This paper describes the challenges faced, and rules devised, while dealing with bilingual interview data as part of a life history study of a female science teacher's conceptions of the nature of science while teaching in a school in Karachi. The interview data generated was both in Urdu and English, which underwent a number of processes (transcription, translation, and transliteration) to evolve into "interim texts," to finally become a part of the data analysis process. I have called these translated materials "transmuted texts," as they reflect the original, but have been recreated. This paper is significant because as globalization connects diverse societies, more research studies have to deal with research data in more than one language. (Contains 1 footnote.)