a response from brian geytenbeek, the ngarla songs translator
Re the comment on 17 August about pronouns,
(1) Ngarla often omits them once a discourse has been commenced and the "stage" set. From then on, cultural understanding of the context often makes them unnecessary, and usually they are only added if one wishes to emphasise them. E.g., "THEY are the ones at fault, not me!". Thus if you build them into ordinary Ngarla sentences you will skew the meaning considerably (from their point of view), and the readers would be quite puzzled.
(2) In the second verse about the Koolinda the composer talks to the ship in the first line, then describes what happens after the ship does what he told it to do. In the third verse the 'he' and the 'they' are the bosun and crew, then in the final line the composer talks to the ship again.
(3) In Piyuwiki/Buick all three mentions of 'he' in the second verse refer to the driver, not to the car. The pronouns are in fact non-existent in the Ngarla, but they have to be provided in English or the English would not make sense to an English speaker (though they would to a Ngarla speaker who knew a bit of English) -- an interesting example of the basic principle of translation:
"Translation must be of the meaning, not of the words."
Most people just assume that you have to translate each word in the "source" language with an equivalent word in the "receptor" language, but that is not the case at all. Figurative speech illustrates this well. For instance, if someone literally kicked the bucket on the back porch you would never translate the information in the form of "He kicked the bucket", because that would communicate the idea that he had died. Instead you would use some other circumlocution, maybe, "He bumped the bucket with his foot." It's the meaning that counts, not the words used to convey the meaning.
Actually whether you use half as many words or twice as many words in the receptor language may not be very relevant. The important things are that
(a) the result in the receptor language is so natural that the hearers/readers will not be conscious of the fact that it is a translation, and
(b) the hearers/readers in the receptor language will get the same message, and feel the same impact, as did the original hearers/readers in the source language.
That's what a translator is supposed to aim at, anyway. Whether he can always hit it or not is another matter!
Thus genre styles need to be genuine for the receptor language too. There is a marked difference in Ngarla between narrative genre and poetic genre. In some songs I had to work hard to express the poetic meanings that were in the Ngarla in suitably poetic styles in English, especially with a few songs which were expressed not in full intelligible sentences but in one-word "short-hand" poetic images. In these the Aboriginal hearers understood perfectly all the cultural clues and implications just from one word, but English readers wouldn't have had a clue what was going on!
Someone spotted Alex Beeton's Aboriginal nickname. Probably Alex Beeton never even knew that that is what they called him. (Actually his name happens to be in the Nyangumarta language that I've spent 30 years on, not the Ngarla one.) It was very descriptive. He had a huge paunch, and when he laughed he had a huge "belly-laugh", and his belly bounced up and down. Hence his nickname, "Dancing Stomach".