Abstract
William Carey’s historic role in Bible translation is widely recognized. That Carey’s actual translations were of an inadequately low quality is not so widely known. This article, while not undermining Carey’s importance as a pioneer, points out five reasons why Carey’s translations were never widely used. Modern understandings of translation inform this paper, and Carey’s historical context explains many of his weaknesses. Not only is this article historical, but it concludes with the modern repercussions of inadequate Bible translations, calling for new translations in all major India languages that focus on people outside the church.
William Carey (1761–1834) holds iconic status as a central catalyst of the modern missionary movement. The wide range of the work that he and his Serampore colleagues undertook is a marvelous example of the holistic concern that should characterize every disciple of Jesus. 1 Yet, Carey’s fame is most closely associated with his efforts to translate the Bible into many of the languages of South Asia. Indeed, this undertaking became the central focus of Carey’s life and work. The quality of his translation work was poor, however, as has been noted in scholarly analysis, even while being overlooked in more popular literature. 2
Two studies focused on Bible translation provide brief but insightful looks at Carey and the problems with his translations. J. S. M. Hooper and Wesley Culshaw present a broad overview of Bible translation in the Indian subcontinent, which includes an assessment of Carey’s translations into nearly three dozen languages. They point out the “enormous waste of energy” involved in producing translations that proved terribly inadequate. 3 William Smalley also considered Carey’s work in his assessment of Bible translation in the modern missionary movement. Smalley concludes that “Carey’s massive language and translation program was seriously flawed. If there had been competent Bible translation consultants in Carey’s day, they would not have approved for publication much, if any, of the translation work done by Carey and under his supervision. More than that, during Carey’s own lifetime translations appeared, or were under way, which quickly supplanted ones produced at Serampore.” 4
These critics of Carey’s translation work are careful to esteem the man and his vision, and this writer shares that esteem. This article will cull academic studies of Carey’s translations to account for the reasons his efforts fell short. The aim is to be scrupulously fair to Carey and his context, while seeking to understand the reasons for his poor results and their application to translation issues in India today. 5 To that end I present five reasons for the poor quality of Carey’s translations.
Limited linguistic knowledge
First, and most broadly, Carey was in a pioneering situation with only the most basic understanding of linguistics. Smalley notes “Carey’s inadequate understanding of the nature of translation, of language and of communication.” 6 Hooper and Culshaw recognize these shortcomings in commenting on “the state of philological knowledge” at that time. 7 Nothing like modern linguistics even existed then in Europe, although it would develop and come to maturity in close relationship with India and Sanskrit. 8
Reflecting the linguistic naïveté of his time, Carey wrote in 1803 that “we have it in our power, if our means would do it, in the space of fifteen years to have the Word of God translated and printed in all the languages of the East.” 9 Over the following fifteen years Carey and his colleagues did not even succeed in getting translations into known languages in their region, let alone “all the languages of the East.” The task of Bible translation was much larger and more complex than originally assumed.
One specific linguistic misunderstanding pretty much doomed most of Carey’s translations: a bias for Sanskrit, India’s classical language, which was used in the oldest sacred texts and which for centuries had poisoned Indian intellectual opinion against vernacular languages. 10 The learned opinion of the time in Bengal is summarized by Sisir Kumar Das: “Sanskrit scholars continued to despise Bengali, despite the fact that it was the mother tongue of the majority of them. They refused to regard it as worthy of use as a vehicle of literary composition.” 11
Clearly, Carey did not share this opinion, but he did make Sanskrit too central to his vernacular translation processes, based on a false understanding of how, and how significantly, Sanskrit influenced the vernacular languages. Two statements from Anand Amaladass and Richard Young explain the problem:
Carey was convinced, erroneously as we now know, that Sanskrit was the “parent” of the modern Indic languages that he called “secondary cognate languages.” Estimating that at least three-fourths and in some cases as much as seven-eighths of the words in these languages were “understood through the Sanscrit,” Carey believed that “there is little variation throughout the whole of the Indian family” in construction, idiom, and figures of speech.
12
[Carey] labored under the misapprehension that modern Indic languages were directly derived from Sanskrit. . . . [This perspective’s] domination over his method of translating the Bible into other languages was complete; using his Sanskrit Bible as an explicit point of reference, pandits would render the text into their own language and then submit to him a draft for revision, at which point once again Carey had reference primarily to his own [Sanskrit] version, as well as the King James Bible and other texts in the biblical languages. In this process, a remarkable degree of uniformity became evident in the productions of the Serampore Mission. Subsequent versions [by other translators] in turn, especially until the later nineteenth century, often mirrored the earlier, highly Sanskritic models developed by Carey at Serampore.
13
William Carey at age fifty, courtesy of the Center for Study of the Life and Work of William Carey, D.D. (1761–1834), William Carey University, Hattiesburg, Mississippi, USA.
Carey wrongly assumed that vernacular words and phrases carried the same meaning as their Sanskrit cognates and that Sanskrit terms and constructions could be introduced meaningfully into Indian vernaculars. These assumptions help explain why he would try to translate the Bible into languages he did not know that were spoken over a thousand miles away from where he lived, but they also crippled his translations. Das states the result of introducing Sanskrit words into the Bengali Bible: “mixed vocabulary of this sort was bound to be displeasing to a native reader.” 14
In essence, Carey invented dialects for his translations, and his readers found his new language strange and often incomprehensible. Lack of linguistic understanding at the broadest level clearly had a negative impact on the details of Bible translation undertaken by William Carey and his cohort.
India’s undeveloped regional languages
A second reason for the weakness of Carey’s translations is that the vernacular languages in India during his time had not yet been standardized. Vernacular prose writing hardly existed; literature was by definition poetic. 15 Literary (poetic) language was also quite different from spoken language, and tensions remain to this day related to this reality in Indian vernacular languages. 16
Even the scripts into which languages were printed were in flux. Robin Boyd shows that Carey’s Gujarati Bible was poor in every way, not least in using the wrong script. “The [Gujarati] pundits . . . were out of touch with current Gujarati, for the book [New Testament, 1820], printed as it is in Devanagari rather than Gujarati type, was little used in Gujarat, and was superseded almost immediately by the better translation published by the LMS missionaries in Surat in 1821.” 17 Bengali was Carey’s best language, and he is widely recognized for his contribution to Bengali prose. 18 But modern Bengali prose is generally dated to the 1870s and differed from Carey’s earlier attempt. 19
A similar, even more complex, dynamic is in play related to the various dialects of Hindi. Hooper and Culshaw single out these translations for criticism, referring to volumes that “had never been circulated, and it was doubtful whether some of them were intelligible to any portion of the population. This applies particularly to a number of dialects of Hindi into which the New Testament was translated, chiefly between 1820 and 1827.” 20 Since even today the status of Hindi and its dialects is controversial and the focus of politicized debates, 21 it is no surprise that Carey’s work in this realm proved inadequate. 22 The undeveloped state of the languages into which Carey was translating the Bible certainly helps account for the poor quality of his work.
Too much attempted
A third reason Carey produced poor translations is that he tried to do too much. 23 At this point some difficult questions need to be raised about the entire Serampore translation program, particularly how greatly it might have been motivated by competition with other translators who were already at work in various languages. This concern is most clearly the case with the translation into Chinese that Joshua Marshman worked on; even his son strained to justify that project. 24 Carey’s Gujarati translation of 1820 was released barely ahead of a local production in Gujarat (as was the case with the Chinese version). Similarly, Carey’s Telugu New Testament came out the same year (1818) as a locally produced version that was clearly superior. 25
As early as 1806 William Ward, whom Christopher Smith regarded as the best missiologist among the Serampore Trio, raised concerns and objections about the translation project. “As to making Bibles for other missionaries, into languages which we ourselves do not really understand, I recommend them [Carey and Marshman] to be cautious, lest they should be wasting time and life. . . . I tell them that the Jesuit missionaries have made grammars, dictionaries and translations in abundance, which are now rotting in the libraries of Rome.” 26 Ward was prescient, although the Serampore translations did not sit idle on European shelves but in Serampore itself.
Failings of assistants
A fourth reason for the poor quality of Carey’s Bible translations is his overly confident dependence on untrained or poorly supervised local assistants. Related to his misunderstanding of Sanskrit and to his laudable goal to translate the Bible into many languages, Carey clearly trusted the predilections of some of his local informants more than he should have. The Marathi language translation provides a good example of a failure to do sufficient field research.
Carey’s Marathi version of the Bible (1819, NT in 1806) proved to have been developed in “a dialect peculiar to a district near Nagpur, and the language thus proved useless for general circulation in the Marathi country.” 27 This assessment finds concurrence in a bicentennial history of the Bible Society in Mumbai. “With the help of Pandit Vaijanath of Nagpur, he [Carey] undertook the translation of the New Testament in Marathi. The Serampore press published the Marathi New Testament in Modi script in 1806. Unfortunately, the Marathi dialect used in the Serampore version was prevalent in the Nagpur region only [from where Pandit Vaijanath came] and the Modi script could not be read easily. Hence this version was not popular.” 28
This case is of particular interest because on the surface Carey could not have found a more reliable translation assistant. 29 But no matter how qualified an individual might be, trying to translate the Bible from a place hundreds of miles from where that language is spoken, particularly in an era that predates travel by train, is not a viable undertaking. Aspersions should not be cast on Carey’s assistants, yet there were many complex cultural nuances involved between the British Christian employer and the Indian (usually Hindu) employee. J. C. Marshman provides insight here with a rather stinging comment on the first helpers involved with the first Bengali Bible translation. “The construction of the sentences in the first edition, which the flattery of the pundits had pronounced to be perfect, was so entirely at variance with the idiom of the language that the work was barely legible.” 30
Hooper and Culshaw, as quoted above, identify one of the crippling factors for Carey’s work to be “the resources of Indian scholarship then available.” 31 In no way should Carey’s informants be blamed for failing to transcend their historical context, but it is now evident that Carey’s translations were published without first getting feedback from readers on the intelligibility and literary quality of the texts.
Misplaced focus on words and word order
The fifth reason for the weaknesses of Carey’s translations finally gets to the actual theory of translation. Carey placed an undue focus on words in his translation work. Modern linguistics has demonstrated that words vary in meaning according to contexts, as do even sentences, so the central units in communication are sentences and paragraphs. But Carey assumed that words were central and that words in one language have counterparts in other languages, which convey the same meaning. J. C. Marshman reported that Carey, in his first Bengali translation, “presented a simple translation of the English words arranged in the English order of collocation.” 32 A similar assessment is available for a later translation (1819) into Siraiki. 33
Sisir Kumar Das identifies the error of Carey’s approach:
It was extremely difficult for Carey to translate English sentences into Bengali and to get the right equivalent in the proper context, because translation is not merely a transference of equivalent words from one language to another, even when these are known, which was not always the case, it is also a transference of the spirit of the content from one language to another. Literal translations generally fail to convey the spirit of the original and often lead to complete meaninglessness. . . . [Carey] wanted to be faithful to the words of the original text and his strict adherence to the exact [wording and structure of the] text crippled his style and eventually [he] failed in many cases to convey the spirit of the Bible in Bengali.
34
A focus on words is indeed crippling, and the problem is compounded when married with Carey’s assumptions regarding how directly local terminologies were linked to and dependent on their Sanskrit roots. It is no surprise that unintelligible translations often resulted.
Carey clearly worked on the assumption that the syntax of the New Testament should be transferred rather than translated, as if it were an untranslatable part of the inspired text. Amaladass and Young point out this issue in their analysis of Carey’s Sanskrit Bible. “What rankles most is that Carey so strictly adhered to the syntax of the King James Bible, as if to deviate from it would have detracted from the gospel itself.” 35
This syntax issue is also at the root of the problems in Carey’s Bengali work. Das presents a quite detailed analysis of Carey’s failure to adjust to Bengali syntax, highlighting five features. The English order of subject-verb-object was followed rather than the Bengali order of subject-object-verb. The English pattern of a relative clause following its antecedent (“the missionaries who translate the Bible”) was used instead of the Bengali pattern of the clause preceding (“the who-translate-the-Bible missionaries”). The regular use of the conjunction “and” is a mark of English and particularly of the English Bible, but Bengali does not contain such a proliferation of conjunctions. Bengali generally does not use verb copulas (“the translations [were] disappointing”) at all. Finally, adverbial clauses, like relative clauses, precede the main clause in Bengali rather than following the main clause. Illustrations of all these issues are provided. 36 In summary, “The word order in the sentences of the Bengali Bible is of an English pattern; words are used in collocations which are foreign to Bengali and in consequence the language as a whole presents an unidiomatic appearance.” 37
Similar problems are identified also in Shackle’s analysis of Carey’s 1819 Siraiki New Testament. “This early translation is stylistically very peculiar. Since it is the work of a Hindu pandit, many Sanskritic words are used. . . . Even allowing for this, however, the syntax is very forced and clumsy—indeed, it often hardly reads like Siraiki at all. This is in fact little more than a word for word paraphrase which keeps unnaturally closely to the English version.” 38
Smalley provides an excellent summary of this point. “Carey believed that ‘accuracy’ was enhanced by as literal a translation as possible, with the wording and the grammatical structure of the translation geared to the wording and grammatical structure of the Greek or Hebrew original. Ironically, as with many other translators, his overwhelming drive to make the Scriptures accessible in the languages of India and beyond was thwarted partly by the distortion he introduced through fear of distorting the Bible by making it truly accessible.” 39
Conclusion
The weakness of William Carey’s Bible translations is not merely a concern for historians. The style and terminology of the Bible has a massive impact on the Christian community in every language area, and in much of India the church is linguistically isolated from the mainstream of society because it has learned a strange dialect from the Bible itself. Das explained this reality in Bengal. “Carey’s Bible became the parent of a specialised and restricted type of the language which may be called Christian Bengali. . . . It underwent change in the hands of later writers, but its essential character was stamped on it by Carey. . . . Christian Bengali has never become established as a definite style of Bengali but is nevertheless different from other styles of Bengali Bengali. Within its own linguistic limits, it is consistent in respect of both vocabulary and grammar; and what is perhaps most important, Bengali Christians are used to it and proud of it.” 40
A “Christian” dialect that Christians are comfortable with and proud of is not a problem as long as the focus is on the Christian community and they adequately understand it. 41 But if communication of the message of Christ beyond the Christian community is in focus, this language problem (both the language of the Bible translation and the language of Christians) becomes a major challenge. All the major languages of India need new Bible translations that resonate with people outside the church. Most languages need at least two different translations, one for Muslim peoples and one for Hindu peoples, with special attention also to many so-called dialects that are the heart languages of people.
India is in the midst of massive transition, with emigration, urbanization, and globalization all impacting social and linguistic trends in complex and profound ways. The task of Bible translation into India’s languages that Carey began is far from ended. May this article inspire fresh concern and effort toward a better fulfillment of the legacy of William Carey.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.