Discourse Analysis and Translation: an introduction

I’ve been doing lots of ‘discourse’ study recently and some people have asked how they can find out more about it.

The people to read:

• Teach-yourself courses for discourse discovery for source and target languages and translation between them
  • Steve Nicolle: Narrative Discourse and Translation – example from Acts 16 (Free direct download PDF)
  • Steve Runge (Logos):
    • ntdiscourse.org
    • Discourse Grammar of the Greek New Testament ($50): A practical introduction for teaching and exegesis
    • Free videos: on word order, discourse function of Greek Perfect indicative.
  • Stephen Levinsohn: Discourse features of New Testament Greek (Amazon, $20)
• Discourse discovery for a target language:
  • Steve Nicolle: Discourse in Digo, and Movement verbs, Grammaticalisation of tense markers.

My summary:

It’s all about the wood and the trees, but especially the wood.

We all know the meaning of a sentence is greater than the sum of its words, and there’s more to the packaging of words together than syntax. We know you can’t safely match a word in one language with one word in another, or to do that with syntactic constructions.

Choice implies meaning: What we want to do instead is understand the choices an author had, and reconstruct why the chose the options they had rather than another. Every language has its own mix of syntactic requirements and permitted ‘stylistic variation’. We vary the style to help our hearers or readers understand well what we’re trying to say, how the parts connect, what our main point is and what’s part of the background or the basis.

A discourse discovery method:

• Choose a feature to study, work through a text looking for that feature.
• For each feature…
  • Work out what choices an author has.
    • Where there is no choice we move right along – nothing to see here.
    • Where there’s a choice work out what’s normal/unremarkable, and what is unusual/notable.
  • Then reason out the motivation behind one choice rather than another.

Some of the features that we want to ask about are

• Verbal marks of new paragraphs / episodes
• Clause connections (and, but, yet, when)
• Participant reference – eg noun phrases, pronouns, definites, demonstratives etc.
• How we report speech – how much of a speech margin there is
• How we indicate movement and location
• Highlighting by changing word order (moving things to start or end)
• Marking the main event line vs background material (or less significant main events)
• Signs of significant ‘peak’ episodes or other parts of a narrative

We learn about the norms of the languages involved for 2 reasons:

• To understand how the writer expected his audience to take in the text (in the original language).
• To predict how people we communicate with (in the target language) are likely to interpret what we say.

This matters if we respect the original author’s motivations and intentions. If we don’t care about the original author’s purpose, then we can totally ignore it and just mix up all the ingredients used to make the text mean whatever we like.

We really have to study natural, untranslated texts, otherwise we cannot be sure whether the features we study are really from the language of the translation or the original and thus which rules are in play. However, once we have studying the source and target languages and understand how different motivations are expressed, we can evaluate how adequately translations communicate the original author’s purpose.

The purpose: Discourse analysis is important on the source end to make sure that we’re understanding the source text right the way the author intended the audience he had in mind to understand it. It’s important on the target end (whether we’re a preacher or a translator) so that we communicate the author’s intended meaning adequately for a new audience.

Some of the feature-specific questions we might ask

• How do we know we have a new paragraph?
  • eg. NT Greek often has a Verb + δε at a new paragraph/episode. English rarely begins a new paragraph with ‘But…’ or ‘And…’, though ‘Now…’ might work. Usually English will adopt a different strategy, which we need to find out. Ishɛ tends to use participant reference to indicate new paragraphs.
• What’s the role of a particular connector?
  • Beyond just matching up και = and, δε = but, γαρ = for, οὐν = therefore, can we figure out what the purpose is for using each, since there’s a choice.
  • Also consider the significance of using nothing (juxtaposition).
  • What’s normal and what is unusual?
    • eg we notice that Greek δε is very common but in natural stories Hausa amma or Ishɛ amma or English but rarely begins a new sentence. This suggests that simply translating δε with ‘but’ is likely wrong.
• What’s the significance of lots of words introducing speech?
  • Especially when it’s a closed conversation and we know who is talking with whom.
  • NT Greek’s use of ἀποκρινομαι ‘answer’ is interesting, especially in the gospels, because it often seems to indicate a surprising or at least significant turn in the conversation.
• (Loads more could be added, but hopefully that’s enough to give you an idea.)

You can actually ask questions about features or about motivations. Really we want to match them both up, so that when we see a particular way of doing things in the original language, we can deduce the author’s motivation for that choice. Then, when we understand the original author’s motivation we reach for the appropriate way of expressing that in the target language.

A note on ‘Bible handling skills’

I have started wondering how many English ‘Bible handling skills’ are remedial skills for avoidable problems:

• We want to take the Bible seriously, and
• We get used to the Bible not always meaning what it appears to mean, because it doesn’t use English correctly, because
• However good the study of Greek/Hebrew behind the translation, the translators either haven’t considered sufficiently how English really works so get sucked into one of two temptations:
  • Using English words in Greek/Hebrew style/idiom. (ESV, CSB, )
  • Using natural English idioms here and there somewhat inconsistently and only when something sounds really out of place.
• Finally, for those influenced by the KJV, old-fashioned Christianese feels more spiritual or acceptable in the Bible, and translators and readers like a little artificial archaising to add ‘weight’.

All this means that we suddenly need special tools for reading and interpreting the Bible properly, because there’s a special kind of English used in it, just as Shakespeare doesn’t always mean what modern readers/hearers might think it means. If the Bible were originally written in English then there might be some benefit in studying a version in old-fashioned English, but it wasn’t! Really we’d be best served by modern English Bible translations that use modern English well, and then learning the Biblical languages well to read it in the original.

NIV, NLT, ICB and GNB have in different ways tried to remain faithful to English and faithful to the original meaning, but to the best of my knowledge the translators have done almost no systematic study of how English discourse really works in story, teaching, and poetry. The hope has been that by getting eminent English-speaking scholars to work on it and stylistic editors to check it before publication, that something natural will naturally fall out. Our experience is that is far to naive an expectation. A good translator needs to be able to lean on a body of natural texts to remind themselves of how each language really works, so that they don’t fall into the trap of abusing the target language.