Bible (Translation) Programs: ParaTExt

Extracted text from Paratext_MdL.pdf:

1

ParaTExt: Closed Book and Open Screen?1 Marijke H. de Lang, mhdelang@solcon.nl

In April 2012, the Nida Institute of ABS organized a seminar in Misano, Italy, on intersemiotics. One of the topics we discussed was the computer program ParaTExt and the possible effects it has on the people who use it. Questions such as the following rose to the surface: how is the information in ParaTExt, such as the source texts and the dictionaries, presented? Does the way in which this information is presented have any implications for the user? Does ParaTExt give the translation teams more independency in terms of information access and ways of working? What kind of expectations in general do computers generate with their users? Do computers raise expectations of democracy and openness or maybe even open endedness, or do they confirm old patterns of power and authority? Do they change our perception of what text is? The questions I would like to raise in the present article are just a selection from these and in most cases the answers are not supplied. I merely attempt to flag some of the concerns that were raised at the intersemiotics seminar, some of which have also been discussed in earlier issues of The Bible Translator.2 This could give the impression that the bulk of this article is negative. But I would like to emphasize that the fact I am mostly asking questions does not imply automatically negative answers.3 The shifts in translation caused by the developments in technology are still under debate. In his book Translation and Globalization, Michael Cronin sees computer technology not as a break with the old tradition of translation, but as a further step in the development of objects people have used for writing and storing information (clay tablet, scroll, book, press, typewriter, etcetera). But he points to the fact that even though technology has become an integral part of the world of translation, translation studies has focused more on issues such as translation and text and translation and translator but much less on the implications of the interaction between translation and technology.4 In the case of Bible translation the program ParaTExt has become an indispensable tool of translation projects. There is generally a strong tendency to perceive technology as something neutral and objective. However, since technology is thought out by humans and used by 1

I would like to thank Robert Hodsgon, who came up with this idea for me to work out and helped to set out the first outlines, and Sarah Lind and Phil Towner for their comments and suggestions for improvement. 2 The April 2011 issue of The Bible Translator was devoted to the theme Computers as Translators: Translation or Treason? and discussed some of the implications of the use of computer programs such as ParaTExt and Adapt It. In a later issue (TBT 62-4 2011) Andy Warren-Rothlin added to the discussion with his article Source Texts, Source Languages, Technology, and Values. He emphasizes the fact that computer programs cannot subsitute working from the original languages and that translation from the source texts remains vital for Bible translation. 3 Lind 2. 4 Cronin 10. The main point of Cronins book is that in a globalized world with advanced technology, the task of translation has not been made redundant but has become more urgent as a safeguard for linguistic and cultural diversity.

2

humans, it is far from neutral and objective. In the case of ParaTExt our thinking about the program tends to revolve around its technical possibilities rather than its wider (ethical, sociological, philosophical) implications. In other words, as Cronin states, the tendency to view tools almost exclusively in the domain of technology leads to predominantly descriptive readings of their use (what they do) and a subsequent neglect of the wider implications of their presence in the world of translation (what they represent).5 Indeed, in ParaTExt training events the focus is on the technical side of the program. We learn about new features and what they can do, and through best-practice exercises we are shown how to use the program in the most efficient way. But perhaps more attention should be given to the relation between program, text, and user, and to some of the more invisible implications of the use of the program. ParaTExt was originally designed as a tool for storing translated texts and preparing them for publication, and not so much as a Bible study program. The emphasis was on post-translation processes. Over the years in response to feedback and suggestions from users in the field a variety of study applications were added and gradually the purpose of ParaTExt shifted from a pure Bible storage and production tool to a modest study tool. In its present shape users have access to the source texts in Hebrew, Aramaic and Greek and a large number of resource texts (Bible translations of all kinds) and to dictionaries and Translators Handbooks. ParaTExt allows for relatively advanced and detailed searches in the source languages in the feature called Source Language Tools. In the same window, users have the option to view glosses, lemmas, and a transliteration of the source texts. ParaTExts feature of the Interlinearizer allows for lemmatization and morphological analysis of the target languages and is helpful in producing back translations and interlinear versions of any target language and any intermediate language. The ParaTExt website describes the program and its related tools as a collection of software programs for Windows developed by the United Bible Societies which allow you to input, edit, check, and publish a translation of the Scriptures, based on the original texts (Greek, Hebrew), and modeled on versions in major languages.6 There are two things here that raise questions. The first is the implied suggestion that major languages automatically function as models, presumably for minority languages. I do not think that the developers of ParaTExt intended to suggest a one-way traffic from minor to major languages, but the introduction could enforce the idea that major languages such as English or Spanish have priority over minor ones. The other thing that attracts attention is the seemingly unproblematic notion of the original texts. There are of course theological reasons for not letting go of the notion of 5 6

Cronin 28. See the website: http://paratext.org/about/pt (last seen on 06.05.2013).

3

the original text, but we also should ask ourselves whether the concept is in reality as stable as theology would like to see it. We no longer have access to the original. And the sources we use are at best a hypothetical reconstruction or approximation of what we assume was the original. Textual and redaction criticism have shown that the biblical texts were used by the community as objects to be changed and modified as circumstances required. We now know that the books of Joshua and 1-2 Kings have been reworked and revised to make them accord with the book of Deuteronomy, that the composition of the book of Isaiah is the result of a long and complicated history during which new passages were added and older ones modified, and that 2 Corinthians is more likely to be a compilation of several (parts of) letters of Paul written at different moments in his complicated relationship with the Corinthian community than a composition aus einem Guss. It is hard to say what in these cases is the original text: is it the reconstructed first version, or the final version as we have it now? In this respect, as Pym indicates, the volatile character of pre-print texts such as Bible texts can be compared to that of post-print text.7 Within electronic technologies the notion text has become highly changeable. Texts have become far less stable and more interactive than they ever were in the period of print. Just so in the times of the Old and New Testament texts were not fixed but fluent notions, to be used as the context required. David Parker has argued this for the text of the Gospels.8 For translation these questions seem perhaps less urgent, since for our translation projects we use as our sources the printed editions which offer us a reconstructed original. But even if we take the source texts as we have them in our printed editions for granted, enough uncertainty remains. They are not simply equivalent with the original. The Old Testament editions can be based either on the Leningrad or on the Aleppo Codex and the critical edition of the New Testament is a reconstruction. And all editions reflect the vast number of alternative readings and interpretations. Similarly, the introduction to Source Language Tools simply states that it provides quick and easy access to the Hebrew, Aramaic, Greek and Syriac source language texts.9 ParaTExt 7 8

Pym 22. Parker 7: Almost all writers on theological, ethical and historical issues have taken it for granted that the New Testament evidence on which they are drawing for support is the original text of the witnesses. The importance of scrutinising this assumption cannot be stressed too highly. It is the principal theme of this book. Parker 204 argues that texts did not have an existence independent of scribal activity and their use in the churches so that the character of all manuscript copying meant that there was a continuing interplay between Scripture the text copied and the tradition the person engaged in the process of copying in and for the church. According to Parker the recovery of the original text is a task that remains beyond all of us (204). 9 The Syriac text is no longer part of SLT. The Greek of the Septuagint and the Syriac are in fact versions (early translations), but also function as source texts in some churches. For some individual books they are indeed the only available source texts, for instance the Greek of some of the deuterocanonical books and Syrian Baruch, which is part of the Syrian canon. Strictly speaking, Latin should have been mentioned as well: the Vulgate is considered source text in the Roman Catholic tradition, and for e.g. 4 Ezra, the Latin text is the only available text we have.

4

refers here to the texts they use from the printed editions. As far as the text of the Old Testament is concerned, some would say that the Masoretic text is the source text for Bible translation projects. But manuscripts show that the Masoretic principles are represented in many different ways and that the Masoretes made their own specific choices in readings, chapter and verse division, and other features.10 Moreover, the Masoretic text represents just one text type next to other text types and in many places either the Septuagint or the scrolls of Qumran reflect better and older readings. The Septuagint sometimes offers a considerably shorter text than MT, for instance in Jeremiah and Esther but who reflects the more original text? The Qumran scroll of the book of Isaiah shows many different and often better readings than the Masoretic text of Isaiah but which text should be used as the original? Even the printed editions of the Old Testament, the Biblia Hebraica (BHK), Biblia Hebraica Stuttgartensia (BHS) and Biblia Hebraica Quinta (BHQ), all based on the Codex Leningradensis, vary in e.g. textual annotations, the number of emendations suggested for difficult readings, in chapter and verse division, in prose and poetry layout, or in the order of the biblical books.11 ParaTExt has added the notes of the Report on the Hebrew Old Testament Text Project (HOTTP) but these only represent a selection from the countless textual difficulties in the Hebrew text. And without sufficient training, translators do not know how to handle the textual differences in any case and often enough isolated decisions are being made usually based on other arguments (e.g. tradition, convention, or on whether one or the other reading is easier to translate) than textual ones. The situation with the New Testament is not much better. The scholarly editions that are mostly used, Novum Testamentum Graece and The Greek New Testament, are not diplomatic editions based on one particular manuscript, but eclectic ones in which an original text has been reconstructed on the basis of previous editions and manuscript evidence.12 Printed

10

Tov 11 clearly states that the Masoretic text does not reflect the original text of the biblical books and that it is not a uniform textual unit, but is itself represented by many witnesses. The discovery of the Qumran scrolls has only underlined the fact that a search for an original text of the Hebre w Old Testament is a complicated and in many cases impossible one. Some scholars prefer to speak of a plural source texts. 11 R. Kittel and P. Kahle, Biblia Hebraica, Stuttgart 1929-1937 (BHK), of which the seventh edition (1951, with an apparatus of Qumran readings) was printed until the 1970s. The revised edition was BHS: K. Ellinger and W. Rudolph, Biblia Hebraica Stuttgartensia, Stuttgart 1967-1967. Of the new Biblia Hebraica Quinta (BHQ) so far only several fascicles have been printed., the first one in 2004. The Biblia Hebraica and Biblia Hebraica Stuttgartensia differ in many respects even though they are both based on Codex Leningradensis. In both these editions the critical apparatus does not offer the user much help in assessing the different readings. The new project of the Biblia Hebraica Quinta offers an extended and improved critical apparatus. It is still a hotly debated issue whether diplomatic editions of the Old Testament should be replaced by critical ones but because of the large degree of uncertainties so far diplomatic editions are still preferred. 12 In the 26th edition of Nestle-Aland the claim was made that the text was reconstructed on the basis of all the manuscript evidence available. However, in reality it was a revised version of NA25 which in turn relied heavily on the editions of Tischendorf, Westcott & Hort, and Bernhard Weiss. Only with the full edition of the Editio Critica Maior will the promise of a text that is reconstructed on the basis of manuscript evidence become a reality.

5

editions usually contain forewords in which the principles of the textual reconstruction are explained. In ParaTExt however such explanations are missing. Perhaps the designers of ParaTExt rely on this to be self-evident for the average user of the program, but one can ask whether it is correct to assume the awareness of a less stable source text. References to the original text and the source text can be very misleading in so far as they suggest that there is an objective verifiable source text ready for us to use and translate. Moreover, the information added to the source texts, such as the lemmas, glosses, and tags, are not objective computer-data, but reflect choices made by individuals and groups who contributed data to the program the program. On the website of ParaTExt Source Language Tools (SLT) is described as follows: The ParaTExt Source Language Text window provides quick and easy access to the Hebrew, Aramaic, Greek and Syriac source language texts. It allows users to view a source text at the current verse, along with an interlinear gloss and morphological analysis of that verse. Users can easily search the source texts, and see how a particular source word is used in different parts of the Bible. Words in the source text can also be looked up in a collection of dictionaries and lexicons.13 Indeed, the great advantage of SLT is that it allows the user to do relatively detailed searches in the source texts. For users who are not familiar with the source languages, SLTs option to show glosses, lemmas, and forms is helpful. However, danger lurks in the use of these tags as soon as the user perceives them as objective information and not for what they are, as information based on the choices of people behind the program. A Hebrew verb or noun may be tagged in one way by the programmers of ParaTExt, but may be defined in quite a different way by others in dictionaries or commentaries. The same applies to the glosses. In many cases the form or meaning of a word is disputed and the choice for one tag or gloss excludes other options. And since the computer and the technology supporting it are often disconnected from the people behind them, the choices can be seen as objective and neutral information rather than the result of human choices.14 The tags, glosses, or lemmas used in ParaTExt are indeed helpful for those who do not know the original languages well, but at the same time can become dangerous exactly for the same reason: when a user does not know the original languages well, the tags and glosses become the ultimate source of information and stop the translator from looking for answers elsewhere.15 So what do we need to be aware of when we work with teams? Apart from the fact that the information given in the program can be understood as objective and neutral, there are perhaps also some other issues that we need to take into account when we use ParaTExt. Of 13 14

See the website: http://paratext.org/about/pt (last seen on 06.05.2013). The same applies to the use of e.g. dictionaries. 15 Warren-Rothlin 241: In ParaTExt too, wrong use of resources such as source-text interlinearisation and Translators Handbooks may enslave a project to particular reading ...

6

course not all the things listed below will automatically occur in all projects, but they are at least potential dangers. 1) The designers of ParaTExt initially sought to provide a working environment that would help translators in their work on translation. But the website also claims that the program can help translation teams to produce quality translations from the point of view of both format and content. 16 Without doubt ParaTExt has become indispensable when it comes to typing and formatting the text of a translation, and for running all kinds of checks. However, one can be hesitant about its claim that it helps to produce quality of content. ParaTExt is a great tool for translators, but in itself does not provide the translators the encyplopaedia necessary to their work. This remains something for the translator to explore elsewhere and demands training and above all patience.17 2) ParaTExts focus on the source texts with the additional information in lemmas, tags, glosses, or model texts, creates the potential risk that translators will concentrate too much on the micro-level of the text, forgetting its overall composition.18 A problem that sometimes occurs is that translations pick one example from one model, another example from a second model while in fact the models represent different types of translation strategies. 3) The focus on details in the text, and a working method that is concentrated too much on the micro-level of the source-texts, creates the risk of restricting the task of the translator to the transfer of words or phrases from one language into another, or even of bringing translation back into the arms of the paradigm of natural equivalence.19 4) The way ParaTExt is set up has the potential danger of emphasizing exegesis as an exercise of explaining the Bible within the limits of the books of the canon, or perhaps more accurately, the canons that are available within the program. It is of course impossible to include the vast number of non-canonical texts (Jewish and Christian) in the program or add the endless non-biblical texts that shed light on the texts of the Old and New Testament. But the present set-up of ParaTExt may enforce the idea that the Bible is an isolated phenomenon and that, to use an old adage, scriptura scripturam interpretatur. Moreover, the lines between the occurrence of certain words in different books on the one hand, and the phenomenon of intertextuality on the other, become fuzzy. It is important to understand that the authors may not always have been influenced by books or texts per se, but much more by ideas current in the world they lived in. This world did not consist of texts only. If translation teams work within the restricted parameters of ParaTExt, they may run the risk of thinking in terms of text to text influence only. 16 17

See http://paratext.org/about/pt (last seen on 06.05.2013). Cronin 21. 18 Lind 13. 19 According to Pym 22, natural equivalence works with the idea of equal value, and presupposes that different languages do or can express values that can be compared in some itemized way and that they are somehow on the same level.

7

5. ParaTExt has made quite a number of modern translations available. When translation teams struggle with difficult passages they have the possibility to look for solutions in existing translations. No doubt this can be very helpful. But there is a danger that translators tend to restrict themselves to the options that are present in the available resources and then stay within parameters of traditional renderings. When a team comes up with a creative solution that is not found in the available resources of ParaTExt, they tend to switch back to an existing and therefore safer translation. Again, working solely within the parameters of a Bible environment as ParaTExt, sometimes has the odd effect that translations result in conformism and traditionalism rather than creativity and innovation. Moreover, the use minority languages make of these majority language translations may bring back the idea of the superiority of major languages over minor languages. Again, this does not automatically have to happen, but the potential danger is there. In a similar way translators can let themselves be restricted and confined by the versification mode of the program. The program offers the possibility to create a new project with an empty window without any verse numbers. For practical reasons most projects, however, prefer to use the window in which the numbers of chapters and verses are already inserted. But once teams start using this window, it often stops them from restructuring verses or passages, because they then feel obliged to follow the traditional verse numbering. In a sense, these pre-set data are often not used as tools but as parameters. What is designed to help the translator in his or her task then becomes a rule to be followed. In some cases it takes constant training to convince translators that these data are not there because they are rules but as helps and tools. 6) The development of ParaTExt software continues to progress rapidly. But the pace at which software (and hardware) develops does not always coincide with the environment people work in. Users are required to keep up with their internet connections and hardware. But the reality is that not everywhere can people keep up with the technology that is required to use the newest and latest version. Moreover, in many cases the average ParaTExt user, wherever they work in the world, needs time to understand even the most basic features of the program and needs practice to implement them systematically. Consolidating the use of the basics of the program is perhaps more needed than adding new options and possibilities to the program. The danger is that the much anticipated democracy that computers are supposed to bring is being frustrated because of technological limitations. A too rapid development of a computer tool may lead in the end to the establishment of a new elite: those who have the fastest internet connection, the fastest computer and the best training options can do more with the program than those who do not.

8

7. Many assume that improving technology leads to faster results. In the case of ParaTExt, people assume that translation work has become faster and better.20 ParaTExt does indeed speed up certain parts of the work of Bible translation. However, the program cannot do everything. In the end the quality of the output depends on the quality of the input and as with many computer programs the quality of the input is still a human factor. Translators themselves have not become more accurate, faster or qualitatively better by using a computer and translation remains a highly creative and complex process. But in a world in which time is money, the problem is as Cronin states: We are now in an age where we are spending more time than ever before pointing up the intrinsic complexity of translation but we have less time to make that complexity a textual reality.21 Information retrieval made possible by computers is indeed not the same as historical knowledge of languages and cultures. Even if we assume that information is gathered more efficiently than before, the process of moulding this information into knowledge is still a process that is completely dependent on human efforts. Often this is forgotten by stakeholders who are intrigued by the idea that computers can potentially speed up the tedious and painstaking work of Bible translation. How much better it would be to sell to the donors the idea that a translation project of a full Bible will take only 7 or 8 years instead of an average of 15 to 20 years. But again, the idea that a Bible translation project will proceed faster than before does not take into account that for a good understanding of the ancient texts careful historical and linguistic research is needed and that knowledge is not the same as information retrieval. It does not take into account the local conditions of electricity supply and internet access. It does not take into account working restrictions because of political conditions. We can share knowledge and we can exchange experiences to avoid repeating mistakes. But to assume that later generations in translation can do the work faster because they can learn from a previous translation generation and because of ever growing technology is tantamount to thinking that your children with their computer skills can do their schooling in half the time you did as a parent. Without doubt the technology of ParaTExt has made it much easier for translators and translation projects to input, review, store and print their work. In fact, the tool has become indispensable in doing otherwise time consuming and tedious jobs such as spell checks and consistency checks. But it remains a tool, which has to be used with care and only for the purposes it has been designed for. A hammer is not designed to drive in a screw. ParaTExt is to be used in a similar way: with an understanding of what it can, but also of what it can not do.

20

Cronin 109: The difficulty is the assumption that translation is not a qualitatively distinct operation from the mechanism of delivery, a difficulty compounded by the reality of MT packages or services on the Web. 21 Cronin 121.

9

References Cronin, Michael. 2003. Translation and Globalization. Cambridge: Routledge. Hodgson, Robert. 1999. Semiotics, Fidelity, and New Media Translation. Fidelity and Translation. Communicating the Bible in New Media. Edited by Paul A. Soukup and Robert Hodgson. New York: American Bible Society: 233-248. Lind, Sarah. 2005. Ethics and Computing Technology. Paper presented at the UBS TTW, Mombasa. Parker, David. 1997. The Living Text of the Gospels. Cambridge: Cambridge University Press. Pym, Anthony. Exploring Translation Theories. London and New York: Routledge. Tov, Emanuel. 20012. Textual Criticism of the Hebrew Bible. Minneapolis: Fortress Press, Assen: Royal Van Gorcum. Warren-Rothlin, Andy. 2011. Source Texts, Source Languages, Technology, and Values, The Bible Translator 62.4: 240-246.