Introduction

The spatial turn of recent years has brought a number of novel landscapes into focus for scholars of East Asia. One such frontier-located at the intersection of urban development, state power, and territorialization-provides the conceptual ground for the papers guest edited by You-tien Hsing in this issue of Cross-Currents. Another-the domain of imperial cartography-undergirds the present collection of articles.

Old maps have gained new life in the academy. No longer read solely for locational data (or evaluated in terms of scientific accuracy), maps are increasingly seen as cultural artifacts that bear on a wide spectrum of social and political problems. From the worldviews and spatial imaginations of their makers to the economic and ideological projects they advanced, historical maps speak to fundamental issues of both social scientific and humanistic inquiry. Informed by new interpretive questions from cultural geography and visual studies, and armed with new techniques of digital visualization and analysis, curious scholars from across the disciplines are turning their attention to historical maps. In the process, cartographic archives from Siam to Siberia are coming into public view.

One of the latest such archives to make its way into the public domain is the corpus of Japanese military and imperial maps known as gaihōzu (外邦図), or "maps of outer lands." Starting in the early Meiji (1868-1912) era, the Land Survey Department of the General Staff Headquarters (the former Japanese army) was charged with an ambitious mandate: to map select territories beyond Japan's borders. Beginning with secretive surveys conducted in areas where the government was contemplating military action, this cartographic commission steadily expanded to encompass delineation of inter-imperial boundaries, cadastral surveys of the colonies, and detailed drawings of strategic cities and fortifications. By 1945, the lands that had fallen under the umbrella of the gaihōzu ranged from Alaska and Siberia in the north to Australia in the south, and from Micronesia in the east to India, Pakistan, and even Madagascar in the west. The long-running effort to map this vast territory eventually resulted in a massive, heterogeneous corpus.

It also gave rise to a taxonomic conundrum. The category of gaihō, or "outer lands," was anything but simple. In theory, the distinction between the domestic and the foreign may have been straightforward, but in practice, Japan's boundaries were highly unstable. Both the dramatic expansion of the Japanese empire between 1895 and 1945 and the assimilationist conceit that animated its ideology ensured that the distinction between inner and outer lands was constantly in flux. Consider the case of Korea. Prior to 1910, the peninsula belonged unambiguously to the realm of the outer. But once it was forcibly annexed to Japan, Korea was notionally brought within the compass of the inner. At that point, the status of Japanese surveys on the peninsula-as well as the level of resources they could bring to bear-changed fundamentally, yielding colonial cartography rather than "outer-lands maps" per se. The same was true wherever formal governors-general were established; provisional, small-scale sketch maps hastily produced behind enemy lines were replaced by systematic, large-scale surveys, yielding standardized topographic sheets of a uniform size and scale. Yet, in common parlance, the category of "outer-lands maps" continued to encompass the full range of these productions, embracing materials produced both before and after formal colonization. This disjuncture is one reason the term gaihōzu defies easy translation or characterization.

A second source of gaihōzu diversity, however, springs from the production process itself. Overseas cartography was an opportunistic affair, with frequent recourse to makeshift methods. The earliest Japanese maps of coastal China, for instance, were patched together from widely divergent sources of information. Observations made by Japanese officers on the ground were superimposed on existing Chinese and European maps, which themselves were of incommensurate types and scales. Nor did this patch-work quality disappear as the empire expanded. On the contrary, wherever Japan's cartographic ambitions ran ahead of its formal empire, the military mapping enterprise continued to make room for eclectic, ad hoc efforts. The resulting archive embraced maps made by disparate means from disparate materials, subsuming Korean, Chinese, Russian, English, Dutch, German, and other sources. One of the lingering challenges for scholars working with such documents today is to assess the provenance-and the reliability-of their putative content.

Equally daunting for postwar scholars was the challenge of access. Only a fraction of the original gaihōzu survived the war. Politically charged as they were, many maps were burned by the retreating imperial army before they could fall into enemy hands. On the home front, a few caches of gaihōzu were rescued from incineration by fast-acting Japanese academics, Allied intelligence officers, and civilian collectors. A major set of etched plates was also captured by the Occupation, allowing many lost maps to be reprinted and deposited alongside remaining originals. Yet, the surviving set of gaihōzu was not only fragmentary; it was also deliberately scattered among more than a dozen repositories, where many of these maps remained sequestered in basements or attics for decades. While major libraries began cataloguing their collections years ago, the contours of the archive as a whole are only now becoming clear, as the last surviving maps are located, catalogued, and accounted for.

The articles featured here grew out of an international symposium on the gaihōzu held at Stanford University in October 2011.¹ The occasion for the conference was the belated discovery that Stanford is among the half dozen universities in the United States to harbor an as-yet uncatalogued collection of Japanese military maps.² Bringing together librarians, geographers, and historians from both sides of the Pacific with generous support from the Japan Society for the Promotion of Science, the Stanford symposium had two fundamental aims. The first was to inform a wider scholarly public about the origins, character, and coverage of the little-known gaihōzu corpus. To that end, Shigeru Kobayashi of Osaka University, whose pioneering research on these maps has recently appeared in both monographic and popular form in Japan, was invited to serve as keynote speaker for the symposium. His address-to our knowledge, the first comprehensive introduction to the gaihōzu in English-is the lead article in this issue of Cross-Currents. Offering a magisterial overview of the surviving collections, as well as a deeply informed discussion of the chief institutions and procedures through which the main subsets of these maps were produced, Kobayashi's essay lays essential groundwork for the essays that follow.

The remaining articles address the second aim of the symposium: namely, to showcase the utility of outer-lands maps for East Asian history. Representing the diversity of conference participants, who ranged from graduate students to senior scholars, the authors of these papers offer a suggestive pair of case studies. Each takes up a different subset of maps in pursuit of a fundamentally different problem.

Yoshihisa Tak Matsusaka leads off by retracing the delineation of Russian and Japanese spheres of interest in Manchuria and Inner Mongolia in the early 1900s. Matsusaka's mission at one level is resolutely empirical: to graphically reconstruct the spheres of interest described in the Russo-Japanese accords of 1907 and 1912. Despite widespread recognition of these accords' importance, the actual location of the dividing line that they established has remained elusive. As a result, the first contribution of this thoughtful essay is to sift through the contradictory cartographic and toponymic evidence-starting with maps of Manchuria and Inner Mongolia produced by cartographers of the Kwantung garrison-to try to nail down where the boundary was meant to run. But in Matsusaka's hands, this exercise in interimperial border delineation also serves as a window onto a more elusive process: the imaginative and ideological work through which the Japanese summoned into existence the land that they called "Manmō."

The other case study, by David Fedman, investigates Japanese land surveys in colonial Korea from 1910 to 1918. In a context in which maps were called upon to serve as tools of economic as well as administrative planning, he notes, mapmakers stood at the front lines of empire. To a striking degree, however, their enterprise was a multinational one, requiring close and continual work with members of the colonial population. Fedman draws our attention to the many ways in which Korean laborers, farmers, and bureaucrats interacted with the triangulation survey parties that produced the baseline measurements of the peninsula. His story entails both biographical and technological vignettes, and it is enriched by diagrams of the theodolites that the survey teams carried into the field and photographs of the clerks who plotted their results. Such close attention to mapmaking as a site of imperial interaction gives substance to Fedman's rhetorical analysis of the resulting documents.

As this brief synopsis suggests, Japanese military and imperial maps can speak to the fields of social, diplomatic, and economic history alike. Whether interrogated as evidence for the mentality of their makers, the process of their production, or the content of their data, gaihōzu offer a wealth of scholarly riches. If this forum has one take-home message, it is that those riches have only begun to be tapped. Uneven and fragmentary though they may be, the surviving outer-lands maps promise grist for the colonial historian's mill for years to come. Given the increasing visibility of spatial questions across the disciplines, as well as new developments on the digital front, one can easily imagine them assuming a more prominent role in the colonial archive of the future.³ It is the editors' hope that this issue of Cross-Currents may advance that prospect in some modest way.