The Appearance and Disappearance of Ryukyu: The Historical Views of Tō Teikan, Motoori Norinaga, and Ueda Akinari

Abstract

Two of the renowned figures of Edo-era Kokugaku (National Learning), Motoori Norinaga and Ueda Akinari, famously debated the merits of their scholarly approaches to Japanese antiquity during the latter half of the eighteenth century. Their intellectual dispute was the result of the radical conclusions reached by Tō Teikan in his Shōkōhatsu (An Outburst of Provocations; 1781) in which he argued that the Korean peninsula and China influenced ancient Japan, and that Japan’s first emperor, Jimmu, was from Ryukyu. While Akinari supported the notion of continental influence on ancient Japan, Norinaga did not, and while the former was mostly agnostic about Jimmu’s Ryukyuan roots, the latter opposed that as well. Norinaga, however, was not opposed to the idea of ancient ties between Ryukyu and Japan, an issue with which Akinari’s silence seemed to signify some degree of agreement. This commonality between these two intellectual giants demonstrated the extent to which Japanese intellectuals of the Edo period viewed the Ryukyu Kingdom (now Okinawa Prefecture) as occupying an ambivalent geopolitical space, in which it was neither fully foreign nor fully native. At the same time, Akinari’s historiographical approach to Japanese antiquity, which emerged in his debate with Norinaga, exerted an influence on nineteenth-century depictions of Ryukyu’s historical and cultural ties to Japan, chiefly Kyokutei Bakin’s Chinsetsu yumiharizuki (Fantastic Tales of the Moon Bow; 1811).

1. Introduction

In one of Iha Fuyū’s earliest works, Ryūkyū jinshu ron 琉球人種論 (Theories of the Ryukyuan Race; 1911), he describes how Ryukyu was prominently featured in a controversial Tokugawa work, the Shōkōhatsu 衝口発 (An Outburst of Provocations; 1781), by a somewhat obscure scholar named Tō Teikan 藤貞幹 (also known as Fujii/Fujiwara Sadamoto 藤井/藤原貞幹, 1732–1797).1 What intrigued Iha, one of the founders of Okinawan Studies (Okinawagaku 沖縄学), the most was Teikan’s conclusion that Japan’s legendary first emperor, Jimmu 神武天皇 (711–585 BCE), was born on the Ryukyuan island of Iheyajima 伊平屋島. Iha observes how such an exciting conclusion, made nearly a century and a half earlier, would have made (then) Ryukyu the “homeland” (kokyō 故郷) of the Japanese people (Fuyū 1975, p. 3). Such a radical idea, however, did not go unchallenged, drawing the critical scorn of Motoori Norinaga 本居宣長 (1730–1801), the most famous Kokugaku 国学 scholar of the Edo period and one of the era’s most celebrated intellectuals, who refuted and even ridiculed it. The idea that Jimmu was born in Ryukyu did not take root during the Edo period, but Iha argues that modern (for him) scholars were beginning to reach similar conclusions regarding the prehistoric connections between Japan and Ryukyu, which made Teikan a scholar who was simply too far ahead for his time. Iha predicted that the modern methods inherent in the disciplines of linguistics, anthropology, archeology, and others, would find the proof for these connections upon which Teikan could only have speculated.

The appearance of the Shōkōhatsu not only elicited a harsh rebuke from Norinaga, but it also drew another prominent Kokugaku scholar to respond, Ueda Akinari 上田秋成 (1734–1809), a figure whose literary standing was perhaps the equal of Norinaga’s scholarly one. Akinari was partially motivated to defend Teikan’s work, but he was also energized by a desire to put Norinaga in his place. Akinari rushed his response to Norinaga’s refutation of Teikan and sent it directly to him. The views of both intellectuals were later summarized and presented side-by-side in a text called Kakaika呵刈葭 (Cutting the Reeds, first part; ca. 1787), a work that later appeared in the collected writings of both Akinari and Norinaga. The debate recorded in the Kakaika was one of the most famous internecine disputes in the history of Kokugaku, pitting as it did two intellectuals who believed themselves to be the heirs to the teachings of the great Kamo no Mabuchi 賀茂真淵 (1697–1769). The two scholars were unable to disguise their feelings of contempt for one another, and the debate evolved into their competing views of ancient history and the ancient Japanese language, leaving Teikan’s views, which were collectively the initial cause of the debate, behind.

The fact that Ryukyu, which had been so vital to the arguments made in the Shōkōhatsu, receded and eventually disappeared in the ensuing debate between Akinari and Norinaga is critical for two reasons. First, it formed the intellectual context within which Kyokutei Bakin 曲亭馬琴 (1767–1848) composed his own monumental work of fictional history, the Chinsetsu yumiharizuki 椿説弓張月, which prominently featured the Ryukyu Kingdom, along with its culture and history (see Takizawa 1958, 1962). Bakin was inspired by Akinari’s blending of standard historical writing with fiction in the conceptualization and composition of his yomihon, and the Chinsetsu yumiharizuki became one of the greatest examples of the genre ever written in the Edo period. However, the fact that Bakin was the one to connect fictional writing to Ryukyuan history, and not Akinari, leaves us wondering how and why Ryukyuan matters had faded from the Kakaika debate in the first place. Put simply, Bakin’s work demonstrated to a more mass audience how Ryukyu was neither completely foreign nor was it fully native. For both Akinari and Norinaga this was the case, and so they understandably shifted their scholarly attention away from Ryukyu and toward the more solidly foreign regions of the Korean peninsula and China.

The second reason for the importance of Ryukyu in the Shōkōhatsu and the Kakaika debate it helped to inspire is because of its connections to Kokugaku. Perhaps no other group of scholars and intellectuals during the Edo period were more concerned with matters both foreign and native to Japan and its people. The fact that two of the most celebrated figures within Kokugaku, while focusing on Japan’s foreign interactions in antiquity, momentarily focused on Ryukyu only to lose sight of it for good, tells us about Tokugawa attitudes toward Ryukyu. If Kokugaku scholars could accept Ryukyu as part of Japan, even if not enthusiastically, then just how foreign was it in the minds of Japanese people in the Edo period? The fact that Norinaga, the most renowned of the Kokugaku scholars, supported the view that the father of Ryukyu’s first historical king was the famed Japanese warrior, Minamoto no Tametomo 源為朝 (1139–1170), a view Akinari did not refute, is indicative of his overall attitude that Japan and Ryukyu had profound historical ties.2

2. Results

2.1. The Shōkōhatsu

Tō Teikan’s treatise on ancient Japan is mostly concerned with the kamiyo 神代 (“age of the kami”) chapters in the Kojiki 古事記 (Record of Ancient Matters; 712 CE) and the Nihon shoki 日本書紀 (Chronicles of Japan, also called Nihongi 日本紀; 720 CE3), with a special reliance on the latter (see Kurano 1996; Ienaga et al. 1994). Teikan was a former Buddhist monk who had returned to lay life as a young adult, and so the influence of Buddhism might explain his rather skeptical approach to kamiyo, an approach that was perhaps more tolerant of Confucianism, as Norinaga later observed, which might account for his preference for the Nihongi over the Kojiki. Teikan sought to analyze Japan’s ancient creation stories as myths, treating references to the various kami of high antiquity as metaphors for actual historical figures, and the stories themselves as narrative traces of significant interactions among the peoples of East Asia, what are today China, the Korean peninsula, and Japan. What he discovered in his analysis is that Japan’s ties to China and to the Korean peninsula were older and more profound than most people of his time had thought, but its ties to Ryukyu were just as old and even more important; Japan’s connections to Ryukyu were older than even the Japanese (Yamato) state itself.

Jimmu’s parents were both described as kami. His father was Hikonagisatakeugayafukiaezu-no-mikoto 彦波瀲武盧茲草葺不合尊 and his mother was Tamayoribime 玉依姫, according to Teikan’s interpretation of the kiki shinwa 記紀神話 (“myths of the Kojiki and Nihongi”). They resided in the Watatsuminomiya 海宮 (“Palace of the Sea [kami]”), where Jimmu was said to have been born. This palace was associated with the kami of the sea and so was thought to be located within the sea itself. Teikan argues that Watatsuminomiya was not literally in (or under) the sea but on top of it, specifically as a location on an island. Since Watatsuminomiya had to be a real place, for Teikan, and located on a real island, he chose Iheyajima 恵平也島 (伊平屋島), a small island about forty kilometers northwest of Okinawa island (Tō 1930, p. 230). Teikan avoids elaborating on the implications of his conclusion, which was more an inference, an educated guess, than it was the result of textual analysis, but it invites either of two further conclusions. Teikan refers to Iheyajima as belonging to Ryukyu (Ryūkyū no 琉球の), an officially independent realm in Teikan’s time, allowing his readers to conclude that Jimmu was born outside of Japan (or, that “Japan” itself was meaningless at that time, since the Yamato state, which Jimmu founded, did not exist). Alternately, one could conclude that Ryukyu, as the birthplace of Jimmu, was an important part of Japan, or what would later become Japan. This conclusion resonates with the ways in which Arai Hakuseki 新井白石 (1657–1725) had described Ryukyu a little more than half a century earlier, namely, that regardless of Ryukyu’s geopolitical twists and turns over the centuries, it was always and originally part of Japan (Wa 和) as a region.

Although Teikan relies on an unspecified source in identifying Watatsuminomiya with Iheyajima, he claims that the latter is the same island as Anmijima 天見島, citing passages from both the Nihongi and the Shoku Nihongi 続日本紀 (The Chronicles of Japan, Continued; 797 CE) as evidence for the association, since their narratives contain references to islands with very similar names, namely, Amami 阿麻美 and Anmi 奄美, respectively (Tō 1930, p. 230). It seems that Teikan may have been confused by Ryukyuan geography. While Iheyajima is a small island that sits quite close to the island of Okinawa’s northwest, Amami usually refers to a group of islands (the Amami shotō 奄美諸島) today that lie a little further (about 100 km) away to the northeast of Okinawa Island, although the name “Amami” signified the main island of the group as well, Amami Ōshima 奄美大島. Teikan, following the ancient historical narratives, treats Amami or Anmi as one island, and so it is reasonable to assume that he had Amami Ōshima in mind. Teikan describes the island as having a sacred mountain called Amamigatake 天孫嶽 (“the mountain of the heavenly descendants”), and that this was the place where the kami descended from Takamagahara 高天原 to the earth. Specifically, it was the place upon which Fukiaezu-no-mikoto’s father, Hikohohodemi-no-mikoto 彦火火出見尊, descended from Takamagahara in order to take up residence on his earthly abode, according to Teikan. While Teikan conflated two distinct islands, the fact that both islands were connected to Ryukyu was true. During the Edo period, the Amami islands were officially recognized as belonging to the Ryukyu Kingdom, even though they had been under direct control by the Satsuma domain following its invasion of Ryukyu in 1609. Of even more ideological importance is the fact that Teikan also conflates the Amami 天孫, “heavenly descendants” of Japanese mythology, with the Tenson 天孫 (also, “heavenly descendants”) line of Ryukyuan kings, by connecting their descent from Takamagahara to Ryukyu (whether as Iheyajima or as Amami).

A little historical context is necessary in order to understand the relationship between Ryukyu and Japan. The rather rocky economic relationship between the Shimazu daimyo of Satsuma and the Ryukyu Kingdom, in the wake of the Kakitsu Revolt of 1441, was made even worse by an equally contentious political relationship between the two over where the border was between their respective areas of control. An official Korean entry from 1453 in the Chosŏn wangjo sillok 朝鮮王朝実録 (조선왕조실록 [Veritable Records of Chosŏn]) relates an account by Korean castaways who had washed ashore on the small island of Gajajima 臥蛇島, one of the Tokara islands 吐噶喇列島 south of Kyushu. One should note that Smits argues that this was likely “Gaja,” which signified the Tokara islands in general, rather than Gajajima itself (Smits 2019, p. 49). In any case, they observed how both the Ryukyuans and the Japanese of Satsuma claimed the island as theirs, which was evidence of this frontier dispute between the Shimazu and the Ryukyu Kingdom (Turnbull 2009, p. 9). During the First Shō Dynasty (1406–1469), especially the first half of the fifteenth century, the Ryukyuans brought the islands north of Okinawa under their control, chiefly the Amami islands 奄美諸島, which brought them into direct contact with the forces of the Shimazu, who were similarly interested in expanding their power throughout both the rest of Kyushu and the islands south of it. The Shimazu nearly conquered all of Kyushu until their defeat by Hideyoshi in 1586 (Smits 2019, p. 208). The Shimazu attempted to seize what was the jewel of these islands, Amami Ōshima, in 1493, but King Shō Shin 尚真王 (1465–1526) and his Ryukyuan forces defeated and drove them out, thereby cementing their control over the Amami islands for the next century. Thus, one of the major justifications of the Shimazu for their invasion of Ryukyu was to seize the Amami islands as repayment of the kingdom’s debts incurred during the 1590s.

As if Teikan’s assertion regarding Jimmu’s birthplace in Ryukyu was not controversial, or sacrilegious, enough, he even challenges the view that Ugayafukiaezu-no-mikoto, which Teikan refers to as simply Fukiaezu-no-mikoto 葺不合尊, was Jimmu’s father (Tō 1930, p. 229). Teikan observes how Jimmu could not have been Fukiaezu-no-mikoto’s son, because Jimmu was older than his purported father was, citing an unspecified text (aru ki ga iu 或記云) as proof. As the son of Toyotamahime 豊玉姫 (his mother) and Hikohohodemi-no-mikoto, Fukiaezu-no-mikoto was also born in the Watatsuminomiya palace, ultimately sharing the same Ryukyuan birthplace as Jimmu, even if they were not directly related to one another. Having eliminated the kami who was Jimmu’s father by tradition, Teikan was confronted with having to offer an alternate paternal candidate in Fukiaezu-no-mikoto’s place, and his response deepened even further the controversial nature of Teikan’s work. Jimmu’s “real” father, he asserts, was a descendant (sue 裔) of Wu Taibo 呉太伯 (11th c. BCE) (Tō 1930, p. 230). The claim that the Japanese were descended from Wu Taibo had ancient origins in China, dating to as early as the third or fourth centuries CE. From the Muromachi period until Teikan’s time, the idea that the Japanese people were connected to Wu Taibo had various supporters and detractors, and Teikan clearly belonged to the former, a group among whose more illustrious members included Hayashi Razan 林羅山 (1583–1657). Consequently, the view of kamiyo crafted by Teikan was that the first Japanese emperor, and the founder of the imperial line (and, by extension, the Yamato state), was born in Ryukyu to a Chinese father, and so was Chinese via patrilineal descent.

As further proof of Ryukyu’s primordial links to Jimmu and to the Japanese people, Teikan turns to a body of evidence that we would likely refer to today as anthropology, specifically, the tradition of tattooing. He cites the fact that the Ryukyuans practice tattooing, as did the Ainu, which leads him to conclude that it was Jimmu who learned the craft in Ryukyu and then spread it throughout Japan as he established the Yamato state (Tō 1930, p. 230). Since the practice itself was an ancient one that predated the “civilizing” influences of Buddhism and Confucianism, the Japanese eventually abandoned it, which explains why it continued to endure in the less advanced areas on Japan’s periphery, Ryukyu and Ezo (Tō 1930, p. 237). By linking Jimmu to Ryukyu and Ezo, Teikan echoes Hakuseki’s claim that the former constituted Nanwa 南和 and the latter Hokuwa 北和, with the same biased interpretation, namely, that the Ryukyuans was culturally/ethnically linked to the Japanese people while the Ainu were not. In other words, both Ryukyu and Ezo were rightfully under Japanese domination, with the key distinction between the two being that the residents of the former more closely resembled the Japanese than those of the latter. It was for this reason that he associates tattooing with ancient Ryukyu as he links it with contemporary Ezo. In reality, the practice of tattooing was very active in all three cultures during the Edo period, whether Japanese, Ryukyuan, or Ainu.

Not only did the practice of tattooing unite the various peoples and cultures of what Hakuseki had described as a greater Yamato realm, it was, for Teikan, also a vital connection between Japan and China, since the ancient Chinese from the state of Wu had also engaged in the practice of tattooing, bringing it with them as they migrated to Ryukyu where Jimmu was born, and it was he who then transmitted the custom to Japan (Tō 1930, pp. 230–231). Jimmu’s parentage and tattooing were not the only historical links between ancient Japan and ancient China in Teikan’s narrative, since he claims that there was another such link with the people of Qin 秦. Specifically, the connection with the Qin was the indirect counterpart to the direct connection with the Wu because of the mediating role of the Korean peninsula. Refugees fled the state of Qin and went to the Korean peninsula, Teikan describes, where they founded the state of Chinhan 辰韓 (진한) (Tō 1930, p. 228). The first ruler of Chinhan, he then claims, was a kami of the sea, Susano-o-no-mikoto 素戔雄尊, the traditional younger brother of the kami of the sun, Amaterasu Ōmikami 天照大神. As the leader of this early Sino-Korean state, Susano-o cultivated close ties with Japan, and subsequent migrations from Chinhan to Japan ensued.

For Teikan, the strongest evidence for these ancient links between Japan and the Korean peninsula were cultural and linguistic. Teikan does not refute the linguistic ties between Japan and China, chiefly in the area of borrowed words, but he believed that many of the words in the Japanese language that were believed to be of Chinese origin were, in fact, either originally Korean or were Chinese words that had changed under Korean influence (Tō 1930, p. 232). Rather than argue that China’s linguistic influence on the Japanese language was the rule, and Korea’s the exception, he advocates exactly the opposite view, namely, that upwards of ninety percent of Japanese words were connected to the Korean peninsula, with the rest imported directly from China (Tō 1930, p. 231). While tattooing linked the various peoples of Hakuseki’s greater Japan to one another, and to the Chinese, language and culture bonded the Koreans to the Japanese. For example, Teikan notes how clothing styles were strikingly similar between those of the Koreans and those of the Japanese, not just for elites but especially the clothing for commoners (Tō 1930, p. 246). Of all the controversial, even intellectually jarring, observations made by Teikan, perhaps the most upsetting one, and the one that was sure to draw the ire of Kokugaku scholars, was his insistence that poetry, or uta 歌, composition was also a Korean import (Tō 1930, p. 251).

Although Teikan’s narrative began with Ryukyu, it ultimately celebrates what he calculates as 700 years of Korean connections to ancient Japan. However, by the end of his text, he seems to lament the inevitable demise of Korean influence at the expense of the ascending influence of Tang 唐朝 China (Tō 1930, p. 259). He observes how Emperor Kammu 桓武天皇 (737–806) was finally able to stem the tide of Sinification that had been ongoing for nearly a century and a half, allowing the ancient Japanese to chart their own cultural course away from India, China, and the Korean peninsula.

2.2. Norinaga’s Response to Teikan: The Kenkyōjin

Four years after the appearance of the Shōkōhatsu, Norinaga revealed his very detailed critique of the work under the colorful title of Kenkyōjin 鉗狂人 (Restraining the Lunatic; 1785). Not surprisingly, Norinaga tried to downplay the significance of foreign influences on Japanese antiquity, as the wellspring of authentic Japanese culture before the importation of Buddhism and Confucianism. He was especially annoyed that the author of the Shōkōhatsu, whom he refers to as “the author in question” (konronsha 今論者), and never by any actual name, had ventured into the sacred era of kamiyo in an attempt to relate its stories to actual historical events and people. For Norinaga, not only did this analysis of kamiyo subject it to outright misinterpretation, but it also represented an affront to the divine nature of the kami. Consequently, the Shōkōhatsu was an intellectual betrayal of Japan. Although Norinaga was especially keen to refute Teikan’s claims about the depth and breadth of Korean influences on kamiyo, he did not ignore Teikan’s remarks about Ryukyu. While he did not support the ideas about Jimmu’s birthplace or parentage, he agreed with Teikan that Ryukyu had significant connections to ancient Japan, views that further contributed to an image of cultural ambivalence for the Ryukyu Kingdom during the Edo period.

2.2.1. Norinaga’s Views of the Korean Peninsula

Naturally, Norinaga countered Teikan’s enthusiasm for all things Korean by issuing blanket denials of Korean influences as “manufactured theories” (setsu wo tsukureru nari 説を作れる也) (Norinaga 1930, p. 279). He inverted Teikan’s assertion that most of the words of the Japanese language were of Korean origin by arguing that most Japanese words were formed on Japanese soil, completely free of foreign contaminations, and were therefore authentically Japanese. He does not deny that some Chinese and Korean words were absorbed into the Japanese language, but it was irrefutable that the Japanese language itself had already formed during kamiyo prior to any interactions with the Asian continent, with Korea (Kankoku 韓国) in particular (Norinaga 1930, p. 287). So, out of thousands of words within the Japanese language, Norinaga admits that “it is not the case that there were no words that moved into” Japanese from Korea and China (Sankankan no kotoba no utsuritaru mo naki ni wa arazu 三韓漢の戎言のうつりたるも無きにはあらず), but such a phenomenon happened only “very rarely” (maremare) (Norinaga 1930, p. 287). Since the scholarly identity of Kokugaku as an intellectual institution was so closely tied to the analysis of what its adherents believed to be the spoken language of Japanese antiquity, it is not surprising that Norinaga felt compelled to deny Teikan’s claims about the presence of foreign words in the Japanese language as holdovers from that era. Similarly, he could not allow Teikan’s claim that Japanese poetry (waka 和歌) was a derivation from a Korean original to go unchallenged, since the origins of Kokugaku itself were in the ancient verses of the Man’yōshū 万葉集 (The Collection of Ten-Thousand Leaves; 759 CE). He observes how Teikan based this claim about poetry on the legendary figure of the Korean scribe Wani 王仁, who was traditionally credited with introducing the Chinese writing system to the Japanese, as well as The Analects (Rongo; Ch. Lunyü 論語) of Confucius. Wani was thought to be the figure behind a waka masterpiece called the Naniwazu no uta 難波津の歌, which later scholars and poets regarded as foundational for the waka tradition. Norinaga flatly denied all of it: “That poem was not the work of Wani. There is no doubt that it was the work of a later person” (ka no uta wa Wani ga saku ni wa arazu. Nochi no hito no saku naru koto utagai nashi かの歌は王仁が作にはあらず、後の人の作なること疑ひなし) (Norinaga 1930, p. 301). In other words, the famous waka verse attributed to Wani, a foreigner, was the work of a Japanese poet.

Norinaga faulted the erroneous conclusions of the Shōkōhatsu on its author’s reliance on flawed sources, beginning with his faith in the veracity of the Nihongi. Norinaga’s fame among Edo-period intellectuals was the result of a variety of achievements, but among the more important of these was his work analyzing the Kojiki 古事記, the massive tome he entitled Kojikiden 古事記伝 (On the Kojiki; completed in 1798; published in its entirety by 1822). Although the Kojiki was completed a mere eight years before the Nihongi, Norinaga believed that it accurately captured the spoken language of Japanese antiquity, and it did so in a way that clearly demonstrated the lack of any foreign influence.4 The same could not be said of the Nihongi, however. Unlike the Kojiki, whose peculiar use of Chinese characters was proof of its connection to spoken Japanese, the Nihongi was composed in the more standard classical Chinese language that was dominant throughout East Asia, including the Korean peninsula. For this reason, the Nihongi was contaminated with both Chinese, because of its form, and Korean influences, because of its contents. For Norinaga, Teikan’s esteem for Korean influences was directly related to his reliance on the Nihongi for information on ancient Japan and kamiyo (Norinaga 1930, p. 302).

2.2.2. China: Norinaga’s Implacable Foe

Even though the Nihongi was a Japanese historical source, it could not be trusted because its compilers were under the sway of Korean influences, as far as Norinaga was concerned. Norinaga was similarly skeptical of historical sources that contained information about foreign realms, chiefly Chinese sources with information on their peripheral lands. Chinese sources were first and foremost composed to record Chinese history, and the same was true of the connection between Korean sources and Korean history. Why would scholars trust Chinese sources regarding Japan, or even on the Korean peninsula for that matter? Yet this was exactly what Teikan had done, Norinaga argues, citing the example of Susano-o. Rather than reject Teikan’s interpretation by simply invoking the fact that Susano-o was a kami, and not an earthbound person, Norinaga instead chooses to impugn Teikan’s historiographical methodology. The Nihongi describes how Susano-o had descended from Takamagahara to the realm of Shiragi 新良,5 or in Korean, Silla 신라, which Teikan believed was actually a reference to the Korean realm of Chinhan. Norinaga corrects the Nihongi’s rendering of Silla/Shiragi into its more conventional one (新羅), conceding that the passage does indeed refer to the Korean peninsula, but he rejects its identification with Chinhan (Norinaga 1930, p. 268). He points out how Teikan’s interpretation was based on a political connection between Silla and Chinhan, but this was a grave mistake; Silla’s connection was actually to the realm of Pyŏnchin 弁辰 (변진). In a demonstration of his mastery of Chinese historiography, his ideological contempt for it notwithstanding, he cites a passage from the Weishu 魏書 (Book of the Wei) section of the Sanguozhi 三国志 (History of the Three Kingdoms) that describes Pyŏnjin as composed of thirty-six realms or provinces, one of which was called Shirokoku 斯盧國, which, he says, was a reference to Silla (Norinaga 1930, p. 269). Moreover, in the Tangshu 唐書 (Book of the Tang), it refers to Silla as belonging to Pyŏnhan 弁韓 (변한). He observes how Pyŏnchin was another name for Pyŏnhan, and that it was sometimes rendered as Pyŏnchinhan 弁辰韓; thus, the confusion between Pyŏnchin and Chinhan was the result of an error made by Chinese scholars who at some point dropped the first character from the name of Pyŏnchinhan, which produced the erroneous result of Chinhan, along with the incorrect connection between it and Silla. In this way, Norinaga was able to correct Teikan’s mistaken identification of Silla with Chinhan and replace it with its actual historical connection to Pyŏnchin. And, of course, this misidentification aside, Norinaga resoundingly denies that Susano-o was ever the lord of Chinhan, since, at the very least, there were no historical references to him as either the king of Chinhan or even of Silla (Norinaga 1930, p. 276).

Norinaga never denies that ancient Korea and ancient Japan were connected to each other, but he was frustrated that scholars, like Teikan, assumed cultural influences flowed only in one direction, namely, from the Korean peninsula to Japan. For Norinaga, there must have been Japanese influences over developments in Korea, “since the imperial realm dominated most of the states/provinces of ancient Korea” (inishie Kara no kuniguni wa, ōku Kōkoku ni fukuzoku shite aritsureba 古韓の国々は、多く皇国に服属して在つれば) (Norinaga 1930, p. 287). Norinaga undermined the scope and depth of these connections between the Korean peninsula and Japan in antiquity as part of a frontal attack on the scholarly strength of the Shōkōhatsu. In an important way, however, refuting ancient Korea was never his ultimate goal, since it was the repudiation of Chinese influences on Japan that was already, by the 1780s, a hallmark of Kokugaku scholarship. For this reason, he devotes more than a little ink in his critique of Teikan to the anti-Chinese themes that he and his “mentor,” Kamo no Mabuchi, had been exploring for decades.6

While Teikan’s take on ancient Korean history was hopelessly flawed because of shoddy Chinese historiography, for Norinaga, there was another problem inherent in the general use of Chinese scholarship, which was also true of Chinese-inspired works, like the Nihongi, namely, the fact that “true principle” (makoto no kotowari 誠の理) had never revealed itself to the Chinese (Norinaga 1930, p. 266). This comment was especially targeted at Japanese Confucians, chiefly those engaged in the search for principle that characterized Song (or Neo-) Confucianism, a form that was a mere five years away, at the time that Norinaga had completed the Kenkyōjin in 1785, from having its status as Japan’s intellectual orthodoxy re-affirmed by Matsudaira Sadanobu 松平定信 (1758–1829). Not surprisingly, Norinaga assumed that Teikan was either a supporter of, or a sympathizer with, Song Confucianism. Principle was important to the Song Confucians because it was ultimately a moral concept that was tied to every and all things, including people, and even to the cosmos itself. This is why Confucian scholars devoted their lives to the elucidation of principle, whether doing so via an internal path, like the followers of the great Chinese Confucian, Wang Yangming 王陽明 (1472–1528), or via an external path, namely, the study of nature, as Kaibara Ekiken 貝原益軒 (1630–1714) famously did with his study of medicinal plants in Japan. Despite the best efforts and good intentions of the Confucians, however, the Japanese kami, who were responsible for the creation of the world and all its constituent realms, including China, had deemed at the beginning of time that “true principle” was only to be revealed to the ancient Japanese people, Norinaga believed. Specifically, it was Sukunabikona-no-mikoto 少彦名命 who, while assisting Ōkuninushi-no-mikoto 大国主命, had created the various realms of the world; as a mischievous kami, Sukunabiko-no-mikoto, Norinaga relates, had withheld “true principle” from everyone but the Japanese people, leaving even the great Sages of Chinese antiquity in the dark (Norinaga 1930, p. 307). The tragic irony of all this was not that the Chinese Confucians had been misled, but that Confucianism ultimately had taken root in Japan, leading to the intellectual corruption and moral downfall of the Japanese people.

What was this secret that the ancient Japanese had known, of which the rest of the world, even the majority of Norinaga’s countrymen, were ignorant? In order to understand kamiyo, one merely had to read the Kojiki, absorb its wisdom, and accept its knowledge as it was without further inquiry or investigation (Burns 2003, p. 111). People had to grasp the idea that there were phenomena beyond their intellectual grasp, notably knowledge of kamiyo, and simply take the Kojiki at face value. “All of human wisdom,” he writes, “…has its limits, and [the number of] things we cannot know is truly vast” (subete hito no satoi wa…kagiri arite nao eshiranu koto wa ito ōki すべて人の智は…限有てなほえ知らぬ事はいと多き) (Norinaga 1930, p. 264). Moreover, Norinaga pessimistically estimated that less than one percent of what people could grasp had actually been apprehended (hito no yoku hakarishiru tokoro wa, wazuka ni sono hyakubun ga ichi nimo oyobubekarazu 人のよくはかりしる所は、わづかにその百分が一にも及ぶべからず) (Norinaga 1930, p. 263). So, in a way, Norinaga had perhaps some sympathy for the intellectual curiosity of the Confucians (and their ilk, like Teikan), but he was critical of their pursuit of knowledge because they lacked the originally kami-endowed wisdom of the ancient Japanese people, which he had devoted himself to reviving and restoring in his own time, namely, that some things were simply beyond human understanding. Thus, the Confucians thought they had understood things, via their investigations into principle, that they really had not; their pursuit of knowledge, chiefly in their interpretations of kamiyo had been in vain. Confucian interpretations were both wrongheaded and subjective for Norinaga.

2.2.3. Norinaga on Ryukyu: Ancient Connections

The negativity of Norinaga’s assessments of the Korean peninsula and China, as they were portrayed in the Shōkōhatsu, was, for the most part, not true of his comments on Ryukyu. There was no question that all three realms were officially foreign as far as the Edo bakufu was concerned, but he was careful to evaluate them differently. China, of course, elicited his harshest assessment, and this reflected the high regard for Confucianism in Edo society, despite the fact that the Edo bakufu had no formal diplomatic relationship with Qing 清朝 China. The bakufu did have such formal state-to-state relations with both Chosŏn 朝鮮 (조선) Korea and the Ryukyu Kingdom, and the processions to Edo carried out by the representatives from both realms elicited a sense of deference, if not subservience, to the bakufu that fostered far fewer threatening images among interested people, and there were many, during the Edo period. Thus, Norinaga was more dismissive of the Korean peninsula and its connections to ancient Japan than he was contemptuous of them. His comments on Ryukyu, however, were neither dismissive nor contemptuous, and he acknowledged that ancient Ryukyu and ancient Japan had, in fact, been connected in important and remarkable ways.

As famous as Norinaga was during the Edo period for his harsh assessment of historical Chinese influences on Japan, especially Confucianism, the fact that he dismisses any Japanese connection to the Wu Taibo legend should come as no surprise. Although some of his contemporaries saw some merit in the legend, like Teikan, Norinaga points out how the locus classicus for it came from a Chinese source, the Jinshu 晋書 (Book of the Jin; 644 CE): “[The Wa people] themselves said that they were descendants of Taibo” (自謂太伯之後) (Norinaga 1930, p. 278). If the ancient Japanese people had said such a thing to Chinese visitors, why did they not say that to one another, and why was it never recorded in any Japanese historical source, he asks. Norinaga concludes that the Chinese had simply made the story up, but that its appearance in a seventh-century Chinese source gave it a sufficiently ancient provenance that ultimately convinced his contemporaries, living more than a millennium later, of its legitimacy. The legend of Wu Taibo and his connection to Japan was simply too outlandish for Norinaga to accept, but the fact that others in his time supported it did not surprise him, since there were other stories circulating in Japan during the Edo period that were equally preposterous, such as that Ninigi-no-mikoto 邇々芸命, Amaterasu’s grandson, was one of the founders of Wu, or that Amaterasu was Wu Taibo. So, by undermining the legend of Wu Taibo and ancient Japan, Norinaga was able to dismiss any connection to Jimmu as well, without having to muster any special evidential or philological effort (Norinaga 1930, p. 277).

In order to refute the claim that Watatsuminomiya is Iheyajima, he fell back on his maxim that the Kojiki had to be taken literally, which meant that Watatsuminomiya, rendered by Norinaga as the “Palace of the Sea Kami” 海神宮 (instead of Teikan’s “Palace of the Sea,” following the Nihongi), was at the bottom of the ocean (kaitei 海底). There were other incorrect interpretations, such as placing Watatsuminomiya on the island of Tsushima 対馬 (near the Korean peninsula), but the inconvenient truth, for those under the sway of Confucianism, was one that their “small principle of the ordinary” (jinjō no shōri 尋常の小理) made them “unable to believe” (kore wo shinzuru koto atawazu これを信ずることあたはず) (Norinaga 1930, p. 280). So, for Norinaga, Jimmu was from Watatsuminomiya, but not from Iheyajima, and so Jimmu never was able to venture from there to establish the Yamato state in Japan. In fact, even Teikan did not claim to find any textual reference to Jimmu leaving Ryukyu and going to Japan. Instead, he claimed that the unspecified text (aru ki ga iu), which was so critical to his analysis in the Shōkōhatsu, had a wormhole in it where the phrase “from Ryukyu” should have been. For Norinaga, the whole thing was an invented farce: The unspecified text did not exist (which is why it was unspecified), as was true of its nonexistent wormhole (Norinaga 1930, p. 278). One begins to see why Norinaga titled his refutation of the Shōkōhatsu in the way that he did.

Norinaga’s refutation, or more precisely, correction, of Teikan’s interpretation of Amamigatake is especially interesting and important. While he does not deny that such a mountain exists on the island of Iheyajima, it was certainly not the point upon which the kami descended from Takamagahara during kamiyo. Instead, he argues that the amami 天孫 referenced in the mountain’s name were not the kami of kamiyo, like Sukunabikona-no-mikoto or Ōkuninushi-no-mikoto, but the imperial descendants of a much later era, namely, the Heian period. Specifically, the term amami in the context of Iheyajima signified the arrival of Minamoto no Tametomo (Norinaga 1930, p. 281). Norinaga found the justification for his very different interpretation of amami from a Chinese source, the Hongjianlu 弘簡録 (Records of Hongjian, 1688), which briefly recounted the story of Tametomo: “In the fourteenth year of Chunxi 淳熙, Shuntien [Shunten] became king. Shuntien was the son of Lord Weichao 為朝公 [Tametomo]…The people suffered epidemics mostly because of Yingzu 英祖 [Eiso]. Yingzu was a descendant of the Tiensun 天孫 [amami/Tenson] clan” (Norinaga 1930, p. 281).7 As an imperial descendant, Tametomo could be described as an amami, Norinaga believes, since the imperial line was thought to have been unbroken since the time of Jimmu, and Tametomo was a direct descendant of Emperor Seiwa 清和天皇 (850–881), the founder of the Seiwa Genji line of the Minamoto clan. Of course, the Chinese passage itself links [King] Eiso to the Tenson, not Tametomo, so Norinaga’s interpretation only works if Eiso was a descendant of Tametomo (which he was not). Ironically, Norinaga had relied on a Chinese source on Ryukyu, rather than on a Ryukyuan one, which he admonishes his readers not to do, and for which he criticizes Teikan; and, since he had not used sources like the Chūzan seikan 中山世鑑 (1650; see Chōshū 1933, 2011) or the Chūzan seifu (see O. Sai 1972; T. Sai 1998), he ended up misinterpreting the Chinese passage (so that rather than the Hongjianlu being inaccurate, his misreading of it made it inaccurate). However, it is important to keep in mind that Norinaga did not question the story of Tametomo siring the first historical (named) king of Ryukyu.

Finally, Norinaga makes two brief comments in the Kenkyōjin that are revealing of his attitude toward Ryukyu. One was directly connected to Ryukyu, while the other was curiously not. Teikan tried to calculate the temporal length of kamiyo in human years, arriving at the figure of 1,792,470 (Norinaga 1930, p. 228). Norinaga argues that the temporal length of kamiyo was simply beyond human comprehension. It was as if someone tried to describe the size in kokudaka of Kaga domain 加賀藩 to someone in Ryukyu (with its roughly 89,000 koku): “A certain [Japanese] person would go to Ryukyu and describe the daimyo of Kaga as controlling lands of one million koku, which a very clever [Ryukyuan] person would hear, before bursting out in laughter, saying that the Japanese tell only lies. This is very much the same kind of thing” (Norinaga 1930, p. 264). Such a hypothetical encounter imagined by Norinaga was not beyond the realm of imagination during the Edo period, since Satsuma essentially controlled the Ryukyu Kingdom as a kind of vassal state. While Ryukyu was certainly much smaller in terms of its kokudaka than Kaga, it was in no way large in terms of its own kokudaka. As a domain, it would have been on the small side, but the fact is Ryukyu was not a Japanese domain in the eighteenth century, even though it became one roughly ninety years after Norinaga finished the Kenkyōjin. However, Norinaga wrote as if Ryukyu was one of Japan’s smaller domains.

Norinaga’s second comment was connected to Ryukyu but only indirectly. When describing the foreign lands created by Sukunabikona-no-mikoto, he leaves the Ryukyu Kingdom off the list: “When looking through the eyes of Ancient Learning [Kokugaku], we come to think that all foreign realms, even India, even China, even Korea, all the various realms got their start, and everything about them, from Sukunabikona-no-kami” (Norinaga 1930, p. 307). In a text in which Norinaga mentions Ryukyu specifically on more than one occasion, it is interesting that he chose not to mention it in this passage, which comes at the very end of his treatise. When considering the fact that he admits that he believed the story of Tametomo and Ryukyu to be true and wrote about Ryukyu as if it were just another domain in Japan, it seems as if Norinaga himself was caught up in the rhetoric of Ryukyu’s geopolitical ambivalence vis-à-vis Japan. This is somewhat surprising, since he was otherwise quite keen on demarcating what was foreign in Japan from what was authentically Japanese. The case of Ryukyu, as described in the Kenkyōjin, seems to have given him a reason to suspend, even if momentarily, that intellectual imperative.

2.3. Akinari and Norinaga: The Kakaika Debate

As Teikan likely knew even as he wrote it, the Shōkōhatsu would elicit responses, and it certainly did, notably drawing the attention of Motoori Norinaga, an intellectual celebrity in his own time. However, there were readers sympathetic to Teikan’s arguments as well, including Ueda Akinari. Like Norinaga, Akinari had made a name for himself as a scholar of Kokugaku, since both men considered themselves to be followers of the teachings of Edo’s Kamo no Mabuchi. Akinari approached the study of Japanese antiquity very differently than did Norinaga, however. As a writer of fiction, Akinari realized that historical writing was an inherently flawed undertaking, as it only reflected an official view at the expense of suppressing all other views, chiefly those of ordinary people. Naturally, he was unimpressed with Norinaga’s insistence that the Kojiki had to be interpreted at face value, and that its narrative of events in antiquity had to be accepted as it was written. Thus, his defense of the Shōkōhatsu quickly evolved into a series of attacks on Norinaga’s scholarship, revealing the fundamental rifts between the methodology and, especially ideology, of Norinaga and his students and those of the rest of Mabuchi’s posthumous students. The works of the two scholars refuting one another were summarized and published together as one work, the Kakaika, in 1790, as people around Japan were eager to assess for themselves whose work had more merit between the two intellectual giants of the time.

2.3.1. Akinari and Kokugaku

Straddling the years before and after his family business went up in flames (from 1766 to 1772), Akinari studied Kokugaku as a student of the prominent scholar, Katō Umaki 加藤宇万伎 (1721–1777) (Fujii 1919a, p. 12). Umaki was born into a bushi 武士 (warrior) household serving the Nagai 永井 family.8 At the age of twenty-five, he became a student of Mabuchi in Edo at roughly the same time as he married a daughter of the Kawazu 河津 family, the heads of which served as the personal physicians to the lords of the Ōgakishinden domain 大垣新田藩, the Toda 戸田. Following the death of his wife, he was adopted into a family of bakufu retainers, the Katō, spending the rest of his adult life as a kinban 勤番 (“rotating official”), which required him to spend periods of time serving at both Osaka Castle and at Nijō Castle in Kyoto. Although Mabuchi had many students in Edo, Umaki stood out as one of his very brightest, and people later spoke of him as one of Mabuchi’s four greatest disciples, what they referred to as the shitenō 四天王 (“Four Heavenly Kings”), along with Katō Chikage 加藤千蔭 (1735–1808), Murata Harumi 村田春海 (1746–1811), and Katori Nahiko 楫取魚彦 (1723–1782) (Fujii 1919a, p. 12). Akinari met with Umaki during his stints in the Kamigata region and, when Umaki returned to Edo for good in 1772, the two exchanged letters until Umaki’s death about five years later. Thereafter, Akinari became an “independent” scholar, refusing to attach himself to another Kokugaku mentor (Fujii 1919a, p. 18).

Mabuchi is considered one of the greatest scholars of the Kokugaku tradition, along with his predecessor, Keichū 契沖 (1640–1701), and also his self-styled “successor,” Motoori Norinaga.9 Both Keichū and Mabuchi focused their efforts mostly on analyzing the verses in the ancient poetic anthology, the Man’yōshū. Mabuchi was convinced that the Man’yōshū was a record of the sincere feelings of the ancient Japanese people, before the foreign influences of Buddhism and Confucianism introduced artifice and deception, thereby corrupting what he called the magokoro (“true heart”) of the ancients (Tatsuo 1981, p. 136; Marceau 2004, p. 128). Inspired by Ogyū Sorai 荻生徂徠 (1666–1728), perhaps Tokugawa Japan’s most famous Confucian scholar, Mabuchi sought to recreate the conditions of antiquity in his own day in order to recapture the magokoro of the ancients, which he admonished his students to demonstrate within their own ancient-style waka poetry. After his death, his students, who comprised what Japanese scholars have called the Kenmon 県門 (or Agataimon 県居門, which I refer to as the Mabuchi School), vowed to continue to propagate Mabuchi’s teachings on the subjects of ancient history and literature, always with an emphasis on poetry. Although Akinari never met Mabuchi, he thought of himself as an adherent of his teachings via the guidance and mentorship of Umaki.

The fact that Akinari was an accomplished poet, in addition to his achievements in scholarship and in literature, speaks to the emphasis on poetry composition that was so vital to the collective identity of individuals within the Mabuchi School (Marceau 2004, p. 8). It also helps to explain how and why Akinari’s views of ancient Japan diverged so radically from Norinaga’s, even though Norinaga considered himself a fellow adherent of Mabuchi’s teachings. Members of the Mabuchi School engaged in the intense study of ancient waka for the purpose of composing their own waka; in other words, they were supposed to be both scholars and poets. The majority of Kokugaku scholars were “amateurs,” in the sense that they pursued their studies, composed their poems, listened to (or gave) lectures, and wrote their books, all in their spare time. With very few exceptions, and Mabuchi was one of them, they were not paid in any official capacity for their scholarly efforts. Some scholars, like Umaki, had jobs that were unrelated to their Kokugaku work, while others, like Akinari for a brief time in his youth, had family money to support their studies. As serious and committed as these amateur scholars were, an intellectual subculture arose for such people during the Edo period, namely, that of the bunjin 文人 (“literati”). The bunjin were people (usually, but not always, men) who dabbled in several intellectual and artistic pursuits, notably painting, calligraphy, prose writing, and poetry composition. Although there were Confucians in the ranks of the bunjin, elite Confucian scholars were often employed in some capacity, either at the national (Edo bakufu) or provincial (or local) levels, for their scholarly expertise. This was mostly unheard of for the majority of Kokugaku followers, and so a pairing of Kokugaku with the subculture of the bunjin was perhaps natural, and this was especially true of the mostly Edo-based members of the Mabuchi School.

As a writer, poet, and Kokugaku scholar, Ueda Akinari not only exemplified the prototypical Edo-period bunjin, he was also one of its most outstanding figures.10 Although he seemed to have “changed careers” at various points in his life, emphasizing haikai 徘徊 poetry in his youth, then exploring ancient waka, then turning his attention to ancient history later in life, he always maintained his varied interests without abandoning any of them. To a certain extent, Norinaga was the rural counterpart to the urban bunjin like Akinari and Umaki. Norinaga was also a poet and scholar, like Akinari, but he believed that one had to be undertaken in the service of the other, namely, that the composition of ancient poetry was useful for scholarship on antiquity, especially the apprehension of what he called the kodō 古道 (“ancient way”), the cultural essence of ancient Japan. Norinaga believed that the elucidation of the kodō was congruent with Mabuchi’s teachings. For the most part, his Mabuchi School colleagues did not agree with either Norinaga’s intellectual hierarchy or his insistence on the importance of the kodō, believing, in fact, that Norinaga was misguided, and they criticized him for abandoning Mabuchi’s teachings. This disagreement marked the beginning of a split within the Mabuchi School between those who emphasized the primacy of ancient verse, as they maintained their bunjin identities, and those who supported Norinaga’s idea of the kodō, the numbers of which eventually grew to overtake those who had supported Mabuchi, a fact that also irritated Akinari about Norinaga.11

The issue of the kodō was one of the sources for the ideological split within the Mabuchi School. Another source of friction among its members was the issue of anti-foreignism, one for which Akinari and others expressed some ambivalence, but which Norinaga famously espoused. Mabuchi’s criticisms of Buddhism and Confucianism, and his general refutations of Chinese culture are famous, especially as articulated in his essay, Kokuikō 国意考 (Thoughts on the Meaning of the Realm; 1765). As foreign forms of knowledge, Buddhism and Confucianism had, for Mabuchi, obscured the pure and innocent minds of the ancient Japanese people, in an era when they had governed themselves in accordance with the will of the kami (namely, Shinto) and with nature in general. Mabuchi proposed a kind of archaism as the antidote to the cultural ravages of Buddhism and Confucianism in order to recover the magokoro of the ancients.

Akinari and other leading figures of the Mabuchi School, such as Murata Harumi and his student, Oyamada Tomokiyo 小山田与清 (1783–1847), did not interpret Mabuchi’s views of Buddhism and Confucianism as utter condemnations of India or China, respectively, or even of their cultural traditions. None of them were especially anti-foreign, but they were all strongly in favor of composing the best ancient-style waka, like those that were reminiscent of the Man’yōshū, that they could (Uchida 1981, p. 93). Norinaga was, however, more ideologically intolerant and doctrinaire than his colleagues in the Mabuchi School, like Akinari, were.

2.3.2. Akinari’s Views in the Kakaika

Even though books were published at an unprecedented historical pace during the Edo period, for the most part people borrowed books more often than they purchased them. Usually, scholars hand copied the books they borrowed, which gave them working copies for their own personal libraries, and this was the case with Akinari’s intellectual encounter with Norinaga’s works. Akinari’s friend and fellow waka enthusiast, Arakida (Kikuya) Suetomo 荒木田 (菊屋) 末偶 (1736–1802), joined Norinaga’s private academy, the Suzunoya 鈴屋, in 1784. As a member of the Suzunoya, Suetomo had access to Norinaga’s works, likely as reading materials for meetings at the academy, and he lent some of them to Akinari (whether at Akinari’s request or not is unclear). Specifically, Akinari managed to obtain copies of volumes twelve and thirteen of Norinaga’s Kojikiden, a work that Norinaga would continue to write for the next fifteen years, and also his Gyojū gaigen (also Karaosame no uretamigoto) 馭戎慨言 (Mourning the Mastery of the Foreign; 1778) (Fujii 1919a, p. 22). This was also about the time that he came across both the Shōkōhatsu and Norinaga’s Kenkyōjin, which appeared shortly after he borrowed Norinaga’s works from Suetomo. He brushed his response to Norinaga’s scholarship in a work entitled, Ueda Akinari ronnan dōben 上田秋成論難同弁 (Ueda Akinari’s Refutation and Reply), which he sent to Norinaga (Fujii 1919a, p. 23) and which the latter published as the first part of Kakaika in 1788, followed by a specific refutation of the Kenkyōjin, called Kenkyōjin Ueda Akinari hyō dōben 鉗狂人上田秋成評同弁 (Ueda Akinari’s Evaluation and Reply to the Kenkyōjin), which he also sent to Norinaga, who then added it as a second part to the Kakaika in 1790.

As Akinari’s attention was drawn to the Shōkōhatsu and also to Norinaga’s response, as well as to Norinaga’s scholarship in general, there was an unexpected (what we would call archeological) discovery in the Shikanoshima 志賀島 area of the province of Chikuzen 筑前国 in 1784 that had a profound impact on his thinking. A golden seal for the king of Wanona (Wanona no kokuō no kin’in 委奴の國王の金印), with an inscription that indicated it had been conferred by the Han emperor (Guangwu 光武帝 [5 BCE-57CE]) of China, was excavated, and Akinari felt compelled to find out in what ways this discovery confirmed or denied what he and his contemporaries knew (or thought they knew) about Japanese antiquity.12 He immediately composed his historical analysis of the seal in his Kan no Wanona no kokuō kin’in kō 漢委奴國王金印考 (On the Gold Seal of the Han for the King of Wanona; 1784), in which he argued that the seal was an authentic remnant of Japan’s ancient past (Fujii 1919a, p. 21). Although Akinari did not agree with Teikan’s conclusions in the Shōkōhatsu (Burns 2003, p. 110), he at least was sympathetic to Teikan’s willingness to look beyond the textuality of the kamiyo accounts and uncover the materiality of Japanese antiquity, and his intense interest in the golden seal likely solidified in his mind that his sympathies for material inquiries into the past were legitimate.

Although Akinari was skeptical of Norinaga’s scholarship in general, there were three particular areas upon which he concentrated his criticisms. The first of these fundamental disagreements between the two titans of Kokugaku was connected to the issue of phonological (onbin 音便) changes to the Japanese language over time. Norinaga had famously argued that there was no sound in pre-Nara Japanese for the hiragana character ん (“n”), so that words that contained the sound “n” in Tokugawa Japanese had ancient counterparts that contained the sound む (“mu”). Akinari ridiculed this view: “To say that there was no sound for “n” in the language of the ancients is an astonishingly subjective thing. So, he teaches that shinpū (kankaze?) 神風 should be read as kamukaze 加牟加是, etc.” (inishie no hito no gengo ni n no oto nashi to iu wa, watakushi no hanahadashiki mono nari, shinpū/kankaze wo kamukaze to yomubeshi to oshiemu ni unun 古の人の言語に「ん」の音なしといふは、私の甚しき物也、神風を加牟加是と讀むべしと教へむに云々) (Norinaga and Akinari 1926, p. 61). For Norinaga, the ancient Japanese created native words, using native sounds, yet they yielded to the limitations imposed by written Chinese. Akinari believed that the ancient Japanese were not as constrained by the use of written Chinese as Norinaga thought, and they wrote their native words phonetically in Chinese characters anyway, even the ones with the “n” sound, as he “questioned the assumption that writing was capable of accurately and minutely recording the sounds of the ancient language” (Burns 2003, p. 108). In this way, Akinari tried to deny what for many intellectuals at the time thought was one of Norinaga’s signature linguistic breakthroughs.

A second contentious issue between the two intellectuals was Norinaga’s assertion that Amaterasu was not only the kami of the sun, but the sun itself. Akinari demonstrated how Norinaga’s view was based on a flawed reading of the Kojiki; Norinaga misunderstood a key term in the relevant passages about Amaterasu, which led him to think that Amaterasu’s dominion over Japan meant dominion over the world, since the sun’s light shines over the whole world: “On matters related to the kami of the sun, how can [Amaterasu] shine over the four seas and the myriad realms?...[The kamiyo passages] refer to rikugō 六合 as signifying the four directions of Heaven and earth (tenchi shihō 天地四方), but it was applied to the august realm, and did not mean the four seas and the myriad realms…Moreover, in what texts do we see that [Amaterasu] shines over not just the august realm but over all foreign realms in Heaven and earth?” (Norinaga and Akinari 1926, p. 88). Although Akinari does not match Amaterasu to some actual historical figure, as some Tokugawa intellectuals did by claiming that Amaterasu was Wu Taibo, he is clearly sympathetic to the idea that Amaterasu had a material existence in Japan in remote antiquity. In this way, his indignation over Norinaga’s refutation of the Shōkōhatsu becomes easier to understand. At the same time, one can more readily see how Akinari’s observation was felt so acutely by Norinaga, since Akinari was doing nothing less than impugning Norinaga’s philological interpretation of the ancient sources.

The third source of friction and rivalry between the two, as recorded in the Kakaika, is related to the status of Amaterasu as the sun. For Akinari, to claim that the sun had some original connection only to Japan made Japan more cosmically central than could possibly be the case in reality:

Even for realms without writing, they have their own fantastic legends, and [so] we cannot ignore foreign matters. The people of this realm must rectify their steadfast belief in the fantastic legends of antiquity like those of the Great Man [Norinaga]. Theories that have been propagated about other realms, such as those in the Gyojū gaigen, are things we must approach with a skeptical eye. Classical texts [that speak of] one certain realm [as the head of] one world, inclusive of other realms, with analogous evidence from [their own] proverbs, are impossible for both sides [native and foreign] to accept.

(Norinaga and Akinari 1926, p. 89)

In other words, Akinari argues that foreign peoples have their own views of the cosmos and creation that must also be considered; the Japanese people, even those of remote antiquity, had no special knowledge of or insight in such matters. Such a remarkably cosmopolitan view is surprising coming from such a renowned Kokugaku scholar, since Western scholars have categorized Kokugaku as Tokugawa Japan’s version of “nativism.” It is possible that Akinari was merely the lone figure within Kokugaku for espousing such views, or, more likely, that there was a cosmopolitan strain that ran through the writings of others within the Mabuchi School. If the latter, then this would constitute even more proof that the blanket identification of Kokugaku with nativism is problematic.

Norinaga, on the other hand, held a view that was decidedly in favor of “[Japan’s] imperial supremacy” (kōkoku shijō kan 皇国至上観), and Akinari was determined to expose him for it (Tsutsumi 1981, p. 77). Specifically, he attacks another of Norinaga’s signature concepts, the Yamato damashii, a view that was inspired, Akinari observes, by Sinocentrism:

Having to decide which is better, robes on the left (sajin 左衽) [Japanese-style] or robes on the right (ujin 右衽) [Chinese-style], is certainly not something we should have to do. [We think that] how could there not be a reason why, when all other realms wear their robes on the left, that only one realm wore them on the right? If we were to investigate the customs of the myriad realms, then we might [just] discover that most of them [actually] wear their robes on the right, and that there were never any legends about only one realm that wears its robes on the right. So, to the extent that we seek to place the imperial realm above the myriad realms, we should also denounce the Way of China that reveres [only] itself by using this insight. The bias (katayoru toki wa 偏よるときは) within the Nihon [or Yamato] damashii 日本魂 is the same as the mind [found within] Chinese books (Kanseki’i 漢籍意).

(Norinaga and Akinari 1926, p. 93)

For Akinari, Norinaga had thought that he had rid himself of the Chinese mind only to stand that concept on its head by asserting the primacy of the Yamato damashii. Akinari’s response to both the Chinese mind and the Yamato damashii was to abandon hierarchical thinking itself, since the logic of pitting one against the other made no sense to him. Akinari had a rather straightforward explanation for Norinaga’s perplexing views. Norinaga’s interpretations were eminently provincial because he lived in the provinces, whereas Akinari’s more nuanced and sophisticated (in other words, cosmopolitan) views were directly linked, in his own mind, to his urban upbringing and residential setting:

When one looks through a one-thousand-league mirror, called a zonglas [telescope], [one can tell that] the sun is very hot, and that the moon is boiling, and so [see that Amaterasu] is not like [either of them]. That old country bumpkin’s (inaka hito no futokoro oyaji 田舎人のふところおやぢ) theory can only be believed by the country people (inakamono) who hear it. Someone from the capital hears only disrespect for the king [emperor] (ōsama no fumenboku 王様の不面目).

(Tandai shōshinroku, quoted in Fujii 1919a, p. 23)

Akinari explained the rather wide ideological gulf that separated him from Norinaga by invoking the fact that he himself was a city dweller, while Norinaga was not. In this way, he could rationalize to himself why Norinaga had so many students, since most (but certainly not all) of them were also rural folk like Norinaga. Akinari also seems to imply that he also had a legion of supporters, especially fellow urbanites like himself, even if they were not necessarily his students.

2.3.3. Norinaga’s Responses to Akinari

Akinari’s pointed and legitimate observations and criticisms notwithstanding, Norinaga did not alter his positions on the language, culture, and ancient history of Japan, positions that had made him one of Tokugawa Japan’s scholarly authorities. While he and Akinari both thought of themselves as adherents of Mabuchi’s foundational teachings for Kokugaku, it was clear to Norinaga that Akinari never really understood what those teachings meant and so was never able to recover the magokoro of the ancients. Instead, he remained a prisoner of the Chinese mind or Karagokoro, a sad truth of which Akinari was unaware (Norinaga and Akinari 1926, pp. 90–91). This was the same criticism that Norinaga had made against Teikan as well, even though Akinari’s conclusions were not in complete agreement with Teikan’s. Norinaga cites a prime example of Akinari’s Chinese mind at work when Akinari uses Japan’s relatively small size as proof that it is not the world’s most supreme realm. Akinari asks how it could be possible for a realm that was so much smaller than others, notably nearby China, have cosmic truths revealed only to its people and to no others, when surely other peoples, living in larger realms, thought the same of themselves? For Norinaga, the validity of this revelation was never meant to be understood by people. In a way, the proof of Japan’s status as the world’s “foundational leader” (ganpon sōshu 元本宗主), despite its small size, was precisely because it seemed to make no sense at all. The fact that Akinari sought a rational explanation for this fact was the problem, and that was the irrefutable proof that his Chinese mind was at work. Norinaga, however, does invoke history to bolster his argument even further. Japan, he describes, has “never been violated by foreign realms” (gaikoku ni okasarezu 外国に犯されず), not even by the Mongols (Norinaga and Akinari 1926, p. 91). Humans could simply not understand how such a small realm, in very close proximity to a larger and more powerful one, could successfully resist subjugation for the entirety of its history, without invoking Japan’s divine dispensation, the same dispensation that had made it the world’s “foundational leader.”

Norinaga was puzzled and disturbed by Akinari’s refusal to recognize Japan’s superiority in the world, and the superiority of authentic (archaic) Shinto over the foreign creeds of Buddhism and Confucianism (Burns 2003, pp. 121–122). Akinari agreed with Norinaga that the introduction of Buddhism and Confucianism in antiquity had changed the values of Japanese elites in certain ways that were not necessarily beneficial for the realm as a whole (Tatsuo 1981, p. 136). However, this fact did not mean that the two foreign intellectual imports were in and of themselves pernicious, or that the cultures and peoples who had produced them were morally inferior. Norinaga met this nuanced attitude of Akinari with his own absolutist one, namely, that there was nothing redeeming about either Buddhism or Confucianism, and that his wholesale rejection of them was congruent with Mabuchi’s own criticisms. While he spent most of his efforts refuting the stories and legends of the Indians and the Chinese, the Europeans, as the people associated with the ever increasingly popular Rangaku, had their own as well and theirs were no more valid: “Every realm [in the world] has its own legends (densetsu 伝説), but they are all wrong, [as they] conceal their imperfections and vainly set out to deceive their stupid peoples. To use an [example] from realms without Chinese writing, it is like the Catholic Church which is revered in the realms of the far west: They are all fabricated deceptions” (Norinaga and Akinari 1926, p. 91).

Norinaga countered Akinari’s rejection of the interpretation that Amaterasu was the sun, by appealing to one’s own firsthand observations, with the stern admonition that the stories of kamiyo in the Kojiki had to be understood literally and not be read metaphorically. Akinari mistakenly expected to find in the kamiyo stories, relying mostly on the Nihongi instead of the Kojiki, a reference to Amaterasu as the sun, which he did not, and so concluded that such an identification of one with the other was untrue (Norinaga and Akinari 1926, p. 89). Instead, the key to unlocking the secrets of the stories of kamiyo was the application of a careful, philological method to them. It was only through such approaches and techniques that the sublime truths of kamiyo could be revealed, according to Norinaga, whose view of history was closely tied to his reliance on the connection between it and documentary, textual evidence. Of all the important intellectual divergences between these two great Kokugaku scholars, this fundamental difference in their historiographical views was perhaps the biggest one, and certainly the most relevant one to Japanese views of Ryukyu in the Edo period.

2.3.4. Akinari’s Literary Approach to History

Although Akinari studied ancient Japanese history under Umaki’s tutelage, he was drawn more to the subject of waka at that time in his life than anything else. It was not until he reached middle age, namely, after the age of forty, that he turned his attention toward a more serious look at Japanese history (Tatsuo 1981, p. 135). He was doubtlessly aware of Mabuchi’s own interest in the Kojiki late in life, but Akinari was still somewhat skeptical of the enigmatic text, viewing the Nihongi as the more reliable of the two ancient histories of Japan (Burns 2003, p. 102). About a decade after he began looking into Japan’s ancient period, his debate with Norinaga began, drawing the attention of other scholars with Norinaga’s composition of the Kakaika. Thus, Akinari’s historiographical outlook took shape in these three stages, namely, pre- and post-debate, with the Kakaika in the middle.

Akinari viewed the Nihongi as a prime example of orthodox historical writing, and so it was a somewhat accurate representation of the way events unfolded in the past in the eyes of those who had compiled it, a team led by the son of Emperor Tenmu 天武天皇 (?–686), Prince Toneri 舎人親王 (676–735) (Fujii 1919a, p. 28). The problem was, however, that the Nihongi was obviously biased in favor of Emperor Tenmu, since it was compiled by his own descendants, which meant that there must have been other narratives, and likely other stories, that were suppressed in the compilation of the Nihongi (Burns 2003, p. 114). This was also true of the Kojiki, which Akinari dismissed as even more biased than the Nihongi, since it was based on the recollection of one individual, Hieda no Are 稗田阿礼 (?–?), and its transcription by another, Ō no Yasumaro 太安万侶 (?–723). Consequently, both the Kojiki and the Nihongi were flawed (Burns 2003, p. 118; Tatsuo 1981, p. 137), because their success as orthodox histories came at the expense of the elimination of alternative narratives, chiefly those of non-elites.

The elites of antiquity had the obvious advantage of literacy, which non-elites did not have, and so they were easily able to assert their own historical narratives over those of commoners because of it, even if they had to compete with one another to do so. Akinari was not only skeptical of the content of these elite historical narratives, but he also felt the same way about their written forms, expressing serious doubts that historical writing was able to capture the fullness and complexity of the past. In a way, the Nihongi had a certain intellectual integrity to it that the Kojiki did not have, because it was composed in literary Chinese, which made its putative representation of Japan’s ancient past rather obviously flawed, a fact upon which Norinaga and his students had seized. On the other hand, the problem with the Kojiki was that it was supposed to be the transcription of the spoken language of Japanese antiquity, which Norinaga argued brought the reader closer to the minds of the ancients, but which Akinari thought was simply Norinaga’s own naïve illusion brought on by his misplaced faith in philology (Burns 2003, p. 115). For Norinaga, philology was the solution that unraveled the mysteries of the Kojiki, which was no solution at all for Akinari. Orthodox historical writing was the problem, in Akinari’s mind, the solution to which was to apprehend the past in different ways.

Akinari wrote the fictional work for which he is best known, the Ugetsu monogatari (1767), before his intellectual sojourn into ancient history, yet while he was actively studying Kokugaku poetics under Katō Umaki. As a collection of fantastic stories, replete with ghosts and monsters, the Ugetsu monogatari represents Akinari’s exploration of themes that held for him a special interest, namely, the supernatural and history, and in the case of the latter, the Heian era in particular (Tatsuo 1981, p. 136). For him, ghosts were real, a phenomenon that affected everyone, elite and non-elite alike, and the same was true of the kami, both of which were spiritual beings with connections to living, human beings (Fujii 1919a, p. 14; Fessler 1996, p. 5). Encounters and experiences with the spiritual were an important aspect of being human, and so Akinari viewed his fictional accounts in the Ugetsu monogatari as one ordinary person’s emotional expression of their humanity (Matsuda 1981, p. 24). This accounts for why the story, “Shiramine,” is so important within the Ugetsu monogatari, since it allowed the ghost of Emperor Sutoku, a tragic figure in Japanese folklore during the Edo period, to speak for himself (Ueda and Whitehouse 1938, p. 255, n. 2), something that was unheard of within orthodox histories.

The Ugetsu monogatari is a masterpiece of a genre of Tokugawa literature known as the yomihon. The genre was characterized by its historical and supernatural themes, as well as its exploration of moral ambiguity (Marceau 2004, p. 127, n. 31; p. 128). Akinari’s work embodied all three of these aspects especially well, since he felt that the genre of the yomihon gave him the creative freedom to demonstrate the serious ideas that he had, but in a way that defied the intellectual standards of the day. As a result of his debate with Norinaga, Akinari decided that fictional writing was the dynamic answer to the issue of the staid historical writing that had produced works like the Kojiki and the Nihongi (Tatsuo 1981, p. 137; Marceau 2004, p. 128).13 In the words of Susan Burns, Akinari “attempted to make fiction the form of writing that allowed for the questioning of the ‘official’ representations of the social and political world” (Burns 2003, p. 125). He applied these insights to his other great yomihon, the Harusame monogatari 春雨物語 (Tales of Spring Rain; 1808/9), which he completed shortly before his death.

3. Conclusions

Akinari’s confidence in fictional writing as an alternative to the biases inherent in orthodox histories became an important intellectual and methodological contribution to the ways in which Japanese views of Ryukyu developed in the Edo period. Akinari’s literary approach to the imperial histories was an offshoot from that of Teikan, who embraced a more material and historical position, but Norinaga’s literal and scriptural approach to the imperial texts was opposed to both. Akinari’s works, especially the Ugetsu monogatari, exerted a considerable influence on contemporary and subsequent scholars and writers, including Kyokutei Bakin, whose nineteenth-century work, Chinzei yumiharizuki, was another masterpiece of the yomihon genre, but one which dealt with Ryukyu directly (Fujii 1919a, p. 13; Marceau 2004, p. 128). Akinari’s Kakaika debate with Norinaga played no small role in nudging Akinari to his conclusion about the efficacy of fictional writing, but it was Tō Teikan’s Shōkōhatsu that had spurred Akinari into openly expressing his reservations about Norinaga’s ideas in the first place. Teikan’s work placed Ryukyu in a very prominent historiographical position, but this centrality eroded with the emergence of Norinaga’s Kenkyōjin, disappearing entirely by the end of the Kakaika debate.

The fact that Teikan even included Ryukyu at all was a sign that Arai Hakuseki’s works still held sway during the second half of the eighteenth century. By the time the Shōkōhatsu appeared, Ryukyuan processions to and from Edo had been a fact of life for more than a century and a half, generating excitement among the Japanese people in general, but also curiosity among intellectuals in particular. While the former were able to indulge themselves in the spectacle of Ryukyuan foreignness, the latter were confronted with the Ryukyuan cultural ambivalence that Hakuseki had described. To figures like Teikan, Ryukyu was already sort of Japanese, a view that perhaps is not especially striking in and of itself. However, the fact that it drew a response from Norinaga, who ended up affirming Ryukyuan historical ties to Japan, chiefly in the form of Tametomo’s siring of Shunten, does seem to invite at least some explanation, since Norinaga was the key figure in Kokugaku, a movement that was founded on the distinction between what was native and what was foreign. Even Norinaga found nothing especially controversial about Ryukyu, and he was evidently comfortable with the kingdom’s geopolitical position between Japan and China, and its cultural position between the native and the foreign.

The Kakaika debate was important in the development of Japanese views of Ryukyu, not only because it encouraged Akinari to develop an alternative historical epistemology, but also because it demonstrates how Ryukyuan matters were not a source of friction between Norinaga and Akinari. The two Kokugaku scholars found ample material upon which to disagree; the list of points of agreement between the two was much shorter than the list of contentious matters between them. The histories of China and the Korean peninsula, as the other two foreign realms mentioned in the Shōkōhatsu, generated a range of problematic issues that arose in the debate between Norinaga and Akinari, but the same was not true of Ryukyu. Given how much the two intellectuals seemed to dislike one another, especially Akinari’s contempt for Norinaga (Tsutsumi 1981, p. 77), one would think that one of them would have raised a Ryukyuan issue to disagree with the other if such an issue had ever presented itself, as was the case with issues connected to the histories of China and the Korean peninsula. Since no controversy over Ryukyu ever emerged between them during their debate, one might reasonably conclude that there were no disagreements between the two over Ryukyuan matters, whether serious or trivial. So, if Norinaga seemed at ease with Ryukyu’s ambivalent status, then Akinari likely was as well.

Just as Hakuseki had juxtaposed Ryukyu with Ezo, inviting subsequent generations of scholars to develop connections between the two more fully, the debate between Norinaga and Akinari tied Ryukyu to China and to the Korean peninsula. Rather than encourage comparisons between China and Ryukyu, their debate pushed Ryukyu and the Korean peninsula into the same conceptual category, namely, as close foreign Others set against China as the more distant foreign Other. It likely seemed natural to Tokugawa scholars, intellectuals, and government officials that Ryukyu and the Korean peninsula belonged in the same category as foreign Others, since envoy processions from both had been arriving in Japan since the first half of the previous century, fundamentally distinguishing them from China. While the Kakaika debate stimulated an interest in more closely analyzing the conceptual connection between the Korean peninsula and Ryukyu, a connection that had existed in an unexamined way for more than a century, it was not until Ban Nobutomo, laboring more than a generation later on a history of Japanese foreign interactions, that a rigorous comparison between the Korean peninsula and Ryukyu finally emerged (see Ban [1838] 1988). In this way, the paradigm of the Korean peninsula as a model by which to compare Ryukyu gave way to one in which Ryukyu became the model for Chosŏn Korea during the Meiji period.