“Were Buddhist Scriptures in China Translated for Chinese?”: A New Investigation into the Development of Early Buddhist Chinese from the Perspective of the Buddhist Speech Community

Abstract

Modern scholarship focuses on the lexical or syntactical features of Buddhist Chinese used in Chinese translations of Buddhist scriptures; however, the origins of the language have attracted relatively little attention. Our article explores the issue from the perspective of the speech community, and we argue that the community’s ethnicity played an important role in the pre-fourth-century development of Buddhist Chinese. Buddhist scriptures were mainly introduced to inland China via the Western Regions and the Han people were not officially allowed to be monks. In addition to the translated scripture’s readership, considerable numbers of scripture transmitters and translators were not Han Chinese, and Han Chinese translation assistants were mainly grassroots literati. We combined an analysis of language learning with an examination of the universal features of translated languages and views on translation, and we argue that the early Buddhist community generally could not write and did not need to read refined literary Chinese. The lifting of a ban on Han peoples converting to Buddhism from the fourth century onwards accelerated the development of Han Buddhist communities, and dramatically promoted the localization of Buddhism and scriptures, which marked a new era of translation and translated language. Our investigation into the framework of the Buddhist Chinese speech community provides new perspectives compared with philological studies on the Buddhist language and explains the early historical development of Buddhist Chinese.

1. Introduction

The texts and languages used in Chinese translations of Buddhist scriptures are a much-discussed topic (Ji [1947] 1998a, [1959] 1998b; Brough 1962; Zürcher [1977] 2013a; Mair 1994; Boucher 1996, 1998; Nattier 2008; Zhu 2008). However, the early people that accepted Buddhism and their language(s) have attracted relatively little attention. Buddhist scriptures were translated into Chinese; however, before the fourth century, a considerable proportion of Buddhist monks, believers, and followers in China were not the Han people, and they spoke different dialects and languages, with different competency levels of colloquial Chinese and literary Chinese (文言 wenyan).

Buddhist monks, lay followers and other believers formed a speech community, as they were people who shared a language variety in a regionally or socially definable human group (Crystal 2008, p. 446), and these communities developed through prolonged interactions among those who operated within these shared and recognized beliefs and value systems regarding forms and styles of communication (Morgan 2014, p. 1). In these speech communities, Buddhist Chinese (BC) was used in religious activities, such as reading, chanting, preaching, annotating, and translating.1

A language in a speech community is not only shared by all as a medium—it is also linguistically and ideologically affected, shaped, and constructed by all members. However, previous scholarship has primarily focused on individual translators as BC creators and, to some extent, neglected the roles and contributions of other Buddhist speech community members, such as translation assistants and common readers. BC is widely assumed to be a language mixed with vernacular, dialect, and transliteration elements; however, few scholars have explained BC as a Chinese language variety from the perspective of the ethnic features of the early Buddhist speech community. The influence of linguistic and translation views on BC development has not been fully explored either.

In our article, we examine the early development of BC by focusing on the speech community, shedding new light on its nature and causes of its hybridity, and deepening our understanding of early Buddhism in China, as well as its language and translation history. First, we contextualize the development of the Buddhist speech community through the early spread of Buddhism and scriptural translations in China. Second, we explain BC’s hybridity as a translated language. Third, we explain how post-fourth-century changes in the ethno-demographic structures of Buddhist communities influenced BC.

2. Role of Speech Community in Early Spread of Buddhism and Scriptural Translations

Translations of Buddhist scriptures in medieval China can be generally divided into three or four stages, and the first stage is the period before the Eastern Jin Dynasty (317–420 CE) (Liang 2001, pp. 212–16; Ma 2006, p. 69; Wright 1990, p. 4). The period is called archaic translation (古譯 guyi) in Buddhist catalogues and denotes Chinese scriptures produced before the introduction of a more sophisticated translation idiom and the elaboration of a much more diversified technical vocabulary in the last quarter of the fourth century (Zürcher [1982] 2013c, p. 261).

During this time, most scriptures were introduced by monks from India, or the Western Regions (西域 Xiyu), a name generally used in early medieval China to designate the regions west of the Yumen Pass, including Central Asia. The slow infiltration was helped by foreign merchants, refugees, envoys, and hostages in China (Zürcher [1959] 2007, p. 23; Chen 1964, p. 19). Because foreign monks did not have mastery of adequate Chinese to perform translation by themselves, they had to rely on translation assistants. Hence, early translations were usually conducted through the cooperation of foreign missionaries and Chinese participants. When translators tentatively began to use Chinese to express Buddhist elements, the language used in their translations was generally straightforward, awkward, or obscure. Though far from being satisfactory, it was not only key to the spread of Buddhism in China, but also a medium for Chinese speakers in their religious activities. Through preaching, reading, translating, and other activities, a missionary endeavored to attract new disciples and believers—new members of his Buddhist group and speech community. With the migration of missionaries and the circulation of their translations, Buddhism spread in some scattered centers and grew in a fragmented and regionalized fashion, meaning the religion and its language were limited to certain marginal groups and communities in China.

The second stage ranges from the Eastern Jin period to the end of the sixth century, when Buddhism began to be further involved in religious and intellectual life. In the early fourth century, restrictions on Han people’s conversion were removed, which resulted in huge increases in Han monks and monasteries, rapidly expanding the Buddhist speech community (see Section 5 for details). Many officials, gentry, and literati became monks or laymen, which increased the general literary and intellectual levels of Buddhism. Indigenous believers favored Chinese literary tradition and called for a more refined language and high-quality translation of scriptures; therefore, the localization of the religion and its language began to be an important issue. In the meantime, Buddhist groups obtained more governmental, official, or private support. Translations continued to be mainly conducted through cooperation in translation studios (譯場 yichang); however, the Chinese began to exert more influence (Liang 2001, p. 173). Learned monks made efforts in exegesis and translation studies, attempting to improve BC to cater to Buddhism in the new era.

Translations in the Sui (581–618 CE) and Tang (618–907 CE) Dynasties (or after) can be regarded as works in the third or fourth periods. Now, translation studios could be supported by central governments, and large-scale translations of Buddhist scriptures continued as labor divisions in translation studios matured and became institutionalized.2 Chinese translators usually dominated translation studios, and chief translators, such as Xuan Zang (玄奘, 602–664 CE), not only mastered the languages needed for translations, but also had a deep understanding of the Buddhist doctrinal system. Products in this period were generally regarded as mature or new translations (新譯 xinyi). Buddhism furthered its assimilation, and both the religion and its language began to be an integrated part of Chinese culture.

Studies of Buddhist translation history and its relevant languages are not just an issue of text and language. Except for text analysis, more factors should be taken into consideration, such as translators and their assistants, patrons and financial support, views on translation, literature norms and traditions, translation organizations, translation procedures, quality of original scriptures, as well as the readership, reception, and circulation of translations. All these complex factors are related to each other in a speech community, which is a “field” (Bourdieu and Wacquant 1992) for the production and circulation of scriptures and the exchange and appropriation of religious knowledge and expertise. The language used in the field acts as a habitus, which is related to practices in both individual and collective ways. Hence, we value all the important parties involved in the making of scriptural language.

The creation of BC through translation involves a collision and compromise between translation participants with different cultures, ideologies, and language competencies:

初則梵客華僧，聽言揣意，方圓共鑿，金石難和，椀配世間，擺名三昧，咫尺千里，覿面難通。

(Zan Ning 1987, pp. 52–53)

In the initial stage of sutra translation, the visiting monks speaking Sanskrit and the Chinese monks speaking in the Chinese language listened to each other and tried to guess at what the other meant. The communication was as difficult as pairing square and round articles or striving for a resonance of different bells and chimes, since the issues involved understanding the highest wisdom in the deepest sense. They sat face to face, but the barriers in communication were quite insurmountable.

(Cheung 2006, pp. 177–78, with minor revisions)

Without qualified translators with adequate bilingual competence, translations had to be conducted in cooperation. The visiting monk as a chief translator (譯主 yizhu) recited or read out a text, which was then interpreted and explained by a bilingual assistant or interpreter (傳語 chuanyu). Then, scribes (筆受 bishou) wrote the information down, possibly for further polishing or editing. The chief translator provided the scriptures; therefore, they decided on the contents to be rendered. The Chinese assistants could only process the texts provided by the foreign monk; however, they could influence the written form and contents. Therefore, the translated scripture’s final language was the product of reconciliation.

However, the reading, understanding and evaluation of translated texts encountered another process of collision and compromise. BC’s development required the efforts of generations. It was first tentatively used or created by separated groups in different places and at different times. In different time periods, there were different requirements and needs for Buddhism and BC because monks and lay believers had different mother tongues, Chinese levels, and literary traditions, as well as ideological and cultural backgrounds. The varieties of scriptural languages and styles in early scriptures revealed translators’ efforts and compromises with the whole speech community, which ultimately determined whether a text and its language could be accepted.

In the following sections, we are going to examine in detail the features of the early Buddhist community and demonstrate how its development influenced BC and scriptural translations.

3. The Western Regions as Frontier and Non-Han Peoples as Pioneers in the Spread of Buddhist Scriptures in China

As for the route of Buddhism’s introduction to China, there are two different views. One holds that the religion was introduced from the northwest and north. In line with this view, Tang ([1938] 2000, p. 63) argued that: “According to records, most people go abroad by land since the opening to the Western Regions in Emperor Wu of Han’s era… For states such as Kusana (月氏 Yuezhi), Sogdiana (康居 Kangju), Parthia (安息 Anxi) and others, land routes were undoubtedly mostly chosen. Even for Indic monks in the Western and Eastern Jin eras, sea routes were not preferred”. Another view points out the existence of Buddhism in coastal areas, south China and southwest China based on archaeological discoveries.3 We believe that religion can spread through several channels. The artifacts could imply that Buddhism was accepted by common people as a kind of vulgar belief in some areas, whereas in other areas, it predominately spread among the learned and cultured people through scriptures.

Considering the characteristics of Buddhism’s spread in China (Zürcher [1985] 2013d, p. 384; [1990] 2013e, p. 366), the Western Regions are the frontier and transfer station for the dissemination of Buddhist scriptures, and it was even to some extent regarded as a replacement for India because of its proximity to inland China. For example, Zhu Shixing (朱士行, 203–282 CE) travelled to Khotan for scriptures (Hui Jiao 1992, p. 145). Dao An (道安, ca. 312–385 CE), an eminent exegete who encountered difficulties in exegesis, had to wait for new scriptures from the Western Regions (Seng You 1995, p. 265). Furthermore, some scriptures were first translated to a local language in a specific region before being introduced to Chinese as original and genuine Indian scriptures. Compared with scriptures written in an Indic language (梵本 Fanben), those written in other languages or brought from the Western Regions were imprecisely called Hu scriptures (胡本 Huben)4. Subsequently, for the Chinese who were lacking in awareness or knowledge, or perhaps just for the sake of habit, Buddhist scriptures were called “Hu scriptures”, and “Hu” and “Fan” were used indiscriminately in many cases. The problem was not noticed or corrected until the sixth to seventh centuries, as revealed in Yan Cong (彥琮, 557–610 CE)’s comments.5

Credible historical records such as Records of the Emanation of the Chinese Tripitaka (出三藏記集 the Chu sanzang ji ji) and Biographies of Eminent Monks (高僧傳 the Gaoseng zhuan) provide evidence that the majority of pre-fourth-century transmitters, translators, and many assistants came from the Western Regions, and were not Han Chinese. According to the first part of “Monks as translators” (譯經上 Yijing shang) in the Gaoseng zhuan, about two-thirds of chief translators came from the Western Regions.6 From studying the second and third parts, i.e., the biographies of monks since the fourth century, we can see that the proportion of translators from India and China increased. As Zürcher ([1999] 2013g, p. 553) pointed out: “The earliest phase, from the second to the late third century, is characterized by the activity of Indo-Scythian, Parthian and Sogdian missionaries, to such an extent that the period from ca. 150 to 270 CE may be called the era of western Central Asian dominance”. Because those chief translators were the founders and leaders of Buddhist groups, their mother tongues and Chinese proficiency cannot be underestimated as factors in BC development.

Moreover, the majority of the early Buddhist community comprised various Hu peoples, which means that the majority of the translated scriptures’ readership was not Han Chinese. In fact, before the early fourth century, the Han people were not officially allowed to be monks. It is recorded that:

佛出西域，外國之神，功不施民，非天子諸華所應祠奉。往漢明感夢，初傳其道。唯聽西域人得立寺都邑，以奉其神，其漢人皆不得出家。魏承漢制，亦修前軌。

(Hui Jiao 1992, p. 352)

Buddha, having been born in the Western Regions, is a foreign god. His merit does not help the people, and he is not one whom the emperor and the Chinese should sacrificially worship. Formerly when the Emperor Ming of the Han had his dream and they first passed on this teaching, they only permitted men of the Western Regions to build temples in the capitals wherein to worship their god. No Han Chinese of this period were permitted to leave lay life. Wei adopted the Han regulations and followed the precedent.

(Wright 1948, pp. 355–56, with minor revisions)

The ban was still effective in the Eastern Jin Dynasty. One record in the Treasury Literature of Dharma (法苑珠林 the Fayuan zhulin) states that: “In the era of Taikang, the Han Chinese were not allowed to be monks”.7 Another in The Collection for the Propagation and Clarification of Buddhism (弘明集 the Hongming ji) records that: “Previously in the Jin Dynasty, monks originated from different Hu peoples, and there were very few Chinese worshippers. As they were not allowed to be exposed to Han people, their customs and conventions can be maintained.”8 Furthermore, there are even examples of Chinese subjects being killed for not being willing to resume secular life (Hui Jiao 1992, p. 27).

The above materials indicate that, before the fourth century, it was impossible for large numbers of Han people to officially become monks. We cannot deny that there must have been secret support for religious activities, nor can we neglect the fact that high-rank officials or literati can still communicate with monks and learn about Buddhism, as has been recorded in works such as A New Account of Tales of the World (世說新語 the Shishuo xinyu). However, for them, Buddhism represented no more than fresh ideas, a topic for social interactions, or even a kind of magic.

Considering the situations of the Han people, we deduced that early monks were dominated by various Hu peoples living in China. According to the Gaoseng zhuan and the Chu sanzang ji ji (also see Yan 2007), during the third to early fourth centuries, most scriptures were translated in large cities such as Luoyang (洛陽), Chang’an (長安), and Jianye (建業), or in northwestern cities, such as Liangzhou (涼州) and Dunhuang (敦煌). It was sensible for foreign monks to bring scriptures to the Buddhist communities in areas that would support their living, studying, translating, and preaching. However, due to a lack of credible sources, it is very difficult to give detailed descriptions of the sizes, distribution, and daily operations of these communities, although we still found some valuable information. In the biography of Zhi Qian (支謙, fl. ca. 229–252 CE), there is a record that his grandfather and about 100 expatriates came to China together in the late second century (Seng You 1995, p. 516). Additionally, records show that more than 100 Sogdiana aristocrats from Samarqand lived in Central China in the early fourth century (Wang and Li 1997, p. 73). According to regulation and governance traditions, those foreigners living in China as expatriates or diplomats had their own settlements.9 Furthermore, non-Han merchants in China also had their own congregations and were allowed to keep their customs, and as Buddhist believers, they could establish temples and hold their rituals (Ye 2010).10

The various Hu communities were important incubators for the spread of Buddhist scriptures. We use the example of Zhu Shixing (Hui Jiao 1992, pp. 145–48) to demonstrate how scriptures were spread among those communities. We hypothesize that Zhu was very likely to be a naturalized Khotanese. Though he was described as a preacher, there were no records of him learning a foreign language or Buddhism. If he was the first Han Chinese monk, his learning experience must be worthy of recording. By contrast, we can see detailed descriptions of the conversion of Bo Yuan (帛遠, ?–305 CE), a Han Chinese subject who was determined to be a devoted monk, despite all the associated difficulties (Hui Jiao 1992, p. 36). Moreover, Zhu had Khotanese disciple(s), so it is sensible to believe that Zhu was of Khotanese origin and could speak Khotanese.

According to Zhu’s biography, the scriptures obtained in Khotan were brought by his followers to Luoyang, Xuchang (許昌), and a temple in Chenliu (陳留), and the manuscripts were kept for a long time in each place. After several years, the scriptures were translated into Chinese and requested by other monks together with their original texts. We determine two key points from this record. First, unlike Chenliu, there were several Buddhist communities or temples in large cities, such as Luoyang and Xuchang, so not all the details were recorded. Second, various communities required the newly brought scriptures, so they were kept in each place for a period of time to provide a sufficient duration for copying and research.

In this case, if Zhu were a naturalized Khotanese, then it is plausible for him to have had a passion for Buddhist learnings, to have travelled to Khotan for the original scriptures, to have had Khotanese disciples, and to have stayed in Khotan for dozens of years until his death. If he were Han Chinese, the record would show the existence of various non-Han communities that could read and study original scriptures, and that the different needs of the scriptures were arguably influenced by the linguistic competencies of each specific speech community. In both situations, we can see that Buddhist communities in and out of inland China were connected through scriptures. When arriving in China, it was natural for foreign monks to spread scriptures and study Buddhism in communities belonging to their own ethnic group and communicate with them in their own ethnic language.

In this section, we demonstrated the main routes for scriptures to be introduced to inland China and subsequently translated. As the Han people were not legally allowed to be monks, we argue that early Buddhist speech communities in China were dominated by various non-Han peoples, whose ethnicity influenced the development of BC.

4. Language Used in Early Buddhist Speech Community and Its Causes of Formation

Many scholars have conducted textual and linguistic analyses regarding BC. Zhu (2008) argues that it has two basic features: blending Chinese with foreign elements and blending literary Chinese with oral elements. Zürcher ([1977] 2013a, pp. 29–30) points out its specific divergence from standard classical Chinese, including vulgarisms and oral elements, the simple and artless syntactic patterns, the limited vocabulary, extreme frequency of binomes, a considerable number of verbal compounds, a strong reduction of forms in the pronominal system, the frequent use of “plural” suffixes, the comparative rarity of the nominal sentence with “也”(ye), and a great many other points by which it deviates from the largely standardized literary Chinese of the same period; thus, he assumes that the language of some texts might be a kind of pidgin used by a foreign master with inadequate native assistance. Philologists (Cao and Yu 2016) also highlight other grammatical phenomena, such as the use of vocatives, reflexive pronouns, postpositive attributes, etc. Mair (1994, p. 710) argues that medieval Buddhist texts were decidedly more vernacular and colloquial than their non-Buddhist counterparts from the same period. Considering the status and reputation of literary Chinese, BC, as a new social dialect with unique features could not be attributed to Han Chinese but to non-Han members in Buddhist speech communities.

A comparative analysis of an original Hu-language text and its translated Chinese equivalent specifies the influence of the non-Han members; however, there are many difficulties regarding these studies. The first is the possible existence of vernacular “literature” in oral realm (Mair 1994, p. 717), which means it would be very difficult to find sufficient texts for research. According to the Gaoseng zhuan, there were about 36 oral and written languages in the Western Regions (Hui Jiao 1992, p. 23). It is not easy to determine the specific mother tongue of a particular monk, especially for a language that had no writing systems at that time. The second is that only a small fraction of the early translations have survived, though possible future unearthed texts may help promote our studies, and change our knowledge of the actual language or scriptures used in early medieval China. The third is determining paralleled texts, including the authorship of each text involved and their reliability and relevance. The fourth is the possible editing performed by scribes during the translation process. It would be difficult to determine who actually influenced a specific linguistic feature in a text. The fifth is the possibility that the original texts from which the first translators were working may have differed substantially from those available to us today (Nattier 2008, p. 44). The sixth is possible pseudotranslation, adaptation, and borrowing in translation, which could make it difficult to perform a linguistic analysis to determine the style of an author or the authorship of a scripture (Nattier 2008, pp. 103, 125, 152). The final difficulty is that the influence of ethnicity is usually combined with other factors, which makes it very difficult to determine typical linguistic phenomena that were only influenced by the mother tongue of the translators or interpreters.

Considering all the difficulties, we will study the formation of BC in a macro fashion instead of using linguistic analysis in the following sections. From the speech-community perspective, to contribute to an understanding of BC linguistic features, we will analyze the issue regarding linguistic competence, the universal features of translated language, and perspectives on translation.

4.1. Linguistic Competence of Chief Translators

Chief translators brought scriptures to China and were in charge of translation. As authorities in exegeses and leaders of Buddhist groups, their varied degrees of linguistic competence influenced translations considerably.

First, some were of foreign origin and had lived in China for generations. They usually could speak one Chinese dialect and learned foreign languages before translating. A prominent example is Zhi Qian. According to his biography (Hui Jiao 1992, p. 15), he began to learn literary Chinese when he was 10 years old. At the age of 13, he learned Hu languages, mastering six of them. As a multilingual, he began to collect scriptures for translation. Other examples are Kang Senghui (康僧會, ?–280 CE), Dharmaraksa (竺法護 Zhu Fahu, ca. 230–316 CE), Zhu Shulan (竺叔蘭, fl. the third to fourth century), and Zhu Fonian (竺佛念, fl. the fourth to fifth century), who acquired foreign languages and Buddhist learnings via travelling or in monasteries. High-level bilingual monks, such as Zhu Fonian, were invited by various groups to perform translations, even if their Buddhist learnings were limited, which demonstrates how precious a qualified translator was at that time (Hui Jiao 1992, p. 40).

Second, some foreign monks began translating soon after they arrived in China, even though they did not have adequate Chinese proficiency. Some examples are Zhu Falan (竺法蘭, fl. the first century), An Shigao (安世高, fl. the second century), Lokaksema (支讖Zhi Chen, fl. the second century), Zhu Foshuo (竺佛朔, fl. the second century), Vighna (維祇難Wei Qinan, fl. the third century), and Zhu Jiangyan (竺將炎, fl. the third century). Whereas some were described as mastering Chinese in a short time, such as An Shigao, others were requested to perform translations despite not being qualified, such as Vighna and Zhu Jiangyan. An Shigao’s translations have been evaluated as plain and sometimes not expressive (Seng You 1995, p. 254), whereas Vighna and Zhu Jiangyan’s translations represent an almost literal transition with many transliteration elements (Seng You 1995, p. 273).

Third, some foreign monks gradually learned Chinese upon arrival in China, including An Xuan (安玄, fl. the second to third century), Buddhararksa (佛圖羅刹, fl. late fourth century), and Sanghadeva (僧伽提婆, fl. the fourth to fifth century).

Fourth, some monks seemed to be unable to speak Chinese, and instead contributed to translations by providing sources, either through providing physical scriptures or reciting by memory. Some examples are She Moteng (攝摩滕, fl. the first century), Srimitra (帛屍梨蜜多羅, fl. early fourth century), Sanghabhadra (僧伽跋澄, fl. late fourth century), Dharmanandin (曇摩難提, fl. late fourth century), and Sangharaksa (僧伽羅叉, fl. late fourth century). Concurrently, there were monks of foreign origin that were born in China and only spoke Chinese. Arguably, those monks had little direct influence on the translation process. Moreover, very few Han people before the fourth century are recorded as having acquired foreign language(s) to become a translator.

Generally speaking, despite some unrealistic records of tremendous memory or grasping Chinese at an amazing speed (Hui Jiao 1992, p. 4), it is difficult to command literary Chinese in a short amount of time. Thus, we shall be cautious regarding the different descriptions of monks’ linguistic competencies because some materials just reflect peoples’ expectations and desires for qualified translators, and thus cannot be literally understood. A good Chinese proficiency and translation competence required time and practice. For example, Tan Wuchen (曇無讖, ca. 385–433) spent three years learning Chinese before beginning to perform translations (Hui Jiao 1992, pp. 76–77), some of Dharmaraksa’s late works were directly interpreted into Chinese by himself (Seng You 1995, p. 274), and Nie Chengyuan (聶承遠, fl. the third to fourth century) began to independently translate scriptures after years of experience as an assistant to Dharmaraksa (T.55.2154. p.501a).

Furthermore, it is almost impossible for a foreigner to grasp literary Chinese as well as Han literati. Many so-called “translators” acted as an interpreter, such as An Xuan and Sanghadeva. Even for Kumarajiva (鳩摩羅什 Jiumoluoshi, 344–413 CE), whose translations were widely acclaimed, an evaluation from his disciple Huiyuan (慧遠, 334–416 CE) was that: “My master is not good at (literary) Chinese and can just translate literally. Though he speaks fluently, he has not commanded written language well” (法師于秦語大格，唯譯一往，方言殊好，猶隔而未通) (Seng You 1995, p. 387). Even Kumarajiva himself frankly admitted to his disciple Sengzhao (僧肇, ca. 384–414 CE) that “My understanding is not inferior to you, but my expressions are less rhetorical” (吾解不謝子，辭當相挹) (Hui Jiao 1992, p. 249).

4.2. Translation Assistants and Their Linguistic Competence

Most early translations were conducted via translation groups, which later evolved into institutionalized translation studios. In a group, interpreters and scribes were the main assistants. In the Chu Sanzang ji ji, more than 30 scribes are recorded, and in A Catalogue of Translations of All Times (古今譯經圖紀 the Gujin yijing tuji, T.55.2151), there are about ten interpreters and more than 60 scribes. Furthermore, some that were recorded as translators in the Gaoseng zhuan, such as Sanghadeva, Zhu Fonian, and Zhu Fadu (竺法度, fl. the fourth to fifth century) (Hui Jiao 1992, pp. 33–34, 38, 67), were actually interpreters.

Arguably, there must have been additional, unrecorded, and anonymous translation assistants. For monks that only provided source scriptures for translation, such as She Moteng, Srimitra, Sanghabhadra, Sanganandin, and Sangharaksa, there must have been translation assistants. Additionally, there were no records of Chinese learning for others such as Zhi Yao (支曜, fl. late second century), Kang Ju (康巨, fl. late second century), Kang Mengxian, Dharmakala (曇柯迦羅, fl. the third century), Kang Sengkai (康僧鎧, fl. the third century), Tan Di (曇帝, fl. the third century), and Bo Yan; therefore, they were possibly helped by interpreters. For example, in the Gujin yijing tuji, Zhu Dali (竺大力, fl. late second century) is recorded to have come from the Western Regions and translated the Xiuxing benqijing (修行本起經) in two volumes in Luoyang with Kang Mengxiang (康孟詳, fl. the second century) as his interpreter (T.55.2152. p.350b). In contrast, the record in the Gaoseng zhuan was that Mengxiang translated the scripture (Hui Jiao 1992, p. 11). One more example is Dao An, who was helped for several scriptures by the interpreter Zhu Fonian and the scribe Hui Song (慧嵩, fl. the fourth to fifth century); however, the two assistants were not acknowledged in both the Chu sanzang ji ji and the Gaoseng zhuan11.

As for scribes, most were grassroots literati with mediocre linguistic competence and taste. They were not recorded in official histories and have left behind very few acclaimed works12. The languages they handled were in many cases crude or plain because they had been orally transmitted by foreign monks. With a limited understanding of Buddhism, it was not easy for scribes to polish, process or manipulate the intermediary languages and make them more refined, especially when translating large quantities of religious concepts and culture-specific terms. In addition to the lack of cultured elites in translation teams, we also consider the lack of Chinese speakers in the Buddhist speech community, as well as the time needed for the growth of a mature religious language. Hence, we cannot expect to see the emergence of great masters of the BC language at that time.

Compared with chief translators, who were responsible for understanding of scriptures and content, scribes were generally more responsible for the expression of scriptures, as a comment from Seng You noted: “Translators can decide whether the contents are right, while scribes influence the style of expressions” (義之得失由乎譯人，辭之文質系乎執筆) (Seng You 1995, p. 14). In conclusion, scribes’ writings can influence BC and the linguistic features of scriptures, which may help to explain the disparities in a translator’s style.13

4.3. Common Readers’ Influence

Common readers of Chinese scriptures had an overall majority in the Buddhist speech community. As we discussed in Section 3, foreign communities in China formed the main core of Buddhist believers in China, and included professional clergymen, devout believers, and common followers, who were involved in religious activities in different ways. Those foreigners or expatriates living in China were legitimately allowed to become Buddhist monks, and they were the actual major recipients of early Buddhist translations in China. They may have been Sinicized to some extent and had different levels of Chinese proficiency. Their linguistic tastes, preferences, and requirements for Chinese were different from the Han people or cultured elites; therefore, they influenced and shaped BC as a hybrid language.

Some philologists have also partly attributed the language’s hybridity to reader influence. For example, Nattier (2008, p. 18) argues that there are extremely difficult multisyllabic transcriptions in the vernacular-oriented translations of Lokaksema that would not be comprehensible to an ordinary Chinese reader (or listener) of the time, whereas in Kang Senghui’s elegant compendium of stories, there are relatively few transcribed proper names, and they have generally been domesticated to conform with the standard maximum of three characters for a Chinese name. Thus, she argues that there are different audiences: on the one hand, an audience of Chinese literati, who would have expected a fine classical style and would have found long foreign terms inelegant and coarse, and on the other, an audience of immigrants of various nationalities that were able to use a modicum of spoken Chinese, despite lacking a Chinese literary education, but with a higher tolerance for foreign terms, even if the language was not their own (ibid.). Nattier’s observation rightly points to the influence of readers. Whatever style a translator or scribe chose, it was the general readers in the speech community that finally determined whether a scripture was accepted and circulated.

In the early multilingual monastic communities, non-Han members learned Chinese as a second language or foreign language, and their Chinese was mixed with elements from their respective first languages. From the perspective of sociolinguistics, their Chinese could be regarded as pidgin, with simplified grammar and word usage to facilitate communication (Bussmann 1998, pp. 904–5). The findings of historical linguists also support our argument. Yu (2008) points out that the Chinese used in scripture translation demonstrates the performance of foreign monks as second-language learners, whose mother tongues interfered with their expressions; therefore, their Chinese is to some extent an interlanguage. We can see the transfer of features from translators’ first languages into their Buddhist Chinese, such as plural forms, the passive voice, and other grammatical or lexical features (Cao and Yu 2016). The translated Chinese with Hu or Indic features was not a problem and was even useful for readers with a knowledge of those languages. However, it was not convenient for Chinese readers without a foreign-language background, especially for Han literati who were accustomed to reading refined language. This can partly account for the prevalence of primary BC before the fourth century and its change after that, as will be discussed in Section 5.

The readership’s role can also help to explain the unexpectedly and unintelligibly high degree of vernacular content in Buddhist texts (Mair 1994, pp. 712–13), especially for non-Han Chinese readership. Through history, we can see many translations that have been adapted for readers, such as Yan Fu (嚴複, 1854–1921 CE)’s choice of literary Chinese in his translation of western social science classics in the late 19th century for traditional Chinese intellectuals, and the vernacular translations of the Bible in early modern China aimed at the grassroots. Therefore, we cannot afford to ignore the vast number of readers influencing translation.

4.4. Influence of Translation Practices and Views

The influence of translation on the development of BC can be demonstrated in two ways.

First, BC reflects some universal features of translated language. As Mair (1994, p. 712) has pointed out, BC includes various types of grammatical and syntactical discrepancies with literary Chinese, whereas texts composed by native authors approximate literary Chinese more.

In translation studies, many scholars have investigated the universals of translated language and revealed that when compared with texts written in one language, the language of the translated text will show simplification, explication, or normalization (Baker 1993; Kenny 2001; Olohan 2004). This means a translated text tends to use more common words and syntax (Baker 1996), and implied and inferred information tends to be explicated (Vinay and Darbelnet 1958; Shuttleworth and Cowie 1997, p. 55). Thus, translated language is typically less sophisticated, cannot fully convey the rhetorical or rhythmical features of source texts, and is evaluated as plain or awkward by cultured readers.

Features can also be exaggerated by the translation process. Distortion may occur in the process of translation, whether from a chief translator’s understanding or oral explanation of the source scripture, an interpreter’s understanding of his expression, or the interpretation of a scribe’s writing. Furthermore, distortion can be amplified through relay translation, such as translating Indic texts to Chinese via Hu texts, because the process would create difficulties for translators attempting to maintain the aesthetic elements of source texts.

Second, BC demonstrates the influence of views on translation. Philological studies tend to focus on texts and seldom take writers’ views into consideration. However, the problem of language began to haunt people from the beginning of scriptural translation when translators were bewildered by all the intricacies of creating a new language for Buddhism in the Chinese cultural context. They conducted different experiments from literal translation, transliteration to free paraphrasing, and tried a variety of combinations of different methods. Early in Zhi Qian’s era, a discussion on the language and style of scripture in Chinese was recorded (Seng You 1995, p. 273).14

Though there were different practices, in theory, it is usually better to strictly adhere to the original texts in religious translations, because scriptures are deemed to be holy words left by saints (Seng You 1995, p. 290). A typical view is that translators only need to convey the contents of scriptures and ignore the issues of language and rhetoric; therefore, the best way was to translate as literally as possible (Seng You 1995, p. 382; Cheung 2006, p. 85). This logic can be analogized to Saint Jerome’s (ca. 347–420 CE) proposal that a holy scripture shall be rendered word for word, because even the syntax contains mystery (Robinson [1997] 2002, p. 25). The theoretical preference for the “Zhi” style (though not necessarily practically (Ren 1981, p. 174)) can help to explain how BC, as an immature language, is tolerated and used in so many Buddhist texts.

In this section, we analyzed the features of BC from the viewpoint of the linguistic competence and language learning of non-Han people, as well as the universal features of translated language and the influence of views on translation. Our analysis not only consolidates observations regarding the nature of BC based on textual analysis, but also demonstrates why BC can only be such a language at that time.

5. Change of the Speech Community behind Development of Buddhist Chinese

Because Buddhist scriptures were produced, spread, and used in Buddhist speech communities, changes in speech communities can influence language use and translation practices.

When non-Han people dominated the Buddhist speech community, the lingua franca was a variety of Chinese. At that time, Buddhism was a “heterodox” cult adhered to by more or less Sinicized foreigners, a part of the illiterate population and a comparatively small number of literate and semi-literate individuals (Zürcher [1959] 2007, p. 72). Though some Chinese subjects participated in translation as assistants, in general, the Chinese elites and literati could not join the Buddhist speech community to fully influence BC development.

In the late third and early fourth centuries, an intellectual clerical elite formed that consisted of Chinese or naturalized foreign monks (Zürcher [1959] 2007, p. 71). Furthermore, the final collapse of the Western Jin signaled the conquest of northern China by different non-Han peoples, leading to larger numbers of non-Han migrants, who settled in the north and interacted with local Han people. New rulers and their people brought new social values and customs to inland China. With the breakup of Han Chinese life and culture in the north, Buddhist domestication was promoted, and it found acceptance among both the surviving gentry and the mass of the peasantry (Wright 1990, pp. 15–16). In the 330s, an edict was issued under the influence of Fo Tucheng (佛圖澄, ca. 232–348 CE), a monk and state adviser:

朕出自邊戎，忝君諸夏，至於饗祀，應從本俗。佛是戎神，所應兼奉，其夷趙百姓有樂事佛者，特聽之。

(Fang 1974, p. 2488)

We were born out of the marches, and though we are unworthy, we have complied with our appointed destiny and governed the Chinese. As to sacrifices, we should follow our own customs, equally with those of the Chinese. Buddha being a barbarian god is the very one we should worship. Now regulations are carried into effect. As for I, the Chao, and the myriad barbarians, if there are those who abandon their unauthorized worship and take pleasure in worshipping Buddha, we hereby permit all of them to become adherents.

(Tsukamoto 1979, p. 261, with minor revisions)

The edict, of comparable significance to The Edict of Milan (313 CE), officially removed the restrictions on Han people joining monasteries. Considering the population difference between the Han and non-Han peoples, as well as the impact brought by the new policy, it is rational to suppose that the new believers and monks were predominately Han people. If we further compare the growth of Buddhism with that of Christianity in the first several centuries, even a conservatively estimated growth of 40 percent per decade (Stark 1997, pp. 6–7) means that Han monks and believers easily outnumbered their Hu peers at the end of the fourth century15.

The demographical structure of the Buddhist community began to change when huge numbers of Han monks were allowed to be monks. From the fourth century, many Buddhist groups emerged that were led by Han monks. For example, Dao An was once a disciple of Fo Tucheng when he was young. Later, he became a leader of a Buddhist group of hundreds of followers. Additionally, he dispatched his disciple Hui Yuan (慧遠, 334–416 CE) to spread Buddhism in southeast China, successfully establishing new monasteries and developing new Buddhist groups.

The proliferation of Han Buddhist monks means that their specific needs and requirements began to be valued. For the newly emerging cultured monks that combined Buddhist doctrine and traditional Chinese scholarship (Zürcher [1959] 2007, p. 6), all Buddhist activities were practiced using their mother tongue. They needed more fluent, expressive, and refined writing conforming to traditional writing norms, and their concerns were demonstrated in their scriptural studies.

Dao An was one of the most notable of these cultured monks. Born in a family of Confucianists, he had very prominent secular learnings. As a scholar, he had many reflections on, and criticisms of, previous translations, and he gradually proposed the idea of Five Deviations (五失本 Wu shiben), i.e., the five occasions when translations can deviate from source scriptures. The first and second instances concerned grammatical and stylistic disparities between Chinese and foreign languages, which reflected Dao An’s doubts regarding the prevalent assumption that Buddhist languages could only be translated in a “Zhi” style. His proposal was repeatedly criticized by his patrons, his own disciples or other participants in translation studios.16 The debate of “Wen” and “Zhi” was started by Zhi Qian; however, Zhi Qian’s and Dao An’s preference for the “Wen” style failed to garner wide support. One of the most important reasons was that they could not refute the misconception that the Buddhist language was inherently “Zhi”, and consequently could not convince people that it was religiously acceptable to translate Zhi-styled scriptures to “Wen”-styled Chinese writings.

Scholars generally agree that the arrival of Kumarajiva helped to change people’s views. He pointed out that “it is a convention for people in India to pay much attention to stylish diction and writing” (天竺國俗，甚重文藻) (Seng You 1995, p. 53). His argument was accepted because of Kumarajiva’s identity, experience, bilingual capability and reputation. We think that the crucial factor behind the change also involved changes in the Buddhist Chinese speech community. In Kumarajiva’s time, the Buddhist community had become a group dominated by the Han people. Though their reading capabilities regarding the original texts were inferior to foreign monks, the Han could learn from their heritage of hundreds of years of scriptural studies, annotation, and translation. Combined with their advanced literary Chinese and traditional cultural learnings, they helped to promote the development of BC and the translation quality of Buddhist scriptures. Furthermore, at that time, Chinese elites began to intellectually accept the religion, which is reflected in the emergence of different schools and sects that fused Daoist metaphysics and Buddhist philosophy (Hurvitz 1975). It was a time for Buddhism in China to develop from a period learning and imitation to localization and integration. Thus, behind Kumarajiva’s personal translation preferences and views, we also consider his adaptation to the accelerating trends of Buddhist assimilation. As increasing numbers of local elites began to read and write Buddhist works in Chinese, Kumarajiva cooperated with his learned assistants to create and produce a new language and new translations, fulfilling locals’ needs for a high-quality BC that corresponded with a deeply domesticated Buddhism.

Kumarajiva’s contributions to creating a standardized scriptural style was to remain exemplary for several centuries (Zürcher [1995] 2013f, pp. 452–53), and he was actually a speaker of the new Buddhist speech community at that time. Behind him, there were hundreds of participants in his translation studio (Seng You 1995, p. 293), including dozens of eminent assistants, such as Dao Sheng (道生, 355–434 CE), Seng Rui (僧叡, fl. late fourth to early fifth century), and Seng Zhao (Hui Jiao 1992, p. 142), who were also important contributors to the new speech community.

Ultimately, the development of Buddhist Chinese is inextricably linked to the development of Buddhism in China. The religion would not have been successful without an effective religious language. On the one hand, Buddhism had to adapt itself to Chinese culture, with Buddhists learning to express their teachings and ideas in Chinese. On the other hand, the Chinese language adapted itself to Buddhism, and provided sufficient vocabulary to successfully absorb Buddhist culture. When the domestication of Buddhism widely occurs, such as in thought and philosophy, religious beliefs and practices, social values and behaviors, and literature and art, BC, as a media and carrier, has to grow to a corresponding level for believers to utilize it in the new era.

The growth and maturity of BC is a cross-generational effort. Everyone, from chief translators, interpreters, and scribes to the more numerous general Buddhist Chinese users, can contribute to the circulation and development of the language in a speech community. Thus, studies on speech communities can help explain the historical features and development of BC.

6. Conclusions

Buddhist scripture translation is a complicated issue that cannot be explored with unidimensional scholarship. In our paper, we examined how BC historically developed to help clarify some language features. We utilized a speech-community perspective to re-evaluate and reinterpret neglected materials on ethnicity and translation history. First, we considered the dissemination of Buddhist scriptures and argued that their pre-fourth-century circulation was primarily in non-Han-dominated communities. Second, we investigated the influence of early Chinese scripture translators, translation assistants, and readers and discovered that they did not need, or could not write, high-quality literary Chinese. Third, we argued that early BC was a hybrid language affected by the language-learning conditions of various non-Han peoples; therefore, BC conforms to some common features of translated languages. Furthermore, BC is also closely related to the “Zhi” and “Wen” controversy in translation history, reflecting the Buddhist speech community’s choices and preferences. Moreover, fourth-century changes in the speech community vigorously promoted the religious and linguistic domestication of Buddhism, which helps to explain the coming of a new era of translation and BC since Kumarajiva’s time.

Compared with discussions of, or investigations into, BC hybridity that focused on very specific materials and processes (Zürcher [1978] 2013b, p. 85; Mair 1994; Nattier 2008, p. 18; Zhu 2008), we provided a more comprehensive explanation from the speech-community perspective and focused on the influence of ethnicity—an important factor in BC development. In the future, more specific investigations and case studies should be performed, exploring particular Hu communities, translation teams, or texts produced by translators or interpreters of a particular ethnicity. To conclude, we hope our analysis can shed some new light on early medieval Chinese Buddhism.