"The syntax of discourse structure" (John Roberts)

Some languages typically have the verb before the object in a clause (eg. 'I kicked the ball', Subject-Verb-Object), others typically have the object before the verb (eg. 'I the ball kicked', Subject-Object-Verb). 

In this article (attached below), John Roberts argues that this pattern at the clause level correlates with patterns at a higher level, concerning relationships between propositions.  For example, whereas an SVO language typically uses the order RESULT-reason (e.g. 'they were amazed, because they heard them speaking in their own languages') an SOV language typically uses the order reason-RESULT (eg. 'they heard them speaking in their own languages, so, they were amazed').  

This has huge implications when translating the Bible from Greek (or Hebrew) into languages that are verb-final (i.e. SOV languages), as sentences that stretch across several verses may need to be reversed to achieve a natural translation. 

This is especially important for MEANS-PURPOSE relationships. In SVO languages, the MEANS typically comes before the purpose: 'they prayed (MEANS) so that they might receive the Holy Spirit (PURPOSE)'. In an SOV language, it is likely to be more natural to say 'they wanted them to receive the Holy Spirit (PURPOSE), and so they prayed (MEANS)'. This is particularly important because often if the original order is maintained, the translation ends up either being unnatural or may read like MEANS-RESULT rather than MEANS-PURPOSE: 'they prayed and as a result they received the Holy Spirit'. (Further explanation of this can be found in my youtube video here).

This article gives a broad selection of examples from Amele, a Papuan SOV language, to demonstrate the importance of these issues, with suggestions about broader implications.

Please feel free to use the comments section below to give examples where you have seen this to be useful in other languages, or where you think it is not useful, or to ask further questions.

[This is one of a series of posts pointing to seminal articles and resources about the importance of understanding discourse structures in both source and receptor languages when translating the Bible]

---

Extracted text from SyntaxofDiscourse_2010.pdf:

THE SYNTAX OF DISCOURSE STRUCTURE John R. Roberts SIL International Notes on Translation 11.2, 15-34 (1997)1 Introduction The thesis presented in this article2 is that the order of certain headed constituents at the syntactic level of NT Greek has a profound influence on the ordering of certain headed semantic propositional structures in that language. To substantiate the thesis this study is presented in terms of a contrast between the discourse structure of NT Greek with that of the Papuan language, Amele (see Roberts, 1987, for example), which happens to have opposite headed structures at both the syntactic and discourse levels. Word Order Universals The research conducted by Greenberg (1963) established that the order of certain pairs of grammatical elements correlates with the order of verb and object. For example, VO languages tend to be prepositional, placing the adposition before their objects and OV languages tend to be postpositional, placing the adposition after their objects. Greenberg took a sample cross-section of 30 languages covering all the major language families of the world as his database. The results of the research showed that a Basic Order Typology could be set up on the basis of three sets of criteria: a) prepositions vs. postpositions b) the relative order of S(ubject) O(bject) V(erb) c) the relative order of N(oun) and A(djective) From the sample of languages taken it was found there were three main orders for Subject, Object, Verb, viz. VSO, SVO, SOV Greenberg also suggested that there are two basic word order patterns and that any given language is moving towards one of these basic patterns: VSO Pattern Verb Prep Aux Noun Noun Noun Main Cl Obj of Comp Comp/Conj

SOV Pattern Object Object Main verb Adj RelCl Genitive Subord Cl Standard of Comp S

Object Object Main verb Adj RelCl Genitive Subord Cl Standard of Comp S

Verb Postp Aux Noun Noun Noun Main Cl Obj of Comp Comp/Conj

Since Greenberg more detailed research by Dryer (1992) has shown that some of these word order correlations are not valid. Specifically, in the list above the Noun1

Notes on Translation is a publication of the SIL. Much of what is presented in this article is developed from the work of Terry Borchard (Borchard, 1991). 2

16

Syntax of Discourse

Adjective pair does not pattern as Verb-Object in the larger database of languages that Dryer uses. In fact, Dryer proposes the Branch Direction Theory3 to account for the correlation pairs found in his research. THE BRANCH DIRECTION THEORY (BDT): Verb patterners are non-phrasal (nonbranching, lexical) categories and object patterners are phrasal (branching) categories. That is, a pair of elements X and Y will employ the order XY significantly more often among VO languages than among OV languages if and only if X is a nonphrasal category and Y is a phrasal category. In other words, languages tend towards one of two ideals: right-branching languages, in which phrasal categories follow nonphrasal categories, and leftbranching languages, in which phrasal categories precede nonphrasal categories. By selecting one of these options the difficulties of processing centre-embedded structures are avoided. The full set of category pairs that Dryer determines follow a right-branching or left-branching pattern are: Verb Patterner

Object Patterner

verb verb adposition copula verb „want‟ tense/aspect auxiliary verb negative auxiliary complementizer question particle adverbial subordinator article plural word noun noun adjective verb verb

object subject NP predicate VP VP VP S S S N N genitive relative clause standard of comparison PP manner adverb

Head-first vs. Head-last Languages In current generative grammar (Government and Binding Theory) there is also a parameter relating to constituent order in the phrase structure of languages called Head first

vs.

Head last

A language will tend to have the heads on the same side in all phrases. For example, English is considered to be a head-first language, kick the ball (the verb is to the left of the NP) on the table (the preposition is to the left of the NP) Amele, on the other hand, would be considered to be a head-last language, sab jeia [food he.ate] (the verb is to the right of the NP) jo na [house in] (the postposition is to the right of the NP)

3

Dryer (1992:114) modifies the BDT to account for phrasal categories that are not fully recursive and to distinguish between major and minor constituents.

John R. Roberts

17

The notion of head described here relates to the head of the phrase and its complement. However, Dryer (1992) argues that this notion of head does not give an adequate explanation of the correlation pairs found. For example, the article in the pair article+N is not normally considered to be the head of the NP. Propositional Orders of NT Greek and Amele Contrasted In the remainder of this article I will present evidence that the order of semantic propositions in language can be correlated with the verb-object and adposition-NP patterns in that language. I will do this by contrasting NT Greek with the Papuan language Amele. Constituent Order in NT Greek According to Funk (1961) NT Greek follows the first Greenbergian word order pattern given above with respect to the constituents in the clause almost entirely. In the unmarked case the orders are: (conjunction) verb, subject, object preposition, noun noun, attributive adjective (anarthrous) noun, attributive genitive noun, relative clause main clause precedes subordinate clause. With reference to Dryer‟s set of correlation pairs NT Greek also has the following set of VO correlates in the unmarked case: complementizer (such as ὅτι „that‟), sentence adverbial subordinator (such as ἵνα „that‟), sentence article, noun/adjective adjective, standard of comparison verb, prepositional phrase These constituent orders are not fixed, however, and NT Greek allows a great deal of variation in word order. Neither are these constituent orders necessarily the most common to occur in a given NT book. Levinsohn (1992), for example, notes that in the book of James, out of the 80 sentences which contain both a subject and an object, the frequency of the different orders is as follows: SOV (34); SVO (24); VSO (12); OSV (2); OVS (6); VOS (2). The unmarked order of VSO occurs in only 15% of the possible instances. Even so, the critical VO order occurs in nearly 50% of instances. It is also the case, according to Funk, that the preposition invariably occurs before the nominal it is governing. NT Greek can therefore be defined as a right-branching and a head-first language, the same as English. Indeed, in NT Greek the head-first pattern also applies to the headdependency relationship of noun followed by adjective. So NT Greek is a near perfect example of Greenberg‟s VSO word order pattern and Dryer‟s VO, P[NP] pattern. Propositional Semantics In this contrast of Greek and Amele propositional structure I will focus on some of the logical propositional relationships found in these languages. The system I will adopt is that used by Larson (1984) and in SIL‟s Semantic Structure Analyses (SSAs) (see chart below).

18

Syntax of Discourse

Chart of Relations between Communication Units EQUAL NATURAL

chronological (time in focus)

sequential simultaneous

nonchronological

conjoining alternation orienter - CONTEXT circumstance - HEAD introduction - HEAD opening - BODY moven - HEAD stepn - GOAL occasion - OUTCOME HEAD - equivalent n GENERIC - specific generic - SPECIFIC contraction amplification HEAD - comparison HEAD - illustration HEAD - manner contrast - HEAD

PROMINENCE

orientation

chronological

restatement UNEQUAL NATURAL PROMINENCE

clarification nonchronological

RESULT -

reason means MEANS - purpose condition RESULT -

logical

CONSEQUENCE

concession CONTRAEXPECTATION

• grounds
• comment HEAD - parenthesis CONCLUSION

associative

CONCEPT - COMMUNICATION UNIT RELATIONS

HEAD

CONCEPT

• identification

CONCEPT

• description

In Larson‟s framework of binary relationships between propositions a key notion is that of support-HEAD, where one proposition is deemed more „prominent‟ semantically than the other proposition. Larson divides propositional relationships into those having equal natural prominence and those with unequal natural prominence. This is similar to the syntactic notion of exocentric versus endocentric constructions. Endocentric constructions are headed constructions where one of the constituents of the construction is functionally equivalent to the construction as a whole and functions as the head of the construction. Exocentric constructions, on the other hand, do not have a constituent that functions as the head of the construction. The support-HEAD relationship is one of unequal natural prominence. In this relationship the more prominent proposition is labelled HEAD and the second proposition supports this HEAD proposition by orientation, clarification or logical argumentation. Larson (1984) does not appear to offer an explanation or define on what basis the HEAD proposition is determined. In most cases the support-HEAD relationship seems to correspond to the subordinate clause - main clause distinction and the proposition functioning as the main clause is the semantic head of the relationship. Thus the notion of semantic head appears to be based on the syntactic notion of headedness. Beekman and Callow (1974), on the other hand, do define the notion of HEAD proposition in semantic terms as the „development‟ proposition, i.e. the proposition

John R. Roberts

19

that develops a semantic unit of a discourse in some way. They also say that support propositions can be classified on the basis of their semantic function in a discourse according to three different functions:

1. The support proposition clarifies another proposition, by explaining or highlighting it.
2. The support proposition argues for another proposition by giving its logical antecedent or consequent.
3. The support proposition orients another proposition by giving its setting relative to time or space or other Events. Logical Propositional Relationships Contrasted Perhaps the most striking examples of differences in propositional ordering between NT Greek and Amele is in the case of the logical relationships. In most of these cases the unmarked arrangement in NT Greek is for the HEAD proposition to come first, whereas the unmarked order in Amele discourse structure is for the HEAD proposition to come last. In this respect Amele follows the more natural logical order of cause-EFFECT, where cause precedes EFFECT and EFFECT is the HEAD of the relationship. Some examples of logical restructuring in the Amele translation are given below. These are taken mainly from the book of Acts and cover the following logical relationships: RESULT-reason RESULT-means MEANS-purpose MEANS-neg purpose

reason-RESULT means-RESULT purpose-MEANS neg purpose-MEANS

The first set of examples involve the RESULT-reason relationship. The unmarked order for this relationship in NT Greek is that of RESULT preceding reason and such clauses are typically conjoined by [ὅτι] „because‟. However, the unmarked order in Amele is of reason followed by RESULT. Such clauses can be conjoined by the subordinating conjunction nu „so, because‟ and indicates that the preceding clause is the reason. Alternatively the reason-RESULT relationship can be indicated by the result clause beginning with eu-nu „that-for‟. In this case the demonstrative eu functions as a pro-clause substitute for the preceding reason clause(s). It should be pointed out that it is possible to express the RESULT-reason relationship in that order in Amele by having a question construction in between the two propositions. However, this is the marked order when the speaker wants to frontshift the RESULT clause for pragmatic purposes. RESULT

Eeta-nu? „What-for?‟

reason

20

Syntax of Discourse

GREEK

AMELE

[RESULT] [ὅτι] „because‟4 [REASON]

[REASON] [nu] „so‟, [Eunu] „Therefore‟ [RESULT]

Acts 2.6 happening δὲ this sound came together the multitude and [RESULT] were.confounded [ὅτι] because [REASON] they.heard one each in his own language speaking them

When they heard that sound a large group of men gathered. [REASON] They each heard them speaking in their different languages [nu] so [RESULT] they were all confused.

Acts 6.1 [RESULT] there.was murmuring of.the Hellenists against the Hebrews [ὅτι] because [REASON] were.overlooked in the service daily the widows of them. Acts 8.20 [RESULT] the silver of yours with you may.it.be into perdition [ὅτι] because [REASON] the gift of God you.did.suppose through money to get

[REASON] the Greek speaking Jews said, “Our widows at the time of the daily food distribution they always overlook them.” [Eunu] Therefore [RESULT] the Greek speaking Jews argued with the Hebrew speaking Jews. [REASON] you think that you can buy with money the gift thing God has given us [nu] so [RESULT] may you and your money go to hell!

With respect to the means-RESULT relationship it would seem to be more difficult to establish RESULT-means as the default order in the Greek. For one thing, there do not appear to be many examples in the NT where the means proposition is expressed by a clause instead of a nominal. Even so, some examples are given below taken from Acts and Luke where it was necessary to switch the order of the propositions found in the Greek in order to produce the appropriate order in Amele. As with the RESULT-reason relationship the RESULT-means relationship can be expressed in that order in Amele. However, this is also a marked structure with the RESULT clause front-shifted for pragmatic purposes. RESULT

means

eu-na „that-by‟

When the preposition ἐν occurs with the instrumental/means function in conjunction with an immediately following clause then we have the unmarked propositional order of RESULT-means. This applies to the first two examples given below. However, where ἐν is followed by a pronoun with a relativising function, such as τίς or ὅς, then the propositional ordering is one of means-RESULT. This is illustrated by two examples given below. Levinsohn (1992: 186) notes that in continuative relative constructions the information preceding the relative pronoun characteristically becomes the ground for what follows. In these examples of ἐν with a relativising pronoun it may be the case that the means clause is backgrounded. Finally, there is an example with δια in Acts with the order means-RESULT, contrary to what the theory predicts.

4

It is perhaps significant that οὐν is seldom used for a true reason-RESULT relationship, but most often to resume the argument after a reason for a previous assertion.

John R. Roberts

GREEK

AMELE

[RESULT] [ἐν] „by‟ [σύν] „with‟ [MEANS]

[MEANS] [euna] „that by‟ [RESULT]

Acts 4.29-30 [RESULT] give to the slaves of you with boldness all to speak the word of you [ἐν] „by‟5 [MEANS] the hand stretching out of you

Luke 24.35 [RESULT] he.was.known to.them [ἐν] „by‟ [MEANS] the breaking of.the loaf Acts 7.35 [RESULT] this man God both a ruler and a redeemer has sent [σύν] „with‟ [MEANS] the.hand of.the.angel appearing to him in the bush

[MEANS] [ἐν τίνι] „by what‟ [ἐν οἷς] „by which‟ [RESULT] Acts 4.9 [MEANS] a good deed done to an infirm man [ἐν τίνι] „by what‟ [RESULT] this man has.been.healed Acts 11.14 [MEANS] who will.speak words to you [ἐν οἷς] „by which‟ [RESULT] may.be.saved you and all the household of you Acts 8.18 [δια] „through‟ [MEANS] the laying.on of.the hands of.the apostles [RESULT] was.given the Spirit

21

[MEANS] as you show them your strong doing make people well … And then strong doing [euna] „that by‟ [RESULT] we your work men to speak your word boldly may you help us [MEANS] the Lord broke the bread for them (and) [RESULT] their eyes could perceive [MEANS] God he the angel he saw standing in the flame of a small burning tree his hand [na] „by‟ [RESULT] sent him to free them and rule over them („by the hand of‟ is idiomatic in Amele) [MEANS] [euna] „that by‟ [RESULT]

[MEANS] we made a lame leg man well, making well [euna] „that by‟ [RESULT] that man stands well about that if you are asking us … [MEANS] he will speak words to you [euna] „that by‟ [RESULT] you and all who live in your house God will save. [MEANS] Peter and John placed their(dl) hands on their(pl) heads (and) [RESULT] the Holy Spirit came down onto them

We found in translation that the MEANS-purpose relationship can often be confused with the means-RESULT relationship. This could be doubly confusing to the Amele reader since a propositional relationship is presented that on the face of it appears to

5

This assumes the instrumental reading of ἐν τῳ which is controversial in this instance.

22

Syntax of Discourse

be logically correct with the HEAD occurring last, but when reinterpreted as a MEANSpurpose relationship the HEAD element is now first. GREEK

AMELE

[MEANS] [ἵνα] „that‟ [PURPOSE]

[PURPOSE] future tense [nu] „so that‟ [MEANS]

Acts 5.15 [MEANS] they.brought.out the ailing and to.place on pallets and mattresses [ἵνα] „that‟ [PURPOSE] coming of.Peter if.even the shadow might. overshadow one of.them

Acts 9.21 [MEANS] here for this he has come [ἵνα] „that‟ [PURPOSE] having.been.bound them he.might.bring before the chief priests

Acts 16.30 [MEANS] Sirs, what me behoves.it to do [ἵνα] „that‟ [PURPOSE] I.may.be.saved?

[PURPOSE] men (when) Peter will come and pass them by his shadow will hit them and they will become well [nu] „so that‟ [MEANS] they carried the sick and brought them on their beds and woven mats and put them on the road. [PURPOSE] he will people from here arrest them and bind them and take them to the chief priests and take them there to Jerusalem [nu] „so that‟ [MEANS] has he not come here? [PURPOSE] Two big men, God will save me [nu] „so that‟ [MEANS] what will I do?

John R. Roberts

GREEK

AMELE

[MEANS] [ὅπως] „so that‟ [PURPOSE]

[PURPOSE] future tense [nu] „so that‟ [MEANS]

Acts 8.15 [MEANS] who going.down prayed concerning them [ὅπως] „so that‟ [PURPOSE] they might receive the Holy Spirit

Acts 8.15-16 [PURPOSE] when they arrived there those believers that the Holy Spirit had not come down on and had only received water baptism in the name of the Lord Jesus for them to receive the Holy Spirit [nu] „so that‟ [MEANS] they prayed

Acts 9.12 [MEANS] and saw in a man in a vision Ananias by name coming in and putting on him hands [ὅπως] „so that‟ [PURPOSE] he may see again Acts 23.15 [MEANS] inform the chiliarch with the council [ὅπως] „so that‟ [PURPOSE] he may bring down him

[PURPOSE] he in a vision a man named Ananias will come to him and he will see again [nu] „so that‟ [MEANS] he will put his hand on top he saw [PURPOSE] they will send a word and it will go up to the Roman soldier head and he will send Paul and he will come down [nu] „so that‟ [MEANS] pretending that we will ask him whether his talk is true send a word

GREEK

AMELE

[MEANS] [ἵνα μὴ] „that not‟ [NEG PURPOSE]

[NEG PURPOSE] [dainnu] „lest‟ [MEANS]

Acts 5.26 [MEANS] brought them not with force for they feared the people [μὴ] „lest‟ [NEG PURPOSE] they should be stoned Luke 16.28 [MEANS] so that he may witness to them [ἵνα μὴ] „that not‟ [NEG PURPOSE] also they come to place this of torment

Luke 22.46 [MEANS] rise up and pray [ἵνα μὴ] „that not‟ [NEG PURPOSE] you enter into temptation

23

[NEG PURPOSE] they might hit them with stones [owainnu] „lest‟ they feared. [MEANS] Therefore they brought them without violence [NEG PURPOSE] may they not come here to this place of great pain for the body to feel [owainnu] „lest‟ [MEANS] send Lazarus to my father‟s house rise up and [NEG PURPOSE] not go into temptation [owainnu] „lest‟ [MEANS] pray

A comparison of the propositional structures of NT Greek and Amele revealed that it is not only the case that certain logical propositions are reversed between these

24

Syntax of Discourse

languages. It is also the case that the whole reasoning process is reversed. In NT Greek a deductive reasoning process is usually followed, whereas in Amele expository text material an inductive reasoning process is usually followed. These two reasoning processes can be contrasted as follows: Deductive reasoning THESIS [argument(s)]

Inductive reasoning [argument(s)] THESIS

In the deductive method you present the thesis first and then follow this with the supporting argument(s).

In the inductive method you present the supporting argument(s) first and then follow this with the thesis.

Where there was a section of extensive deductive reasoning in the NT Greek source text we usually found that Amele readers could understand this material better and be be better able to answer comprehension questions if we reordered the logical argumentation of the text to follow an inductive reasoning structure. Some translation examples requiring logical restructuring are given below. In the 1 Corinthians 2.10-11 example a simple reversal to give an inductive reading is sufficient. For the Romans 8.1-8 example, however, it was necessary to adopt an inductive „sandwich‟ strategy with the thesis mentioned at both the beginning and the end of the section. 1 Corinthians 2.10-11 Deductive reasoning THESIS [argument(s)]

Inductive reasoning [argument(s)] THESIS

10

10-11

But [γάρ] God by his Spirit has shown us his hidden word and it has become clear to us. For [γάρ] that Spirit searches and sees all things. And he also searches and sees God‟s thinking that lie hidden. 11 For [γάρ] a man does not know the thinking of another man. A man‟s own spirit knows well all his own thinking. In the same way too only God‟s Spirit knows God‟s thinking well.

But a man does not know the thinking of another man. A man‟s own spirit knows well all his own thinking. In the same way too only God‟s Spirit knows God‟s thinking well. And God‟s Spirit searches and sees all things. And he also searches and sees God‟s thinking that lie hidden. Therefore God by his Spirit has shown us his hidden word and it has become clear to us.

John R. Roberts Romans 8.1-8 Deductive reasoning THESIS [argument(s)]

Inductive reasoning (sandwich) [argument(s)] THESIS

1

1-8 (1)

There is therefore now no condemnation for those who are in Christ Jesus. 2 For [γάρ] the law of the Spirit of life in Christ Jesus has set me free from the law of sin and death. 3 For [γάρ] God has done what the law weakened by the flesh could not do: sending his own Son in the likeness of sinful flesh and for sin, he condemned sin in the flesh, 4 in order that [ἵνα] the just requirement of the law might be fulfilled in us, who walk not according to the flesh but according to the Spirit. 5 For [γάρ] those who live according to the flesh set their minds on the things of the flesh, but those who live according to the Spirit set their minds on the things of the Spirit. 6 For [γάρ] to set the mind on the flesh is death, but to set the mind on the Spirit is life and peace. 7 For the mind that is set on the flesh is hostile to God. It does not submit to God‟s law, indeed it cannot; 8 and those who are in the flesh cannot please God.

25

Therefore we men gathered together with Christ Jesus for us to receive a punishment for our sins a talk does not exist. (7) If man‟s thinking follows the fashion of the old stomach then he does not follow and do God‟s law talk. (8) He is not able to do like that. Therefore that man is God‟s enemy. (5) Therefore those men that walk in the fashion of the old stomach their thinking only thinks about things of the old stomach. But men that walk in the fashion of the Spirit their thinking only thinks about the fashion of the Spirit. (6) If a man‟s thinking only stands in the fashion of the old stomach then that man will die. But if a man thinks only about the fashion of the Spirit then that man will receive peace sitting and life sitting. (3-4) Therefore we with the old stomach are without strong doing so that thing the law talk is not able to do God did. That law talk that tells people to do righteous fashion we that do not walk in the way of the old stomach fashion lifestyle but walk in the lifestyle of the Spirit so that we are able to do that righteous fashion God sent his own Son and he took and clothed himself in a sinful body like our sinful bodies and came down by that he did to put an end to the sin in our old stomachs. (2) And then the law talk of the life sitting Spirit that has given life to us all that sit gathered together with Christ Jesus has freed me from the law talk of sin and death. (1) Therefore we men gathered together with Christ Jesus for us to receive a punishment for our sins a talk does not exist.

Restatement and Clarification Relations Contrasted For the other nonchronological relations of unequal prominence under Larson‟s system there is evidence that the order in which these propositions are presented in discourse is also influenced by the syntactic configuration of the language in question. So, for example, with respect to the HEAD-equivalent relationship under the restatement category, when all things are equal there is a preference in Amele discourse structure for a command to follow an assertion and for a positive proposition to follow a negative proposition. The effect of this in a HEAD-restatement relationship is typically to place the HEAD proposition last.

26

Syntax of Discourse HEAD

Believe!

restatement

Don‟t doubt!

restatement

Don‟t doubt!

HEAD

Believe!

It is not clear if the reverse holds in NT Greek discourse structure, i.e. that there is a preference for a command to precede an assertion and for a positive proposition to precede a negative proposition. However, some relevant translation examples of this requiring restructuring into Amele are given below. GREEK

AMELE

COMMAND - assertion

assertion - COMMAND

Luke 10.3 [COMMAND] „Go your way. [ASSERTION] Behold, I send you out as lambs in the midst of wolves.‟

[ASSERTION] „See, I send you like they send sheep amongst wild dogs. [COMMAND] Go your way.‟

Matthew 26.26 [COMMAND] „Take and eat. [ASSERTION] This is my body.‟

[ASSERTION] „This is my body. [COMMAND] Take and eat.‟

John 9.23 [COMMAND] „You ask him and see. [ASSERTION] He is not a child, he is able to tell you.‟

[ASSERTION] „He is not a child, he is able to tell you. [COMMAND] You ask him and see.‟

POSITIVE - negative

negative - POSITIVE

Acts 18.15 But since it is a matter of questions about words and names and your own law, [POSITIVE] see to it yourselves. [NEGATIVE] I refuse to be a judge of these things.

But because you only argue about your talk and your names and your law talk [NEGATIVE] I will not sit as judge of these things. [POSITIVE] Make it straight yourselves.

Acts 22.22 … “[POSITIVE] Away with such a fellow from the earth! [NEGATIVE] For he ought not to live.”

… “[NEGATIVE] Such a man is not worthy to live. [POSITIVE] Take him away! Kill him!”

COMMAND - prohibitive

prohibitive - COMMAND

Matthew 17.7 [COMMAND] „Arise. [PROHIBITIVE] Do not be afraid.‟

[PROHIBITIVE] Do not be afraid.‟ [COMMAND] „Arise.

Luke 18.16 [COMMAND] Allow the children to come to me [PROHIBITIVE] and do not prevent them. [REASON] For [γάρ] to such belongs the kingdom of God.

[REASON] It is the case that God‟s kingdom was made for such children [PROHIBITIVE] Therefore do not prevent them. [COMMAND] Let the children come to me.

With respect to the clarification relationships of HEAD-comparison and HEADmanner again there is a preference in Amele discourse structure for the HEAD to come last, in contrast to NT Greek discourse.

John R. Roberts HEAD

His hair is white

comparison

as snow (is white)

HEAD

so his hair is white

HEAD

He will return

manner

As he went

manner

as he went

HEAD

so he will return

comparison

27 As snow is white

Some relevant translation examples are also given below to illustrate these contrasts. GREEK

AMELE

OBJECT of comparison [ὡς] „as‟ STANDARD of comparison

STANDARD of comparison [we] „as‟ [odi] „likewise‟ OBJECT of comparison

Acts 7.51 [OBJECT] you always the Spirit Holy oppose. [ὡς] „as‟ [STANDARD] the fathers of you also [OBJECT] you ___

[STANDARD] your gt. gt. grandfathers used to obstruct the Holy Spirit [we] „as‟ [odi] „likewise‟ [OBJECT] you also always obstruct the Holy Spirit

Acts 11.17 [OBJECT] the equal gift gave them God [ὡς] „as‟ [STANDARD] also to us ___

[STANDARD] now God this gift he gave us [we] „as‟ [OBJECT] he has given these men also

Acts 23.11 [ὡς] „as‟ [γάρ] „for‟ [STANDARD] you did solemnly witness the concerning things about me in Jerusalem [OBJECT] so you it behoves also in Rome to witness

[STANDARD] you witnessed about me in Jerusalem [OBJECT] Therefore there in Rome too [odi] „likewise‟ very about me you will witness

GREEK

AMELE

HEAD [ὃν τρόπον] „in the way‟ MANNER

MANNER [odi] „like that‟ HEAD

Acts 1.11 This Jesus, [HEAD] who was taken up from you into heaven, will come [ὃν τρόπον] in the way [MANNER] you saw him go into heaven.

This Jesus [MANNER] that you saw God took him and he went up into heaven [odi] like that [HEAD] he will return and come down.

Acts 15.11 we believe that [HEAD] we shall be saved through the grace of the Lord Jesus [ὃν τρόπον] [MANNER] just as they will

[MANNER] As God will save them by the grace of the Lord Jesus [odi] like that [HEAD] he will save us too by the grace of the Lord Jesus that we believe

28

Syntax of Discourse

The Syntax of Propositional Structure In the preceding section I have contrasted the order of certain propositional relationships between NT Greek and Amele on the basis of head-first versus headlast. These orders, in fact, follow the ordering of certain syntactic constituents which can be understood as also being in a head-first versus head-last correlation, viz.

Greek

Amele Syntactic Correlations

VSO V [object] V [subject] P [NP] V [PP] N [RelCl] N [genitive] Comp [S] Verb [object nominal clause] know [ὅτι ‘that’ … ] see [ὅτι ‘that’ … ] Subord Conj [S] MAIN CLAUSE [subordinate clause]

SOV [object] V [subject] V [NP] P [PP] V [RelCl] N [genitive] N [S] Comp [object nominal clause] Verb [ .. ec] dogina [ .. ec] fegina [S] Subord Conj [subordinate clause] MAIN CLAUSE

Semantic Propositional Correlations COMMAND [assertion] POSITIVE [negative] COMMAND [prohibitive] OBJ OF COMP [stand of comp] X is as Y is HEAD [manner] X does like Y does RESULT [reason] X because Y RESULT [means] X by Y MEANS [purpose] X so that Y MEANS [neg purpose] X lest Y CONCLUSION [grounds] X because/for Y Deductive reasoning THESIS [argument(s)]

[assertion] COMMAND [negative] POSITIVE [prohibitive] COMMAND [stand of comp] OBJ OF COMP as Y is so is X [manner] HEAD as Y does so does X [reason] RESULT because Y, X [means] RESULT by Y, X [purpose] MEANS so that Y, X [neg purpose] MEANS lest Y, X [grounds] CONCLUSION Y therefore X Inductive reasoning [argument(s)] THESIS

Thus the discourse structure of propositional relations in each language follows the syntactic structure in that language. This has major implications for anyone translating into a language with SOV word order and postpositions from NT Greek or English based source texts.

John R. Roberts

29

Implications for Translation The main implication of these findings for translation is that if the translation into Amele is to be fully idiomatic in that language then it has to take account of the opposite logical structures of NT Greek and Amele. In fact, as we have gone along in the translation process I have realised how critical it is that we follow Amele logical structure, especially in the expository material, if we are to produce a translation that can be readily understood and followed by Amele native speakers. In the beginning of the translation process we were not so careful to follow Amele logical structure. This was partly because those who did the first-drafting were Amele-English bilinguals and they were influenced by English logical structure, which is similar to NT Greek. Even though they also used the Tok Pisin translation this did not help in this area since it too follows English logical structure. It was also the case that in the early stages I was not so aware of the significance that this difference makes between NT Greek and Amele. This only came home to me in checking sessions as we got more into epistle material. I noticed that the Amele translation checkers would often be looking backward through the text searching for the supporting arguments to answer the question that had been put to them on the thesis. On such occasions I would usually have to point out in the text that what they were looking for actually followed (wrongly for them!) the thesis. Conclusion The thesis that has been presented is that the order of the syntactic constituents verb + object and adposition + nominal determines the ordering of certain logical propositions in discourse structure. These orders happen to be completely opposite for NT Greek and Amele. Greek has the orders of VO and P[NP] and Amele has the orders of OV and [NP]P. This makes Greek a head-first, right-branching language and Amele a head-last, left-branching language. Not only does the order of these syntactic constituents affect the ordering of certain headed propositional relationships, such as HEAD-equivalent, HEAD-comparison and reason-RESULT, in these languages it also influences the type of reasoning process used in their respective discourse structures. So a VO, P[NP] syntactic arrangement produces a preferred deductive reasoning process, where the thesis is presented first followed by supporting arguments, and an OV, [NP]P syntactic arrangement produces a preferred inductive reasoning process, where the supporting arguments are presented first followed by the thesis as a conclusion to the arguments. Since the vast majority of Papuan languages are SOV and postpositional it is probably the case that the same facts about the discourse structure of Amele outlined here apply to these languages too. Ellis Deibler hints at some of these characteristics of Papuan languages in Larson (1984) and Borchard (1991) confirms these characteristics in an extensive comparative study of the discourse structure of NT Greek and the Papuan language Ipili. Levinsohn (personal communication) also advises that similar orders of propositional structures apply to the Columbian language Inga as apply to Amele. Inga is also a language with SOV word order and postpositions. So it may be the case that the principles of discourse structure outlined in this article have a universal application.

30

Syntax of Discourse

References Beekman, J. and J. Callow (1974) Translating the Word of God. Grand Rapids, Michigan: Zondervan. Borchard, T. H. (1991) Discourse Level Functional Equivalence Translation. PhD Thesis. Fuller Theological Seminary, Pasadena, California. Dryer, M. S. (1992) The Greenbergian Word Order Correlations. Language 68.81138. Funk, R. W. (1961) A Greek Grammar of the New Testament and Other Early Christian Literature. Cambridge: Cambridge University Press. Greenberg, J, (1963) “Some Universals of Grammar with Particular Reference to the Order of Meaningful Elements” in Greenberg ed. Universals of Language, 73113. Cambridge, Mass.: MIT. Larson, M. L. (1984) Meaning-based Translation: A Guide to Cross-language Equivalence. London: University Press of America. Levinsohn, S. H. (1992) Discourse Features of New Testament Greek. Dallas: SIL. Roberts, J. R. (1987) Amele. London: Croom Helm.