MAP — Modular Aggregation of Principles for Bible Translation

1999The notion of genre and the nature of Bible translations. Notes on translation. Vol.13, number 2,26-42.

THE NOTION OF GENRE AND THE NATURE OF BIBLE TRANSLATIONS

Lourens de Vries

Lourens de Vries worked as a missionary linguist and as a Bible translator with the Mission of the Reformed Churches in Irian Jaya, Indonesia, from 1982 to 1991. Presently he is a professor of Bible translation at the Free University, Amsterdam and a translation advisor with the United Bible Societies for Irian Jaya projects. He received a PhD degree in linguistics from the University of Amsterdam in 1989.

Introduction

The purpose of this article is to discuss the nature of Bible translations from the point of view of an anthropological and pragmatic notion of genre: what kind of texts are Bible translations and what is their relationship to other genres distinguished in a speech community? First, I introduce the genre notion of Bakhtin (1981, 1986) incorporated by Foley (1997) in a framework of anthropological linguistics. Then I apply that notion of genre to Bible translations in Irian Jaya, Indonesia. My contribution should be seen as tentative and is meant to provoke discussion of the very complex issue of the genre nature of Bible translations.[1]

Genre

Building on the work of Bakhtin (1981, 1986), Foley (1997) stresses that genre is not an abstract, closed linguistic object like the grammatical notion of sentence but rather a historically given way to produce and interpret texts in culture specific settings and institutional frameworks. When new institutions develop, new genres are born and genres change when society and its institutions and social roles change. People may use genre patterns creatively for their own purposes. The ability to produce (oral or written) texts according to a specific genre pattern is in many societies an important element in defining social roles, for example the pastor and the sermon or the scholar and the scientific article.

This notion of genre moves away from an a-historical, autonomistic and structuralistic notion of genre and comes close, in several regards, to the genre notion of biblical scholars like Gunkel, Bultman and Dibelius, for example in the link between genre and social and institutional frameworks in which texttypes function, the so called Sitz im Leben (Gunkel 1906). The Sitz im Leben, the place of a texttype in the life of a society, for example a wedding ceremony or a harvest feast, was very much a diachronic notion for scholars like Gunkel. Embedded in 19th century thinking, the Sitz im Leben was the place where oral phases or Vorlagen of the biblical texts functioned.

The notion of genre of Foley (1997) is synchronic, focussing on the way a texttype reflects the social and institutional setting in which it functions. The diachronic perspective enters with the notion of recontextualization of genres. Foley (1997: 376) gives examples of modern recontextualization of traditional genres like the didong genre of the Gayo of Sumatra. The didong was a poetic duel between individuals used in the context of for example marriage negotiations. After 1940 the didong genre developed into duels between teams of around 12 persons in the context of entertaining the public but at the same time focussing on modern issues of a sociopolical and religious nature.

The historically and culturally determined place of a texttype can be illustrated by the genre of lamolaup 'history of the world' of the Korowai of Irian Jaya, Indonesia (Van Enk and De Vries 1997). Only adult Korowai males may listen to and transmit these sacred genesis stories. They are taboo for women, children and outsiders. They are only told in times of crisis like famines or solar eclipses. The lamolaup stories relate how the cosmic order of things came into existence and when the Korowai feel that their world is falling apart, the narration of these sacred stories is thought to restore the cosmic order of things.

Whenever humans produce language, their utterances are recognized by other humans of the same speech community as belonging to a certain genre, whether the utterance is part of an informal conversation between neighbours or of a religious creed. Genre analysis tries to determine the factors that are constitutive for a specific genre of texts in a certain place and a certain time and it also tries to link the way language is used in texts to those factors. Thus the linguistic study of pragmatic patterns in texts is linked to the cultural and institutional place of texts.

The genre of Bible translations: the case of Irian Jaya

I will now apply the notion of genre introduced above to Bible translations of Irian Jaya, Indonesia. In predominantly oral societies of the Irian Jaya interior, Bible translations are a a relatively recent genre, or rather a recent macrogenre containing subgenres like psalms and apostolic letters. This new genre emerged with the birth of a new institution in such communities, the christian church. Missionaries bringing the Gospel planted churches, opened air strips, founded schools and clinics, a whole range of new institutions, each triggering new texttypes such as 'the lesson', 'the sermon', 'the two-way radio talk' and 'the Bible translation'.

Among the many factors that constitute the macrogenre of Bible translations in Irian Jaya, three factors come to the fore: the (still relatively) new institutional and cultural setting in which these texts function, the christian community and its activities, the secondary nature of translated texts and the new medium, writing. Each of these three factors contributes to the pragmatic patterns[2] found in Bible translations of Irian Jaya as I will now proceed to show.

Institutional and cultural function

Let us first discuss the pragmatic consequences of the Sitz im Leben, the cultural function in institutional frameworks. The translated text functions as the religious base text in a Irian Jaya community of Papuan christians, in the liturgy but also during occasions like christian marriages and burials in a way entirely different from the way traditional sacred genres function(ed).

Consider the difference between reading parts of the book of Genesis about the first couple, Adam and Eve, during a christian marriage ceremony and the narration of the Korowai lamolaup story about the original couple, Older Brother and Younger Brother, to the adult males of a clan during a crisis situation, to overcome that situation (see Van Enk and De Vries 1997: 163-173). These differences in cultural function and institutional setting are reflected by the way language is used. An example would be the performance conditions of Korowai lamolaup texts on the one hand and translated Bible stories of origin on the other hand: whereas lamolaup texts are told in their entirety in long sessions of many hours, the Biblical stories of the translations, like the Genesis narrative, are divided into chapter and pericope units which are publicly read to groups of christians (and people interested in becoming a christian) in readings of no longer than ten to twenty minutes. The pericopes have headings guiding the listeners and readers thematically and identifying the short Bible fragments that are read aloud in church or elsewhere. In the 'performance tradition' of most churches and christian communities that I visited in Irian Jaya, there was no place for reading very substantial parts of Bible books like Genesis during a church service or a wedding or when visiting a fellow christian in his home.

The fragmented reading of the Bible books as opposed to unbroken, complete narration of oral sacred texts leads to significant differences in the pragmatics of both types of texts, for example in the patterns of participant handling. The pericope divisions in translations tend to lead to repeated, relatively strong identifications (full nominal phrases, pronouns) of given major participants in the first paragraph of each new pericope unit. Chapter 4 of the Angguruk Yali translation of Luke has 6 pericopes with pericope headings; each heading and each initial paragraph re-establishes Jesus as the major participant (by explicit identification as Yesus). Since the pericopes often function as units to be read in church and elsewhere, as a kind of mini-stories, there is a tendency to establish the time, place and participant framework anew at the start of each pericope, irrespective of the preceding context. The Yali Angguruk translation follows a pattern of participant handling common in Irian Jaya translations. Below it will be seen that the secondary nature of translations also pushes participant handling away from traditional oral genre patterns.

Participant tracking in traditional oral genres like lamolaup texts is done very differently: weak identification of activated major participants is the norm, by verbal (as opposed to nominal) means (agreements and switch reference, see Van Enk and De Vries 1997: 106-138). The repeated explicit re-introduction or re-establishing of the place, time and participant framework is absent in most oral narratives of Irian Jaya. The tendency is to devote the opening paragraph to setting up the framework for the story and from then on keep it as implicit as possible. Therefore opening paragraphs tend to contain relatively much nominal and adverbial phrases, marked by topic or frame markers but the rest of the text tends to be verby: long chains of simply structured verbal clauses, with detailed expression of the events and sub-events and their temporal relationships (sequence, simultaneity). Of course, these are broad generalizations but, whatever the variations in this pattern, the way participant and place/time frames are handled in Irian Jaya Bible translations falls outside these basic oral narrative patterns.

Although both traditional origin stories like the lamolaup and the translated Bible have a religious function, the lamolaup is sacred and secret whereas the Bible functions as a sacred text (in the sense that it is believed to be the Word of God) and as a public proclamation at the same time. Therefore, Bible translations in Irian Jaya avoid specialized vocabulary and aim at clarity, trying to use a language common to all; sacred traditional texts of Irian Jaya aim at restraining access to the secret, magic knowlegde hidden in the texts. Indirection, avoidance of names, wordtaboos and specialized vocabulary are the pragmatic reflex of that function. Thus the different cultural place of Bible translations and sacred oral genres is reflected by and expressed in style and discourse patterns, from registers and levels of language to participant handling.[3]

Translations as secondary texts

The second key factor determining the genre nature of Bible translations in the Irian Jaya context, is the status of the translation as a secondary text, secondary in the sense that the text represents another text, the source text which presupposes a different world, the world in which it emerged. Translated texts are, in the words of Hatim and Mason (1996: 1) 'an act of communication which attempts to relay, across cultural and linguistic boundaries, another act of communication'. Although the language of the secondary text is a familiar, Papuan language, the world behind that text is not the familiar Irian Jaya world of the secondary audience but the strange world of the ancient near east. The participants in the translation (Paul, Moses and so on) speak a Papuan language but they move in that strange world and they behave according the ways of that strange world. No Bible translation, however transculturating, can cut the ties with the primary biblical worlds presupposed by the source texts: there will always be elements in the translation pointing to the primary worlds of the Bible.

It has often been remarked that any text, whether a translation or a primary text, is always and necessarily understood in terms of the receptor culture (see for example Wendland 1996: 130). This does not, however, alter the essential distinction between primary and secondary (or translated) texts, both in production and perception. Even in remote areas of Irian Jaya people realize that the stories of the Bible originate in another world. The Kombai of southern Irian Jaya for example did not expect the people in the Bible to behave like Kombai, just as they did not expect me to behave as a Kombai man. Secondary audiences process the information of secondary texts in terms of how they conceive of the world of the source text: they try to create a secondary world, different from their own primary world, on the basis of what they have been told or instructed on the world of the source text, on the basis of the translated text itself and on the basis of projections of their own primary world into this secondary world. The creation of such a secondary world is needed to make sense of the (secondary) text that gives too many signals that it did not originate in the primary world of the readers of the translation. The great tension between the familiar language of Bible translations and the unfamiliar world behind the language of the text is a genre characteristic of Bible translations in general but a characteristic with a special significance in the Irian Jaya context where the discontinuity between primary and secondary world is so great.

The institution of (payed, professional) translation in itself is new as are the secondary texttypes that are the output of that institution. Of course, multilingual persons in Irian Jaya may act as (oral) interpreters representing in language A what someone else said in a neighbouring language B but then there is not this wide gap between primary and secondary world and this casual interpreting activity is not an institutionalized and payed professional activity. Oral religious texts (origin stories, totem texts), usually considered to be the sacred and secret possessions of clans, are never transmitted by (oral) interpreters to outsiders.

The fact that the ideas which secondary audiences develop on the world(s) of the Bible are all too often not correct does not alter the crucial fact that they process a primary text essentially different from a secondary text and translations anticipate the problematic aspects of secondary text understanding, both in the translated text and in metatext (footnotes, introductions, glossaries). For example, since the participants move in a strange world and do strange things, it is much harder for the audience to keep track of who is doing what to whom and with whom. Anticipating this, translations tend to use explicit nominal phrases and pronouns to refer to participants in ways totally absent in traditional oral genres of Irian Jaya. Culturally unknown or unfamiliar biblical referents like the Sanhedrin or phylacteries are often described in complex nominal phrases. Thus for both participant handling and referent identification nominal means are applied by translators. Above it was remarked that the pericope divisions and fragmented liturgical reading led to more strong identifications of participants. Now we see that not only the cultural place (for example the liturgical function) but also the secondary nature leads to a shift into nominal direction. The tendency to use more nominal phrases and pronouns is in some cases reinforced by the influence of Indonesian translations that function as source texts for translators without knowlegde of biblical languages.

The shift into nominal direction in its turn leads to heavier, more elaborate clause structures. These heavier clauses, the building blocks in the chaining languages of New Guinea, lead to shorter clause chains.[4] Thus the first two factors constitutive of Bible translations as a macrogenre, the cultural function and the secondary nature, have a tremendous effect on the way language is used, setting Bible translations pragmatically apart from any traditional genre.

Missionaries, schoolteachers, preachers in church try to instruct the communities in Irian Jaya on the world of the Bible in the context of explaining the translated Bible texts: the very fact that translations are texts that cannot be understood without some degree of instruction on the world presupposed by the text sets them apart from primary texts that presuppose a familiar world shared by the readers or listeners. Of course, it is possible to a certain extent to work the biblical worlds into the translated texts themselves in the framework of communicative translations by explicating presuppositions and implicatures associated with the source text but even for the most extreme type of transculturating communicative translations it is impossible to build the whole primary world of the Bible source texts into the translation itself and for this reason translations of the Bible can never function as primary texts. Translations will always give out mixed signals, with some elements in the translation suggesting that the translated text originated in the world of the readers of the translation and other elements pointing to the world of the source texts. How readers react to such mixed signals depends on culture specific norms for faithfulness. For example, the Chinese translators of Hofstadter's 'Göder, Escher, Bach' (1979) acknowledged that the Chinese idiom 'Mention Cao Cao and Cao Cao then shows up' meant exactly the same as the English idiom 'Speak of the devil' but had strong objections to using the Chinese idiom in the translation, even against the strong preference of the author Hofstadter for using the Chinese idiom in the translation. The Chinese translators antipicated a negative reaction to such mixed signals: 'How would an American know about Cao Cao?' (Cao Cao is an important character in a famous Chinese historical novel). For the Chinese audience it is hard to accept a translation as faithful when a translated American text 'falsely' suggests that it originated in China.

Isolated communities in Irian Jaya, where the Bible is the first or one of the first translated texts, tend to react more positively to mixed signals in secondary texts. However, when in the course of time national language translations of the Bible become well-known with significantly less mixed signals, the community may change its attitude towards elements in translations that 'falsely' suggest that the text is part of their primary world, even when the element objected to occurs in a parable or in another 'non-historical' biblical genre.

Notice that my emphasis on the secondary nature of translations should not be taken to mean that translated texts (should) function as dependent texts, as a kind of second class texts. Translations are texts in their own right that can and should stand on their own feet. Translations of the Bible that gain acceptance in communities of believers represent Holy Scripture to these communities: they (should) function as the Word of God. The secondary nature (the Bible translation as a translation) and the religious function in communities (the Bible translation as the Bible) sometimes

have an uneasy relationship. There is an inherent tension between these two aspects and history has shown that time and again religious communities have 'solved' the tension by suppressing or denying the secondary nature of translations through total identification of certain translations with the Word of God.

The written medium

The third key factor behind the new pragmatic patterns found in Bible translations is the new medium, writing. In discussions of Bible translations in oral settings the pragmatic innovations found in translated texts but not found in the traditional oral genres are often ascribed to this third factor, the written medium. However, as we saw above, other factors like the secondary nature and the cultural function of translated texts, also have a dramatic impact on the way language is used in translations. Therefore, when we observe pragmatic innovations in Bible translations (like shorter sentences, less topic markers, more and more complex nominal phrases, new patterns of participant handling and so on) in newly literate settings, these changes should not automatically be viewed as part of a developing 'written style'.

Of course, some changes can be linked to the new medium. The processing and production constraints in the oral and written medium differ very much. Readers can go back to the passage just read and read it again if need be. The repair strategies available to speakers are very different from those of writers who can rewrite their sentences at will.

The two most striking effects of the new written medium in Irian Jaya translations (and other written genres) are the sharp decrease and redistribution of topic markers and of tail-head linkage linkages of sentences or clause chains. Both topic markers and tail-head recapitulative linkage have their primary functionality in enabling speakers and listeners to produce and process information in the oral medium (Van Enk and De Vries 1997; De Vries 1993b, 1995).

Tail-head linkage is the term used in the study of Papuan languages for recapitulative linkages of clause chains in which the final clause of the previous chain is partially or wholly repeated as the first clause of the next chain. Consider the following examples from Kombai of Irian Jaya (De Vries 1993a: 67-68):

(1)

U-ra fali mimaye-fo fali-ra

hit.SS-and carry.SS go.out.SS-SEQ carry.SS-and

mimaye-fo dodof-o da rofadef-o

go.out.SS-SEQ wood-CONN surface put.1SG.NF-CONN

leina ge gubino u-ruma-de

after next neck hit-cut-1SG.NF

'I hit it and carried it out and after I had put it on a piece of wood I cut its neck.'

(2)

U-rumadef-o leina fo-ra fali-ra

hit-cut.1SG.NF-CONN after take.SS-and carry.SS-and

maru-me-fo-ra me khalu

go.up-come.SS-SEQ-and come.SS go.down.SS

reme da me rofadefe.

concrete surface come.SS put.1SG.NF

`Having cut it, I took it upwards and then put it down on the concrete floor.'

The clause chains (1) and (2) are linked by tail-head linkage recapitulation. The recapitulated initial clause is pronounced slowly and with a rising intonation. The chain-medial clauses tend to consist of verbs marked for agreement and switch reference (same subject (SS) or different subject (DS) in next clause) with as little explicit nominal phrases as possible. The addressees are expected to identify participants by combining verbal morphology (agreements, switch reference) with context, situation and general cultural information. Since the chains easily contain up to 20 verbs and have a high rate of information, the highly redundant tail-head linkages form conventional resting points for listeners to process the information just given in the previous chain and for speakers to produce their next chain. Kombai type 'verby' chaining patterns with tail-head linkages are found in hundreds of Papuan languages, especially in narratives.

Consider now a narrative section of the Kombai translation of the Gospel of Mark (1.30-31) where chains tend to be linked with other means than tail-head linkage like the conjunction mana 'and (next)':

30. Simon khe khuni yambomara

Simon his mother.in.law ill.with.fever.and

rumofora khunelei Yesus

lay.down.and she.slept Jesus

khu kho khende khu bürü mofena

and people other and place that

khurungenona ya luwano, Simon

they.entered they they.said Simon

khe khuni yamboma neneno.

his mother.in.law ill.with.fever they.spoke

'The mother-in-law of Simon lay down with fever and the people that had entered that place with Jesus, said (to him), Simon's mother-in-law is ill with fever.'

31. Mana Yesus kho lara Simon khe khuni

And Jesus go and Simon his mother.in.law

i lefara ondumana rangane maruna

hand take.and lifted.up stand.up she.raised

yamboma ro mofena emukhe,

ill.fever thing that it.vanished

dunoro andüwara ya rowalefa.

food cook.and them she.put.down

'And Jesus approached the mother-in-law of Simon, took her hand and lifted her up and she stood up and the fever was gone and she served them food.'

The change in linkage patterns from tail-head linkage to other patterns (conjunctions, generic verb linkage, anaphoric linkage) observable in the Kombai translation can be observed in varying degrees and forms in Bible translations all over Irian Jaya.

Not only the decrease in the use of tail-head linkage but also the decrease of topic markers is the result of the new medium. Bromley (1981) notices a tendency towards a new constraint on the use of topic markers in Dani written genres: one topic marker per sentence. Dani is a Papuan language of the Balim valley of Irian Jaya, in the mountains north-west of the Korowai area. In oral genres, the Dani clitic -he (-te after vowels) occurs most frequently in sentence-initial position, at points for possible pause; especially when speakers hesitatingly build up their sentence, the clitic is frequent. In the beginning of discourses, it is not uncommon to find multiple occurrences of -he, also within single-verb clauses. But -he may occur also sentence-medially and sentence-finally with 'afterthought-topics'. If it does occur sentence-medially or finally, it always marks thematic material.

Consider the following Dani example (Bromley 1981: 328):

(3)

(a) Lakakusik-he,

we.went.up.long.ago.prior-TOP

(b) ikke-pa lakaku lakaku lakaku plateau-on we.went.long.ago

lakaku lakaku lakaku-te

we.went.long.ago-TOP

Huehenema(place) we.slept.long.ago.prior-TOP

(d) moso sit appik-en-he

rain drizzle much-source-TOP

(e) we sit appik-en-he

idle drizzle much-source-TOP

(f) we welakaku welakaku welakaku-te

idle we.were.long.ago (repet.) we.were.long.ago-TOP

(g) likke pvlem-at kisukusik-he

midday middle-PREDICATOR we.entered.long.ago-TOP

(h) Jvkheakeima noko-usa'-mekke-te

at.Jukeagima sleep.ought.to-ADVERSATIVE-TOP

(i) it, mel lak,

they,dummy and.associates

(j) Waesom mel

Waesom too

(k) Wenakasvky mel,

Wenakasugi too

(l) it pykkyt-at lakoukwha.

they directly-PREDICATOR they.went.long.ago

'After we climbed up, we kept going on and on along the plateau, and after we slept at Huehenema, because of the heavy drizzle, we just waited idly, waited and waited; we had gone in at midday and should have slept at Jukeagima, but they, what's-their-names, Waesom and Wenagasugi, they went straight on.'

In his comment on this example, Bromley (1981: 329) notes that some of the units in (3), segments (b) and (f), have finite verbs not marked for dependence on a following verb and that 'grammatically, the units marked with -he/-te are disparate, but the occurrence of this clitic has the effect of levelling them as units of information or text.' The following example illustrates the multiple occurrence of -he in a simple sentence with a single verb:

(4) Nen hat-he hamelaik-he

then you(sg)-TOP your.bailer.shell-TOP

helan

take.it.yourself

'You take your bailer shell yourself!'

It seems to me that the topic clitic -he has both a speaker-oriented and an addressee-oriented function. The speaker-oriented function is the processual function as a pause/hesitation marker. The addressee-oriented function is the thematic function, to signal to the addressee the units of information which form a framework for what follows.

On the basis of this analysis, one would indeed predict a decrease of the use of the topic marker in written texts, where the processual pause/hesitation function of the marker is not relevant. It is interesting that Dani written texts show a sharp decrease in frequency of the marker -he and especially the multiplied occurrences of the marker prompt negative reactions from Dani readers, the tendency being not 'to eliminate all instances of -he but rather to limit them frequently to one per sentence, where the clitic occurs attached to an element that may be considered as topical for the sentence' (Bromley 1981: 329).

The change in the use of topic markers is not just a quantitative change. I have observed Papuan Bible translators editing their translations and in the process they did not only edit out topic markers in many places but also inserted them in places where they had not been in the earlier drafts. Whereas the processual function (pause marker) becomes irrelevant in the written medium, the thematic function of topic markers remains but is adapted to the pragmatic constraints following from the factors that determine the genre nature of Bible translations: the written medium, the secondary nature and the social and cultural function. Since Bible translations in Irian Jaya often function as texts to be listened to, it is important that translators, when deciding where to use topic markers, realize that the translation is a special type of written text: written to be listened to.

Little is known about the effects of topic markers and tail-head linkage on the ease with which a translation can be read (aloud). Renck (1990: 87) who compares Bible translations in two dialects of Yagaria of Papua New Guinea, one translation with many topic markers and one hardly using them, remarks that the translation with many topic markers is hard to read.

Bible translations and primary genres in oral societies

Questions regarding the relationship between Bible translations (as a secondary texttype) and primary genres are crucial for the practice of Bible translation in predominantly oral societies. Can traditional oral genres be recontextualized to form discourse models for translated Bible texts? Is the relationship between the way language is used in oral genres and in translations just a matter of the development of a written style? If translations in these contexts function primarily as texts that people listen to (with a minority of the addressees functionally literate), should translators just insert a few pragmatic patterns from oral genres (topic markers, tail-head linkage) to facilitate easy listening?

From the point of view of genre analysis proposed here, there are at least two dangers. The first is to approach the relationship between translation and oral genres in terms of one factor, the oral/written medium. Primary texts and translations are then placed on one continuum of texts, a continuum with typically oral style and typically written style as extreme poles. By doing so other crucial factors like the secondary nature and the cultural function of translations are obscured. Stylistic changes observed in translations that reflect those other factors are wrongly attributed to the oral/written factor. Furthermore, such an approach creates a false sense of continuity between primary oral genres and the genre of Bible translations, as if their only difference would be the medium. The norm for naturalness would then be the style of primary genres except for some adjustments in the direction of a 'written style'.

The second danger is an universalistic approach to genre, the idea that there are universal categories like 'origin narrative', 'love poem', 'proverbial wisdom'. Origin narratives like the Korowai lamolaup and a Papuan translation of the Hebrew Genesis are then seen as instantiations of an universal genre: creation narrative. Such an approach again obscures crucial factors like the secondary nature of translations and the unique historical role of genres within social institutions. Above I pointed out the pragmatic and stylistic discontinuity between Irian Jaya origin texts and translations of the Genesis story; such discontinuities reflect the discontinuity between primary and secondary texts and between the Sitz im Leben, the cultural function, of Irian Jaya origin texts and the translated Genesis story.

The genres that develop in translations of the Bible, like apostolic letters or psalms, are subgenres within the macro genre 'Bible translation'. The genre nature of Bible translations forms the textual and pragmatic framework for those subgenres. For example, since Bible translations are a secondary texttype, the subgenres found in translations are also secondary genres. Secondary genres mediate between primary genres of the source cultures and the primary genres of the target cultures but secondary genres never co-incide with those primary genres. For example, a translation of the Song of Songs in a Papuan language will always have certain genre continuities with Israelitic love songs and with love songs of Irian Jaya but at the same time the translated text will be essentially different from both. This is because elements from primary (source or target culture) genres cannot be mechanically copied into the translation but rather they are transformed by translators in the process of creating new secondary genres. Such elements from source and target culture genres are transformed and integrated in a textual and pragmatic framework dominated by the factors which set Bible translations apart as a genre: their institutional role in christian communities, their secondary nature and their medium, writing.

The realization that Bible translations have their own distinctive genre nature of mediating, secondary texts, written documents that function as the religious base text for communities of believers, may help to prevent translators to set the wrong goals in terms of genre. For example to aim at translations that are styled as primary texts of the target culture or translations that are styled as the primary texts of the Biblical source.

There are no general rules for the way in which secondary genres mediate between primary genres of source and target cultures. Important factors are the function which a translation has in speech- and church communities, the cultural place of receptor genres and the degree of overlap between source and target culture genres. Such partial overlap may result from overlaps in institutional frameworks. If both source and target cultures have kings and courts, there may be overlaps in the pragmatic patterns that developed in those institutions ('court language'). Such genre continuities support the mediating role of secondary genres.

An example of genre continuity with special relevance in oral contexts is the fact that several genres in the source texts of the Bible are written genres but written to be listened to. If Achtemeier (1990) is right, many source texts in the Bible were written to be read aloud to audiences and accordingly contain 'acoustic clues', audible signals structuring the information. Berger (1987: 212) remarks that the apostolic letters correspond in a significant sense to formal, public letters from hellenistic officials to the ekklesiai of the cities under their authority. The ekklesia

in the case of the apostolic letters is the christian community of a certain place and the authority is apostolic and prophetic. Public reading of such letters to listening christian communities during meetings would fit that Sitz im Leben of apostolic letters. Repetitions such as the three times paredoken autous ho theos 'God gave them up..' in Romans 1 verses 24, 26 and 28 would be examples of acoustic echo's in the sense of Achtemeier (1990). In most predominantly oral societies of Irian Jaya, where most people remain functionally illiterate after literacy programs have been carried out because writing and reading remain largely irrelevant for their daily lives, there is a small group of literati which reads written genres such as government announcements, sometimes newspapers and Bible translations to collective audiences, from the whole village to their families and relations in the houses. Thus there is overlap in the way certain written genres function in source and the target cultures of Irian Jaya.

Conclusion

The Bible translations of Irian Jaya that I am familiar with have a specific stylistic profile that sets them apart from other genres in the target cultures. Some of the aspects of that profile have been discussed in this article, for example relatively few topic markers and tail-head linkages, markedly strong identifications of given participants, shorter clause chains. This pragmatic profile is recognizable all over Irian Jaya. I have argued that this profile is the expression of the factors that distinguish Bible translations as a genre: their liturgical and ecclesiastical function in christian communities, their nature as secondary texts and the written medium.

If Bible translations have their own distinctive genre nature, if they are texts of a specific texttype, it would follow that translations that violate the pragmatic constraints inherent to that genre, would fail in certain respects. An example would be an 'oralizing' translation that copies participant tracking strategies of primary oral genres. This copying would result in leaving major given participants implicit most of the time, identified just by verb agreements and switch reference. Of course, that is fine in a primary text where participants move around in a familiar world but in a secondary text readers or listeners would be at loss to identify who is doing what.

One of the ways a translation is pragmatically adjusted to the constraints of its genre as a written, secondary, religious text is by carrying out tests of audience understanding and acceptance and by revising the translation to accomodate the results of the test; such procedures would mitigate the effects of translational strategies that are 'wrong' from the point of view of the genre analysis proposed here: translational strategies that force the translated texts into the models of primary genres, whether primary genres of the biblical sources or of the target cultures.

Many important factors, like translational strategies, the role of national language style models, the roles of expatriate translators/consultants and of mother tongue translators, were left out in this discussion. Further research is needed to arrive at a more complete picture of the pragmatics of Bible translation in Irian Jaya. The three fundamental factors formulated here (cultural function, secondary nature, written medium) interact with those other factors that were left out of consideration.

Whereas in the English language community and in other major language communities, the genre 'Bible translation' has dozens of instantiations, with influential older translations, with different Bible translations having different functions in a wide variety of churches and groups of believers, the picture in Irian Jaya tends to be much more simple: small language communities with usually just one (or more rarely two) Bible translation. Rather than having to meet audience expectations for the genre 'Bible translation', the pioneer translation, once well established and accepted, creates expectations for future translations. Only when national language versions become widely known, awareness grows of alternative translational strategies and models.

References

Achtemeier, P. J. (1990). 'Omne verbum sonat: The New Testament and the oral environment of late Western Antiquity'. Journal of Biblical Literature 109/1, 3-27.

Bakhtin, M. (1981). The Dialogic Imagination: Four Essays. Austin: University of Texas Press.

Bakhtin, M. (1986). Speech Genres and other Late Essays. Austin: University of Texas Press.

Berger, Kl. (1987). Het Nieuwe Testament als literatuur. In: A.S. van der Woude (ed), Bijbels Handboek, deel III, 194- 221. Kampen: Kok.

Bromley, Myron H. (1981). A Grammar of Lower Grand Valley Dani. Pacific Linguistics C-63. Canberra: Australian National University.

Enk, Gerrit J. van & Lourens de Vries (1997). The Korowai of Irian Jaya. Their language in its cultural context. Oxford/New York: Oxford University Press.

Foley, William A. (1997). Anthropological Linguistics. An Introduction. Oxford: Blackwell.

Gunkel, H. (1906) Die israelische Literatur. In: E. Schmidt (et al.), Die orientalischen Literaturen, 52-102. Berlijn: Teubner.

Hatim, Basil & Ian Mason (1996). The Translator as Communicator. London: Routledge.

Hofstadter, Douglas R. 1979. Göder, Escher, Bach: an eternal braid. New York: Basic Books.

Longacre, Robert E. 1972. Hierarchy and universality of discourse constituents in New Guinea languages.

Washington: Georgetown University Press

Renck, Günther. (1990) Contextualization of christianity and christianization of language: a case study from the highlands of Papua-New Guinea. Erlangen: Verlag der Ev.Luth. Mission.

Vries, L. de (1993a). Forms and Functions in Kombai, an Awyu language of Irian Jaya. Pacific Linguistics Series B, no. 108. Canberra: Australian National University Press.

Vries, L. de (1993b). 'Clause combining in oral Trans-New Guinea languages'. Text 13:4.481-502.

Vries, L. de (1995). 'Demonstratives, referent identification and topicality in Wambon and some other Papuan languages'. Journal of Pragmatics 24, 513-533.

Wendland, Ernst R. (1996). ' On the relevance of "Relevance Theory" for Bible translation'. Technical Papers for The Bible Translator 47.1:126-137.

Bible translations cited:

Allah Wene. Shorter Old Testament and New Testament in Today's Yali Angguruk Version. Jakarta: The Indonesian Bible Society (1998).

Markus khe Lu Yaferambo. The Gospel according to Mark in the Kombai language. Wanggemalo: Mission of the Reformed Churches (1991).

Abbreviations

CONN connective

DS different subject(agent) in next clause

NF non-future tense

SEQ sequence

SG singular

SS same subject (agent) in next clause

1999 The notion of genre and the nature of Bible translations. Notes on translation. Vol.13, number 2,26-42.

0 comments