The European model and the archive in Japan

Introduction

Japan, as is well known, turned itself into a modern nation following western models from the mid-19th century onwards. History played an important role in this process by giving legitimation to the changes. The government that succeeded the Tokugawa shoguns following the Meiji Restoration in 1868 had barely consolidated itself when in 1869 it issued an imperial rescript ordering the compilation of an official history of Japan. While this move represented a continuation of the tradition of the Chinese dynastic histories, the formation of the academic discipline of history, 20 years later, paralleled similar developments in European countries and North America – indeed, it happened almost contemporaneously. Like in other fields, the Japanese observed western countries and followed their example, sending students abroad and inviting foreign experts, in this case the German Ludwig Rieß (1861–1928). Japanese scholars have often described him as a disciple of Leopold von Ranke and made much of his supposed influence in bringing the tradition of Ranke to Japan. ¹ This narrative serves to legitimize the discipline of history as it has been practised since the establishment of the Department of History at the Imperial University of Tokyo in 1887 and the Department of Japanese History in 1889, the same year that the Meiji Constitution was proclaimed. Just as the modern nation-state with its constitutional monarchy was legitimized through historical narratives, so the modern discipline of history was legitimized through a narrative of its formation. Like all historical narratives, however, this one too needs to be examined critically.

For one thing, neither the government-sponsored official history of the early Meiji years nor the practice of history at the Imperial University (now Tokyo University) ever had a monopoly. I will be focusing on historians in government employ, but writing the history of Japan was always contested territory (Mehl, 1998b; Tanaka, 1993: 31–67). For another, the whole notion of the modern discipline of history as it was established at Tokyo University being primarily a western import needs to be abandoned once and for all. Modernity has all too often been closely associated with European and American history with the result that any sign of modernity in Japan is assumed to be a result of ‘western influence’ and somehow alien to Japan (Goto-Jones, 2009: 1–2). Recent trends in historical scholarship, particularly the attention to transnational history, ‘entanglements’ and local responses to global change (Bynum, 2009; Hilton and Mitter, 2013; Müller and Torp, 2009), may well enable scholars to develop more nuanced approaches. The fundamental changes in the treatment of the past in Japan and other non-western countries were not merely a result of westernization. Seen in the context of global history, they represented a response to broader, complex social and political transformations and geopolitical power relations (Conrad, 2013: 37–41). In order to examine how far the European archival ethos was exported to and appropriated by historians in Japan, I will discuss two aspects. First, I will address the complexities of cultural borrowing by examining the relationships between Japan’s historiographical tradition (historiography is here defined broadly as ‘doing’ history, including both research and writing) and the imported practices, bearing in mind that ‘mere parallelism of expression or production proves nothing’ (Gibb, 1964: 157). All too often in the case of modern Japan ‘parallelism of expression’ has been assumed to reflect western influence. Second, I will look at the wider historical context, particularly globalization and state-building, ‘the two great forces that … shaped the world after 1890’ (Darwin, 2007: 422), and were already in evidence by the time of the Meiji Restoration, when the new government embarked on a course of state-building in response to the encroachment by the western powers.

Particular attention is given to Kume Kunitake (1839–1931), one of the founders of the modern historical discipline and a pioneer of komonjogaku, the study of ancient documents. The term komonjo, translated by Kenkyûsha’s New Japanese–English Dictionary (4th edn, 1974) as ‘ancient [antique] documents; a pal(a)eograph; diplomas’, is central to the study of history in Japan. For now we will call them ‘primary documents’. Kume’s career and work provide a suitable case study, because of the way they link the old and the new. A native of the domain (now prefecture) Saga, he was educated at the domain school and at the Shogunate’s school in Edo (Tokyo), and his education centred on the Chinese classics as was usual at the time, although it did include the study of Japanese and Chinese works about the West. In the years before 1868 he was actively engaged in education and politics in his domain. After 1868 he worked for the central government and from December 1871 to 1873 he travelled to the United States and Europe with the ‘Iwakura Embassy’, of which he subsequently wrote the official report, published in 1878. The following year he joined the Office of Historiography in the government and when it was transferred to Tokyo University in 1888, he was appointed professor, together with his colleagues Shigeno Yasutsugu (1827–1910) and Hoshino Hisashi (1839–1917). Following his dismissal in 1892, he taught from 1895 at Tôkyô Senmon Gakkô, later Waseda University. Much has been made of his ‘western education’ (Mayo, 1973), the proximity of his home domain to Nagasaki with its Dutch trading colony, and his two years abroad. Indeed he was the only first-generation historian in the Department of Japanese History with experience of the West. Nevertheless, his historical scholarship developed on the basis of his traditional education and his work in the Office of Historiography. His direct encounter with western historical scholarship, like that of his colleagues, was limited.

If Kume was indeed the founder of komonjogaku we should expect that the study of komonjo as a fundamental element of the historical discipline grew largely from Japanese roots. Certainly, Japanese scholars had been collecting and working with primary documents long before Rieß arrived. They became self-made experts acquiring their expertise on the job; this can be said to apply to the classification and decoding of manuscripts in general (Barzun and Graff, 2003: 82). Systems of classification are never given, but neither are they completely independent of what is being classified. Japanese komonjo are products of Japan’s society through the ages and their characteristics reflect this (Kondô, 2003: 10–11). The techniques for compiling collections used by the scholars employed by the Meiji government owed much to methods practised even before the Meiji Restoration. ² Indeed, Japanese historians have long argued that Japanese practices significantly influenced the modern discipline (Iwai, 1963). Certainly, the question of western influence must be answered differently for the fields of western, Japanese and East Asian history (Ôkubo, 1988: 56–7). Arguably, Japanese practices were most significant in the field of Japanese history, the most relevant one for the subject of the archive in Japan.

This is not to deny the importance of western models. The emergence of history as an independent discipline with its own canon of methods and practised by specialists at institutions of higher education undoubtedly did owe much to European and especially German models, as did the entire nation-building project of which historiography was a part. The direct influence of European models is evident in the institutional organization of the discipline and the interpretative strategies used by Japanese historians well into the 20^th century in constructing narratives of the national past (Conrad, 1999). My point is, that explaining similarities between Japan and Europe with the import of European models fails to take into account other processes that were at work.

The complexities of cultural borrowing

Cultural borrowing and ‘influence’ are highly complex phenomena, and scholars have sometimes neglected this complexity. George Akita, an eminent historian of Japan who describes himself as a positivist, has pointed out recently just how problematic it is to make facile assumptions about western influence on modernization; for example, the Prussian influence on the making of the Meiji Constitution (Akita, 2008: 118–24). In his discussion Akita refers to three generalizations made by Sir Hamilton Gibb already in 1964 in his discussion of Islamic influence in medieval Europe (Gibb, 1964). Gibb’s view is indeed highly relevant to the case of modern Japan. His first general principle is that cultural influences are always preceded ‘by an already existing activity in the related field’ of the culture doing the borrowing. This is what provides the incentive to borrow in the first place. I believe that in some cases it may be better to speak of a perceived need rather than a ‘related activity’. For historical scholarship existing activities and new needs were equally important. Gibb’s second principle states that that which is borrowed will develop only if it adapts to and blends with the ‘native forces’. This too applies; it has become a cliché to state that Japan has always modified what it adopted from abroad and adapted it to its own needs (not that Japan is exceptional in this respect). Gibb’s third generalization is that the recipient culture ‘disregards or rejects all elements in other cultures which conflict with its own fundamental values, emotional attitudes or aesthetic criteria’. This is closely related to the second in that conflicting elements will not generally lend themselves to adaptation. Here, the adoption of the Rankean historical tradition provides a revealing example if we look at the elements Japanese historians did not adopt. Whether in historical scholarship or any other field, Japanese borrowing was always selective and motivated by Japan’s own values; foreign models were ‘not imposed, but truly self-selected’ (Akita, 2008: 121). No one told the Japanese to adopt western methods of historical scholarship, just as no one forced them ‘to listen to Gounod or Verdi’, as Jürgen Osterhammel observes.

The reason for emphasizing perceived need over ‘existing activity’ is that we have to distinguish areas of borrowing where there was no significant domestic competition and areas where the ‘already existing activity’ motivated the borrowing but at the same time was displaced by it. In many areas of natural science and technology, for example, the presence of related activity is less significant than the need. Even in a field like medicine, however, where western practices were studied since the Tokugawa period, where the Meiji government privileged western medicine from the start and practices of traditional East Asian medicine were discarded by the dominant, western-trained physicians in government institutions, traditional approaches continued to inform medical practice. Western-trained physicians combined new practices with traditional techniques informed by traditional views on the nature of disease (Bay, 2012: 4, 9). If even western medicine, which included fields where there was no significant existing activity (particularly public health and hygiene, vaccinations and knowledge developed in the laboratory rather than through clinical practice), intermingled with rather than displaced existing practices, we should expect something similar in the field of historiography. We might even ask, did western-style historiography have anything useful to offer at all, given the long tradition of historical writing? The answer is not immediately obvious, but I suggest three lines of inquiry. One is that legitimating the modern nation calls for a different kind of historical account than legitimating the ruler in the Chinese tradition. Perhaps the most significant difference in our context is that Japan as a modern nation needed to legitimize itself not only to its own citizens but to the other modern nations it so desperately wanted be accepted by; in other words, both nation-state-building and globalization called for a new kind of historiography. The second is the conception or reconception of relics from the past, including historical documents as national heritage rather than as the property of powerful families or temple and shrines. The third, possibly the most significant, is the notion of historical scholarship as a ‘scientific’ discipline along the lines of the natural sciences; the same relics thus also achieve new significance as scientific evidence. ³ In Japan as in the West before the 19^th century, history was primarily perceived as an art of storytelling in order to entertain and to educate. This changed in the decades following the Meiji Restoration.

History as a government project and the first professional historians

The transition from history in the service of the ruler to history in the service of the nation can here be treated only briefly. ⁴ The Meiji Restoration was initially justified as a return to the past [fukko], namely to the Chinese ideal of the imperial bureaucratic state adopted by Japan from China in the 7^th century. The first political institutions created by the Meiji government were modelled after those of the Nara and Heian periods (8th to 12th centuries), the only time the Chinese ideal was a reality in Japan. One of those institutions was the Office of Historiography, which was intended to compile a history of Japan in the tradition of the Six National Histories [Rikkokushi] compiled between the early 8^th century and the beginning of the 10th, which in turn were modelled on the dynastic histories of China. This indicates how the Meiji leaders in the early years after 1868 were operating under the assumption of an order in East Asia still characterized by Chinese cultural and economic hegemony rather than by western dominance. ⁵ Only gradually did they adjust their thinking and policies to the changed situation created by western encroachment in the region.

Government institutions, including the Office of Historiography, were reorganized several times before the promulgation of the Meiji Constitution in 1889 settled the question of how Japan’s government was to be organized. In 1888 the Office of Historiography was transferred to Tokyo University and in 1895 it became what is now the Historiographical Institute, a research institute dedicated to the collection, organization, study and publication of historical documents (Mehl, 1998a).

As the above illustrates, the Office of Historiography owed nothing to western models. Nevertheless, those employed in the office soon expressed an interest in western methods of historiography. The first clear evidence of this is the memorandum entitled Shûshi jigi [Right Historiography], issued in 1875 and most probably drafted by Shigeno Yasutsugu, one of the most important members of the Office of Historiography at the time and a future pioneer of historical scholarship. The memorandum mostly discusses the best way to compile a seishi, i.e. an official history of Japan in the tradition of dynastic histories. The reference to western historiography is vague, presumably reflecting Shigeno’s limited access to relevant information.

Kume Kunitake, on the other hand, who joined the Office of Historiography in 1879, travelled to the West, and his observations recorded in the official record of the Iwakura mission are revealing (Kume, 2002). We can assume that his views counted, for once appointed to the office, he became the most influential member after Shigeno. The members of the Iwakura mission attempted to learn as much as possible about the countries they visited, and their hosts were usually more than pleased to show off what they perceived as their country’s most impressive achievements: factories, schools, government institutions, judicial courts, prisons, hospitals, dockyards, naval bases, military and naval academies, stock exchanges, churches and libraries, but also zoos, botanical gardens, parks and theatres. Naturally the Japanese visitors were particularly intent on studying the institutions and practices that made the western nations so powerful in the modern world. Kume’s report shows how much he and (presumably) his fellow travellers were impressed not just by the most modern inventions, but also by the care devoted to preserving the past. Already after describing the mission’s visit to the British Museum, he observed: ‘Progress does not mean discarding what is old and contriving something which is entirely new’ (ibid.: II, 109–10; original emphases). Describing their visit to the Bibliothèque Nationale, he makes the point yet more forcefully:

Whenever we visited a library or museum in the West, we saw such comprehensive arrangements made that even objects from far-away countries in the East were gathered and catalogued, no matter what the cost or effort involved.… At the root of the march of progress in the West is a profound love of antiquity.… It is the accumulation of knowledge over hundreds and thousands of years which gives rise to the light of civilisation. (Kume, 2002: III, 59–60)

Kume again stresses the connection between progress and the preservation of knowledge when he reports the mission’s visit to the Biblioteca del Reale Archivio di Stato in Venice on 29 May 1873. Kume writes:

When knowledge and skills are developing, there is nothing at all to be thrown away, and where the literary arts are flourishing, books and documents are treasured. It is not the case that the benefits of machinery have only been appreciated since the discovery of steampower or electricity, or that books are only seen as treasures after the craft of politics or scientific knowledge have been elaborated in their pages. Just as the carpenter’s square used by our artisans is based on the principles of geometry, and the bellows used by blacksmiths are based on pumps, so even in the smallest devices there are scientific principles, and the development of those principles is called ‘progress’. Thus, the books and ledgers of towns and cities are the origins of commercial and civil law, and the peace and good order of any country depend on them; this is probably as a result of those books being valued. The official documents of the courts are precious, and if they are treated lightly, then in the end the laws of that country will inevitably fall into disuse. This is the natural course of affairs. In the West, they have museums where they keep even the most insignificant objects and libraries where they store even discarded scraps of paper. This can be seen as the essence of civilized behaviour. (Kume, 2002: IV, 352–3)

I have deliberately quoted the entire passage because of the light it sheds on Kume’s thoughts about scientific knowledge in the West. First, Kume repeats the observation he made before about the link between western civilization and progress and the care taken to preserve relics from the past. Second, he passes easily from books and documents to technology and scientific principles and back to documents. Third, he links ‘the books and ledgers of towns and cities’ to peace, law and order; in other words, their preservation has a clear practical purpose: that of preserving order. The passage needs to be seen in the context of the overall agenda of Kume’s report and of several of the Meiji leaders. It was generally acknowledged that ‘restoration’ [fukko] had become ‘renewal’ [ishin]. But these leaders emphasized the need for selective and gradual reforms while paying due attention to Japan’s own traditions. The need to take into account Japan’s own history was stressed in connection with the Meiji Constitution of 1889 (Akita, 2008: 89–90; Mehl, 1998a: 28–34).

The Venetian archives also held a more specific interest for the Japanese delegates; in them were kept seven letters from Japanese rulers and their representatives of the late 16^th century, when a legation of Japanese Christian converts travelled to Europe. The mission requested that the letters be copied and Kume was asked by Iwakura Tomomi to copy the signatures on them. Kume observed that since the Japanese documents relating to the feudal lords’ dealings with the Portuguese were lost, evidence was only to be found in the western historical documents. ⁶ We can only imagine the thrill Kume and his fellow-travellers must have felt at seeing these documents. Their experience suggests that when 16 years later Rieß advised the Japanese to collect documents relating to Japan in foreign archives, the ground was already prepared.

It is surely safe to assume that once the Iwakura mission returned to Japan and Kume completed his report, the members of the Office of Historiography learned about how in Europe knowledge was preserved in libraries, archives and museums, and disseminated through universities and schools. Certainly Meiji Japan saw the creation of public libraries and museums and the collection and preservation of primary documents, although the government did not establish national archives governed by legislation ruling what documents were to be passed to them. Several institutions were engaged in collecting documents, including documents about the past, as part of the Meiji government’s quest for control over the entire country. Even the early activities of the Office of Historiography were just as much for practical purposes of government as for writing history. This is not the place to discuss institutions in detail, but when we speak of ‘the archive’ in Japan, it is more meaningful to see the emergence of archives (or at least collections of documents by government institutions) as a response to specific needs in a given historical situation – namely the needs of newly established government institutions to secure and organize knowledge, including knowledge of the past, for the purpose of political control and introducing the reforms needed to fulfil the vision of a strong nation that could compete internationally. In other words, collecting documents relating to the past was an integral part of state-building.

Still, the members of the Office of Historiography wished to learn more about western historiography. In 1878 they saw their chance. They entrusted Suematsu Kenchô (1855–1920), who was on his way to London as a secretary to the Japanese Legation, with ‘the investigation of English and French methods of historical compilation’. ⁷ The result of these vague instructions was a work Suematsu commissioned from the scholar and lecturer George Gustavus Zerffi (1821–92), an impressive tome of 773 pages entitled The Science of History and published in 1879. Rather than a kind of practical handbook, which is likely to have been what the members of the Office of Historiography had in mind, the book was a chronological survey of European historiography from ancient times to the present. The translation was not completed until 1887, the year Rieß came to Japan, and the work’s influence was limited. There is, however, evidence that Shigeno at least studied the early translation.

If the book itself was of minor importance, the title points to the third aspect of western historiography of interest to the Japanese: the idea of history as a science. Whether and to what extent the idea influenced the actual practices of historians is not easy to answer, but it certainly provided another source of legitimation. It encouraged them in their critique of the moralizing historiography of the Confucian tradition and in their tendency towards empiricism (Tanaka, 1999: 409). Tsuboi Kumezô (1858–1930), who – significantly – trained as a natural scientist before he was sent to Europe to study history in June 1887, coined the terms ‘applied history’ and ‘pure history’ in his article ‘On History’, published in 1894. Kume Kunitake made a similar distinction when he spoke of those who applied history in contrast to the specialists (Kume, 1895; Tsuboi, 1894). Needless to say, Tsuboi, Kume and their colleagues at Tokyo University regarded themselves as ‘pure’ historians or ‘specialists’. They devoted themselves to the collection and organization of sources. Both Tsuboi and Kume were writing after the Historiographical Institute’s temporary closure in 1893, following the so-called ‘Kume Affair’ the year before. The closure put an end to the compilation of an official history. When the Historiographical Institute reopened in 1895 the collection of and research into historical documents became its exclusive task (Mehl, 1998a: 126–46).

This new departure reflected the change in priorities among the scholars working at the institute, which had already begun in the Office of Historiography in the 1880s when the members concentrated most of their efforts on collecting documents. Initially the aim was still to produce an officially sanctioned history of Japan, and, for decades after the move, people referred to the institute as the ‘History Office’ [shikyoku]. Kume, Shigeno and Hoshino were appointed professors of history when the Office of Historiography was moved to the university, but it is said that Hoshino at least keenly felt a severe loss of status and hoped to return to government service once the history was completed. Following the move, a Department of Japanese History was established in 1889, the year the Meiji Constitution was promulgated. In November 1888 Ludwig Rieß had submitted a proposal suggesting the establishment, but its timing and the fact that it was solicited by the president of Tokyo University suggest that it was in reality a foregone conclusion. If Rieß’ opinion counted at all, it was presumably concerning questions of detail. ⁸

Naturally, Rieß was expected to teach European history and methodology. ⁹ The history of China and Japan had previously been taught in the Department of Classics, which existed from 1882 to 1887 and produced several notable scholars. From 1890 onwards Shigeno, Kume and Hoshino taught courses in the history of China and Japan, although Kume later admitted that their work at the Historiographical Institute took priority.

The former government officials were now members of the academic community at the Tokyo University, which included Rieß and other foreign professors. In November 1889 the Historical Society (Shigakukai) was founded which published its own journal (Shigakukai zasshi; from 1892 Shigaku zasshi), and Rieß became a member, as did his foreign colleague Basil Hall Chamberlain (1850–1935). Shigeno became the president, a post he held until his death in 1910. There is some evidence of exchange between Rieß and the scholars of the Historiographical Institute, but it was limited. For one thing, they had no language in common. Shigeno and his colleagues, moreover, were busy professionals, not students. Even if they had been, Rieß could not have taught them how to use his own methods to work with Japanese primary documents, which he could not study in the original himself. For his research based on Japanese sources he had to rely on translations by his students. They, on the other hand, had to apply the general principles he taught them to Japanese documents. But then they studied with the Japanese scholars as well as with Rieß who surely realized that Japanese scholars would have to adapt his approach to the particular characteristics of their empirical material.

Collecting and studying primary documents

Modern historical scholarship is characterized by its emphasis on the study of primary sources. This was also true in Japan, where the collection of documents at the Historiographical Institute, the former Office of Historiography, predates any significant western influence. In the last years of the government office, its members travelled around the country and systematically collected documents; it is not an exaggeration to say that this activity began to constitute an end in itself. Earlier field trips appear to have been undertaken to look at known collections in order to clarify specific issues, but this changed in the 1880s, when work on the chronological history began in earnest. As early as 1882, a member of the office drew attention to collections of hitherto unknown documents and urged that they be examined and secured immediately before they were lost. Local officials, he asserted, could not be relied upon. The task was so important that money for it should be saved elsewhere, if necessary. ¹⁰ From 1885 onwards members of the office conducted extensive field trips in order to survey and collect documents. This was partly because in 1884 the task of collecting local documents had been transferred from the prefectural authorities to the central government and the Office of Historiography. On that occasion the office had stressed the necessity of collecting the documents and asked that the office receive the money previously allocated to the prefectures to be used for this purpose. In 1885 the office issued a memorandum, which clearly illustrates the importance its members attached to the search for primary documents. Historiography, the memorandum stated, relies on two types of sources: journals and primary documents [nikki monjo] which provide the basis [konkyô] and military and war tales [gunki senki monogatari] which are consulted for reference [sankô]. The tales are recorded by later generations and are less reliable; but being easier to obtain, they are referred to more. The office wished to send representatives out, both to negotiate the transfer to the office of the work formerly done by the prefectures and to investigate new sources, and the memorandum included a list of proposed locations, and a budget (see Komonjo sôhô ikensho, Shimatsu 12 [1884]).

That year, 1885, Shigeno led a field trip to several prefectures around Tokyo lasting 81 days. In 1887, Kume travelled in Kyushu for 150 days. In the process of their fieldwork the members of the office used their authority as government officials to obtain access to documents held by temples, shrines and powerful families as their private property. As a result of this work, the Office of Historiography in 1887 published two volumes containing facsimiles of primary documents with explanations, the first publication of its kind. ¹¹ The second volume, published in 1889, included documents collected by Kume, and the commentaries, published separately, are probably his own.

The increasingly high priority given to the collection of primary documents went hand in hand with a highly critical appraisal of the narrative histories [monogatari] that had formed the basis of previous historical writing. Members of the Office of Historiography met regularly to discuss issues arising from their work, and at several meetings Shigeno, Hoshino and Kume discussed examples of these histories. In March 1885 Kume gave a presentation on the Taiheiki [Chronicle of the Great Pacification], one of the classics of historical writing in the genre of war tales. ¹² Its particular significance lay in the fact that it chronicles the war arising from the retired emperor Godaigo’s attempt to restore imperial supremacy in the 14^th century, which ended with the schism into the Northern and Southern Courts. Because of the place the emperor-centred ideology had in legitimizing the Meiji state, this period in Japanese history was particularly problematic. It had already confounded the scholars compiling the Dai Nihon shi [History of Great Japan] for the domain of Mito in the Tokugawa era, a history in the Chinese tradition, begun with the aim of legitimizing shogunal rule by showing how the shoguns received their mandate from the unbroken line of emperors. As Kume Kunitake later pointed out, the editors of Dai Nihon shi relied mainly on narrative sources, particularly the Taiheiki. They did collect primary documents, but only for reference [sankô]. If we take Kume’s word for it, then the reversal of priorities by him and his colleagues was indeed revolutionary. ¹³ In his critique of the Taiheiki, Kume stated that it was widely regarded as a reliable history, although only 20 to 30 per cent of its contents were supported by other evidence. Kume presented many examples of its unreliability, anticipating the content of his famous (or infamous) article ‘Taiheiki wa shigaku ni eki nashi’ [The Taiheiki is of No Use for Historical Scholarship] (Kume, 1891).

As mentioned earlier, the members of the Office of Historiography acquired their historiographical skills on the job. While collecting, transcribing and organizing the primary documents that were to provide the basis for the official history, Shigeno, Kume and their colleagues refined traditional methods of textual criticism, known as kôshôgaku, a school of textual criticism that originated in China and was developed further by scholars of the Tokugawa period (1600–1868), who applied its principles to Japanese texts. Shigeno’s role in the formation of the historical discipline and the nature of his contribution to scholarship need not detain us here (Mehl, 1998a: 87–93; Mehl, 2000; Tao, 1999). Suffice it to say that Shigeno was known to his contemporaries primarily as a scholar of Chinese learning or kangaku. Throughout his career he looked upon China as the source of true learning, but that does not mean he was a rigid conservative. He contributed to the renewal of Chinese studies as well as historical scholarship, but he did so from a firm basis in the kangaku education he had received in the Tokugawa era. This does not mean that he did not learn from the West as well, and scholars differ in their assessment of his scholarship as ‘traditional’ or ‘modern’, indebted to kangaku or to western scholarship. ¹⁴

Shigeno described kôshôgaku as a method consisting of collecting all the written evidence to establish the facts (Shigeno, 1989[1890]). Using the method to study the primary documents that were now perceived as the key sources for historiography was an innovation. We might compare it with the way Leopold Ranke applied methods of textual criticism practised in the study of theology and law to his historical sources. Kôshôgaku became a dominant element in the school of historiography known as akademizumu, which stressed the collection of primary documents and the verification of facts and became the mainstream of historical scholarship at public institutions (Mehl, 1998a: 103–13). Another element was the one-sided reception of Ranke as the ‘quasi-positivist’ (Breisach, 1994; Mehl, 1998a: 161). Akademizumu represents a good example of Gibb’s general principles of cultural borrowing; only those aspects of Ranke were adopted which fitted in well with existing scholarly activities and concepts: the emphasis on writing history based on primary documents and impartially, without moralizing.

But what of the primary documents themselves, the study of which lay at the core of akademizumu? They are Japan’s own; their characteristics reflect Japan’s political, economic, social and cultural developments. To borrow Gibb’s words again, the documents are among the things which ‘retain their distinguishing characteristics within each separate culture’ (Gibb, 1964: 167). Their nature had to determine the methods by which they were studied and ultimately these methods could be developed only from within. Here we turn to Kume Kunitake’s contribution to the study of komonjo.

Kume Kunitake and komonjogaku

Although he is not always given credit for it, there is good reason to describe Kume Kunitake as the primary founder of what is known as komonjogaku, the study or even the science of komonjo. ¹⁵ So far I have translated komonjo as ‘primary documents’, but I will now retain the Japanese term komonjo, for reasons which will soon become apparent. The word komonjo was already used by scholars in the 17th and 18th centuries, but it did not become a commonly used technical term until the times of Shigeno, Kume and Hoshino.

Kume and Shigeno had much in common, including their education in the Chinese tradition (kangaku) starting with the Confucian classics and including Chinese histories (Hoshino’s background was different, but he too was educated in the kangaku tradition). The most significant difference between Kume and Shigeno was the fact that he personally travelled abroad in his early 30s. Shigeno did not travel outside Japan until late in life. After the Iwakura mission’s return in summer 1873 and while writing the official account of the trip, Kume also organized the numerous documents concerning the mission for the Council of State, as well as arranging the council's classified documents. Once appointed to the Office of Historiography he soon became a leading member.

Kume, was perhaps the most original and certainly the most eclectic of the three scholars, and his experience of foreign travel may well have had something to do with it. Japanese historians have often contrasted the akademizumu school of historiography with the school of Enlightenment history or ‘history of civilization’ [bunmeishi], whose members sought to understand the history of Japan by placing it in the framework of universal history. They were inspired by translated western works like François Guizot’s Histoire générale de la civilisation en Europe (trans. 1874–7), and Thomas Henry Buckle’s History of Civilisation in England (trans. 1875; Buckle’s own dates were 1821–62). The most important early works were Bunmeiron no gairyaku [Outline of a Theory of Civilization, 1875] by Fukuzawa Yukichi and Nihon kaika shôshi [A Short History of Enlightenment in Japan, 1877–82] by Taguchi Ukichi (whose own dates were 1787–1874). These authors were not affiliated with public institutions. Their claim to writing ‘scientific’ history rested on their belief that history was governed by general laws, which applied to all societies, including Japan. Despite obvious differences between the two schools, the representatives of both believed that history should be ‘scientific’ in the sense of rational and progressive and free from moral bias (Tanaka, 1993: 40–1).

Kume had a foot in both camps. His record of the Iwakura mission and his observations about history in some of his later works show that he had absorbed the notion of universal progress, and he had an interest in the natural sciences. He regularly wrote for Taguchi Ukichi’s history journal Shikai, which aimed at a wider audience than the Historical Society’s journal.

It was the publication of Kume’s article ‘Shintô was saiten no kozku’ [Shinto is an Outdated Custom of Heaven Worship] in Shikai that led to his dismissal from Tokyo University in March 1892 which marked the end of his affiliation with public institutions that characterized the akademizumu historians in contrast to the Enlightenment historians. His article on Shinto, moreover, is more representative of the Enlightenment historians’ approach to history as progress. His main contention is that Shinto is a primitive form of worship that precedes the development of religion. The ‘Kume Affair’ needs no further discussion here; the attention it has received as a clash between scholarship and ideology has detracted from evaluating Kume’s importance for historical scholarship. ¹⁶

Kume’s komonjogaku was dogged by one great dilemma. During the first part of his career as a historian, while in government employment he faced the double task of compiling primary documents and producing the official history. As time went on, the sheer numbers of documents unearthed proved overwhelming; Kume later remarked that he could not devote sufficient time to handling and studying them. His dismissal gave him the freedom to pursue his personal interests and the time to reflect, systematize and synthesize. But it also cut him off from the very documents that had fed many of his interests and allowed him to develop his systematic approach to the study of komonjo. Except for the published collections they were not accessible to people outside the institution.

Kume’s major historical works, including his published writings on komonjogaku, were produced during his long tenure at Tôkyô Senmon Gakkô, the predecessor of Waseda University, from 1895 to 1922. Indeed, it is hard to imagine how he could have published as much as he did had he continued to work at the Historiographical Institute. Besides numerous articles in journals, Kume’s works include Nihon kodaishi [A History of Ancient Japan], Nara jidaishi [A History of the Nara Period] and Nanbokuchô jidaishi [A History of the Period of the Northern and Southern Courts]. The books originally appeared as printed lecture notes intended for students taking correspondence courses.

The same is true of Kume’s lectures on komonjogaku which were published serially by the university in autumn 1901. A critique of them appeared in the Shigaku zasshi, to which Kume published a response. In 1904 a revised version Komonjogaku was published in book form (Ôkubo, 1991: 483–520). According to Kume’s own evidence, the manuscript that formed the basis of his published work Komonjogaku originated in notes he made when he and his colleagues from the Office of Historiography ran a research group that held weekly meetings. Members of the office, including Kume, had met to discuss research throughout the 1880s, but the komonjo research group seems to have been formed after moving to the university in order to discuss questions arising from the study of the primary documents and to train researchers. This first manuscript is incomplete, Kume’s dismissal having put an end to his direct involvement (Kume, 1989: 1–82). The manuscript begins with a classification of komonjo; the following chapters deal with periodization, the characteristics of komonjo (including style, language, paper and ink as well as types of contents), copies and forgeries. The next section is the first part of a chronological discussion of documents with examples. Unlike Kume’s lecture notes on the subject, published later, these early notes are clearly not addressed to beginners.

The 1904 version of Komonjogaku represents Kume’s most thorough discussion of the subject (Kume, 1989: 109–549). He begins by stating that to study komonjogaku is to study the core of history; komonjo provide indispensable evidence for historical facts. In Japan, according to Kume, historiography did not progress once compilation by the imperial court ceased. Documents were kept in leading families, temples and shrines and the writing of history, with few exceptions, ceased to be based on documents and became mere storytelling or stage plays. Komonjogaku is a new discipline [gakka], a result of the new and progressive scholarship that emerged after the Meiji Restoration. Kume briefly discusses the collection and study of komonjo in the Tokugawa era and concludes that ultimately scholars did not make them the basis of their historical writing, partly because they had limited access to documents. The work of the Office of Historiography brought many komonjo to light and they formed the basis for systematic research. Having thus introduced his subject, Kume outlines the characteristics of komonjo and provides a classification. The defining characteristic of komonjo is that they were not produced with a view to writing history, but because of a need arising at a given time. Kume uses the term komonjo both as a generic term and for one of the five groups of documents he identifies:

Having explained the different groups, Kume briefly discusses ‘non-komonjo’; inscriptions in stone or metal, he says, are not komonjo, because they are written as a message to posterity rather than arising from current business and thus often manipulate the facts. Although he concedes that the authors of genealogies do this too, he maintains that their work was originally compiled for a different purpose. In the following chapters Kume discusses the language of komonjo, their age, the characteristics of different groups of komonjo in different historical periods, the changes occurring through the ages and their forms and characteristics, the distinctions between originals, copies, drafts and forgeries. Kume concludes with a section entitled ‘Discernment’ [kanshiki]. Discernment is an ability that the researcher of komonjo needs to cultivate and which includes both direct experience or observation, and knowledge [jikken and gakumon] as well as detachment. Komonjo, Kume asserted, are not things to be treated with reverence like antiques or old books. Nor should they be kept hidden away as they were throughout most of Japan’s history.

The idea that komonjo should be made accessible and be scrutinized with scientific detachment illustrates a break with tradition that resulted from or was at least accelerated by the confrontation with the West. During his travels Kume had observed institutions that made knowledge publicly available. Even if access to archives was in practice restricted, the collection and storage of documents in a known location contrasted starkly with Japan, where komonjo were hoarded by families or religious institutions as their private and often secret property. Equally desirable to Kume was the scientific approach to knowledge, and many of his lectures which found their way into print include references to and comparisons with the natural sciences. For komonjogaku a good example is his article ‘Komonjo no kansatsu’, published in 1896 in Shigaku zasshi (Kume, 1989: 83–103). ‘Kansatsu’ [observation] was according to Kume a word much in vogue at the time as an expression for scientific observation. Kume wanted to apply the kind of observation used in the natural sciences to the study of komonjo as the basis for the study of history. Komonjo are to the historian what natural phenomena are to the natural scientists; several times he uses the microscope as an image to illustrate the close scrutiny and attention to detail the historian should cultivate.

Even while Kume advocated the close scrutiny of komonjo, he was critical of the emerging historical school of akademizumu with its positivism based on textual criticism. Textual criticism, as Kume asserted in his lecture ‘Shigaku kôshô no hei’ [The Abuse of kôshô in Historical Scholarship] (1901), was the means, not the end, of historical scholarship and should serve as the starting point for further research and for interpretation by the individual historian (Kume, 1901). He rejected the idea of waiting until all the facts were established before beginning independent research. Kume’s ideal was historiography contributing to the progress of human society by helping to avoid the mistakes of former generations. He rejected the idea that the study of history should serve to emulate the heroes of former times; this would make progress impossible.

 Today Kume’s contribution to historical scholarship, like Shigeno’s, is all but forgotten. His komonjogaku did not go much beyond the classification of historical documents and was soon regarded as outdated by the mainstream historians who came after him. It is usually Kuroita Katsumi (1874–1946), who is credited with having founded the field of komonjogaku. He graduated from the University of Tokyo in 1896, where he was a lecturer from 1902 and professor of Japanese history from 1919 to 1935. From 1919 to 1920 he was the director of the Historiographical Institute; thus he was a member of the third generation of scholars, and the first generation of graduates from Rieß’ seminars to join the Historiographical Institute. As well as with Rieß, Kuroita had studied with Kume’s former colleagues Hoshino Hisashi and Tanaka Yoshinari (1860–1919), and in 1902 took over Hoshino’s lecture course on komonjo. For his doctoral thesis he presented a discussion of Japan’s komonjo (Nihon komonjo yôshiki ron, 1903). Although he refers to Kume, Kuroita’s work is believed to be independent of Kume’s. The various influences on Kuroita’s komonjogaku, however, still await thorough research. ¹⁷

Conclusion

I have argued that we must be cautious in what we conclude from similarities in the development of komonjogaku in Japan and the ‘archival turn’ in Europe in the 19^th century. Further detailed studies are needed into the history of archival studies and practices in Japan, including practices before 1868 and in the period before around 1890, the work produced by those regarded as pioneers of modern historical scholarship like Shigeno, Kume and Hoshino and their relationship with their students and successors, including Kuroita Katsumi who today is credited with having pioneered komonjogaku. ¹⁸ Nevertheless it is clear even now that the collection of primary documents as a basis for writing a national history predated any significant influence stemming from western historical scholarship. The first generation of scholars at the Historiographical Institute amassed vast amounts of documents. At the same time, the similarities, real or perceived, between Japanese and European practices meant that adopting the latter was not necessarily seen as an imposition of alien cultural practices; in fact the foreign models could and often did serve to justify existing practices rather than create something entirely new. There is, moreover, reason to believe that there are significant differences in perception of the archive’s function to this day. Although government institutions in Meiji Japan collected documents, the establishment of national archives was not discussed when the government was reorganized in anticipation of the constitution in the mid-1880s, and the present national archives (Kokuritsu Kôbunshokan) were established only in 1971 (Kondô, 2003).

More significant than the way ‘existing activities’ intermingled with imported practices, we might argue, is that ‘parallelism of expression’ (in our case similar views of history and methodological approaches) also resulted from the changed geopolitical power relations (Conrad, 2013: 40–1). In other words, Japan and European nations were responding to challenges posed by the same global situation. Christopher Hill, comparing Japan, France and the United States, ‘examines the representation of national history … from the point of view of the single modernity of which they were part’ and argues that writing national history ‘was a way to imagine the place of the nation in the world and its relationship to preceding forms of sovereignty and community at a time when the consolidation of capitalist markets and the international system of states was changing the shape of the world’ (Hill, 2008: ix). National histories not only legitimized the individual nation-state, but ‘naturalize[d] the division of the world into a system of nation-states’ (ibid.: 155). Hill (ibid.: 278–9) sees no dichotomy between the global and the local; his framework allows him to accommodate both the asymmetrical political and economic relations on the global level and the local differences. Hill chose France for his European country of comparison, because in France, unlike Germany, the state predated a sense of nationality. For the Meiji leaders, meanwhile, Germany seemed a more apt comparison, because they saw similarities between the process of German unification in 1871 and the abolition of the feudal domains and centralization of Japan in the same year (Mehl, 1998a: 154–66).

Japan escaped formal colonization, although not the threat of losing its independence, nor the injustices of free trade imperialism. Responding to western encroachment and the breakdown of the Sino-centric order in East Asia, it embarked on one of the most successful cases of nation-building in the 19^th century. By about 1900 it had transformed into ‘one of the most densely integrated states in the world’ (Osterhammel, 2009: 596). This process of centralization and nation-building is sometimes described with the terms ‘internal colonization’ and ‘domestic imperialism’ (Bay, 2012: 5–6). In fact there are parallels between the Meiji leaders’ approach to state-building in the home islands and to empire-building by assimilating peripheral areas. Japan’s colonization of Hokkaidô, Okinawa, Taiwan and Korea ‘may be seen on one level, at least, to be an extension of the question of Tokyo’s relationship with non-Tokyo entities’ (Akita, 2008: 78–87). Imperial expansion included the colonization of memory as the assimilated territories were positioned in Japan’s past. In Korea, research into documents and artefacts with the methods of modern scholarship began with Japan’s quest for dominance on the continent. Scholars at Tokyo University collected, compiled and studied Korean texts and published articles in the first issues of Shigaku zasshi and other scholarly journals throughout the 1890s. After Japan’s victory over Russia in 1905, Russian rights over the South Manchurian Railway passed to Japan and in 1906 the South Manchurian Railway Company was founded to develop and exploit the railway zone. The company had a research section (Nantetsu Chôsabu) which employed scholars to conduct scientific studies of Manchuria and Korea, including history and archaeology. Among the researchers were graduates from the history department at Tokyo University. Research was conducted with a view to discovering ‘Japanese origins’ in Korea, a framework that became obsolete once Korea was liberated from Japanese dominance. Nevertheless, the contribution of Japanese scholarship and of Japanese policies to protect and preserve (which included Korea’s first cultural property law in 1916) was immense and lasting. Japanese scholars contributed to the archaeological and historical knowledge of the ancient kingdoms on the peninsula, and to the collection and classification of artefacts and documents. They set methodological standards and trained the first generation of Korean scholars and their influence is still evident (Pai, 2000: 23–56). Thus, as writing the history of the nation became linked with writing the history of the empire, Japanese scholars exported their version of the archival ethos to Japan’s newly acquired territories. ¹⁹

While the development of the modern scientific discipline in Japan should be seen as part of a general trend where nations responded to the same global forces, Japan’s dual position as a state threatened by colonization and a colonizing state makes it unique. Japan ‘remained almost untouched by the expansion of Europe’ (Darwin, 2007: 215) for longer than most non-European states; it escaped colonization, embarking on one of the most successful nation-state building projects of the 19^th century, became a ‘de facto great power’ (ibid.: 356) which had begun to build its own colonial empire by the early 20^th century. Because the speedy and determined nation-building by the Meiji leaders owed much to the fear of being colonized and borrowed heavily from the West, it can be meaningful to compare the encounter between western and non-western practices with similar encounters in other non-western countries, such as India. ²⁰ The fact remains, however, that Japan never was a colony. Japanese historians were never research assistants of western historians, except the students of Rieß during the short time they studied with him, after which several went on to become pioneers in their fields. That their contribution is remembered while that of the Japanese scholars who also taught them is not, says something about the importance attributed to western learning, not just in its own right, but especially as a way of justifying change. But this should not distract us from the importance of indigenous forms of knowledge production. Japanese political leaders and scholars retained full agency throughout the process of shaping modern scholarship. This goes a long way towards explaining why the European archive as a model, while providing both inspiration and legitimation, ultimately was not fully adopted in Japan.