period, and that it would only be a matter of time before they were forced to retreat. Thus, when Kangxi launched his grand geographic survey in Manchuria and fostered imperial worship of Mt. Changbai, the Koreans misinterpreted the purpose of the mission. They speculated that the Qing were preparing a route that could be used by military forces to return to Manchuria through Korean territory, and worried that this would harm their country once again (see, e.g., Sŭngjŏngwŏn ilgi, Sukchong 26–2–26). So the initial Korean response was to cite difficult terrain or harsh weather to avoid assisting the survey missions. This strategy worked in 1711. But when Kangxi urged for a second time, the Koreans felt they had to cooperate, although with the intention of securing more territory to the south of Mt. Changbai (Chosŏnwangjo-sillok, Sukchong 38–2, 3).

In May 1712, the Qing and Chosŏn teams met in Huju, a Korean border town on the Yalu River. They traced the river upstream all the way to Mt. Changbai. A Korean interpreter made Mukedeng believe that both the Yalu and Tumen rivers originated from the Heaven Lake, a volcano lake on the peak of the mountain. Therefore, instead of tracing the other border river, the Tumen, upstream to determine its correct source, Mukedeng decided to search for what he believed to be the shared source of both border rivers at the summit. Mukedeng's team and some of the Korean escorts reached the top on June 14 and started to look for the source. At a site approximately five kilometers to the southeast of the summit, they found a ridge that was quite close to the headwater of the Yalu River. Mukedeng decided that this ridge should be regarded as the “drainage divide” (C. fenshuiling; K. punsurǒng) of both the Yalu and Tumen rivers. This meant that the only task that remained was to find the source of the Tumen connected to this drainage divide. However, this task proved much harder than they had expected: the intricate nature of the river system and the intermittent nature of water flows in this area made it extremely difficult to define a true source. After four days of arduous exploration and debate, they finally agreed to choose one of the small streams that gushed out dozens of ri east of the drainage divide as the true source of the Tumen.Footnote ¹⁰ Explaining that this water had emerged after “flowing underground” from the divide, Mukedeng believed that he had finally accomplished his mission. He ordered a stele to be erected on the drainage divide as the mark and asked the Korean staff to build a row of earthen and wooden barriers in the future to connect the stele with the Tumen headstream (Chosŏnwangjo-sillok, Sukchong 38–5, 6).

Far from being the end of the problem, this was just the beginning. When the Koreans later returned to build the barriers, they discovered that the stream Mukedeng chose was wrong: it somehow flowed north and eventually join the Songhua (Sungari) River, a tributary of the Amur River. However, after intensive internal debate, the Chosŏn court decided not to bother the Qing with such a “trivial” mistake. Instead, the Korean official who supervised the project arbitrarily built a barrier to connect another stream in the south. As a consequence, in Beijing, the Qing government never became aware of the error. Since border security had been relatively stable for more than a century, no Qing official after Mukedeng had ever surveyed the border again. As the earthen and wooden barriers gradually eroded, so did the memory of the demarcation itself. What was worse, even the Qing documents related to the survey mission were lost, possibly in a fire.

In Korea, by contrast, the Mukedeng mission of 1712 was extensively recorded and widely remembered. Since the mid-eighteenth century, geographers and mapmakers have viewed the Mukedeng stele as a marker of the boundary of Korean territory, rather than as an indicator of the source of the Tumen River. However, the Korean knowledge with respect to the boundary was rather incoherent. While the Qing documents referred to the river as 土門 (pronounced “Tumen” in Chinese), the Chosŏn documents used the name 豆滿 (pronounced “Tuman” in Korean) predominantly. Both names, arguably, are transliterated from the Manchu words tumen sekiyen, which mean “the origin of ten thousand rivers” (Yu Fengchun Reference Fengchun2009). After the 1712 demarcation, some Koreans inherited the traditional view that the Tumen/Tuman, being different names of the same river, was indeed the border. Some other literates, however, challenged this view and generated a new theory that the Tumen and Tuman were two rivers separated from each other. According to this theory, the border river was not the “Tuman” but another river that originated in Mt. Paektu/Changbai and flowed further north. Yet, their theory about this “real” border river was even more confusing: some named it “the T'omun River” (土門江) while some others named it “the Division River” (分界江; K. Punkyekang). Some indicated that this river was actually the Hontong River (混同江, Songhua/Sangari River), whereas still others insisted that it located between the Sangari and the Tuman (Nianshen Song Reference Song2017). This caused great confusion regarding the exact location of the boundary during the late Chosŏn period. It was precisely because of this confusion that, more than 170 years after the installation of the demarcation stele, the Korean refugees made a claim that the Tuman River they had crossed was not the real boundary.

Despite interpreting the boundary river differently, however, neither the Qing nor the Chosǒn had ever acknowledged a “neutralized” border zone between them. Numerous Qing and Korean historical materials repeatedly prove that both countries recognized only the Yalu and the Tumen rivers, no matter how they were defined, as their border. Even when there were “unused wildlands” on the north side of the rivers, these lands were evacuated due to the Qing's vigorous seclusion policy, rather than being abandoned by the Qing. That being said, Shinoda Jisaku still insisted that there had been a “no man's land” and handpicked a single European text to support this argument. Why did he do that? What elements of this document did he cite in making his claims? More importantly, how was the information in this document produced?

“No Man's Land,” or Terra Nullius

Shinoda's “no man's land” needs to be examined in a broader historical and intellectual context, starting with the document that he cited. Description de la Chine gained an enormous reputation in eighteenth- and nineteenth-century Europe and was widely considered as one of the foundations of Western Sinology.Footnote ¹¹ But Du Halde never visited China himself. Instead, he composed this masterwork by collecting and drawing upon correspondences and reports written by Jesuit missionaries in China. The description of the Korean border in this book was based on a memoir written by Jean-Baptiste Régis (1663–1738), who had been dispatched by Qing Emperor Kangxi to survey and map Manchuria in 1709 and 1710. Shinoda used this background in making his case that Régis's account was “trustworthy.” Quoting from Régis, Shinoda argued that the western boundary of Korea lay to the east of the city of Fenghuang and that an “uninhabited space” lay between the willow palisade and the Korean boundary. Along with Régis's memoir, Shinoda also included a published map of Korea that had been edited by French cartographer Jean Baptiste Bourguignon d'Anville (1697–1782) (see figure 1). Quoting Régis, Shinoda explained that the Korean boundary was shown on this map using broken lines (Shinoda Reference Jisaku1938, 22).

Figure 1. “Map of Korea” (Du Halde Reference Du Halde and Brookes1741).

Shinoda was not the first Japanese who employed these Jesuit documents. It was possible that he got the information from Naitō Torajirō, later a famous Japanese Sinologist generally known as Naitō Konan (1866–1934). From 1906 to 1908, Naitō, then a correspondent of Ōsaka Asahi shinbun (Ōsaka Asahi news), was commissioned by the Japanese Ministry of Foreign Affairs to investigate the Qing-Korean territorial dispute. During his trips, he gathered numerous historical materials in Han, Manchu, Mongol, and other languages, which later became priceless treasures for the Kyoto School of Sinology headed by Naitō himself. In his survey report submitted to the Ministry of Foreign Affairs in 1907, Naitō mentioned Régis's memoir in order to prove the argument that Kantō was an “uninhabited neutralized zone” and “deserted area” within the Qing boundary (JACAR, n.d.; Naitō Reference Konan1976; Nawa Reference Etsuko2012). But Shinoda had a slightly different interpretation of Régis's description.

Shinoda took this account more seriously because of Régis's use of the precise expression un espace inhabité (uninhabited space), which Shinoda could argue was equivalent to the Japanese term mujinchitai (no man's land). But what was involved here was not simply a matter of translation from one language to another. His ultimate purpose was to prove that the “sovereignty” of this space was “undecided,” that it belonged “to neither the Qing nor to Korea.” In other words, rather than using the phrase “no man's land” in its literal meaning, he instead interpreted it in a specific legal sense.

In the language of international law, the land was to Shinoda Jisaku a terra nullius. This Latin term, despite claims that is derived from the Roman legal term res nullius, is arguably a modern invention (Connor Reference Connor2005; Fitzmaurice Reference Fitzmaurice2007). Usually translated as “no man's land,” terra nullius actually does not refer to an uninhabited or less-populated place but to a land without sovereignty. Appearing first in the form of territorium nullius, the term was proposed and discussed by German professor F. de Martitz at a meeting of the Institut de Droit International (The Institute of International Law) in 1888. Aiming to secure the German share of the European colonies in Africa, Martitz defined the term as referring to “any region not effectively under the sovereignty or protectorate of one of the states forming the community of international law, whether inhabited or not” (quoted in Connor Reference Connor2005, 11).Footnote ¹²

According to the logic of international law, the term terra nullius has a close relationship with one of the basic methods of territorial acquisition: occupation (Khan Reference Khan, Fassbender and Peters2014). A land that belongs to no one is a subject for occupation, provided that the occupier is a sovereign state recognized by the “community of international law.” As many scholars have argued, although the term per se was rarely used in actual operation in history, its logic, which derived from the natural law tradition, was widely adopted by European colonial powers in rationalizing their seizure of indigenous lands beginning in the sixteenth century (Benton and Straumann Reference Benton and Straumann2010; Touri Reference Touri, Fassbender and Peters2014). The natural law tradition recognizes property rights only by effective use. In the context of land and territorial rights, terra nullius concerns not only whether the land had been exploited or not but also done by whom. If necessary, indigenous land rights must be yielded to colonial practices. As Martitiz argued in the 1888 meeting, it is “an exaggeration to talk about the sovereignty of savage or half-barbarian” (Fitzmaurice Reference Fitzmaurice2007, Reference Fitzmaurice, Fassbender and Peters2014). By the late nineteenth and early twentieth centuries, this logic had become firmly embedded in the legal rhetoric used to retrospectively justify what was essentially the encroachment upon and dispossession of the lands of indigenous peoples.

Around the same time when European colonizers manipulated the principle of terra nullius to deny the land ownership of aboriginal peoples in Africa, Australia, and North America, Japan was also quick to adopt this principle in interpreting and justifying its imperial expansion following the Meiji Restoration. As scholars have recently pointed out, during its colonization of the island of Hokkaido (1869) and its expedition to Taiwan (1871), Japan characterized both islands as examples of “no man's land” in legitimizing the expulsion of the indigenous Ainu and Taiwanese aborigines (Kawashima Reference Kawashima, Fassbender and Peters2014; Manson Reference Manson2012; Mizuno Reference Mizuno2010; Yanagihara Reference Yanagihara, Fassbender and Peters2014). Colonization on a global level was unfolding at the same pace, under the same logic. In light of this background, it made perfect sense for Shinoda to argue that on the one hand “Kantō belongs to neither the Qing nor Korea,” while on the other hand “it was fair to regard the lands to the north of the Tumen as Korean territory.” According to Shinoda's reasoning, because the overwhelming majority of the people of this “no man's land” were Korean, and since Korea was no longer a sovereign state but a “protectorate” of Japan, it followed that Japan had every right under international law to govern this region.

The Long Journey of “No Man's Land”

Again, however, Shinoda's ambitions extended way beyond the Tumen River region. Citing Régis's report in Du Halde's book, he argued that the Korean border extended all the way to Fenghuang, a city in the eastern part of the Liaodong Peninsula, meaning that most of the Yalu River region could also be included in his concept of Kantō. The problem was that he was citing Régis outside the original context. To illustrate this point, it is worth quoting Régis's original text at length. The portion Shinoda extracted is marked in bold:

It is obvious that Shinoda obscured crucial information in his selective quotation. Régis never ventured beyond the city of Fenghuang, to say nothing of reaching the Korean border. In fact, the Qing emperor did not permit Jesuits serving in the Qing court to approach Korea. Even if the emperor had allowed it, the Chosŏn court would very likely have forbidden missionaries from entering the country. What is more, the account by Régis and the map of the Korean kingdom included in Du Halde's book were provided by a “Tartar lord” who himself had traveled only as far as the capital city. The Jesuit map of Korea was clearly not based on a direct field survey (as Shinoda somehow implied). In stark contrast to what Shinoda was insisting, Régis himself frankly admitted that he was in no position to offer a “trustworthy” depiction of Korea.

But let us not stop here. Régis's narrative provided us with enough clues to cross-examine the relevant Chinese and Korean documents and investigate how the geographic knowledge he received from the “Tartar lord” was formulated in the first place. The “Tartar lord” that Régis mentioned was Mukedeng, the exact same Manchu official who carried out a survey of the Qing-Chosŏn borderlands and mistakenly erected a stele in 1712 that was to play a key role in the entire territorial dispute. The trip that Régis describes took place in 1713, when Mukedeng was sent to Korea once again by Emperor Kangxi as a deputy envoy. While announced as a diplomatic mission, the real purpose of the trip was to collect further geographical data on Korea so as to complete Kangxi's ambitious project of mapping the entire empire using the Jesuits’ cartographic techniques. Mukedeng and the mathematician in his team carried out measurements en route all the way from the city of Fenghuang to Hansǒng (Seoul). Upon arriving at the Chosŏn court, Mukedeng made a direct request to King Sukchong for detailed hydrographic and geographic data of Mt. Changbai/Paektu, along with a map of the region. But the Koreans rejected by claiming that the court did not possess such information either. Mukedeng was not convinced. Showing that he knew what he was asking for, Mukedeng demonstrated a map produced during his 1712 border survey and urged the Koreans to share their version.

Several days later, Chief Minister Yi Yu made a suggestion to King Sukchong:

Mukedeng made a copy of this map and returned the original.

Exactly which map the Koreans provided to Mukedeng is unknown, but it is this map that constituted the basis of the map Régis created and d'Anville in turn produced. In his memoir, Régis said that he calculated the geographic location and the shape of the Korean peninsula by cross-referencing his survey of Fenghuang city, Mukedeng's measurement en route to Korea, and, of course, Mukedeng's copy of the map from the Korean court (Du Halde Reference Du Halde and Brookes1741, 383).

D'Anville, a famous cartographer and a royal geographer of France, used the Jesuit maps to produce a series of maps for Du Halde's four-volume Description de la Chine. Two years later, the maps were collected and published in The Hague as a single-volume atlas, Nouvel atlas de la Chine, de la Tartarie Chinoise et du Thibet (New atlas of China, Chinese Tartary and Tibet).Footnote ¹³ In the preface to the atlas, the editor praised the quality of the original Korean map:

Of course, the editor could not have known the story behind the Korean map. But the preface nevertheless confirmed that the Jesuit map of Korea published in Description de la Chine indeed relied very much on the flawed map provided by the Korean court.

The quibble between the Qing and the Chosŏn over the Korean map vividly demonstrated J. B. Harley's (Reference Harley1988, 57) observation that “cartography was primarily a form of political discourse concerned with the acquisition and maintenance of power.” In his work “Cartography in Korea,” historian Gari Ledyard (Reference Ledyard, Harley and Woodward1994) argued that the map given to Mukedeng might be very similar to the “Map of the Eight Provinces” made in the seventeenth century, as the two contain much of the same misinformation. Ledyard even suspects that, “given the general Korean inclination to deceive the Chinese and other foreigners about their country (and the map episode only represents a pervasive habit, nourished by foreign invasions over the centuries), it is not beyond possibility that the map given to Mukedeng could even have been doctored in spots” (303).

It was not only the source of Régis's knowledge that was tainted. Régis himself also wrongly understood the geographic location of the Korean boundary. While he did make mention of the Yalu and Tumen rivers, describing them as “most considerable rivers, which are its [Korea's] defense as well as its riches” (Du Halde Reference Du Halde and Brookes1741, 383), he depicted the boundary as lying north of the rivers, implying that the realm across both rivers fell within Korean territory. This description contradicted most Chinese and Korean records and could not have been based on material collected by Mukedeng, who himself surveyed the border in 1712. How exactly he made this mistake is hard to determine, but it is possible that he misinterpreted the willow palisade (“Palisado”) as the Qing boundary instead of a domestic barrier.Footnote ¹⁴ A possible explanation is related to the fact that the Qing deliberately evacuated the area between Fenghuang City and the Yalu River so that no civilian could inhabit the frontier zone. The Korean emissaries who took the tributary journeys to Beijing were received by the Qing only when they reached the fence gate of Fenghuang City on the willow palisade, typically a few days after crossing the border river from the town of Ŭiju. Fenghuang, as opposed to the border river, thus often served as a conceptual division between the two states in terms of communication, psychology, and culture (Wang Reference Yuanchong2015). However, this should not counter the fact that the Yalu River was repeatedly confirmed as the geographic and political border between the two countries.Footnote ¹⁵ In Description de la Chine, there is another document titled An Abridgement of the History of Corea that clearly indicates that the Yalu River was the division between Korea and Liaodong (Du Halde Reference Du Halde and Brookes1741, 4:394). But for Shinoda, Régis's confusion seemed to be more valuable.

To sum up the twisting route of the discourse on the Sino-Korean borderland: in 1713, the Korean court gave Mukedeng false geographic information regarding the country and its frontier in order to safeguard its interests from further Manchu harassment. Based on the material that Mukedeng had collected in Korea, Régis produced a description and a map of Korea that were later reproduced in Du Halde's Description de la Chine in 1735. After nearly two hundred years, Shinoda Jisaku picked out a specific portion of Régis's transcript and distorted it to align with his colonialist ambitions. After being transmitted in various incarnations from one end of Eurasia to the other and back again over a period of almost two centuries, then, the geographic knowledge that Korea had fabricated in order to protect itself was eventually used by the Japanese as evidence that the lands to the north of the Tumen and Yalu rivers could be considered a terra nullius.