Building on a previous experiment to apply Strathern's discussions of Melanesian gift exchange to ethnography relating to Tukanoan societies in Northwest Amazonia, this essay asks why other authors should repeatedly affirm that the gift has no relevance in the Amazonian context. Two answers are proposed. Firstly, the authors concerned tend to assume that a particular type of social formation is characteristic of Amazonia as a whole. The Tukanoans, who engage in ceremonial exchange of food and goods, do not fit this rubric. Secondly, despite their differences, these authors assume that Amazonian societies are bride-service societies where, axiomatically, there can be no gift. The Tukanoans are not bride-service societies and, if anything, tend towards the bride-wealth alternative. This Tukanoan exception serves to re-emphasize the diversity of Amazonian social formations, one that would have been even greater in the archaeological past. It also warns against the dangers of over-hasty theoretical closure.