Recently, I played a short narrative in the language I describe for my PhD to a non-linguist. The person asked what I do with such recordings. I told them I get a translation by a native speaker and then work on them myself, including changing the original translation wherever I feel it to be necessary. Right there, they asked me how I justify correcting a native speaker. I don’t think I did a good job at explaining it back then, so let me try once more in writing.
Native speakers who have not trained as linguists or translators often re-narrate what the speaker said. One of the consequences is that sometimes they simplify what they heard, i.e. they will leave out information and structures that they don’t consider relevant. A person might say that they bought bananas, oranges and mangos but the translator will say they bought fruits. In another instance, somebody might think aloud “I arrived on Friday … was it really Friday? … hmmm, yes”, and all that the translation will contain is that the person arrived on Friday.
At the same time, these translators will add material at other places. If a person does not conclude a sentence but the translator feels certain how it would have ended, they may add their ending to the translation. Also, they may add information that they consider useful for the linguist. E.g., if the speaker said they saw Jola, the translation might be about the nurse Jola. Of course, in the original you won’t find the word “nurse”.
Also, they may mess up the order of whatever was said. I noticed this especially with one native speaker who always listened to long chunks of recording before providing a translation. Hence, there always was a lot of information to remember and apart from leaving out and adding bits, the translator sometimes got the sequence of events wrong.
As I stated in the beginning, untrained native speakers usually re-narrate what the speaker said. Hence, in their version the perspective often changes. So, when a person says “I washed my hands today”, the translator is likely to say “he washed his hands that day”. This phenomenon is quite well known among language documenters. We probably all know the situation where we try to elicit verb paradigms like “I slept, you slept …”, noticing that the language consultant turns it into “you slept, I slept …”.
Sometimes, translators change the meaning of an utterance because there is no exact equivalent in English or because that equivalent is uncommon. An example for this is “egbon”, which in Yoruba means “older sibling (of either gender)”. As “sibling” is a quite technical term and people would rather refer to their sibling as “my sister” or “my brother”, the translator is likely to use “sister” or “brother”, even though it adds gender information that is not included in the original.
Then, there is the issue of Nigerian (Pidgin) English. The translations I sometimes get are not correct or mean something else in British English, the variety I happen to use for writing my thesis. An example is the use of “again”. In Nigeria, this word is used instead of “anymore”. So, when I see a translation like “he doesn’t love her again”, I change it to “he doesn’t love her anymore”.
Finally, native speakers are mere humans. Sometimes, they are unconcentrated or not feeling well. Also, they are not working full time on my project, which means they only look at each utterance for a short time, provide a translation and then move on to the next utterance. I on the other hand sometimes go back to an old utterance and compare it with lexical entries and the alleged meaning of comparable utterances. This means I think about each utterance longer, consult more material and can make sure that I do so at a time when I am able to give the task the full of my abilities.
How do I spot items in a translation that might need change? Usually, I notice a contradiction with something another or the same speaker told me at a different time. E.g. when I have already learned the words “I” and “he”, I will note when the translation doesn’t match the original. Or I suspect that a translation could be improved because I know a lexical item or structure used in related languages. At other times, the translator is very hesitant, which suggests they themselves doubt their own translation.
Once I spot a potential mistake, there are basically three options. Either I simply correct the mistake, I discuss the issue with the original translator or I ask another native speaker for a second opinion. If I speak to the person who translated the piece in the first place or with a second translator, the person sometimes agrees with my own translation (in the end). At times, they even state that I help them to understand their language better. But just as often they convince me of their own version and then I have learned something from them again.
I am grateful to the native speakers who helped me by answering my questions “how do you say this” and “how do you say that”, by telling me stories in their language and by translating them into English for me. There is a reason why we call them language consultants. It’s because they are the experts and owners of their language. It would make me very happy if one of them stepped forward one day to learn the tools of my profession. I’m sure they would do a much better job at analysing and especially at promoting the language than I ever could. Also, I wish to thank the person who raised the question that initiated this post. You made me think, and that’s a good thing.