How should I present data written in another language?

Question

I'm writing a research paper in English but the topic is about a different country and the data is written in a different language, Arabic.

Should I translate the data into English? If yes, should I include a version of the excerpt in the original language? If yes, how?

Apart from using excerpts of data, I sometimes have to mention certain words of the original language such names of participants and research fields, but I do not know whether I should use transliteration, translation or both.

Thank you

Answer 1

It would depend how much your translation of the story makes a difference to your thesis. If it doesn't particularly matter with respect to the finer details, it would be sufficient to provide the entire story in Arabic and then give a synopsis in English. If it's more critical that the audience knows what each line means, then give each sentence with its English gloss.

If you wanted to go to the extreme, you would have an interlinear gloss, which means that each word, or even morpheme, is glossed individually, and then a free translation given at the end. Here's a Hungarian example (from the Leipzig glossing rules):

Gila abur-u-n      ferma  hamišaluǧ  güǧüna  amuq’-da-č. 
now  they-OBL-GEN  farm   forever    behind  stay-FUT-NEG 
‘Now their farm will not stay behind forever.’ 

In linguistics we use these literally hundreds of times in a thesis, because this is where our data really is. It's probably more depth than you'd be willing to go into, and I would advise against it unless the Arabic syntactic and morphological features are absolutely critical to your thesis. The conventions that most linguists use (and tweak to their own subject language) with respect to the names of various grammatical features (OBL and GEN, for example) can be found in the Leipzig Glossing Rules.

I think you'd be fine to select from either giving the entire story followed by a synopsis of whatever detail you'd prefer, or a line-by-line transliteration.

Answer 2

You should be concerned with more than whether readers understand the data: you need to present the data in the rawest form practical, when there is a possibility of error in rendering the data in an altered form. If the data were transcriptions of interviews conducted in Arabic, and was not generally available to readers, then the data should be presented in its original form. This allows diligent readers to verify your conclusions.

For practical purposes (assuming the readership isn't a fluent Arabic-speaking population), a translation should be provided as well. Depending on the amount of material involved, you might present consecutive texts, first the Arabic and then the translation. If there is a lot of material, it may be best to put the original text in an appendix, unless issues of translation need to be discussed in the paper. For Arabic, there is also the question of how to present the text, i.e. in Arabic script or in (augmented) Roman letters. For Classical and Modern Standard, there are standards that lead to interchangeability so it does not matter too much, but for Colloquial dialects it would matter how the data was created (did the subject write it, or was it spoken and someone transcribed it?). In the latter case, questions of accuracy are especially bound to come up (you can't write ث as "th" in a dialect where it is pronounced "t").

There is no need to translate personal names. If you intend to use a term such as مقهى repeatedly, it should be translated the first time and transliterated all times (maqhā).