I'm sure this entire discussion may be old hat to many folks, but I found the Errol Morris blog discussion of photographic manipulation terrific, the sort of thing worth assigning to a class and thought-provoking in its own right. One sentence that stuck out for me:
But doctored photographs are the least of our worries. If you want to trick someone with a photograph, there are lots of easy ways to do it. You don’t need Photoshop. You don’t need sophisticated digital photo-manipulation. You don’t need a computer. All you need to do is change the caption.
In some ways, this is a restatement - in another medium - of the central problematic of documentary meaning, namely that actuality footage carries very little of its own meaning but instead relies on montage and the soundtrack to supplement the "pure plenitude" of indexical representation.
The interview-discussion also touches on the problem of lossy images. Morris asserts that lossiness is the guarantor of authenticity, citing the reality effect of low-resolution documentary and pseudodocumentary filmmaking. However, I'd suggest there's a perceptual difference between low-resolution and lossiness. If we as viewers (and I know I'm being impressionistic here) feel the image failed to capture fully the profilmic (because of distance, obstruction, atmospheric conditions, film stock, etc.), we have a different disposition to the "photographic truth" than if we feel that captured information has been compressed out. This may be a generational difference, though, and certainly digital aesthetics can fetishize the apparatus as much as analog counterparts. But I'd like to suggest that we read the cues of low-res and compression loss differently.
One thing that Morris does not discuss is the political economy of the news industry, which has to be changing its understanding its ability to know photographic provenance.