skip to primary navigationskip to content

Vernacularization in East Asia

Principal investigator: Prof. Peter Kornicki

Research Associate: Dr Rebekah Clements



Project conducted by Prof Peter Kornicki

I have been working for several years on a book provisionally titled All under the heavens use the same writing: Chinese texts and vernacularization in East Asia. The Introduction and chapter 6, ‘Tripitakas and other Buddhist texts’, are  complete, and work is underway on several other chapters. I expect to complete the book by 2015, at least to the point where it can be submitted to a publisher, and regard it as likely to be the most important contribution I shall make to scholarship on East Asia. There follows an outline of the project as put to the Leverhulme Trust for the three-year grant awarded earlier this year.


In many societies across the globe there has been a long process in which a written language valued for the religious, intellectual or political ideas inscribed in it has had an uneasy relationship with local languages, often described as ‘vernaculars’. In the course of time these local languages were literised (ie, made the breakthrough into writing) and increasingly vied with the élite written language, which usually covered a wide geographical area and which has usefully been termed by Sheldon Pollock a ‘cosmopolitan’ language. Much of our thinking on this process hitherto has been determined by the case of Latin and the European vernaculars, but the normacy of the European case has

been challenged by Pollock in his work on Sanskrit and the South Asian vernaculars. Other important cases include Persian in the Mughal empire and Quranic Arabic in the Islamic world. In East Asia the role of cosmopolitan

language was played by the written language known as literary Chinese or Sinitic, which brought not only writing to the surrounding societies but also Buddhism, in the form of Chinese translations of the sacred texts and    commentarial literature, and the whole range of Chinese literary production, ranging from the texts of the Confucian tradition to poetry, histories, medical texts, and, from the Ming dynasty onwards, vernacular fiction. It was thus in

Sinitic that those living in the parts of the world that we now call Japan (including the kingdom of Ryūkyū, now Okinawa Prefecture in Japan), Korea, Vietnam and Palhae (ie Bohai, now part of north-east China) read and wrote

their first texts and in Sinitic that they continued to write their learned texts even after the development of scripts to write their own languages (Japanese kana in the 9th century, Vietnamese nôm in the 11th and Korean han’gŭl in the 15th).


Sinitic in the role of a lingua franca (albeit exclusively a written language) is now dead and there is no common written or spoken language linking those societies. This massive rupture with the past, which was only completed in the 20th century, has a long history, though, and has passed through numerous identifiable stages. Thus local speakers of the vernaculars applied their phonologies to Sinitic script; they used that script to literise vernacular personal and place names; they devised ways of reading imported Sinitic texts in ways that brought them closer to the vernaculars and ways that also changed the vernaculars in consequence; they inscribed texts in Sinitic, usually for

local consumption but occasionally for an ‘international’ audience; using Sinitic script ingeniously, they found ways of inscribing vernacular poetic and other texts; and after the invention of their own scripts they were in a position to inscribe far more vernacular texts, although they did not do so in identical ways. It should be noted, though, that the invention of the vernacular scripts was much delayed. The comparison with Tibet, where a script was devised for Tibetan in the 7th century in order to translate Buddhist texts is instructive, for there was no such development in Japan, Korea or Vietnam, where Buddhist texts were transmitted, copied and studied as Chinese texts, and the only vernacular element was the phonology of recitation, based on ancient vernacular interpretations of Chinese pronunciation, as it is to this day.


By comparison with the small numbers of books which medieval monastic libraries in Europe possessed, the corpus of texts in the Chinese Buddhist canon alone was huge, and to that must be added the vast corpus of other Chinese texts. Scriptoria and the production of manuscripts were on a scale unknown in Europe, and the advent of printing added substantially to the availability of texts. Woodblock printing was undoubtedly developed in China in the 7th or 8th centuries, although the oldest extant examples were single sheets of paper carrying a dharani invocation, which were produced in Korea and Japan in the mid-8th century. In China this technology was used for printing a limited range of secular texts in the Song dynasty, by which time the technology was being put to use in Japan and Korea, and probably Vietnam. Movable-type printing was also invented in China, albeit little used there. Recent discoveries have shown that this technology was used in the Tangut kingdom of Xi Xia in the 12th century, but metal type was pioneered in Korea in the 13th century and there reached the highest levels before modern times. It did not reach Japan until the late 16th century, where it flourished for some 50 years and then woodblock printing resumed its dominance, owing to the demand for printed glosses, which were much easier to produce with woodblocks. Commercial publishing became widespread in Ming China, then in 17th-century Japan and later in Korea and Vietnam. There was little printing and no commercial publishing in Ryūkyū, which relied for its supply of books on its close ties with both China and Japan.


In the light of the vast production of Sinitic manuscripts and printed books and their reproduction in East Asia, this project addresses the mechanisms and motivations underlying the long engagement with the Chinese script and with

Sinitic texts in East Asia, which ended with the massive rupture of the 20th century. Now Vietnam uses roman script, Korean publications are almost exclusively in han’gŭl except for some academic books, and only in Japan do Chinese characters retain a prominent presence: how did we get to this position? This project is driven by a number of questions relating to these issues, amongst which are the following.


1 How was the supply of Chinese books to surrounding societies maintained given the distances, the transportation difficulties and the occasional export-bans placed by Chinese authorities on certain categories of book? What kinds of books were sought and what differences were there in the kinds of books sought over time and between the various societies? To what extent did Sinitic texts composed in one of these societies travel to others or to China? If, as seems to be the case, few did travel, why was this? In the interests of reaching more readers and a bigger market, in Europe many vernacular texts were translated into Latin, from the scientific works of Robert Boyle to literary works, as Peter Burke has shown: in East Asia there are very few signs of any similar reverse translation of this kind – does this result from the slow development of an idea of an East Asian market for books, from hermetic cultural traditions that resisted translation, or what?


2 The evidence from the Dunhuang manuscripts shows that a first stage in making Sinitic texts more accessible was the provision of punctuation, which depended on a prior interpretation of the text. Recent evidence from Korea shows that at least by the 8th-century Koreans were reading Sinitic texts in a new way known now as hundok, which involved using dry-point (kakp’il) glosses or ink glosses to indicate the grammatical particles and inflections necessary in Korean. This technique, it is now recognised, was transmitted to Japan by Korean monks of the Hwaŏm (Ch. Huayen, J. Kegon) sect of Buddhism, where it underwent further development and formed the basis of the reading strategy known in Japan as kundoku, which similarly used dry-point (kakuhitsu) or ink glosses. From the 16th century onwards most Sinitic texts printed in Japan (including Sinitic texts composed in Korea, such as the works of Yi T’oegye) contained printed glosses known as kunten which, according to one of a number of different

kundoku systems, rendered the text accessible as a specialised form of Japanese. This set of techniques, known as kundoku in Japanese and sŏkdok hundok in Korean (as opposed to the later and rather different ŭndok hundok), is currently the subject of intensive study in Japan and Korea, and to a lesser extent in Vietnam. The linguistic operation involved is highly source-oriented and results in a ‘bound translation’ in which the vocabulary of the original is preserved visually in the form of the Chinese characters, even when they are used obscure senses, although they are pronounced in accordance with vernacular phonology. If this is a form of translation, then it is one which shows unusual respect for the vocabulary of the original; but if instead it is a form of reading, as the terms hundok and kundoku suggest, then it appears to be more a method of forcing Sinitic into vernacular straight-jackets. Wherever the truth may lie, there has, however, so far been little assessment either of the underlying understanding of Sinitic such techniques produce or of the gaps in understanding that result from shifting semantic nuances between Sinitic and graphic loans from Sinitic in the vernaculars, and little study of the particular forms of vernacular language produced by hundok/kundoku. Why did these reading traditions develop instead of direct reading of Sinitic texts? If it is a form of translation, as some assert, why is the vocabulary of the Sinitic unchanged even when obscure, and what is the significance of the flattening effect of kundoku in Japan, where some Chinese particles were routinely ignored and where a single Japanese term was used to render a variety of Sinitic terms? How did these practices impact upon the written vernaculars?


3 To what extent was translation resorted to, when and why? The answer, counterintuitive though it is, appears to be that it was resorted to more in Korea and Vietnam, where rigorous examination systems required mastery of Sinitic for advancement, than in Japan where such exams as there were had little career benefit and where, consequently, mastery of Sinitic was less politically essential. In Korea there is a vast corpus of bilingual ŏnhae books containing Sinitic original and Korean translation, but to date they have mostly been studied for the evidence they provide of linguistic change rather than for the evidence they provide of understanding and interpretation of Sinitic texts, while their Vietnamese equivalents have hardly been studied at all. There is also a large corpus of bilingual Manchu-Sinitic books from the 17th century, and further afield evidence of translated Sinitic texts in Western Xia, Mongolian and Tibetan: thus translation was an attested and accepted mode of access to Sinitic texts, especially those of the Confucian tradition.


4 Why was relatively little translation of canonical Sinitic texts undertaken in Japan compared to Vietnam and Korea? Most well known are Japanese translations of Dutch books relating to medicine and science, which began appearing in the 18th century: these have been much studied in the context of the transmission of European science to Japan, although they have been little studied as translations and the whole translation project related to Western writings accessed through the Dutch is still to be brought to light. In the late 18th and early 19th centuries, many translations of vernacular (baihua) Chinese fiction were published in Japan, and in the 1840s and 1850 works published in China based on Western works on military science or the dangers of imperialism, most famously Hai guo tu zhi (1843), were translated into Japanese. The contours of Japanese translations of Sinitic texts, including texts written by Koreans, have yet to be examined in detail. However, it is a fact that editions of classical Chinese texts were published in abundance, with kunten of different kinds and with notes from various sources, and in the Keiten Yoshi (‘Classics without a teacher’) series every effort was made to make texts such as the Classic of Filial Piety accessible even to those with little sinological knowledge, always, however, without actually translating them. When in 1721, for example, Muro Kyūsō was asked to translate the Liu yanyi, he instead rewrote it for a Japanese audience, omitting and adding a great deal, so that his Rikuyu engi taii bears only superficial resemblance to

the original. Although it is clear that, by comparison with Korea and Vietnam, very few texts of the Confucian tradition were translated into Japanese before modern times, in the 17th century a small number of canonical Sinitic texts were nevertheless translated, and one of the most active in this connection was Hayashi Razan, many of whose translations were made in the first instance at the request of prominent individuals and were only published by his heirs after his death. Why was his initiative not followed? Does the availability of kunten editions explain the lack of translations from Chinese classical texts? If not, as seems to be the case at this stage, is this a case of Bourdieu’s notion of ‘cultural capital’ in action, as the educated kept the uneducated at a distance by their knowledge of kundoku reading techniques?


5 On the level of spoken language, in East Asia the vernaculars were never challenged by Sinitic or any form of Chinese, unlike the cases of Latin, Sanskrit or Persian. Spoken foreign languages were not totally neglected in East Asia, however: in 1407 the Ming established an interpreting college called the Siyiguan where Japanese, Korean and Vietnamese but also Persian, Turkish, Malay, Siamese and even Ryūkyūan were taught. In Korea the Sayŏgwŏn, founded in 1393 in succession to earlier institutions, taught spoken Chinese, Japanese, Mongolian and Jurchen (later replaced by Manchu). As the textbooks Nogŏltae and Pak t’ongsa show, the focus was on practical communication for mercantile or diplomatic purposes, rather than intellectual communication; what is more, the low status of interpreters discouraged the yangban elite from showing any interest. There were some exceptions, like Hong Taeyong, who visited China as an envoy in 1766, used his knowledge of spoken Chinese to make contact with Chinese scholars in Beijing, and kept up his correspondence with them later. In Japan Amenomori Hōshū used Korean and Chinese for scholarly purposes (the latter to converse with Ming exiles), and later Ogyū Sorai and others studied spoken Chinese either in order to deepen their understanding of Chinese texts or to read imported Chinese vernacular fiction. The one exception was the Ryūkyū kingdom, which sent a steady stream of students to China for long-term study, with the result that there was a body of scholars there able to travel to and from China regularly and to engage with Chinese intellectuals orally, such as Tei Junsoku (1663-1734). Why was Ryūkyū the exception, why was there no spoken lingua franca, and why were the potential scholarly benefits of dialogue with speakers of other languages not appreciated and realised in practice?


6 Was there anything akin to a Republic of Letters in East Asia? The notion of a Republic of Letters emerged in Europe in the 15th century and survived until local nationalisms overruled the sense of common purpose. It thrived on the publication of books in Latin, on extensive correspondence conducted with people never personally encountered and on face -to-face encounters. The comparison with East Asia needs to be handled with care, particularly because Latin was not sustained or claimed by any of the Italian states or any other European state, while Sinitic enjoyed a difficult relationship, when seen from outside China, with the largest and most powerful state in East Asia. To some extent the conditions were mirrored in East Asia: books were produced in China in prodigious quantities and were much sought after, opportunities for travel were there for Buddhist monks and for members of Korean and Vietnamese diplomatic missions, and some at least maintained a scholarly overseas

correspondence. Yet the flow of books followed centrifugal routes out of China and rarely went in the reverse direction; Korean books travelled to Japan and Ryūkyū but few went the other way; few Chinese scholars left China; and the overseas correspondence was limited. On the other hand, it appears to be that case that the consciousness of distinct languages, literatures and cultures appears to have come much earlier to East Asia than to European states. This is suggested by the changing ways in which East Asian states were referred to both by their own citizens, by others and by Chinese: Ogyū Sorai’s refusal in his vernacular version to refer to the Ming law code as the code of the Great Ming, and the Vietnamese decision in 1428, after the expulsion of the Ming armies, to rename their country ‘Great Viet’ are just two signs of this. So one of the hypotheses to be tested in this book is that the shift from a sense of common purpose, which can be detected in early East Asian Buddhism (eg Wŏnch’ŭk [613-696], the Korean monk one of whose commentaries was translated from Sinitic into Tibetan), to what in Europe is described as the rise of national interests conflicting with the apolitical ideals of the Republic of Letters took place much earlier in East Asia.


As is clear from the scope of these questions, this is an ambitious project which unashamedly seeks to take a broad view, over the longue durée, of East Asian societies on the periphery of China (including, for comparative purposes,

some that did not use Chinese script such as Tibet, the Tanguts, the Uigurs and the Manchus), of their engagement with Chinese texts and of their disengagement. Some topics relating to individual societies have been well studied by others, such as the printing of Sinitic texts in Korea and Japan, while the printing of Sinitic texts in Vietnam and the acquisition of Buddhist and other Sinitic texts in Vietnam remains understudied. The study of translation throughout East Asia is also still largely neglected, in spite of growing interest. On the other hand, the ‘national’ vernacular literatures have been well studied, even if most attempts to relate themes or genres to Sinitic influences have ignored how Sinitic texts were read.


Many questions remain unanswered, not limited to those outlined above, and a conspectus is needed to move the study of these important cultural shifts on to a higher level where the global significance of vernacularisation can be

considered in conjunction with the continuing work of Pollock and others. Another of the hypotheses to be tested, therefore, is Pollock’s suggestion that vernacularization in East Asia was much delayed except in Japan: this only holds true if vernacular literary production is considered, but it appears from the evidence that has so far come to light that Sinitic texts were frequently vernacularised phonologically and often syntactically as well. This was arguably a form of textual appropriation that made it possible for Japanese, Koreans and Vietnamese to consider Sinitic texts emanating from China as part of their own heritage as well.


The methods which are being and will be used to bring this research to a successful conclusion are several.


(A) The question of vernacularization has been explored and theorised by Pollock while others working on questions of diglossia, graphic loans and language contact have examined, for instance, the uses of different scripts and languages in Muslim Spain, and there is also interesting work on graphic loans in an East Asian context. Others, Bourdieu in particular and Lewis in the context of China, have explored the relations between languages and power,

which has an obvious bearing on Korea, Japan and Vietnam, where Sinitic was the language of government and power. This project needs the theoretical insights brought to bear by these writers on the problems to be addressed but also needs to be informed by work in comparable areas (Europe, South Asia, the Islamic world), for the premise is that the overall phenomenon being studied is one of language contact, elite languages and ultimate vernacularization, phenomena of global reach.


(B) The raw data on which the research will be based is two-fold. Firstly there is the documentary evidence relating to the flow of books – Chinese dynastic histories, Chosŏn wangjo sillok (Veritable records of the Chosŏn dynasty), Đại Việt sử kí toàn thư (Complete historical chronicles of Đại Việt), Japanese monastic catalogues and shipping records, etc., all contain information on the genres sought and acquired and the methods for acquiring them.


Secondly, there are extant Sinitic texts themselves: for some of the oldest, excellent facsimiles have been published in Korea and Japan, and the images on the Dunhuang Website give access to early punctuation strategies. The early

texts preserved in Korea and Japan are currently being subjected to intensive research and there is much useful secondary literature on them. This is less true of later texts produced in Korea and Japan as well as Vietnam and Ryūkyū, and I have already spent much time at the Academy of Korean Studies, the Kyujanggak and the National Library of Korea, the Hán-Nôm Institute and National Library in Hanoi, various libraries in Japan, including several in Okinawa for Ryūkyū books, as well as libraries in Britain, USA, Italy and France. The evidence that these provide covers the range of books transmitted and copied/reprinted (for not all genres or texts were), the presentation of

the Sinitic text (local prefaces/postfaces, dry-point or ink glosses, annotation).


I read Japanese, Korean, Sinitic and modern Chinese and have already made use of sources in all these languages in an article published in 2008 in which I examined the vernacularisation of books for women in Vietnam, Korea and Japan. For Vietnamese I have a research collaborator in Hanoi, with whom I wrote the 2008 article and I have already conducted research in Hanoi: Vietnamese texts in Sinitic or Vietnamese editions of Sinitic texts pose no problems but for vernacular nôm texts my collaborator’s assistance is at present invaluable, although I am learning to deal with these and with the limited amount of secondary literature in Vietnamese on my own. I also have extensive experience in the bibliography of woodblock-printed books, initially in Japan but over the last four years in Korea and Vietnam as well.


The objectives are to achieve a comprehensive understanding of the movement, treatment and vernacularization of Chinese texts in East Asia and, as an important part of that, of the role of translation and non-translation in Edo-period Japan (1600-1868). These will be embodied in the book mentioned above; it will consist of a series of preliminary chapters on the comparative and theoretical framework, on scripts and on the forms of books; there will

follow key chapters on the flow of books, on reading strategies and on translation. The book is being written so as to be accessible to scholars such as Pollock and Burke, who have an interest in the themes but no knowledge of East Asian languages, as well as to be of interest to East Asianists, for whose benefit the bibliography and index will include East Asian scripts.



Project conducted by Rebekah Clements

My present research involves a cultural-historical study of translation in early-modern Japan.  I am interested in the question of how translation figured in the lives of those who read and produced texts during the Tokugawa period, and in the forms which translation from a variety of languages into Japanese took during this era.  My work and the monograph I am writing are located within the context of a project on the history of vernacularization in East Asia, overseen by Peter Kornicki and funded by a Leverhulme Foundation grant.


HOME of pre-modern and early-modern studies at Cambridge