16 The Imprint of Buddhist Sanskrit on Chinese and Tibetan: Some Lexical Ontologies and Translation Strategies in the Tang Dynasty

Jens Braarvig

Download Chapter




Braarvig, Jens (2018). The Imprint of Buddhist Sanskrit on Chinese and Tibetan: Some Lexical Ontologies and Translation Strategies in the Tang Dynasty. In: Multilingualism, Lingua Franca and Lingua Sacra. Berlin: Max-Planck-Gesellschaft zur Förderung der Wissenschaften.

When Buddhism was imported into China, it met a complex culture with a rich terminology. Therefore the translation of the sacred scriptures of Buddhism would employ Chinese expressions with already ample religious and philosophical connotations.1 In the Tibetan case, however, the receiving culture maintained a comparatively simple language, with fairly uncomplicated semantic contents.  The terminology coined to receive the rich Buddhist systems of knowledge in Tibet were created without such preexisting semantics, and, as we shall see, the import of Buddhist culture into Tibet took place in a more systematic way than that of the Chinese case.2

In illustrating this movement from one lingua franca to another, namely, from Sanskrit to Chinese, and from a lingua franca to what we may call a national, or local language, namely Tibetan, we focus mostly upon the time of the early Tang period in China, and the same period in Tibetan, because this period shows a highly developed intellectual activity in both countries connected with the transfer and translation of Buddhism, the religion adopted and adapted by similar, but not identical processes by the two countries. The introduction of Buddhism to Tibet began in the seventh century CE, while in the case of China this process started, according to tradition, already in the first century CE.

When systems of knowledge migrate by translation from one language to another, from one culture to another, this may happen in more ordered and systematic ways, or by more arbitrary and individually centered initiatives. Further, the translations may be more verbatim, more in the verbo ad verbum fashion or, on the other side, free translations which try to accommodate the translated text to the receiving culture, or even just retelling the contents of the original text in a free way. Into this picture comes the frequent use of loanwords, balanced with the number of loan translations appearing as neologisms in the receiving language. So our concern is trying to understand the Chinese and the Tibetan case of translation of Buddhist texts in this perspective.

In general it can be said that the Chinese history of Buddhist translating was a more person-centered endeavor, which of course had the blessing of the ruling authorities at the time. The introduction of Buddhism to Tibet was, according to our sources, a strictly controlled process, and Tibetan royal power was the controlling agent and organizing entity in the beginning of the process. Later, with the routines created, the import of everything Buddhist into Tibet followed in general the premises laid down in the initial phase, though further developed and refined. We will start by describing the Tibetan case, because of its relative simplicity, and how the Tibetan chos skad was established—chos skad being a direct equivalent of lingua sacra, as the Tibetan word chos corresponds to Sanskrit dharma and skad denoting language, bhāṣā in Sanskrit. Next, in a comparative perspective, we will look at the dynamics of Chinese translation activities with some representative examples.

First of all, however, we should shortly touch upon how systematically a translation project is carried through. A translation may be undertaken by an individual, who creates much of his translated terminology in an arbitrary way, in accordance with his general understanding of the two languages involved; on one side there is the original and on the other side the receiver language, in which the translator does his best to portray the meaning of the original Vorlage by means of loan concepts, loan translations and loanwords. With the development of translation traditions and general dictionaries with equivalents, the translations become less arbitrary.  The least arbitrary type of translation is where standardization of terminologies is normatively established by comparative grammars and lexica and technical terms have strict equivalents to be employed in the receiving language.

During the long Chinese tradition of translating Buddhist texts—more than a thousand years—we have several phases. In the beginning the terminologies were naturally established by the first translators, but soon a more systematic use of equivalents to Sanskrit words developed, and a certain standardization of terminologies took place. There were, however, no formalized lexica for guiding the translators in the early periods, but earlier translated texts served as guides. This created a situation in the Chinese Buddhist language in which there is sometimes a fairly great number of Chinese equivalents for each Sanskrit term. In the beginning, then, translation techniques were more arbitrary than later, with a stronger established translation tradition, but each translator still had the tendency to develop a personal translation style in the absence of equivalence standards in systematic lexica and word lists. Only in the seventh century do we find lexica for Sanskrit and Chinese, but even these did not establish fixed standards to be followed: the translations remained over all created in the style of the individual translator.

So while we can characterize the Chinese reception of Buddhist thinking and its pertaining conceptual system as moving from an arbitrary process into a situation with more generally accepted translation conventions, the Tibetan endeavor to import Buddhism from India in all its aspects was a much more systematic process, in that a grammar3 and a lexicon for the standardized chos skad, or lingua sacra, were established under royal patronage and authority. In fact, the ordered and planned way that Buddhism was imported into Tibet, with all its disciplines and systems of knowledge, is quite remarkable, if not unique, in linguistic and translation history.

In the seventh century the Tibetan king Songtsen Gampo (Sroṅ-btsan sGam-po, reigned 618–649 CE) decided that Tibet—an inaccessible country, growing in military and political power—should be Buddhist.  He chose to supplant the Central Asian Shamanistic type of religion with the Buddhism of the high cultures that surrounded the Tibet.4 The Tibetan king considered sharing the religious culture of the Chinese empire in the East, but was also impressed by the Buddhist religion in India, so much that in the end it was decided to import the sacred teachings from the South rather than from the East. The king even married a Chinese princess, Wencheng, and a Nepalese one, for the purpose of good relations to possible contributors to the import of Buddhism to Tibet. Thus, during the reign of Songtsan Gampo a Tibetan scriptural system was created on the basis of the India Brāhmī script—an alphabet with fewer letters than the Indian, but adopted to Tibetan. It is not surprising that Tibet at the time chose the Brāhmī writing system, since in fact the letters were already employed west and north of Tibet in the language cultures along the Silk Road. However, an additional reason for Tibet to use this system was the general tendency to lean on Indian Buddhism rather than Chinese tradition in the implementation of Buddhist regimes of knowledge in Tibet. The creation of the alphabet is attributed to Thonmi Sambhoṭa, who, according to tradition, was appointed by the king to construct both the alphabet and the necessary Tibetan grammar to facilitate the Buddhist mission to Tibet.5

This historical fact that Tibet made the choice to implement Indian Buddhist systems of knowledge based on Sanskrit tradition is also borne out by the so-called Samye debate, or Lhasa Council in the 790s.6 Here the Chinese Chan-master Hva-shang Moheyan discussed with the Indian renowned scholar Kamalaśīla whether awakening was attained in an instant or by a gradual path of development, and again the Tibetans favored the Indian side; Kamalaśila and his party won the debate. This took place on the initiative of the Tibetan king Thrisong Detsen (Khri-sroṅ Lde-btsan, reigned 755–797 CE), who, after adopting Buddhism in 772, brought Tibetan power to its pinnacle by conquering Dunhuang in the 780s, and even the Tang capital Changan already in 7627—being chased away only by an alliance of the Chinese and Orkhon Uighurs.

In the period between Songtsen Gampo and 800 CE quite a number of translations from Sanskrit into Tibetan were made, and this earliest phase of translation is amply documented by the Tibetan contents of the Dunhuang library, which came to light again in the beginning of the early twentieth century, discovered by various Western scholars and explorers. However, the now learned Tibetan scholars noticed that the earlier translations were not uniform in style, and that the then existing Tibetan Buddhist terminology should be revised. To the extent it was possible, it should more accurately represent the semantics of the original Indic texts. Under the successors of Thrisong Detsen, and most of all under Thride Songtsen (Khri-lde Sroṅ-btsan, also called Senalegs, Sad-na Legs, who reigned 799–715), such a revision was decided upon, and a committee of scholars was ordered by the king to develop an equivalence lexicon of Sanskrit and Tibetan terms, to be used as a non-deviable norm for the translation work ahead in revising the previous translations and producing new and correct ones. The context of the situation in which these negotiations took place is described by Christina Scherrer-Schaub as follows:

Unauthorized, personal and unbridled initiative, as well as lack of source material, compelled the high authorities to take specific decisions. A chancery procedure, flanked with an increasingly important bureaucracy and deliberative body, was instituted. (Scherrer-Schaub 2002, 314)

Thus, in the year 814, at the time when the Tibetan kingdom reached its greatest extent, king Thride Songtsen issued the following decree, having as its aim to create a bilingual lexicon based on discussions of etymology as developed by Indian linguistic theory and Buddhist commentaries:

(1)In the year of the horse the ruler (btsan po) Khri lde Sroṅ btsan was staying in the palace ’On caṅ do in sKyi. The old warlords of Higher and Lower Tibet, as well as the “robbers” (i.e., the Uighurs) were vanquished, and he received obeisance from Gar log through an envoy. The great ministers Źaṅ khri Zur ram Śag and Maṅ rje lHa lod, these and others brought tribute from the territories. Camels, horses and oxen in huge numbers were offered to the King, and each and every man was recompensed with gifts from the main (źaṅ) minister and downwards.

(2)The King requested the preceptors from the west (i.e., Indian) Ācārya Jinamitra, Surendrabodhi, Śīlendrabodhi, Dānaśīla, Bodhimitra, and the Tibetan preceptors Ratnarakṣita, Dharmatāśīla, as well as the learned translators Jñānasena, Jayarakṣita, Mañjuśrīvarman, Ratnendraśīla and others to translate from Indic languages into Tibetan the terminologies of the Great and the Small Vehicles (two “ways” of Buddhism), to define the terms and make a written word list.

(3)The decree was issued that “One shall never deviate from this list and make it suitable for everybody to learn.”

(4)Earlier, in the time of Father [king Khri sroṅ lde btsan] of the Divine Son [namely, the present monarch, Khri lde sroṅ btsan], Ācārya Bodhisattva, Ye śes dBaṅ po, Źaṅ rgyal ñen Ña bzaṅ, Blon khri gźer Saṅ śi, the translator Jñānadevakoṣa, lCe khyi ’Brug, and the brāhmaṇa Ananta and others, since the Dharma-language was not known in Tibet, coined many terms; some of these were not in accordance with the Dharma-texts and the principles of grammatical theory (vyākaraṇa), and for this reason [it is now prescribed that] the improper [terms] not [well] formed were to be revised.

(5)So analyzing which were deemed the most important terms of the [Dharma-] language, they augmented [the Tibetan vocabulary] and brought the terms into agreement with how they occur in the texts of the Great and the Small Vehicle, how they are explained by the masters of old like Nāgārjuna and Vasubandhu, and how they accord with the principles in the works on grammatical theory. As for the difficult cases, they analyzed the words into their individual parts, provided reasoned explanations, and finally wrote [the results of their efforts] down as an authoritative document.

(6)On the one hand, simple words that need no explanation and that are appropriate to be translated literally (sgra bzhin du) have been fixed taking the literal expression as the main criterion.

(7)Whereas on the other hand, [in the case of] certain words, which are most appropriately fixed in accordance with the meaning, the [equivalent] term has been fixed taking the meaning as the main criterion.

(8)Then, the great scholar dPal gyi Yon tan, the great scholar Tiṅ ṅe ’dzin and all the others assembled before His Majesty, and after they had respectfully addressed the assembly of Lords and Ministers, they codified the methods of translating the Dharma, as well as the fixed terms in the Tibetan language in reference to the Indian language. Then the following decree was pronounced:

(9)“Regarding the way of translating the True Dharma, do not contradict the meaning, and adhere to good Tibetan usage.

(10)Regarding the translation of the Dharma, when a translation into Tibetan that does not deviate from the word order of the Indian original [retains] the connection between the meaning and words and is good usage, one should translate without deviating.”

(11)If one has to deviate [from the original word order] to produce good usage and intelligibility, one can deviate and translate in a pleasing manner, but only within a single stanza. As far as the meter is concerned, it should consist of four lines, or six if required.

(12)In prose, as long as one arrives at the required meaning, one may deviate [from the original word order] for the sake of good usage while taking both the words and meaning into consideration, and one should translate [accordingly].

(13)When it is possible to interpret one [equivocal] word [in Indian] with many [Tibetan] words, one should strive, in the definition, to make [the translation] agree with the [context, namely the words which come] before and after.

(14)As in the word gautamya, from the sound gau one attains a number of entities: “speech,” “geographic area,” “earth,” “light,” “vajra,” “bull,” “heaven,” etc.; and when one interprets a word like kauśika sound-wise as “having to do with kuśa-grass,” “learned,” “liking lotuses,” “owl,” “equipped with a treasure,” and so on, in the process of translating you attain a multiple list [of Tibetan words].

(15)As it is not possible to unite that multiple list [of meanings] into one [Tibetan word]-form in the process of translation, one has to decide for one [equivalent].

(16)But if there is no major reason [to choose any of the possible Tibetan equivalents], one may leave the Indian [term] untranslated, and use the Indian word.

(17)But wherever there is a [Tibetan] expression with a suitable [equivocal] interpretation, you should render it in accordance with its general [and equivocal sense], and not translate [with a generally employed Tibetan expression], deciding for only one direction [of meaning].

(18)If one translates names of countries, living beings, flowers, trees, and so on, and the possible translations are counter-intuitive, not good usage, imprecise, and as such the object of doubt: “Should they really be like this or not,”—which should be avoided—then one may keep the Indian term and preface it with a single Tibetan word “country” or “flower” and so on as applicable to that [case as a classifier]—whatever [class of things] the word [in question] refers to.

(19)When it concerns numbers, if one translates according to the Indian language, one will have “Monk-hundreds, thirteen less a half [hundred = 1250],” but if one translates in accordance with the ordinary Tibetan language “thousand two-hundred and fifty,” it is not in disaccord with the meaning, and it is also in accordance with good Tibetan usage—thus suitable summing numbers should be established in accordance with the principles of the Tibetan language.

(20)If one translates the prefixes pari, sam, upa and the others, which also may have the function of ornaments, the method should be to construe them in accordance with the meaning and translate them in accordance with its [Indian] word, [so as to produce in Tibetan] yoṅs su, yaṅ dag pa and ñe ba. If no additional meaning is attained, it is not necessary to multiply the construction [of the translated term], one should establish the word in accordance with the meaning.

(21)As for words that belong to a list [of related term or synonyms], if the word in question is not closely related, one should establish an [equivalent] term which is a general word in Tibetan and in accordance with good usage. If it is closely related [as a synonym of other terms in the list] one should employ the [one Tibetan] term which designate each [of the Indian synonyms].

(22)Concerning the degrees of respectful expressions, and expressions relating to a particular status, as concerned with the Buddhas, the Bodhisattvas or the Śrāvakas: One should translate [employing] the respectful expressions as related to the Buddha,

(23)while in other cases, when concerned with the middle and lower forms [of respectful expressions], one should translate in accordance with the principles codified by the translations of the Dharma [scriptures] of the Ratnamegha and the Laṅkāvatāra as made earlier by the learned scholars and analysts assembled before the Father of the Divine Son [that is, the now ruling monarch].

(24)Thus it is not granted to anybody individually to construct and after that establish new words which are differing from the principles of language as decreed in this way. However, if there is a need to construct and establish new words in individual schools, then, in the individual schools, one may construct words without establishing them, and state the causes of their origin in accordance with the dharma-books and the principles of word-[formation], and one should investigate how to establish them in accordance with the dharma. Then one should offer them to assembly in charge of the traditions of the Lord in the Palace and to the school of the great revisers of the dharma-translations, make a proper request, and if accepted it is to be added to the word-list [of the Mahā-vyutpatti].

(25)The Tantras with their mantras are to be kept secret in accordance with the scriptures themselves, and it is not proper that they are explained and taught to those not worthy. However, in the meantime they have been translated and given for practice, but their concealed meanings were not the subject of an oral explanation, thus [the words] were understood literally—and false practices have originated. While it is an established fact that selections from among the Mantra-Tantras and translations into Tibetan do exist, henceforth, with regard to dhāraṇīmantras and the Tantras, it has been decreed that unless permission is granted to translate [a specific such scripture], it is not allowed to collect or translate the Mantra-Tantras and the words of the mantras.

(26)The terms of the [Tibetan] language were not codified before, so since the terms were not fixed in the lexicon, one has adhered to what is derived from the books of the Great and the Small Way as well as books on language when explaining. Now the first part is ended.”

Thride Songtsen’s decree serves as the introduction to the work sGra sbyor,8 the etymological treatment of Sanskrit terms to be translated by fixed equivalents into Tibetan. The Tibetan terminology generated on the basis of the discussions in the sGra sbyor fills the standard Sanskrit-Tibetan lexicon named Mahā-vyutpatti, “The Great Etymology.”9

Usually several alternatives were discussed in the sGra sbyor before the final decision when the standardized Tibetan term was agreed upon, probably reflecting real scholarly discussions on the mentioned occasion, as well as previous considerations. The term chosen should first of all represent the original meaning of Sanskrit, but there was still an ideal not to deviate from good Tibetan usage. As we see, there are also rules proposed for when to employ loanwords, though calques or loan translations are preferred to loanwords when there is no question of mantras and sacred formulas. It should also be noted how there is a conscious translation policy on prefixes, that they should be translated only if they contribute an additional meaning, and are not only an “ornament.” However, prefixes in Sanskrit are in most cases faithfully replicated in Tibetan, even though it is quite doubtful whether they contribute with additional semiosis. Unlike the Chinese policy of often not translating titles and names of Buddhas, Bodhisattvas and Śrāvakas, the Tibetans always translate these semantically into Tibetan.

Probably the intention was to create a sGra sbyor article for every lemma in the Mahā-vyutpatti, but given the great number of terms, this was probably given up. The Mahā-vyutpatti was evidently also revised along with revisions of the Kanjur and Tenjur into the thirteenth century.10 We will translate into English the first articles from the sGra sbyor, and refer the interested reader to further reading on the mentioned internet site. The English translations correspond to the Tibetan parts of each article. The first chapter is naturally about the Buddha and his titles and appositions, here we quote the four first ones as examples:

1. buddha, Tibetan saṅs rgyas, “The Awakened”

2. bhagavān, Tibetan bcom ldan ’das, “The One Having Victory and Transcending,” usually translated into English as “The Lord”

3. tathāgata, Tibetan de bźin gśegs pa, “The Thus Come”

4. arhan, Tibetan dgra bcom pa, “The One who has Eliminated the Enemy”


Bacot, J. (1928). Les ślokas grammaticaux de Thonmi Sambhoa, avec leurs commentaires. Traduits du Tibétain et annotés. Annales du Museé Guimet. Paris: Geuthnner.

Bagchi, P. C. (1929–1937). Deux Lexiques Sanskrit Chinois: Fan Yu Tsa Ming De Li Yen et Fan Yu Ts’ien Tsen Wen De Yi-Tsing. 2 vols. Paris, Calcutta: Librairie Orientaliste Paul Geuthner.

Boucher, D. (1996). Buddhist Translation Procedures in Third-Century China: A Study of Dharmaraksa and His Translation Idiom. PhD thesis. University of Pennsylvania. url: http://repository.upenn.edu/dissertations/AAI9636133 (visited on June 11, 2014).

Braarvig, J. (1995). The Phug brag Versions of the Akayamatinirdeśa. Proceedings of the 7th Seminar of the International Association for Tibetan Studies. Graz: Verlag der Österreichischen Akademie der Wissenschaften. Vol. III, Transmission of the Tibetan Canon, Ed. H. Eimer, Wien 1997: 1–9.

– (2012). The Spread of Buddhism as Globalization of Knowledge. In: The Globalization of Knowledge in History. Max Planck Research Library for the History and Development of Knowledge, Studies 1. Berlin: Edition Open Access, 245–267.

Cheung, M. P. (2006). Anthology of Chinese Discourse on Translations. Vol. 1: From Earliest Times to the Buddhist Project. Manchester: St Jerome Publishing.

Demiéville, P. (1952). Le Concile de Lhasa: Une controverse sur le qui étisme entre bouddhistes de l’Inde et de la Chine au VIIIe siècle de l’ère chrétienne. Paris: Imprimerie Nationale de France.

Emmerick, R. E. (1992). A Guide to the Literature of Khotan, Studia Philologica Buddhica. Occasional Paper Series III. Tokyo: The International Institute for Buddhist Studies.

Fuchs, W. (1930). Zur technischen Organisation der Übersetzungen buddhistischer Schriften ins Chinesische. Asia Major:84–103.

Gulik, R. H. van (1974 [1956]). Siddham: An Essay on the History of Sanskrit Studies in China and Japan. Nagpur: Lokesh Chandra.

Hansen, V. (2012). The Silk Road: A New History. Oxford: Oxford University Press.

Ishikawa, M., ed. (1990–1993). sGra sbyor bam po gnyis pa and A Critical Edition of the sGra sbyor bam po gnyis pa: An Old and Basic Commentary on the Mahāvyutpatti, Materials for Tibetan-Mongolian Dictionaries. Vols. 1–2. Studia Tibetica 18. Tokyo: The Toyo Bunko.

Lamotte, E. (1949–1980). La trait é de la Grande Vertu de Sagesse. 5 vols. Louvain: Institut orientaliste, Bibliothèque de l’Université. T. 1509, (Mppś).

Li, Rongxi (1995). The Great Tang Dynasty Record of the Western Regions. Berkeley: Numata Center for Buddhist Translation and Research.

Miller, R. A. (1976). Studies in the Grammatical Tradition in Tibet. Amsterdam: John Benjamins Publishing.

Muradyan, G. (2012). Grecisms in Ancient Armenian. Hebrew University Armenian Studies 13. Leuven: Peeters.

Nattier, J. (2008). A Guide to the Earliest Chinese Buddhist Translations: Texts from the Eastern Han and Three Kingdoms Periods. Buddhica Bibliotheca philologica et philosophica buddhica. Tokyo: International Research Institute for Advanced Buddhology, Soka University.

Panglung, J. L. (1989). New Fragments of the sGra sbyor bam po gñis pa. East and West 44(1):161–172.

Rinpoche, P. (1994). The Words of My Perfect Teacher. San Francisco: Harper Collins Publishers.

Scherrer-Schaub, C. (1999). Translation, Transmission, Tradition: Suggestions From Ninth-Century Tibet. Journal of Indian Philosophy 27:67–77.

– (2002). Enacting Words: A Diplomatic Analysis of the Imperial Decrees (bkas bcad) and their Application in the sGra sbyor bam po gñis pa Tradition. Journal of the International Association of Buddhist Studies 25(1–2):263–340.

Schubert, J. (1937). Tibetische Nationalgrammatik. Das Sum cu pa und Rtags kyi ’jug pa des Grosslamas von Peking Rol pa’i dro rje. Artibus Asiae 1. Leipzig: Artibus Asiae Publishers.

Sen, T. (2016). Buddhism, Diplomacy, and Trade: The Realignment of India–China Relations. Lanham, MD: Rowman & Littlefield Publishers.

Simonsson, N. (1957). Indo-tibetische Studien: Die Methoden der tibetischer Übersetzer, untersucht im Hinblick auf die Bedeutung ihrer Übersetzungen für die Sanskritphilologie. Vol. 1. Uppsala: Almqvist & Wiksells.

– (1982). Refletions on the Grammatical Tradition of Tibet. In: Indological and Buddhist Studies: Volume in Honour of Professor J. W. de Jong on this Sixtieth Birthday. Ed. by J. W. de Jong. Canberra: Australian National University Press.

Takakusu, J. (1896). A Record of the Buddhist Religion As Practiced in India and the Malay Archipelago. Oxford: Clarendon Press.

Takata, T. (2000). Multilingualism in Tun-huang. Acta Asiatica: Bulletin of the Institute of Eastern Culture 78:49–70.

Verhagen, P. (1994). A History of Sanskrit Grammatical Literature in Tibet. Vol. 1: Transmission of the Canonical Literature. Leiden: Brill.

– (1997). Tibetan Expertise in Sanskrit Grammar—A Case Study: Grammatical Analysis of the Term pratītya-samutpāda. The Journal of the Tibet Society 8:21–48.

– (2001). A History of Sanskrit Grammatical Literature in Tibet. Vol. 2: Assimilation into Indigenous Scholarship. Leiden: Brill.

Wang, B. (1999). A Discussion of the 42 Syllabary (四十二字門考論). Chung-Hwa Buddhist Journal 12.

Whitfield, S. (2004). The Silk Road: Trade, Travel, War and Faith. Chicago, IL: Serindia Publications.

Yutaka, Y. (2015). Sogdian Literature i. Buddhist. Encyclopædia Iranica.

Zürcher, E. (1959). The Buddhist Conquest of China: The Spread and Adaptation of Buddhism in Early Medieval China. Leiden: Brill.


The classic on the topic, still relevant, of the Chinese reception of Buddhism, is Zürcher (1959).

On the spread of Buddhism as a system of knowledge, see Braarvig (2012).

At least according to tradition, see below.

See Scherrer-Schaub (2002, 266–67), with references, on the (probable lack of) historicity of this tradition.

On the earliest Tibetan grammar Sum cu pa, see Bacot (1928), Schubert (1937), and cf. Simonsson (1982), and on its historicity, Miller (1976, 2ff.) and Verhagen (1994, 207).

Demiéville (1952), Scherrer-Schaub (2002, 267).

Scherrer-Schaub (2002, 76–77 and notes 46–49), with ample references.

*Śabdayukti, “The construction of words.” The term yukti, which is a probable Sanskrit equivalent of Tibetan sbyor, means originally “joining,” but also logical “consistency.” Therefore we have ventured the translation “construction of words,” since this is the subject of the sGra sbyor, but yukti can also mean syntax, thus “the construction of sentences, or language” in general. For sGra sbyor, see Ishikawa (1990-1993); and further the internet version with English translation in Bibliotheca Polyglotta: https://www2.hf.uio.no/polyglotta/index.php?page=volumevid=263, accessed July 7, 2017.

Both works are available in a number of modern versions, and the introductory sGra sbyor is translated into English in two seminars in Berkeley and Marburg 2011–12, see the internet publication of this work in Bibliotheca Polyglotta, texts and translations in Sanskrit, Tibetan, Chinese and English. The latest and best critical edition of the original text is found in Ishikawa (1990–1993). The second volume has ample references to the Indian “etymologies” employed by the sGra sbyor. The classic of the study of Tibetan translatology is Nils Simonsson (1957), which contains a partial translation of the sGra sbyor into German beside the discussions. See also Verhagen (1994, 9–45).

See Braarvig (1995).

Panglung (1989), see the summary of Panglung’s arguments in Scherrer-Schaub (2002, 270–271), and summary of the events as seen by Scherrer-Schaub (2002, 315–17).

Historical revisions did not end with this, as can be gleaned by historical textual criticism, cf. Braarvig (1995, 8, note 22).

On the complex landscape of Silk Road culture, see Whitfield (2004); Hansen (2012). The Uighurs translated Buddhist texts from Chinese, for example from Xuánzàng’s translations of Indic texts, but they also translated from Sanskrit, and used Indian semantics to coin their Buddhist technical terms in Old Turkish. It is worth noting the founder of the Tang dynasty was half Götürk. Many other languages translated Buddhism, mostly from Sanskrit and its dialects. On Khotanese literature, including numerous Buddhist titles, see Emmerick (1992); there is also an example of a Sanskrit-Khotanese bilingual made for the purpose of learning Sanskrit, Emmerick (1992, 47–48). Sogdian Buddhist texts are mostly found in Dunhuang; one surmises that Sogdians emigrating to the East became Buddhist, see Yutaka (2015). In the earlier periods, all non-Chinese languages, including Sanskrit and the Central-Asian ones, were lumped together as 胡語 húyǔ “barbaric languages,” only later the expression 梵語 fànyǔ “Brahmā language” came into use, see Cheung (2006, 6–7). On the multilingualism of Dunhuang, see Takata (2000).

The Chinese lexica we have commented on are all Tang except the Fānfànyǔ. The four Tang lexica are edited and commented on in detail by Bagchi (1929–1937), see also van Gulik’s comments (1974 [1956], 31ff.. Van Gulik disagrees, correctly so, with Bagchi that T. 2133 was made for Indians wishing to learn Chinese, which is born out by the introduction as read by Bagchi—it is rather the other way around. Van Gulik thinks that the lexicon was made to help traveling traders, rather than students of Buddhism, and that the author cannot be 義淨 Yìjìng because of stylistic reasons (pp. 32–33 and 35), but van Gulik’s argument is not convincing in this respect given the lack of day-to-day language terminology.

The Pañcaviṃśatisāhasrikā Prajñāpāramitā, “Perfection of Wisdom in 25000 verses” was translated into Chinese four times, see Lamotte (1949–1980, vol. I, p. VII ff.), T. no. 1509: 摩訶般若波羅蜜經釋論.

See below on Xuánzàng’s principles for employing loanwords or loan translations.

The latest translation is Li (1995).

Takakusu (1896). See Cheung (2006, 167–8) on his other works.

T. 2132, 1186a5 南天竺般若菩提悉曇, 大唐山陰沙門智廣撰.

On the siddham alphabet, see van Gulik (1974 [1956]) and Wang Bangwei (1999).

On Wǔ Zétiān’s remarkable life, see Sen (2016, 93ff.).

There are numerous translations and versions into Chinese, from the early fifth century to Amoghvajra’s Tang translation in the mid-eighth century: T. 387, 991, 992, 993, 950; Tibetan in Derge mdo sde wa 113a–214b, 250b–263a.

Chinese translations in T. 489, 658, 660, Tibetan in Derge mdo sde wa 1b–112b.

Sen (2016).

天地玄黃宇宙洪荒日月盈昃辰宿列張寒來暑往秋收冬藏閏餘成歲律召調陽, tr. Nathan Sturman.

I thank Jianrong Shi for helping me to understand this introduction.

梵語千字文譯注. T. 2133b.

全真: 唐梵文字, T. 2134, no Chinese character transliteration; T. 2135, 梵語雜名 “Various Sanskrit words,” collected by Lǐyán 禮言, with Chinese transliteration, but another order of single Chinese characters as lemmata; and 唐梵兩語雙對集 (T. 2136) starting out with the parts of the body, but having only Chinese transliterations apart from a small Brāhmī quotation in the end.

慧琳: 一切經音義, between 788 and 810.

T. 2131, 翻譯名義集, see Cheung (2006, 199–200) for his interesting reflections on translation.

A very similar case of the introduction of Buddhism is that of the Armenian choice to translate Greek classical literature as well as Christian texts directly from Greek in the fifth century CE, not basing themselves on Syriac Christian texts which were dominant among the Armenians before that, since Syriac influence was tolerated by the Persians, who constantly sought to dominate Armenia. Thus the creation of an alphabet and various linguistic tools to have direct access to the source of the dominating Greek culture, was a partly, if not mostly, nationalistic project, just as in Tibet at the time of the Tang dynasty. The Armenians, like the Tibetans, did also not have an alphabet before they embarked on the project of gaining cultural autonomy through access to a world culture of learning, in the cases Greek and Sanskrit based. See Muradyan (2012) on the massive influence of Grecisms on Armenian, similar to the Sanskritisms in Tibetan as a result of these important political decisions.

Further in Scherrer-Schaub (2002, 298–304).

Cheung 2006, 166, unfortunately the source for this information is not given.

四十二章經. The language seems more modern, either the texts is revised, or is later construction. There is no exact Sanskrit counterpart.

月支, on the role of the yuèzhí as translators in pre-buddhist environments, see article of Bolzmann in this volume.

See further in Cheung (2006), and Nattier (2008).

As described by Zanning (919–1001), in Cheung (2006, 188–190). See also Fuchs (1930, 84–103); van Gulik (1974 [1956], 29–30); Sen, Tansen, (2016, 97).

Cheung (2006, 167).

The translation is by Diana Yue, Cheung (2006, 157–8), though abbreviated and with some modifications.

See further Cheung (2006, 159) on Xuánzàng stressing that Buddhism and Daoism are separate religions, and that Buddhism should not be linked up with Daoist meanings. This, however, naturally was an important discussion in attempts to reconcile the traditions or to suppress the foreign influence.

See Cheung (2006, 169) on Empress Wu’s decree that Buddhist and Daoist priest should pay respect to their respective sacred spaces.

But cf. below, on Chos-grub, and van Gulik on this problem, the latter, who, in describing the Siddham system of writing, cannot hide his depreciation of Chinese intellectual culture that greatly appreciated Indian calligraphy, but not at all Sanskrit grammar, throughout their tradition.

T. 1733, van Gulik’s (1974 [1956], 19–20) translation with some moderations.

In the following, Fǎzàng’s grammatical note explains vocative as the eight category, and then gender and number. T. 1733 T. 1733 149a28–b16: 第十八聲者依西國法. 若欲尋讀內外典藉. 要解聲論八轉聲法. 若不明知必不能知文義分齊. 一補盧沙此是直指陳聲. 如畫像人斫樹指說其人. 二補盧私是所作業聲. 如所作斫樹. 三補盧崽拏是能作具聲. 如由斧斫. 四補盧沙耶是所為聲. 如為人斫. 五補盧沙是所因聲. 如因人造舍等. 六補盧殺娑是所屬聲. 如奴屬主. 七補盧鎩是所依聲. 如客依主. 瑜伽第二. 名上七種為七例句. 以是起解大例故. 聲論八轉更加補盧沙. 是呼召之聲. 然此八聲有其三種. 一男聲. 二女聲. 三非男非女聲. 此上 畫像且約男聲說之. 以梵語名丈夫為補盧沙故. 又此八聲復各三. 謂一聲. 二聲身. 三多聲身. 則為二十四聲. 如喚丈夫有二十四女及非男女聲亦名有二十四. 總有七十二種聲. 以目諸法可以准知. 然此方多無此例.

Verhagen (1997, 22, references to sources in note 5, 32–33). See further Verhagen (1994, 47ff., and 2001, 5ff.) for a more detailed treatment of Lce-khyi-’brug. He is mentioned as a translator in sGra sbyor §4, see above.

See Verhagen (2001, 29ff.) on this work, bibliography,47 and another Dunhuang grammatical manuscript without title. Further on “The Formation of Bilingual Tibeto-Chinese Communities [in Dunhuang]” in Takata (2000, 62ff.), with a list of “Tibeto-Chinese Bilingual texts,” Takata (2000, 64–65), though produced probably as late as the tenth century.

See Cheung (2006: passim).