This article uncovers some problems involved in culling and translating non-western texts--written in other languages, at particular times, for specific audiences, and rooted in particular local milieus--before assembling them into academic arguments in English in the west. Based on my longterm, evolving endeavour regarding English- and vernacular-language teaching in Gujarat, India, I argue that the seamless ways in which academic knowledge--which draws on translated texts--gets presented belie the murkiness, selectivity, and complexities underlying various stages of the translating and knowledge-building process. Drawing on eight years of extensive translating of Gujarati and Hindi documents and interviews in a variety of educational and community contexts, and on my own roles and experiences as postcolonial researcher who has had to translate herself divergently in across different geographical spaces and in different languages, I argue that uncovering the messinesses involved in selecting and translating particular texts which then get re-inscribed into another language, culture, and disciplinary space--offers interesting insight into the artificial assemblage that constitutes our general disciplinary knowledges. Occasionally dis-assembling aspects of our profession's practices--especially those that we assume as "givens" might prod us into thinking about all texts and knowledges--not just translated ones--as artificial constructions that can be restrung in alternate ways. (Contains 3 figures and 9 notes.)