Complicating Realism

Elliot, in Ethnography, Reform, and the Problem of the Real: James Mooney's Ghost-Dance Relgion, writes: "The meeting with Wovoka provides an opportunity for Mooney to demonstrate his abilities as a realist ethnographer, someone willing to go to great lengths to distinguish fact from fiction and someone able to make the reader feel the authenticity of the scene he describes through the conventions of representational realism." The notion here of "representational realism"  engenders an evaluation of the efficacy and utility of a "realist" ideology when conducting ethnography: does "realism", in its apparent unbiased/ethical nature, actually allow as much intepretative freedom as say, in Elliot's argument, "romantic sentimentalism." Can "realism" and "sentimentalism" really be dichotomized and distinguished so easily? Couldn't we understand "realism" as being a repository of imagined images/symbols to the same degree as "sentimentalism" draws on overly-simplified representations; in a different sense, "realism" is employed towards the same end as "sentimentalism": that is, of convert and transcribing images into representations that are most comfortable and that have certain ontological ties/some characteristic to be easily assimilate-able to the culture to which they are presented. Do things not get lost to the same degree in a "realist" vs. "sentimentalist" ideology in ethnographic work? In other words, "realism" is not totally void, totally detached from the cultural imagination of the ethnography conducting the studying; there is still a creative aspect to the realist doctrine, one which may have an effect to the same degree as the sentimentalist doctrine.