Daejeon, July 27–31
Early modern East Asia generated a substantial body of records concerning shipwrecked Chosŏn subjects who were recovered, detained, and repatriated through Chinese imperial administrative systems. This paper examines these repatriation practices in Ming–Qing China, focusing on how administrative order was enacted through spatial practices rather than articulated solely through normative regulations.
These sources include local gazetteers, administrative memorials, court diaries, diplomatic reports, first-hand accounts by repatriated drifters, and compilations such as P’yohaerok (漂海錄, Records of Drifting at Sea). They contain dense spatial information, including place names, administrative jurisdictions, relay points, and sequences of movement. Yet when the sequential narratives of movement are read in relation to fixed spatial features, whether natural or built, the resulting paths often appear incoherent or spatially “implausible.” Place names along the recorded routes appear misplaced, administrative handovers seem geographically inefficient, and the distances implied by the records often defy cartographic plausibility.
Conventional historical scholarship has largely treated such inconsistencies as problems to be corrected. Discrepancies have been explained as scribal errors, corrupted toponyms, or incomplete transmission of information, while spatial reconstruction has functioned primarily as an illustrative supplement to textual analysis. Similarly, many GIS-based studies of historical mobility prioritize the restoration of accurate or optimized routes, smoothing irregularities in order to produce legible and internally consistent maps.
This paper challenges that corrective paradigm. It asks instead: what if spatial inconsistencies are not defects in the sources, but historically meaningful traces of how repatriation was organized, justified, and governed? By shifting the analytical focus from accuracy to interpretation, this paper reframes GIS-based reconstruction as an annotative practice—one that foregrounds ambiguity, discrepancy, and uncertainty as primary objects of historical inquiry.
In practice, the records of Chosŏn drifters preserved in sources such as P’yohaerok are often imprecise, and even official documents that purport to record the drifters’ testimonies verbatim frequently contain inaccurate geographical information. Yet insofar as the drifters’ arrival on the coast under Chinese imperial jurisdiction and their eventual safe return to Chosŏn are established historical facts, it is equally clear that the paths they traversed must have constituted a route, however fragmented it may appear in the sources. For this reason, this paper approaches place names appearing in drifter-related records not as neutral data but as capta—that is, information taken rather than given, produced and mediated through specific institutional and narrative contexts (Drucker 2011). By distinguishing manifest transcriptional errors from interpretively meaningful misalignments along the repatriation routes, this paper demonstrates that GIS-based reconstruction is not merely a technical exercise, but a means of reconstructing and annotating Chosŏn drifters’ repatriation routes at a finer, micro-historical scale.
A concrete example illustrates this point. Figure 1 presents a partial reconstruction of the repatriation route of Yi Pang-ik, a Chosŏn drifter who was cast ashore on the Penghu Islands after drifting off the coast of Jeju in 1796 (Nam and Bae 2021). As the map shows, rather than proceeding directly northward to Fujian along the shortest maritime route, Yi was first transferred southward via Taiwan before being conveyed to the provincial authorities. This detour was neither the result of mistaken testimony nor of erroneous recording. Instead, it reflected an administrative principle embedded in Qing governance: because the Penghu Islands were administered as a ting (廳) under the jurisdiction of a Taiwanese fu (府), Chosŏn drifters were transferred sequentially from ting to fu, and from fu to sheng (省). This case demonstrates how GIS-based reconstruction makes visible otherwise implicit principles of escort and transfer governing the repatriation of Chosŏn drifters, thereby offering a concrete example of annotation as an interpretive practice.
Figure 1. Partial reconstruction of Yi Pang-ik’s repatriation route.
This paper reconstructs the repatriation routes of Chosŏn drifters during the Ming–Qing period by using GIS not as a tool for cartographic correction, but as a layered analytical environment. The GIS dataset is organized into three levels. First, it draws on data from the Harvard-Fudan CHGIS, a pioneering achievement in the field, to construct the base map. Second, by consulting Ming and Qing local gazetteers and administrative documents, it restores the spatial foundations of drifter movement by distinguishing different types of toponyms: administrative places organized according to the administrative hierarchy of sheng, fu, zhou (州), xian (縣), and military-administrative units; human-geographical toponyms; transport-related names; and natural features such as mountains and rivers. Each place name is assigned a set of attributes, including index number, name, administrative level, source attestation, longitude and latitude, present-day administrative location, and implementation status. Third, the dataset incorporates official records produced by the Chosŏn court alongside records derived from the first-hand experiences of repatriated drifters, including P’yohaerok. On this basis, the movement of drifters is represented not merely through discrete locations, but through route segments linking those places and administrative transfer points along the route.
Through these steps, the study organizes terrain, place-based attributes, and movement sequences within an integrated spatial database and visualizes information drawn from heterogeneous sources as multiple GIS layers. Particular attention is paid to points where spatial information diverges across sources, including mutually inconsistent records or spatially anomalous movements. Rather than resolving such differences in favor of a single definitive route, the study preserves them within this layered GIS environment, enabling systematic comparison among recorded routes, administrative boundaries, and plausible paths.
Within this framework, the paper introduces the concept of “meaningful misalignment.” Meaningful misalignments are defined as spatial discrepancies that recur in patterned ways and correspond to institutional priorities rather than random error. These include detours that privilege jurisdictional continuity over geographic proximity, repeated routing through specific administrative nodes regardless of distance, and the selective representation of inland administrative space at the expense of maritime segments. Such patterns emerge only when inconsistencies are retained and explicitly annotated, rather than eliminated.
Through this approach, the paper identifies a set of micro-level principles governing repatriation practices under Ming and Qing rule. First, repatriation prioritized administrative accountability over spatial efficiency: shipwrecked individuals were transferred along chains of responsibility aligned with bureaucratic hierarchies rather than the shortest routes. Second, jurisdictional boundaries functioned as organizing principles of movement, producing routes that appear geographically irrational but institutionally coherent. Such routes involved not only the transfer of drifters from one administrative point to another, but also episodes in which, in some instances, escorts and drifters, traveling together under official supervision, deviated from prescribed paths for purposes that included detours, informal visits, or excursion-like activities. Third, textual representations of repatriation routes systematically minimized maritime mobility, even in cases where sea travel was unavoidable, reflecting administrative preferences embedded in record-keeping practices.
Crucially, these principles emerge between spatial points rather than at points themselves. Individual locations are often stable and identifiable, but the relationships connecting them—routes, transitions, and sequences of transfer—are where administrative order becomes visible. The paper therefore argues that repatriation order in Ming–Qing China was relational and processual, produced through movement across administrative space rather than through cartographic plausibility alone.
By treating GIS-based reconstruction as annotation, this paper contributes directly to the Annotating track of DH2026. Annotation is understood here not as the addition of explanatory labels, but as an interpretive act that makes visible the epistemic assumptions and institutional logics embedded in historical data. The GIS map functions as a layered argument, in which misalignments are marked, compared, and contextualized rather than corrected.
Methodologically, the paper demonstrates how digital tools can support interpretive workflows that resist premature closure. Instead of resolving contradictions, the GIS workflow holds them in tension, allowing historians to interrogate why certain inconsistencies persist and what they reveal about governance and knowledge production. This approach offers a model for spatial annotation applicable to other historical contexts characterized by fragmented or contradictory spatial data.
Theoretically, the paper intervenes in discussions of early modern governance by showing how administrative order was enacted through practices that tolerated, and indeed relied upon, spatial irregularity. In Ming–Qing repatriation practices, order was not imposed by eliminating inconsistency, but by managing it within bureaucratic frameworks. Spatial misalignment thus appears not as evidence of administrative failure, but as a trace of governance in operation.
In conclusion, this paper argues that annotating misalignment enables a rethinking of both historical mobility and digital methodology. By finding order(s) between spatial points rather than along idealized routes, it demonstrates how GIS-based annotation can reveal the micro-logics of historical institutions. Spatial inconsistency emerges not as an obstacle to analysis, but as a privileged site for understanding how historical order was produced, negotiated, and recorded in early modern East Asia.