Ethnography of the ‘Southern islands’: Arai Hakuseki,Ryukyuan diplomacy,and the making of ethnological knowledge in early modern Japan

Introduction: Knowledge, power, and the ethnographic gaze

In the first decades of the 18th century, Japan entered a moment of intensified curiosity about its surrounding world. Behind the apparent stability of the Tokugawa regime, a new mode of observing and classifying the foreign was taking shape—one that transformed diplomatic encounter into an epistemic act. Among the leading figures of this transformation was Arai Hakuseki (1657–1725), the Confucian scholar and senior political advisor whose writings on the Ryukyuan Kingdom redefined what it meant to “know” the other in early modern Japan. His Nantōshi (Records of the Southern Islands, 1719) was far more than a geography of a tributary domain located in the East China Sea, which functioned as a vital maritime and cultural crossroads. It was a project of knowledge and governance, one that revealed how the shogunate’s bureaucratic order and Confucian moralism converged to produce what might be called an early modern ethnographic imagination.

This paper explores Hakuseki’s Ryukyuan studies as an instance of ethnographic writing before the advent of anthropology. Rather than viewing Nantōshi as a descriptive curiosity or as a mere reflection of foreign policy, it treats the text as part of a broader regime of power and knowledge through which the Tokugawa state made sense of the world. The Ryukyuan Kingdom—politically subordinate yet culturally sophisticated—provided a unique mirror for Japan’s own self-definition. In describing Ryukyu’s geography, language, and customs, Hakuseki was not only compiling information but also constructing a moral hierarchy. His ethnographic gaze rendered difference intelligible and governable, positioning the southern islands simultaneously as Japan’s ancient kin and its civilizational junior. However, it is crucial to qualify that this systematic ordering was not merely a benign intellectual exercise; it enacted a form of structural and epistemic harm by subordinating Ryukyuan realities to Japanese geopolitical and discursive ambitions.

Modern scholarship has long recognized Nantōshi as a milestone in the development of Japanese knowledge about Ryukyu. Early interpreters such as Iha Fuyū, the pioneering Okinawan historian, read the text as a precursor to Ryukyuan studies, emphasizing its claim that Ryukyu descended from ancient Japan—a theory that later served nationalist purposes. Subsequent historians, notably Miyazaki Michio and Arano Yasunori, situated Hakuseki’s work within the context of Tokugawa foreign policy, underscoring its role in articulating a “Japan-centered world order.” More recent scholarship has expanded this frame. He Ziyi has traced how the spectacle of the Ryukyuan embassies (Edo nobori) in 1710 triggered a popular “Ryukyuan boom” in Edo, while Ronald Toby has interpreted these spectacles as performative affirmations of Japan’s status within a modified Sinocentric order. Building upon these foundations, emerging scholarship from the past 5 years has increasingly highlighted the intersection of Tokugawa knowledge practices and early modern classificatory regimes, re-examining the Ryukyuan embassies not merely as diplomatic spectacles but as crucial engines for standardizing ethnological knowledge. Yet despite their richness, these readings tend to treat Nantōshi as the product of political ideology rather than as a form of knowledge in its own right.

To understand Hakuseki’s project anthropologically requires a shift of perspective—from intellectual history to the anthropology of knowledge. This approach asks not simply what he knew, but how he knew: how information was gathered, organized, and authorized; how acts of observation became textual evidence; and how the very concept of “the other” was manufactured within bureaucratic and moral discourses. Consequently, this article reframes Nantōshi as an instance of knowledge production tied directly to Tokugawa governance and moral ordering. To make this thesis explicit, this study is guided by one central research question: How did Arai Hakuseki’s textualization of Ryukyu function as a technology of early modern Japanese statecraft? Furthermore, it addresses three sub-questions: (1) What specific textual mechanics were employed to organize and subordinate Ryukyuan difference? (2) How did the bureaucratic process of gathering information during embassy visits shape the ethnographic output? and (3) How does this early modern case complicate our understanding of the history of ethnography? To maintain conceptual clarity, it is equally important to state what this article does not attempt. It does not seek to reconstruct Ryukyuan society “as it actually was” in the 18th century, nor does it offer an exhaustive diplomatic history of Tokugawa-Ryukyu relations. Instead, the scope is strictly focused on the discursive and epistemological frameworks through which Japan imagined its southern neighbor.

Hakuseki’s inquiries into Ryukyu’s institutions, genealogy, and customs drew upon interviews with visiting envoys such as Zheng Shunzu and Prince Mishin, as well as upon earlier embassy diaries and Chinese gazetteers. But his ethnography was less empirical reportage than a disciplined synthesis of observation and moral reasoning. The foreign, for Hakuseki, was never wholly outside Japan’s order of knowledge. It was a reflection—distorted yet instructive—of what Japan imagined itself to be.

The theoretical frame for such an inquiry draws inspiration from Michel Foucault’s notion of power/knowledge, James Clifford’s critique of Writing Culture, and Johannes Fabian’s concept of temporal othering. Foucault’s insight that every regime of truth is bound to systems of power illuminates the political nature of Hakuseki’s scholarship: his description of Ryukyu was also an act of governance. The ethnographic gaze he deployed was inseparable from the bureaucratic technologies of observation that defined Tokugawa statecraft. Clifford’s analysis of ethnography as a rhetorical practice helps to clarify how Nantōshi constructs authority through its textual strategies—its careful use of citation, classification, and moral commentary. What Hakuseki produced was not a neutral ethnographic record but a morally charged narrative that fused the empirical with the didactic. Fabian’s notion of temporal othering, finally, offers a way to interpret Hakuseki’s claim that Ryukyu represented “ancient Japan” (konan wa). By locating the islands in Japan’s archaic past, he simultaneously included and excluded them—making Ryukyu at once ancestral and subordinate, familiar and primitive. The other was not elsewhere in space, but earlier in time. Where broader comparisons are made in this study, such as tracing parallels between Tokugawa knowledge practices and Enlightenment epistemologies—these are intended to highlight conceptual resonances rather than to suggest exact equivalence or direct influence. Acknowledging these limits is essential to avoid flattening the distinct contours of early modern East Asian thought.

In this light, Nantōshi can be read as a textual technology of order, an ethnography whose structure mirrors the moral geography of the Tokugawa world. Each section—on geography, language, polity, customs, and etiquette—reflects a classificatory impulse to translate difference into hierarchy. The text’s tone oscillates between curiosity and correction: Hakuseki admires Ryukyu’s ritual propriety while lamenting its dependence on “barbarian customs.” Such moral evaluation is not incidental; it is integral to his method. As Kate Wildman Nakai has observed, Hakuseki’s Confucianism transformed moral virtue into a political science of the world. In describing others, he reasserted the universality of the Confucian order—now re-centered upon Japan.

Seen against the backdrop of global intellectual history, Hakuseki’s work reveals striking parallels with the ethnographic impulses of the Enlightenment. It is necessary to state that these regional developments were neither directly influenced by nor strictly equivalent to European paradigms. Just as Jesuit missionaries in China and travelers in the Pacific sought to document and classify human diversity, Hakuseki turned empirical observation into a moral inquiry about civilization itself. Yet unlike European ethnographers, he operated within a world that lacked an overseas expansionist colonial horizon. His ethnography was not an account of conquest but of inclusion. It was an attempt to fold the periphery into a cosmic order defined by decorum, hierarchy, and virtue. In this sense, the Nantōshi anticipates what might be termed an alternative mode of empire-building, one that disciplined alterity through the rhetoric of moral kinship rather than racial difference. However, to romanticize this process as entirely “noncolonial” or as an “order without violence” would be misleading. By systematically reclassifying Ryukyuan customs to fit Japanese ideological needs, Hakuseki’s moral ordering enacted its own form of structural and epistemic harm, effectively stripping the islanders of their autonomous historical agency.

The implications of this perspective extend beyond Hakuseki himself. The Edo nobori of 1710, which first exposed him to the Ryukyuan envoys, also transformed Edo into a laboratory of ethnographic vision. Crowds of townspeople gathered to watch the procession, artists produced illustrated scrolls of the event, and publishers flooded the market with guides to foreign customs. The spectacle of diplomacy became a medium through which Japan learned to see itself in relation to others. Out of this culture of seeing emerged the bureaucratic ethnography that Hakuseki would later codify in text. Knowledge, power, and spectacle were mutually reinforcing. In this political landscape, to govern was to know, to know was to see, and to see was already to order.

This study builds upon that insight. It treats Nantōshi as both a product and a performance of the Tokugawa knowledge regime, serving as a text through which the shogunate’s moral universe found ethnographic expression. Methodologically, the analysis combines close textual reading with the theoretical vocabulary of anthropology, treating the Nantōshi not as an inert artifact but as an active participant in the politics of representation. Its sources, categories, and silences reveal how early modern Japan transformed the immediacy of encounter into the permanence of archive.

Ultimately, to read Nantōshi anthropologically is to provincialize the history of ethnography itself. It challenges the familiar narrative that casts anthropology as a by-product of European expansion and instead foregrounds the polycentric emergence of ethnographic thought. In early modern Japan, the impulse to know the other arose not from colonial conquest but from the moral and bureaucratic demands of an insular empire. Arai Hakuseki’s writings on Ryukyu capture this moment of epistemic transition. During this period, when the spectacle of the foreign became a problem of knowledge, and when knowledge itself became a form of rule.

The diplomatic theatre: “Edo Nobori” and the visibility of the other

In the early 18th century, the streets of Edo became the stage for one of the most elaborate spectacles of Tokugawa diplomacy. This was the ceremonial processions of Ryukyuan embassies known as Edo nobori (江戸上り, “the ascent to Edo”). These missions, dispatched from the Ryukyuan Kingdom to pay tribute to the Tokugawa shogunate, were ritualized performances of loyalty, hierarchy, and cultural difference. Beyond their formal diplomatic function, they served as a crucial apparatus for the production of ethnographic visibility. Through the visual and performative dimensions of Edo nobori, Edo society encountered, imagined, and systematized the image of the foreign, thereby transforming the capital itself into an ethnographic public sphere.

The Edo nobori of 1710, coinciding with Tokugawa Ienobu’s rule and Arai Hakuseki’s tenure as senior counselor, was the most meticulously staged. As Yasunori Arano (1992: 112) and Ronald Toby (1991: 52) have shown, these embassies were integral to the Tokugawa symbolic order. They functioned as a dramatization of Japan’s moral centrality within an East Asian world of tribute and hierarchy. Recent historiography has further enriched this perspective. Building on foundational studies, scholars such as Travis Seifman (2022) and Marco Tinello (2021) demonstrate how the logistical mechanics and visual staging of the Edo nobori actively standardized ethnological knowledge, turning diplomatic protocol into a systematic tool for early modern boundary-making. Nearly one hundred Ryukyuan officials, attired in brilliant robes and bearing banners, paraded through Edo’s main avenues, transforming the city into a theatre of the exotic. Edo sunago (1710) describes the crowd’s astonishment at their “bright garments and foreign songs.”

As Timon Screech (1996) notes, Tokugawa visual culture treated spectacle as a form of knowledge. The encounter between Japanese spectators and Ryukyuan envoys embodied a “gaze of hierarchy”. This was an act of seeing that both recognized and domesticated difference. By orchestrating the embassy as a public ritual of subordination, the shogunate converted vision into governance.

The 1710 embassy also catalyzed what scholars later called the Ryūkyū boom (琉球ブーム). Merchants, publishers, and artists capitalized on the fascination with Ryukyuan culture, producing an abundance of images and objects that circulated among urban consumers. Illustrated scrolls like Ryūkyūjin gyōretsu zu (Processional Pictures of the Ryukyuans) and ukiyo-e prints popularized Ryukyuan figures, while kabuki plays and ningyō jōruri performances featured them as exotic characters. Textile patterns labeled Ryūkyū-mon and “Ryukyuan-style” costumes further commodified the foreign. As Eiko Ikegami (2005) observes, Edo’s spectatorship turned difference into spectacle, and spectacle into commodity.

This commodification of alterity illustrates how visual contact generated reproducible knowledge. The Ryukyuan envoys, once diplomatic emissaries, became icons of exotic familiarity through constant reproduction. Each print or performance distilled difference into a consumable form, enacting what Nicholas Thomas (1991) calls an “active mediation” that constructs the very categories it represents. In Edo, Ryukyu thus oscillated between foreign and familiar, subordinate and stylish. This created a tension central to Japan’s emerging ethnographic imagination.

The Edo nobori also embodied the Tokugawa shift toward ocular empiricism, which was the belief that seeing was knowing. Under the sakoku edicts, such controlled spectacles offered curated glimpses of the external world. The state regulated who could appear, what could be seen, and how it could be represented, creating what Michel Foucault (1977: 194) would call a “regime of visibility.” The Ryukyuan envoys became visible subjects of observation Consequently, their bodies, gestures, and words were recorded in official documents and illustrated scrolls. Seeing thus became a bureaucratic practice.

This visual regime permeated popular epistemology as well. Illustrated encyclopedias such as Wakan sansai zue (1712) and Bankoku zukan (1719) codified Ryukyuan customs and attire, teaching readers how to see and classify the other. As Clifford (1986: 15) reminds us, ethnographic authority begins with acts of inscription. As he argues, “writing culture” starts with visualizing difference.

For the shogunate, however, spectacle was also surveillance. The same processions that delighted commoners reaffirmed Japan’s authority over its tributaries. The envoys’ route, gestures, and speech were choreographed; their deference before the shogun transformed diplomacy into a technology of observation. As Toby (1991) argues, Tokugawa diplomacy was less about negotiation than performance. Specifically,the visibility of the foreign was produced to reaffirm domestic order. The Edo nobori thus functioned as a bureaucratic ethnography of spectacle, where the observation of others reinforced Japan’s hierarchical cosmology.

Arai Hakuseki’s writings emerged directly from this context. His interviews with Zheng Shunzu and Prince Mishin Ōji resembled ethnographic fieldwork conducted under state auspices. Unlike European explorers or Jesuit missionaries (a comparison drawn here to highlight differing modalities of spatial power rather than strict historical equivalence), Hakuseki’s authority derived not from travel but from control of access. The Ryukyuan other came to him, summoned to Edo, where observation unfolded within the shogunate’s moral geometry. The act of seeing thus became an extension of rule.

While the state’s gaze was disciplinary, the public’s gaze was participatory. Townspeople engaged with the embassy as consumers of culture through illustrated guides and gossip sheets (kawaraban). As Harootunian (2000) notes, Genroku-era urban culture thrived on the circulation of signs detached from their referents. In this sense, Edo nobori inaugurated a form of popular ethnography—a collective spectatorship that turned diplomacy into entertainment. The pleasure of decoding costumes and gestures reflected a new semiotic engagement with the foreign.

Johannes Fabian (1983: 31) described ethnography as a temporal practice that situates others in another time. Edo’s spectators imagined Ryukyu as Japan’s “ancient south,” simultaneously archaic and connected. The embassy thus bridged temporal and cultural distances, allowing Edoites to experience history as spectacle. However, this temporal distancing inherently diminished Ryukyu’s contemporary political agency. Through the public’s gaze, ethnographic seeing was popularized, though it remained strictly bounded by Tokugawa ideology and served to justify structural inequalities.

The spectacle of Edo nobori did not end with the parade but was sedimented into texts. Officials recorded the event; scholars like Hakuseki transformed observation into systematized knowledge. His Ryūkyū koku jiryaku (1711) and Nantōshi (1719) represent the textual consolidation of these encounters. This was the moment when visual ethnography became written ethnology. As Toby (1991) suggests, Hakuseki’s works institutionalized the public ethnographic gaze by aligning observation with governance. The embassy’s visual excess was distilled into textual order; spectacle became knowledge.

By mid-century, Edo had become an ethnographic public sphere. It became a space where state power, scholarly inquiry, and popular curiosity converged. The Edo nobori’s visibility exemplified a new circulation of knowledge defining Tokugawa urban modernity. Official records, scholarly treatises, and popular prints together produced categories (such as “foreign,” “southern,” and “tributary”) that structured Japan’s understanding of the world.

Seen in this light, Edo nobori was not a marginal curiosity but a crucial node in the global history of ethnography. Long before modern anthropology, Edo Japan cultivated visual and textual technologies for knowing difference. The performative visibility of Ryukyu, staged through diplomacy and disseminated through print, marked the emergence of an indigenous ethnographic imagination. This worldview defined the Japanese self through the disciplined spectacle of the other.

Arai Hakuseki and the bureaucratization of ethnographic knowledge

In the early 18th century, Arai Hakuseki (1657–1725) embodied a rare fusion of intellectual and bureaucratic authority. As both a Confucian scholar steeped in Chinese learning (kangaku) and a senior advisor to the Tokugawa shogunate, his intellectual work straddled scholarship and statecraft. His encounter with the Ryukyuan embassies of 1710 offered a pivotal moment when diplomacy, textual scholarship, and ethnographic curiosity converged into a single epistemic enterprise.

Hakuseki’s engagement with Ryukyu transformed firsthand diplomatic experience into a form of bureaucratized ethnographic knowledge. Through close reading of Ryūkyū raihin nikki shō (Arai, 1710), Ryūkyū koku jiryaku (1711), and Nantōshi (1719), this study argues that his work marked the institutionalization of ethnographic observation within Tokugawa governance. For Hakuseki, knowledge was never merely descriptive. Instead, it was administrative, a mode of understanding designed to govern.

Educated in the Zhu Xi Confucian tradition, Hakuseki viewed scholarship (gakumon) as inseparable from moral governance (seiji). As Miyazaki (1969) and Arano (2005: 42) note, he regarded the systematic collection of information on peripheral polities as essential to defining Japan’s position within a Japan-centered Sinocentric order. The Ryukyuan embassy led by Prince Mishin Ōji and scholar Zheng Shunzu (Tei Junso) in 1710 thus provided an opportunity for both moral inquiry and administrative investigation. Assigned by the shogunate to interview the envoys, Hakuseki prepared detailed questionnaires on Ryukyuan geography, institutions, and genealogy. His preface to Ryūkyū koku jiryaku later described these exchanges as “an inquiry into the roots of their customs and governance”. This statement reveals an ethnographic investigation embedded in state ritual.

Hakuseki’s practice resembled that of a state ethnographer, collecting data not for contemplation but for classification. His writings illustrate the Tokugawa ambition to convert the fleeting spectacle of diplomacy into an enduring archive of knowledge. The result was a hybrid corpus combining firsthand testimony, archival borrowing, and textual cross-reference. The interviews with Zheng Shunzu provided what might be called the field notes of Tokugawa ethnography, later reworked into the ordered structure of Ryūkyū koku jiryaku. Hakuseki also consulted diaries from earlier embassies compiled by Satsuma officials, as well as Ming–Qing gazetteers on maritime polities. For instance, when recording Ryukyuan mythologies of origin, Hakuseki deliberately cross-referenced oral accounts from the envoys with ancient Japanese chronicles like the Nihon Shoki. By translating Ryukyuan ancestral terms into the phonetic equivalent of Japanese deities, he executed a distinct translation move that subordinated their local history to a Japanese cosmological framework. By juxtaposing these materials, he constructed a comparative narrative linking Ryukyu to both ancient Japan and southern China. This gesture, as Toby (1991: 52) argues, situated Japan between, yet superior to, its regional others.

Completed in 1719, Nantōshi crystallized this process. Organized into sections on geography, polity, language, and customs, it mirrored the classificatory logic of encyclopedic works such as Wakan sansai zue, yet pursued greater epistemic precision. The textual mechanics of Nantōshi vividly demonstrate this hierarchical ordering. Hakuseki employed specific classification headings, such as Fūzoku (customs), to strictly categorize behaviors he deemed divergent from Confucian norms. In his section on etiquette, he approvingly cited Ryukyu’s adoption of formal court dress but explicitly marginalized their indigenous mourning rituals by framing them within a didactic commentary on “barbarian practices.” Each entry included citations and moral evaluation, transforming the transient encounter of diplomacy into ordered bureaucratic knowledge. The text exemplifies what Foucault (1977: 194) later termed a “regime of truth”—where knowledge is produced through systems of verification embedded in institutional authority.

Although Hakuseki never visited Ryukyu, his authority rested on mediated hearing rather than direct vision. Far from a limitation, this method reflected a distinctly Confucian epistemology: truth emerged not from empirical seeing but from disciplined interpretation. In his preface to Nantōshi, he wrote, “I heard from those who came, and I examined their words with the principles of reason.” By filtering Ryukyuan voices through his own analytical framework, Hakuseki did not merely record the other; he overwrote their epistemic sovereignty. As Sakai (1997) and Ivy (1995) note, early modern Japanese thought often privileged interpretive listening as a moral technology for constructing alterity. Through this “auditory empiricism,” Hakuseki converted oral testimony into textual authority, enacting the Tokugawa hierarchy in which written inscription authenticated experience.

From an anthropological perspective, Hakuseki’s approach anticipates later debates on mediation and authenticity in fieldwork. The Ryukyuan envoys acted simultaneously as informants and political subjects constrained by ritual decorum. Their voices, filtered through bureaucratic expectation, produced an ethnography that reflected as much about the Tokugawa will to order as about Ryukyuan society itself. The ethnographic text became both observation and governance As Foucault might term it, this constituted an archive of rule.

Within the broader logic of Tokugawa governance, this mode of knowledge served to stabilize Japan’s hierarchical world order. As Arano (2005: 115) observed, the early 18th century saw the consolidation of Nihon-gata ka’i chitsujo, a Japan-centered Sinocentrism that framed peripheries like Ryukyu, Ezo, and Korea as moral satellites. Hakuseki’s ethnography simultaneously legitimized this hierarchy and asserted Japan’s civilizational virtue. His claim that “Ryukyu was once ancient Japan’s southern province” inscribed the islands within Japan’s temporal continuum effecting a genealogical domestication of difference. Through meticulous description, he transformed Ryukyu into a moral landscape requiring benevolent oversight, recasting governance as civilizational care. However, this rhetoric of benevolence must be critically understood as a mechanism of epistemic violence. By framing subordination as moral care, Hakuseki effectively masked the structural inequities and geopolitical coercion inherent in the Tokugawa tribute system.

This bureaucratic ethnography paralleled contemporary European empires’ classificatory projects but with a distinct moral logic. It is crucial to note that this parallel does not suggest direct equivalence or shared historical trajectories, as Tokugawa Japan operated without the engine of overseas capitalist extraction. Hakuseki translated Confucian virtue into geopolitical order, turning moral hierarchy into spatial hierarchy. His ethnographic labor thus constituted a Confucian anthropology of statecraft—an early modern experiment in knowing and ruling through the same textual apparatus. This was an early modern experiment in knowing and ruling through the same textual apparatus.

By the time Nantōshi circulated, ethnographic documentation had become part of the Tokugawa administrative routine. Satsuma officials continued to record Ryukyuan embassies, and encyclopedists integrated such data into general knowledge compendia. Hakuseki’s methods (interview, collation, and moral evaluation) established a bureaucratic template later applied to studies of Ezo, the Ainu, and the Dutch enclave at Nagasaki. As Howell (2005) notes, such ethnographic practices were central to defining Japan’s “internal others,” shaping the moral and territorial boundaries of the emerging nation. Recent scholarship from the past 5 years has corroborated this perspective, demonstrating how early modern Japanese mapmaking and ethnographic cataloging were not merely scientific endeavors but aggressive acts of spatial and cognitive boundary-making.

From this perspective, Nantōshi was not merely a descriptive text but a technological apparatus of rule. It served as a model for how the shogunate could know its peripheries without leaving Edo. It embodies the bureaucratization of ethnographic curiosity, where observation became institutionalized, standardized, and moralized. Through the transformation of field encounter into textual record, and textual record into administrative knowledge, Hakuseki completed the circuit of Tokugawa epistemic governance.

Hakuseki’s ethnography demonstrates that the origins of anthropology need not be traced solely to European colonial modernity. In Edo Japan, too, the act of knowing the other was inseparable from the project of governing the self. Nantōshi thus stands as both an intellectual artifact and a political technology. It represents an early modern Japanese experiment in mapping the world through the disciplined imagination of difference.

Writing Ryukyu: Textual ethnography and cultural hierarchy

When Arai Hakuseki completed the Nantōshi (Records of the Southern Islands, 1719), he had spent nearly a decade refining the observations gathered during his 1710 interviews with Ryukyuan envoys. The resulting text was not a geographical compendium or diplomatic report but a hybrid epistemic artifact. Being part ethnography, part geography, and part moral treatise, it sought to situate the Ryukyuan Kingdom within the intellectual and spatial order of Tokugawa Japan.

To trace the trajectory of this decade-long refinement, one must examine its immediate textual precursors preserved in archival collections: the manuscript Ryūkyū raihin nikki shō (1710) and Hakuseki’s formal report Ryūkyū koku jiryaku (1711). The preliminary manuscript Ryūkyū raihin nikki shō (1710), compiled directly during the 1710 embassy, served as a raw administrative log. Drawing from the rare copy in the Sakamaki-Hawley Collection (HW443), it reveals how Hakuseki systematically cataloged the material culture, behavioral protocols, and linguistic utterances of the envoys to manage immediate shogunal anxieties regarding foreign status (Arai, 1710, fol. 3a–4b). In the subsequent Ryūkyū koku jiryaku (1711, HW518), presented formally to Shogun Tokugawa Ienobu, Hakuseki performed a critical initial translation move. He began to organize these raw empirical observations into a structured historical and genealogical narrative, forcibly mapping out the island kingdom’s boundaries to fit within a Japan-centered world order (Arai, 1711, fol. 5a–6b). Therefore, rather than appearing ex nihilo, Nantōshi represented the final, weaponized codification of these earlier conceptual scaffolds, transforming a temporary diplomatic encounter into a permanent administrative science of the other.

As a document, Nantōshi exemplifies the bureaucratization of knowledge under the shogunate. Yet as a text, it reveals the subtle rhetoric through which Hakuseki transformed empirical data into moral hierarchy. Organized into sections on geography, institutions, customs, and the arts, its very architecture reflects an ethnographic logic of classification. To illustrate this textual mechanic, one only needs to examine his table of contents. By cleanly segregating formal “Institutions” from local “Customs”, Hakuseki created a taxonomic divide that elevated bureaucratic statecraft to the realm of Japanese equivalence while relegating indigenous Ryukyuan practices to the category of exotic anomalies. Through comparison, evaluation, and linguistic framing, Hakuseki turned Ryukyu from a polity into a conceptual province of Japan’s moral world.

The opening section on geography already establishes this ideological geometry. Ryukyu is placed “south of Satsuma,” yet “originally belonging to ancient Japan.” This double gesture of distance and inclusion encapsulates the work’s central tension: Ryukyu is rendered as both exterior and interior, foreign yet genealogically continuous. The following sections move systematically from geography to polity, law, and custom, mirroring the Chinese zhigong shu (treatises on tributaries) while introducing a new empiricist rigor. Citing oral testimony, earlier embassy diaries, and Chinese sources, Hakuseki constructs an apparatus of citation that signals epistemic modernity avant la lettre. Each description is followed by a moral gloss. For example, his textual authority is weaponized when he pronounces that Ryukyuan laws are “humane yet inconsistent with the ancient way,” or their arts “refined yet effeminate.”

Here, ethnography and moral philosophy intersect. Geography locates, customs describe, and morality judges. What emerges is a totalizing ethnographic system in which empirical observation legitimizes moral evaluation. As Toby (1991: 52) observes, such classificatory knowledge was central to Tokugawa symbolic governance. In this epistemic regime, to know was to order, and to order was to govern.

Comparison lies at the heart of this textual ethnography. Ryukyuan titles are matched to Japanese court ranks; local deities to kami; poetic forms to waka and fu. Comparison, however, does not merely illuminate. It actively hierarchizes resemblance. A concrete micro-demonstration of this translation move is Hakuseki’s treatment of the Ryukyuan language. By systematically identifying Ryukyuan vocabulary as corrupted or archaic pronunciations of classical Japanese words, he strips the local language of its autonomy. Translating Ryukyuan linguistic categories into Japanese origins renders them intelligible but strictly subordinate. Hakuseki’s ethnography thus operates within what Foucault (1970) called the episteme of order which is the classificatory impulse that defined early modern knowledge. Yet his taxonomy is distinctly moral. The Ryukyuans, he writes, “follow Chinese ways but lack sincerity; they adopt Japanese dress yet misunderstand propriety.” This oscillation between praise and critique performs what might be termed the civilizing syntax of Tokugawa ethnography. However, this “civilizing” process must be recognized as fundamentally coercive, actively erasing indigenous epistemologies to fit a prescribed imperial mold.

In Nantōshi, civilization becomes a process rather than a state. Ryukyu’s partial refinement serves as Japan’s proof of superiority. The act of writing itself becomes governance, proving that to describe is to discipline.

One of Hakuseki’s most consequential claims is the hypothesis that Ryukyu was once part of “ancient southern Japan” (konan wa). On the surface, it is a philological conjecture; beneath it lies a genealogical strategy of inclusion. Identifying Ryukyu as Japan’s ancient fragment collapses spatial distance into temporal continuity Consequently, the foreign becomes ancestral. This move reverses what Fabian (1983: 31) termed temporal othering. Hakuseki temporalizes space to domesticate difference. The Ryukyuans emerge as “our southern descendants,” living embodiments of Japan’s moral past.

The political implications are profound. By genealogizing Ryukyu, Hakuseki not only asserted cultural kinship but also legitimated authority. History replaces conquest, while kinship naturalizes subordination. As Sakai (1997) argues, such moral imperialism exemplifies Japan’s capacity to internalize the foreign as a lost part of itself. In Hakuseki’s logic, if Ryukyu embodies Japan’s past, Japan embodies Ryukyu’s future. The relation is pedagogical. It models the dynamic of parent to child, master to pupil. Temporal hierarchy becomes the ethical rationale for rule. However, framing this relationship as pedagogical obscures its fundamentally coercive nature. This paternalistic rhetoric inflicted deep epistemic harm by stripping Ryukyu of its historical autonomy and forcing it into a subservient developmental timeline.

Complementing this genealogical trope is another key formulation: “Southern Domain Ryukyu” (南藩琉球). This phrase marks the epistemic shift from kinship to administration. The term han—the Tokugawa designation for feudal domains—when extended to Ryukyu, a polity outside Japan’s borders, symbolically annexes it into the shogunate’s moral geography. Ryukyu becomes a “domain” in the cognitive sense. It is reimagined as a loyal yet distinct limb of the Japanese body politic.

This linguistic operation exemplifies what Stoler (2009) calls the “epistemic labor of empire”, meaning the work of naming that precedes rule. Nantōshi does not simply describe Ryukyu; it produces it as an object of governance. Each entry, ranging from geography to ritual—functions as a micro-administration of knowledge, transforming descriptive prose into disciplinary taxonomy. Labels such as “southern,” “tributary,” and “civilized yet imperfect” enact a symbolic annexation, turning difference into an ordered category within the Japanese epistemic state.

Hakuseki’s textual domestication of Ryukyu prefigured later imperial ideologies. The Meiji discourse of naichi–gaichi ittai (“unity of inner and outer territories”) echoed the kinship–subordination logic articulated here. Yet in Hakuseki’s context this was still an empire of virtue,but it is vital to emphasize that “virtue” here did not equate to an absence of violence. Instead, it represented a moral cartography of civilization grounded in Confucian benevolence that exacted a profound structural violence. By erasing indigenous epistemologies and mandating ideological conformity, this allegedly non-military order functioned as a highly effective mechanism of cultural subjugation.

For Hakuseki, writing was governance. To inscribe was to impose structure on the flux of the world. Nantōshi thus operates as both document and disciplinary act. Its numbered sections, cross-references, and citations reflect the bureaucratic modernity of Tokugawa information culture, in which ethnography became administration. The bureaucrat, in turn, became ethnographer. A concrete textual mechanic illustrating this is Hakuseki’s systematic cross-referencing of Ryukyuan administrative ranks. By forcibly mapping Ryukyuan court titles onto Japanese hierarchical equivalents, he essentially bureaucratized the other through the mechanical act of translation.

The ethnographic impulse here is inseparable from the moral function of bun (writing, culture). Like Foucault’s notion of power/knowledge, Hakuseki’s bun governed through classification. To describe Ryukyu’s customs was to reaffirm Japan’s civilizational centrality. Precision itself became a disciplinary gesture. The more exact the description, the more complete the subordination. As Geertz (1973) later observed, ethnography is always interpretive, meaning it turns local detail into symbolic order. Nantōshi anticipated this process, reading Ryukyuan culture not for its empirical reality but for its place in Japan’s moral cosmos.

Through this textual labor, ethnography became a mode of cultural governance. Each observation, whether regarding attire, etiquette, or art reproduced the moral gradient separating center from periphery. The Nantōshi thus illuminates the interdependence of writing, knowledge, and power in Tokugawa Japan. It reveals the conversion of observation into record, of record into rule.

By constructing Ryukyu as both “ancient southern Japan” and “southern domain,” Hakuseki inscribed hierarchy into history. His ethnographic prose transformed kinship into subordination, genealogy into governance. In doing so, he enacted what might be termed a scriptural sovereignty. This was a power to define the world through the moral act of writing.

Long before anthropology named itself as a science, the ethnographic imagination was already at work in the archives of early modern empires. In Edo Japan, as in Europe, to know was to govern, and to govern was to write. While comparing Edo Japan to European empires reveals structural similarities in the will to classify, it is imperative to state the limits of this comparison. Tokugawa ethnographic practices were neither direct precursors to nor strict equivalents of Western colonial anthropology, as they were driven by regional diplomatic boundary-making rather than global capitalist expansion. Nantōshi stands as testament to this polycentric modernity of ethnographic knowledge. It represents a Japanese ethnography of the other that, in defining its object, simultaneously defined the self.

Ethnography as political thought

Arai Hakuseki’s Nantōshi was not simply a record of the Ryukyuan Kingdom but an act of political reasoning. Written at the height of Tokugawa intellectual maturity, it crystallized a worldview in which knowledge of the other was inseparable from the governance of the self. To read Nantōshi solely as ethnography is to miss its deeper function. It is a treatise on order, representing an attempt to locate Japan within a moral and geopolitical cosmos that was at once Sinocentric and self-centered. For Hakuseki, ethnography became a mode of political thought: a disciplined representation of cultural difference that imagined hierarchy, civilization, and legitimacy.

The Tokugawa polity rested on what Arano Yasunori terms a “Japan-centered world order” (Nihon-gata ka’i chitsujo). Formally within the Chinese tributary sphere, the shogunate nonetheless claimed an autonomous moral sovereignty. Within this dual framework, foreign polities, such as Korea, Ryukyu, and the Dutch, served as epistemic mirrors through which Japan’s centrality could be affirmed.

Nantōshi captured this duality with precision. By defining Ryukyu as both “tributary to China” and “offspring of ancient Japan,” Hakuseki reconciled the Confucian universalism of the hua–yi hierarchy with a nascent Japanese ethnonationalism. The result was a hybrid Sinocentrism. It involved an appropriation of Chinese categories of civilization, re-centered on Japan.

This maneuver echoed a broader East Asian tendency to localize the universal. Chosŏn thinkers such as Yi Ik redefined Confucian orthodoxy to preserve moral authority without political power. Furthermore, to concretize this regional parallel, one can look to the Qing dynasty’s ethnographic mapping of its own peripheries. As Emma Jinhua Teng (2004) demonstrates in her study of Qing colonial travel writing, Chinese scholar-officials actively utilized geographical and ethnographic prose to imaginatively annex Taiwan, textually transforming a “savage” frontier into a known, governable imperial space. What distinguishes Hakuseki from these continental parallels is that he translated such political doctrine into a distinctly Japanese ethnographic form. Japan’s centrality was not asserted abstractly but demonstrated through observation and classification of a concrete other. Knowledge thus became sovereignty’s medium.

In Nantōshi, describing Ryukyu performs a dual act: the accumulation of empirical data and the conversion of that data into an imaginative map of order. Each detail, including rituals, kinship, and protocol, illustrates a larger cosmological claim that Japan occupies the moral center of a concentric hierarchy of civilization.

This transformation drew on the Confucian doctrine of cultivation and governance (shushin chihō), where the ruler’s self-cultivation was the foundation of rule. Hakuseki extended this outward. Consequently, the moral observation of others became a form of governance. The ethnographic gaze became a governing gaze, translating curiosity into control.

Foucault’s concept of power/knowledge is instructive. In Discipline and Punish (1975), he notes that modern power operates through documentation and surveillance, achieving a “minute writing of the world.” Hakuseki’s ethnography anticipates this logic. His meticulous classification of Ryukyuan institutions functions as textual technology of rule, a moral surveillance rendering the distant knowable and thus governable.

At the same time, Nantōshi exposes the ethnographic paradox. Specifically, the will to know the other reproduces the asymmetry between observer and observed. As Clifford (1986: 15) reminds us, ethnography constructs both subject and object. Hakuseki’s Ryukyu is not an independent voice but a mirror of Japan’s self-understanding. In this process, writing becomes the means by which encounter hardens into hierarchy.

The Tokugawa state’s civilizing project contextualizes this dynamic. The 17th and 18th centuries witnessed a surge of bureaucratic and intellectual institutions. These included domain schools, surveys, encyclopedias, and maps, which produced what Toby (1991: 52) calls the “information state.” Within this system, ethnography both delineated external moral boundaries and disciplined internal imagination.

The shogunate’s policy toward Ryukyu, which allowed visible submission to China while controlling it through Satsuma, was a political performance of dual autonomy. Nantōshi textualized this logic. By portraying Ryukyu as culturally subordinate yet morally educable, it provided ideological justification for the arrangement. Ethnographic writing became a soft diplomacy of the page. However, this symbolic domination still constituted a profound epistemic violence that subjected Ryukyuan realities to strict Japanese archival control.

Hakuseki’s moral tone reinforces his belief that civilization is an ethical vocation. To provide direct insight into the text itself, one need only examine his specific linguistic and cultural assessments in the author’s general preface. In Nantōshi, Hakuseki explicitly writes: “Their language preserves the ancient words of our realm (waga chō no kogo), yet their daily customs have become accustomed to the superficial forms of barbarian lands” (Arai, 1719, fol. 2a). This direct quotation, sourced directly from the manuscript preserved in the Cabinet Library of Japan, reveals the core mechanic of his ethnography: it is not merely descriptive, but deeply normative. His critiques of ritual excess or linguistic impurity were not xenophobic but pedagogical, serving as expressions of Confucian tutelage. For Hakuseki, governance was intrinsically tied to moral education. Therefore, to rule was to cultivate. In this sense, Nantōshi functions as a manual of instruction. It acts as a guide for seeing, judging, and correcting the world.

Placed in global perspective, Hakuseki’s project parallels developments in Europe. To further concretize this comparison, one must look beyond the standard narrative of 19th-century colonialism. As Han F. Vermeulen (2015) argues, the true genesis of European ethnography occurred during the 18th-century German Enlightenment, where scholars developed Völkerkunde (the study of peoples) primarily as a tool for state administration and historical classification rather than overseas conquest. Hakuseki’s contemporaneous project resonates deeply with this specific Enlightenment impulse: he, too, deployed the classification of human diversity as an administrative technology. Yet the contrast is revealing. While comparisons to this 18th-century European matrix are conceptually useful, they must not imply a direct equivalence. Japanese ethnographic impulses evolved independently of the global expansionist drives that fueled much of later European knowledge production. European ethnography emerged from imperial expansion, meaning its knowledge of others was shaped by conversion and conquest. Jesuit depictions of China balanced admiration with evangelism, while Enlightenment thinkers such as Montesquieu or Buffon recast this mission in secular form.

Hakuseki’s ethnography, by contrast, arose without an overseas empire. His gaze was bureaucratic, not colonial; his other was tributary, not alien. Whereas European ethnographers projected difference outward to Africa, the Americas, and the Pacific. Hakuseki folded it inward, locating the foreign within Japan’s moral geography. The Ryukyuans were not savages to civilize or equals to admire, but kin to correct.

This difference suggests that ethnography, as a global form, cannot be reduced exclusively to Western colonial paradigms. The Tokugawa case reveals a distinctively regional genealogy of ethnographic reason, grounded in administrative rationality and moral hierarchy rather than military expansion. Nevertheless, recognizing this difference should not romanticize Tokugawa practices as benign. This regional knowledge regime still inflicted severe structural harm by obliterating local agency to maintain a rigid geopolitical hierarchy. If Enlightenment ethnography justified empire through universalism, Hakuseki justified hierarchy through genealogy. Both, however, transformed observation into order.

Tokugawa ethnography thus performed a double movement. It mimicked Sinocentric universalism even as it displaced it. By reinterpreting China’s hua–yi distinction through a Japanese lens, Hakuseki transformed dependency into autonomy. His assertion that “Ryukyu was once ancient Japan” acknowledged Chinese civilization as model yet relocated moral authority to Japan.

This re-centering exemplifies what Najita (1987) calls the emergence of an “autonomous domain of knowledge”, representing a Japanese epistemology no longer requiring validation from China. Nantōshi marks the moment when Japan ceased to be a passive participant in the hua–yi world and began to imagine itself as arbiter of civilization. Knowledge of the other thus became the knowledge of self-rule.

The text was written amid anxieties over Japan’s international position. These anxieties stemmed from memories of Hideyoshi’s invasions, isolationist policies, and delicate diplomacy with Qing China through Ryukyu. Nantōshi offered a response. By mastering representation of others, Japan could assert the autonomy of its knowledge. Ethnography became a political technology of independence.

Reading Nantōshi as political thought reveals ethnography’s role in constructing the moral infrastructure of sovereignty. Hakuseki’s text articulates a theory of governance grounded in knowledge, detailing nested hierarchies from family to polity, from domain to cosmos, each sustained by proper understanding of relations.

Long before Malinowski or Durkheim, Hakuseki intuited that customs and institutions form a coherent moral system and that to know them is to govern them. His anthropology, however, was normative rather than functionalist, because knowledge served not prediction but hierarchy.

The Japanese ethnographic tradition thus arose from bureaucratic and moral imperatives, not scientific curiosity or missionary zeal. Its impulse was endogenous, born from the need to reconcile diversity with unity. Nantōshi therefore represents an indigenous form of anthropological reason. It is a politics of description that prefigured modern social science while grounded in Confucian ethics of rule.

In Nantōshi, Hakuseki transformed the study of Ryukyu into a meditation on order itself. Knowledge here functions as political act. It defines civilization’s boundaries, establishes cultural hierarchy, and legitimates authority. For the Tokugawa state, to know Ryukyu was to govern symbolically; for Hakuseki, to describe was to legislate.

Comparing Hakuseki’s project with Enlightenment ethnography clarifies that the early modern world produced multiple epistemologies of difference. European ethnography mastered the world through universal reason, while Japanese ethnography stabilized it through moral hierarchy. Both converted observation into sovereignty. Yet where the Enlightenment gaze expanded outward, Hakuseki’s turned inward—toward Japan’s own moral topography.

To synthesize the core argument, Nantōshi demonstrates how Tokugawa knowledge production actively structured moral governance. This dynamic is most vividly evidenced in Hakuseki’s specific textual mechanics. By employing strict taxonomic divisions between Japanese “institutions” and Ryukyuan “customs,” and by systematically translating indigenous Ryukyuan vocabulary into subordinate Japanese etymologies, Hakuseki constructed a discursive hierarchy that effectively annexed Ryukyu into Japan’s epistemic domain.

Nantōshi thus stands as both political document and intellectual milestone. It reveals that early modern ethnography was not a prelude to science but a genre of governance. The knowledge of others was never mere curiosity. Rather, it was the grammar of power. Through the disciplined act of writing Ryukyu, Arai Hakuseki inscribed Japan into being: as a moral center in a world of gradated civilizations, a sovereign of knowledge in an empire without overseas colonies, and a precursor to the anthropological imagination that would later define modernity.

Genealogy and afterlife of Ryukyuan studies

The intellectual genealogy from Nantōshi to Meiji anthropology thus reveals a striking continuity: the transformation of ethnographic curiosity into a state epistemology. Under Tokugawa rule, the collection of knowledge about others served the moral order of the shogunate; under Meiji rule, it served the administrative and ideological order of empire. Both regimes relied on ethnography to imagine unity amid diversity in order to reconcile the contradiction between cultural hierarchy and political inclusion.

At the same time, the modern revival of Hakuseki’s ideas illuminates a paradox central to non-Western modernities. Japan’s turn to anthropology was both imitative and autonomous. As Oguma Eiji and Tessa Morris-Suzuki have argued, Meiji intellectuals borrowed the disciplinary form of Western ethnology while filling it with indigenous content. This included the myths of descent, the rhetoric of civilization, and the bureaucratic desire for order that dated back to Tokugawa scholarship. In this hybrid formation, the Nantōshi functioned as an invisible foundation. It served as a pre-anthropological text that made anthropology imaginable within a Japanese frame.

Such genealogical continuity complicates the conventional narrative of global knowledge diffusion. It suggests that non-Western empires produced their own anthropologies, not by passively adopting Western science but by reactivating preexisting modes of governance through new vocabularies of expertise. Japan’s imperial ethnography, with its fusion of bureaucratic empiricism and moral pedagogy, exemplifies what Dipesh Chakrabarty calls the “provincialization of Europe”, demonstrating the emergence of modernity as a plural phenomenon. The Nantōshi’s transformation from Tokugawa document to imperial genealogy reveals that the anthropological imagination is not the exclusive offspring of colonial modernity but a recurrent response to the political problem of difference.

The 20th-century consequences of this intellectual inheritance were profound. During the colonial period, ethnographers working in Okinawa and Taiwan continued to deploy the idiom of kinship and tutelage. Anthropological fieldwork was often framed as a mission of cultural preservation and moral improvement, echoing Hakuseki’s Confucian paternalism. Even progressive scholars, including the ethnologist Yanagita Kunio, struggled to disentangle their work from the legacy of hierarchical inclusion. When Yanagita traced Japan’s rural folklore to “southern islands” and “ancient kinship routes,” he was, knowingly or not, extending Hakuseki’s geographical metaphor of descent. The island, once a tributary, had become an archive of origins. However, this folkloric romanticization continued to enact epistemic violence by denying Okinawans their contemporary political reality, effectively reducing them to living fossils of a Japanese past.

To read Nantōshi within this longue durée is to confront the entanglement of knowledge and ideology across Japan’s transition from moral to imperial statehood. The text’s endurance owes not merely to its historical curiosity but to its conceptual flexibility: its ability to accommodate both paternalist governance and nationalist assimilation. Its categories (such as ancient Japan, southern domain, and the kinship of civilization) proved endlessly adaptable to new political circumstances. Through these reconfigurations, the Nantōshi continued to perform the work of ethnography: transforming encounter into hierarchy, difference into identity, and knowledge into rule.

Yet the afterlife of Nantōshi also opens a space for critical reflection. By tracing its reappropriations, we can see how Japan’s modern anthropology emerged not as a derivative copy of Western science but as a self-reflexive technology of empire. Hakuseki’s bureaucratic ethnography offered a precedent for imagining order without military conques. While previously idealized as a “hierarchy without violence,” this paradigm must be critically recognized as a mode of power that inflicted profound structural and epistemic harm. It survived in the rhetoric of benevolent assimilation, systematically erasing indigenous cognitive frameworks under the guise of moral inclusion. At the same time, the persistence of his categories reminds us how deeply the structures of early modern thought are embedded in modern institutions of knowledge.

To rethink the history of ethnography from this perspective is to question the geography of its origins. The discipline’s canonical narrative, which begins with European colonial expansion and culminates in the professionalization of anthropology, obscures the fact that other empires, including Japan’s, generated parallel epistemologies of human difference. However, it is crucial to reiterate the limits of this comparison. These parallel epistemologies were not direct equivalents to European colonial science, as they operated within distinct regional logics of tributary diplomacy rather than global capitalist extraction. These non-Western anthropologies did not simply reproduce European Orientalism; they localized it, moralized it, and redeployed it as a strategy of internal cohesion. The Japanese case, with Nantōshi at its root, demonstrates how the ethnographic imagination could emerge from the logic of governance rather than conquest.

In the end, the genealogy of Ryukyuan studies is inseparable from the genealogy of Japan’s political thought. From Hakuseki’s moral ethnography to Iha Fuyū’s racial philology, from Confucian order to imperial science, the thread that binds them is not continuity of doctrine but continuity of desire. This is the desire to know the other as a way of knowing the self. The Nantōshi’s enduring relevance lies precisely in this dialectic. It invites us to read ethnography not as a transparent window onto culture but as a mirror reflecting the power that seeks to describe.

To synthesize the core argument of this study, the assertion that Nantōshi functioned as an apparatus of Tokugawa moral governance is firmly substantiated by its textual mechanics. As demonstrated throughout this analysis, Hakuseki’s taxonomic segregation of customs and his etymological subordination of the Ryukyuan language served as concrete epistemological tools to structurally annex Ryukyu.

If the early modern shogunate used ethnography to moralize the world, and the modern empire used it to nationalize it, the task of contemporary scholarship is to historicize that gaze. We must to recognize in the pages of Nantōshi both the origins of Japanese anthropology and the limits of its imagination. The Ryukyuan other, once a figure of moral tutelage, remains a reminder that every ethnography, whether colonial or postcolonial, is also a political act. In acknowledging that lineage, we begin to see that the anthropology of empire, whether Western or non-Western—is not a monologue of domination but a complex dialogue between power, history, and the written word.

Conclusion: Toward an anthropology of Tokugawa knowledge

To return to Arai Hakuseki is to return to a moment when Japan’s encounter with the world was still mediated through the language of moral philosophy rather than science, and when the urge to know others was inseparable from the desire to order oneself. In his Nantōshi, the Ryukyuan kingdom becomes more than a geographical object; it becomes a mirror through which the Tokugawa state imagined the coherence of its world. This conclusion proposes to read Hakuseki’s Ryukyuan studies as a foundational, though inherently coercive, form of premodern ethnography, a genre that prefigures the epistemological and moral tensions that would later define anthropology.

What makes Nantōshi exemplary is not simply its descriptive richness or its empirical curiosity, but its fusion of knowledge and governance, and of empirical observation and moral evaluation. Long before the professionalization of anthropology in the 19th century, Hakuseki practiced an ethnographic discipline that was both textual and bureaucratic. His method combined interviews, comparative analysis, and cross-cultural translation, yet always within a moral framework that located Japan as the center of civilization. The Nantōshi thus stands as an early articulation of what might be called an anthropology of order. This was a mode of inquiry in which the act of knowing the other becomes an instrument for defining the ethical boundaries of the self, often at the severe cost of the other’s cognitive sovereignty.

In the preceding chapters, we have seen how this ethnographic impulse evolved through multiple registers: from the performative diplomacy of the Edo nobori to the textual codification of Ryukyuan difference; from the bureaucratic rationalization of knowledge to its later transformation into imperial ideology. Through these stages, ethnography emerges not as a neutral practice but as a persistent cultural technique of governance. It functions as a way of imagining hierarchy through the forms of description. As demonstrated through specific textual mechanics, such as Hakuseki’s taxonomic segregation of Ryukyuan customs from Japanese institutions and his etymological subordination of local languages, the Nantōshi did not merely reflect geopolitical realities. Instead, these micro-demonstrations prove how Hakuseki’s contribution lies precisely in having recognized that writing itself is a political act. He understood that the ethnographic text can produce the very reality it claims to record.

This recognition situates Nantōshi within a broader global history of the ethnographic imagination. In early modern Europe, the Jesuit missionaries and Enlightenment travelers were engaged in similar enterprises. They sought to translate unfamiliar cultures into systems of knowledge that reinforced their own moral universes. Yet Hakuseki’s ethnography diverges from these Western precedents in its aims and structure. It is imperative to state that this comparison does not imply direct equivalence or shared historical trajectories. Lacking a colonial frontier, Tokugawa Japan turned its ethnographic gaze inward and southward, transforming tributary encounters into sites of moral reflection rather than military conquest. However, this absence of military expansion must not be conflated with an absence of violence. The “southern islands” were subjected to an epistemic erasure that reduced them to a mirror of continuity, serving merely as a reminder of origins and of what Japan might have been before it became itself.

In this sense, Nantōshi reveals a distinctly polycentric genesis of ethnographic knowledge. The conventional history of anthropology, written from a Eurocentric vantage, traces the discipline’s origins to the Enlightenment’s encounter with the colonial world. But Hakuseki’s project compels us to reconsider that genealogy. The ethnographic impulse, specifically the desire to describe, classify, and understand the other, was never the monopoly of the West. It was an epistemic possibility latent in any society confronted with difference. In Tokugawa Japan, that difference took the form of moral gradation rather than racial hierarchy; its technologies were textual rather than cartographic; its fieldwork was conducted not in remote islands but in diplomatic audiences and bureaucratic archives.

To call Nantōshi “premodern ethnography” is not to impose a modern category retroactively, but to recognize a shared epistemological form: the conversion of encounter into knowledge, and knowledge into hierarchy. Hakuseki’s text enacts this process with remarkable self-awareness. His prefaces insist upon the necessity of accurate description, yet his commentary constantly interprets, judges, and instructs. The oscillation between empirical curiosity and moral certainty mirrors anthropology’s later struggle between relativism and authority. As James Clifford has argued, the ethnographer is never a neutral observer but always a narrator whose authority depends on the performance of objectivity. Hakuseki anticipated this paradox. By documenting Ryukyu with both admiration and admonition, he produced a text that is at once analytic and prescriptive, scientific and moral.

If anthropology is the study of humans by humans, Nantōshi represents an early experiment in self-reflexivity: Japan studying what it imagined as its own other. The Ryukyuans were not a foreign species but a displaced kin, a living archive of Japan’s ancient self. The ethnographic encounter, therefore, doubled as an act of introspection. Through Ryukyu, Hakuseki wrote the moral genealogy of Japan; through the description of the periphery, he stabilized the meaning of the center. The Nantōshi’s ethnographic gaze thus operates as a mirror of cultural self-recognition. It presents a vision of the self refracted through the figure of the other.

This mirror function distinguishes the Tokugawa ethnographic imagination from the Western colonial model. Where the latter produced knowledge to dominate through global extraction the former produced knowledge to discipline through regional assimilation. T. The shogunate’s power was framed as internal and moral rather than external and territorial; it ruled through decorum, precedent, and classification that functioned as a coercive apparatus of cultural erasure. Within such a regime, the ethnographic text became a bureaucratic archive of virtue. To catalogue Ryukyu’s customs was to reaffirm the universality of Confucian civility; to measure its deviations was to instruct Japan in its own perfection. Ethnography, in this context, was a technology of ethical self-maintenance that exacted a heavy toll on the subjugated peripheries.

Yet this self-reflexive ethnography was not without consequences. As later chapters have shown, the same classificatory logic that moralized the other in the 18th century provided the scaffolding for Japan’s imperial anthropology in the 19th. The moral hierarchy of Hakuseki’s Nantōshi evolved into the racial hierarchy of Meiji ethnology; the pedagogical tone of benevolent governance became the ideological language of assimilation. The genealogy from Tokugawa ethnography to imperial anthropology is thus not one of rupture but of transformation. It represents a continuity of epistemic desire under changing regimes of power.

Recognizing this continuity allows us to reframe the question that underlies the study of Nantōshi: what does it mean to write an anthropology of Tokugawa knowledge? Such an anthropology would not merely trace ideas or catalog data; it would examine the conditions under which knowledge itself became an instrument of order. The Tokugawa world, with its intricate bureaucracy, its cosmology of center and periphery, and its moralized conception of civilization, generated its own anthropology—a study of humanity grounded in the ethics of governance rather than the empiricism of science. The Nantōshi exemplifies this anthropology in its mature form: systematic, comparative, morally charged, and profoundly political.

To move toward an anthropology of Tokugawa knowledge, then, is to decenter the modern narrative of the human sciences. It is to acknowledge that the anthropological impulse—the will to describe, the compulsion to classify, the moral need to comprehend difference—arose wherever states and scholars sought to reconcile diversity with order. In the Tokugawa case, this reconciliation took textual form: the ethnography as archive, the archive as empire. The Nantōshi teaches us that the anthropology of others always begins with an anthropology of the self.

Such a perspective offers a corrective to the teleological historiography that equates anthropology with modernity and modernity with Europe. The intellectual world of early modern Japan demonstrates that there were, and are, multiple ways of knowing humanity. The polycentricity of ethnographic knowledge—its simultaneous emergence in missionary Europe, bureaucratic China, and Confucian Japan—invites us to provincialize the discipline itself. If the anthropology of empire has taught us how knowledge serves domination, the anthropology of Tokugawa Japan reminds us how knowledge can also serve introspection, moral discipline, and self-definition. However, it is crucial to reiterate that this moral discipline was not benign; it operated as a potent form of epistemic subjugation that coerced peripheral polities into conforming to a Japanese center.

The relevance of Hakuseki’s project to contemporary anthropology lies precisely in this polycentricity. In an era when the discipline continues to wrestle with its colonial inheritance, the Nantōshi offers an alternative genealogy. It presents a vision of ethnography as moral reflection rather than military conquest, and as a language of order rather than territorial possession. It demonstrates that the act of representing others can also be an act of self-examination, and that the politics of description rarely coexisted with a genuine ethics of restraint. Rather, it exacted a profound structural violence by overwriting indigenous realities to serve state ideology. To recover this alternative tradition is not to idealize it but to expand the field of anthropology’s past. This recovery demands that we recognize the coexistence of multiple modernities of knowledge without sanitizing their respective mechanisms of control.

In the end, Arai Hakuseki’s ethnography stands as both artifact and allegory. It belongs to a world that no longer exists, yet it speaks to the enduring problem of how knowledge and power intertwine. By treating Nantōshi as a premodern ethnography, we glimpse a moment when the boundaries between philosophy, history, and anthropology were still porous. This was an era when to study others was to think politically, morally, and cosmologically. The Tokugawa scholar-official, pen in hand, listening to Ryukyuan envoys and compiling their words into moral order, anticipates the ethnographer of later centuries. While not a direct historical ancestor to the modern Western anthropologist, Hakuseki shares the same fundamental posture. He was disciplined, interpretive, and haunted by the authority of his own gaze.

To write the history of such knowledge is to write an anthropology of anthropology. It requires us to trace not only what was known but how knowing itself became a form of governance. The Nantōshi endures because it reveals that ethnography, at its deepest level, is not merely about the diversity of cultures but about the conditions of selfhood in a world of difference. In the quiet bureaucracy of Edo, as in the restless science of modernity, the ethnographic gaze remains the same. It is a search for the human that inevitably returns to the self who seeks to know. By historicizing this gaze, contemporary scholars are compelled to confront the uncomfortable reality that our foundational archives are inherently compromised by pre-modern power dynamics. Acknowledging this non-Western epistemic violence is the necessary first step toward a more rigorous, self-reflexive anthropology for the future.