Language and the Missionary World Map: Platt’s and Radley’s histories of the BFBS

Two unpublished histories of the British and Foreign Bible Society were written in the 1820s to 1830s (BFBS Archives, Cambridge University Library, GBR/0374/BFBS/BSA/E3/8/1 and E3/8/2). It is unclear to me why there were two, both by BFBS staff, written at roughly the same time; they cover much the same topics, figures, and languages and do not express notably strong or divergent views. What is clearer is why they were never published. Both manuscripts are very lengthy compilations of excerpts, transcripts, summaries, and in the case of the largest manuscript, of literal cutting and pasting from printed BFBS reports. All that material is arranged by language, with a chapter for each language into which the Bible was translated before or during that period, and no attempt at overarching narrative or analysis.

The biggest of the two manuscripts – in 15 volumes and envelopes of some 200 quarto pages each – was compiled by Thomas Pell Platt, the BFBS librarian between 1822-1831 and editor of its Greek, Amharic, and Ethiopic (Geez) versions. By far the largest chapter, filling two half-volumes, is taken up by the Serampore Mission. Serampore was a Danish colony near Calcutta, where a trio of Baptist missionaries churned out the unlikely number of 34 translations between 1800-1837 (i.e. in part before the BFBS was founded). What makes the chapter so large is also that it is largely a collage of the successive printed reports of the Serampore Brethren – reports that are otherwise hard to find even in Cambridge University Library. The same goes for Platt’s chapter about Sinhalese (the main language of Sri Lanka), where disagreements between missionaries turned into a veritable translation war. This recycling process makes Platt’s history a valuable historical source even despite its lack of originality.

Comparative vocabulary from Radley’s History of the BFBS

The other manuscript, though also filling 15 octavo notebooks, is considerably more condensed, enough so to fit into a single archive box. Its author is listed as John Radley, about whom less is known. Still the linguistic information is generally much richer than in Platt’s larger volumes: Radley provides comparative vocabularies and samples of alphabets as well as sketch language maps of Sulawesi and the upper Ganges region. More than Platt, he is inclined to cite and draw his information from recent non-missionary sources; his focus is on the missionary frontier in South/East Asia, whereas half of Platt’s history is devoted to larger and smaller European languages. Accordingly, Radley mixes missionary history with late enlightenment ethnography, taken from the works of British scholar-administrators in India and Indonesia (Colebrooke, Marsden, Raffles, Crawfurd).

What both manuscripts show us is how Bible translation resulted in a linguistic world map. Though written by philologically versed authors, neither was intended as a language encycylopaedia; but they contribute as least as much to our understanding of linguistic dynamics as of missionary history, and with its collection of linguistic ‘specimens’, Radley’s history leans towards a missionary Mithridates. The sheer multilingual scope and – sometimes misguided – optimism of the more industrious translators is staggering. In March 1810, the linguistic prodigy John Leyden promised to deliver gospels in “Siamese, Macassar, Bugis, Afghan or Pushtoo, Rakheng, Moldivian & Jaghatai” (accordingly grouped into one chapter in Platt’s history, although they belong to different regions and language families). With the aid of an unspecified number of “persons who assist Leyden in his literary researches” he estimated that “a year and a half might be sufficient for completing the Afghan, Jaghatai, and Siamese versions, and most probably the Bugis and Macasar” – and true enough, before his untimely death on Java 17 months later, he had pulled off complete gospels in Maldivian, Mark and Matthew in Pashtu, and Mark in Balochi, Makassarese, and Bugis, the latter two delayed by illness of his interpreter.

Leyden’s list of translations, from Platt’s History

Sometimes that optimism was sheer naïveté. Joshua Marshman, one of the Serampore trio of translators, cheerfully announced that he was first translating Confucius with the aid of the Chinese Armenian Joannes Lassar (Hovhannes Ghazarian) and an unnamed ‘Chinese assistant’, and then using the knowledge of Chinese thus acquired in translating the Bible. For all his insistence on method and autopsy, he never set foot in China; it is unsurprising that Chinese converts were rather won by other versions, like that of Karl Gützlaff and Robert Morrison (a work that inadvertently inspired the Taiping Rebellion, a mid-century syncretic millenarian movement that left 10-20 million Chinese dead). William Carey’s Marathi version fell flat for the plain reason that his munshi spoke ‘corrupt’ or nonstandard Marathi. But even Carey’s Sanskrit Bible, though more of a status object than a practical tool for proselytization, had its uses as a matrix for other translations.

 

Triangulation and Translation War

The story of Sinhalese, narrated in detail by both Platt and Radley, is illustrative in this regard, and in other ways. A first translation had been made in the early 18th century by the Dutch clergyman Willem Konijn, which was judged too plain as well as “unintelligible, formed according to the Dutch idiom, and not according to the Cingalese” by the Wesleyan missionaries after the British annexation of the colony. A colonial administrator with a passion for languages, William Tolfrey, undertook a new version with the aid of the converted Buddhist priest Abraham de Thomas and other (ex-)Buddhist clerics. To ensure that the new version was less foreign and more up to Sinhalese literary standards, a parallel version was made in Pali, the holy language of Theravada Buddhism:

To judge of the extent & appreciate the merit of Mr Tolfrey’s labours, it ought to be stated, that he carried forward the Cingalese translation in connexion with a second translation of Dr Carey’s version of the Sanscrit version into Pali; judging it expedient to render every verse into the Pali before it could be revised with effect in the Cingalese. The old Cingalese text was then revisited – it was afterwards compared with the Pali, & also with the excellent Tamul translation of Fabricius; in which the form of expression is so much alike, as to run easily from the Pali into the Cingalese: – but all with continual reference to the original Greek, & our own English version. The Pali though hereafter a work of great utility, only served at present to give precision & clearness to the Cingalese version.

That is from Radley’s History, notebook VI. More detail on the translation process is provided in Platt’s chapter on Pali:

The Pali translation is conducted in this manner. Mr Tolfrey reads from the text of Dr Carey’s Sanskrit Testament a certain number of verses to Don Abraham de Thomas, who writes the whole passage in Pali, as nearly as the idiom of that kindred language will admit. They afterward read over the Pali together, compare it verse by verse with the Sanskrit, and make any correction which in their judgement may be necessary. The Bengalee version is also often consulted in difficult passages, when the Sanskrit phrases are not easily expressed in Pali.

Moreover, their Pali version was used as a matrix by “two learned priests of Matura, Karratote Unnanse, and Bowila Unnanse […] who are both ignorant of English, and totally unacquainted with the Scriptures” for translating several chapters into common Sinhalese, so that it would be “perfectly intelligible to the natives, and free from all improper phrases or expressions borrowed from the English or Dutch languages”. Tolfrey sadly died in early 1817, with half the work done; fortunately for the BFBS, a committee of four missionaries who had been taught by Tolfrey stepped in to set forth his translation, following his ‘style &c.’ as closely as possible, and completed it by 1823.

That was not the end of the story. If Konijn’s version had been too plain, the new version was now criticized for being too difficult for ordinary Sinhalese, who needed a glossary. A remark in the BFBS reports that “The Natives of Ceylon were under the dominion of Europeans for two hundred and fifty years before their conquerors gave them any part of the word of God” provoked an angry letter from the Dutch Bible Society. CMS missionary Samuel Lambrick and three of his brethren protested against the new version because the Word should be for the poor, and because the new version used concepts and honorifics that were Buddhist in origin, thus importing heathenism into the holy writ. The local Bible Society invited Lambrick to provide his own version of six chapters of Matthew, which were not met with approval, after which he ended up publishing a simplified Sinhalese Book of Common Prayer and a Sinhalese Grammar at the Church Mission Press.

 

Time, Souls, and Money

There is no direct correlation between the time and effort involved in these translations and their impact. Although the Ceylon mission proudly claimed to have 10,000 native children in mission schools by the 1820s, with enough demand for 50,000 if there had been enough missionary teachers, Buddhism is still by far the majority religion on Sri Lanka, and most Sri Lankan Christians nowadays are Catholics. The first Javanese translation (1829), by the Baptist missionary Gottlob Brückner, though completed in the early 1820s, was held up by technical difficulties and the Java War (1825-30), and finally printed at Serampore, only to blocked by Dutch colonial authorities who wanted to avoid causing new unrests. The Dutch Bible Society’s own version was two more decades in the making, only to be eclipsed later by a less philological BFBS rival version. On the whole, missionary efforts were much more effective where they sought to replace Indigenous religions, like in Oceania, than when they were up against other ‘world religions’ with literary canons like Hinduism, Buddhism, and especially Islam. Inspired by the success of the LMS mission on Tahiti, and given the great similarities between Polynesian languages, the BFBS sought to save time and effort by using Tahitian as a lingua franca for other parts of Polynesia, but this turned out to be harder than expected.

The Gospels in Amharic, tr. Abu Rumi, ed. Thomas Pell Platt, 1824

A case in which Platt was personally involved as an editor, although he does not mention his own role, was Amharic. In 1820 there were painstaking negotiations (mediated through the French Orientalist mogul Sylvestre de Sacy and the British consul in Cairo) with the ex-priest and dragoman Jean-Louis Asselin de Cherville who wanted to sell a manuscript of a complete Amharic Bible translation for £ 1500. The BFBS offered only £ 750, which was already more than its standard fee of £ 500, in spite of doubts about a translation made by one man and not directly from Greek and Hebrew. In fact it had not been made by Asselin but by the Ethiopian priest Abu Rumi who lived with him in Cairo; in Ethiopia, Amharic was traditionally the language of the court whereas Geez was the language of the Bible. Eventually, Asselin and the BFBS agreed upon £ 1250, and the 9539-page manuscript was inspected by its indefatigable philological factotum, Professor Samuel Lee. (Apart from the details about the negotiations, the story can be read at greater length in William Jowett’s Christian Researches in the Mediterranean (1822), 197-204.) But preparing the manuscript for print took the BFBS nearly a quarter of a century: Gospels in 1824, the New Testament in 1828, and the whole Bible in 1844. Platt does not tell anything about the twenty-four-year editing process, but one can easily imagine him wearily looking up to the skies.

Some insight into our Digital Humanities Stream

by Louis Knölker, Student Research Assistant

As part of our research, one of our goals is to visualize the distribution and development of Bible translations and to make the resulting interactive world map available to researchers on a dedicated website. This will make it easier to analyse such relationships as those between colonial expansion and the global distribution of the Bible. We have sourced a number of books that list various Bible translations. These books provide lots of dates, names, and places associated with the translation of the Bible into about 1400 languages. It would have been conceivable for us to manually enter all the data of interest from our various data sources to the project into a database ourselves. This would have taken a lot of time that we can save with DH tools. Nevertheless, we still have to do a lot of work before the data can be feed into our database. In order to extract this selected data, it is necessary to ensure that the books were in a machine-readable format that can be further processed by the computer. To do this, we needed an optical character recognition tool (OCR). We opted for the open-source program OCR4all from the University of Würzburg. For our process, the following essential steps are required:

  1. Preprocessing
    In this step, each individual page we are interested in is categorized into different sections. This is necessary for a subsequent step of optical character recognition (OCR). In the preprocessing step, for example, a distinction is already made between the year of certain translations and the number of speakers of a particular language. This differentiation of different numbers would hardly be possible for the AI and must therefore be prepared manually.
  2. Text Recognition
    Now the first automated text recognition takes place, as is also known from programs such as Adobe Acrobat or similar. However, this first run is based on a very general model that is not yet tailored to the specifics of the book. Accordingly, there are still many errors, such as the confusion of O and 0. In this state, the data cannot be used for our purposes.
  3. Correction of the automated text recognition
    Any incorrect results from step 2 must now be corrected manually. However, this does not involve correcting the entire corpus that we created, rather only a fraction. These must be meticulously checked to see whether the text recognition matches the original and, if necessary, changes must be made. This quickly reveals typical errors that the program has repeatedly made, such as not recognizing special characters or accents.
  4. Training an adapted text recognition model
    The data resulting from the previous step is now used by the program to create a machine learning model with which the text recognition is tailored to the corpus. This significantly reduces the error rate.
  5. Application of the improved model
    The optimized model is now applied and the text recognition step is repeated.
  6. Correcting the result again
    Once again, a limited number of pages are corrected manually, thereby increasing the data set for further improvement of the AI model.
  7. Repeat steps 3-6 until the text recognition has been optimized
  8. Finalization

Theoretically, this should be the process. Unfortunately, not everything always works as intended and so we also encountered problems that we had to deal with. Unfortunately, step 4 only worked once and since then we have had problems with the program. Therefore, we had to correct many of the passages of the book that were relevant to our regional focus (Arctic, Australia and Oceania, West Africa) completely by hand.

1. Compiling a list of the relevant regional languages

Using a keyword search, I compiled a list of a few hundred languages from the regions mentioned.

2. Linking with the Glottolog

These languages were also linked to their respective data in the Glottolog, an open-access website that lists all the world’s languages and links them to identification codes, among other things. The linking of language and the so-called glottocode will facilitate future research and was planned for the final map. The difficulty in this step lies in the names of the languages. Languages often have different names, which may differ from our base sources and in the “Glottolog” database. Some names are also outdated or may only include one dialect today, meaning that they are no longer considered independent languages.

3. Manual correction of the respective sections

Correcting the individual pages is a relatively time-consuming process in which you always have to compare the original lines with the generated ones. As a result, you have to read every line of the corpus twice. This limits the reading pleasure, as does the often keyword-like text structure, so that you have to make a concerted effort to concentrate.

Nevertheless, there are always exciting passages and amazing anecdotes in our sources. I was often impressed by the great personal commitment of people who often translated the Bible into languages that only had a few hundred speakers. I would like to briefly present one of the best and extreme stories here, namely the story of the Bible translation into the Auca language:

Auca is one such micro-language with just 300 speakers, the Warani, who live in the Ecuadorian jungle. As early as the 17th century, there was peaceful contact between the Warani and a Jesuit priest who lived among them for several years. However, his successor was murdered and contact with the tribe was broken off. It was not until around 300 years later, in 1956, that there was another attempt to Christianize the Warani. A group of five missionaries led by Nathaniel Saint set out to convert the Warani, but all five members of the company were also murdered. Amazingly, this terrible event motivated Rachel Saint, Nathaniel Saint’s sister, and Betty Elliott, widow of one of the murdered missionaries, to try again – this time successfully. Thanks to the great efforts of these women, the entire tribe was evangelized ten years later and the Gospel of Mark was translated. In 1965, the children of the deceased Nathaniel Saint were baptized in the Curaray River at the site of their father’s murder. The baptism was performed by one of the men involved in the murder of their father.

From: Nida, Eugene A. (ed.) Book of a Thousand Tongues. 2nd ed., United Bible Societies, 1972, p. 69.