Thanks--I've seen descriptions of her work but haven't read it myself. The work you describe--essentially an ethical bibliography as part of the research itself--seems like a much more promising solution.

I generally agree with your last point, but I'd sound some notes of caution: (1) The key mistake that occurred with IRBs was to treat things like ethnography as human subjects research, when at most non-medical lab studies (behavioral psych) really fit the model. Is the language "human subjects" with its medicalized distance even a proper fit for ethnography, oral history, sociology, etc.? (2) Defining recursive benefit is really hard, and awareness is really all we can ask. Suppose you conclude in all good faith that (to make up an example) the Amish suck -- in only the most Kantian sense, if at all, can your book be considered to benefit the Amish, though you might be able to claim that you're honoring the interests of some subgroup over another. This may be clearest with biography. (On the other side, some communities may consider "advancing outside knowledge about and thus sensitivity to this community" not a benefit to them at all -- what you and I think are recursive benefits may only be so by our terms, not theirs.)

I'm glad you liked the title!