Chapter 7 case study
Digital manipulation: OK if revealed—and restrained?
Eric Kee and Hany Farid, (Nov. 28, 2011). “A perceptual metric for photo retouching,” Proceedings of the National Academy of Sciences,” available at: http://www.pnas.org/content/early/2011/11/21/1110747108.full.pdf+html. Accessed Feb. 8, 2011.
Representative examples of photo editing [before (top row) and after (bottom row)]. Images from: (First) http://www.flickr.com/photos/manekineko/4526217217/; (Second) http://merengala.blogspot.com/2010/12/sin-photoshop-y-con-photoshop.html; (Third) before image http://www.flickr.com/photos/carreon/3509720855/in/set-72157603947306951, after image http://www.photoretoucherpro.com/; (Fourth) http://www.flickr.com/photos/manekineko/4645113101/; (Fifth) before image http://models.com/v-magazine/v-size-2.html, after image http://www.befter.net/user/Scarione/beft/fat-model-befteredited-by-thescarione/. Sites accessed October 26, 2011.
Each of these photos on the bottom row has been digitally manipulated. And it shouldn’t take you long to see that there’s more digital manipulation in each bottom row photo as you go from left to right. In fact, these photos illustrate a proposed 1-to-5 ranking scale that mass communicators, their audiences and their photo subjects could use to specify the amount of digital manipulation they’re using, according to researcher who developed the scale, Dartmouth University computer scientist Hany Farid. He surveyed 390 people on 468 sets of before-and-after manipulation sets of images like those above and found they were able to distinguish among five levels of manipulation 81 percent of the time. Then he fed their ratings into a computer that correlated them with measurements of the subjects’ facial and body dimensions and texture. The result is a computer program that anyone could use to accurately predict whether most people would rate a set of digitally manipulated photos as a 1, 2, 3, 4 or 5. “Give me a 1 or 2,” models could say to photo editors, “but don’t make me a 5.” Teenage girls could look at a photo of a model that’s labeled a 5 and know that it would be impossible for them to ever look that skinny.
The ethical question is: Is revealing the degree of digital manipulation you've used enough to make it OK, especially when "the greatest number" is doing it surreptitiously all the time anyway? Would Farid’s system, if universally adopted, make digital manipulation ethical? Is transparency the new truth? Well, one might respond, has nutrition labeling on food cans and jars made people healthy? That’s different, you say, although the American Medical Association has issued strong warnings that digitally manipulated photos of models are causing body image problems–psychological and physical—for millions of young people and adults.
Alternatively, does it seem that how the manipulated image is used is a key component of how ethical the manipulation is? For instance, if image #5 above were used in deceptive advertising for a product, that's about as unethical as you can get. If it's used simply to make a woman's image look better in a feature story, it's still unethical, but would it rate a "3" rather than a "5"?
If advertisers and magazine editors clearly labeled every digitally manipulated photo on the 1-5 scale, would that satisfy Kant’s categorical imperatives for The Truth and dignity? Would Rawls see that as protecting the most vulnerable? Would Aristotle call that a golden mean that would benefit advertisers, the audience and models—better than no digital manipulation at all? And would Mill call such a system the greatest good for the greatest number, the audience?
Finally, could such transparency allow photojournalists to use digital manipulation ethically? Would it have been ethical if photo editors had written into the caption that this Kent State photo had been manipulated? (Actually, an editor airbrushed out of this 1970 pre-digital photo a fence post that appeared to be protruding from the head of the kneeling woman.) And look below to the L.A. Times Iraq photo, which is a composite of two other photos. What if its caption said it had been digitally manipulated?
Where, on the 1-5 scale, would you rate those levels of manipulation? Someone might say the degree of manipulation in the Kent State photo is a "5," because it totally changes the impact of the image; i.e., it's definitely more "noticeable" than the manipulation in the L.A. Times composite. But could you still say the Kent State photo wasn't at all an unethical use of this highly manipulated image, because it showed the very moving scene without a distraction--and because the photo wouldn't have needed manipulation if taken from a slightly different angle? The L.A. Times image, by contrast was a deliberate combination of two different portrayals of the same scene—a deliberate violation of L.A. Times policy. But what if the photographer and L.A. Times editors had presented in the caption as a "photo representation" and labeled it a “3” under Farid’s scale? Would that have maintained their credibility?
L.A. Times composite photo: http://victorianvisualculture.wordpress.com/2010/10/15/275/
· For background on Prof. Farid’s research on rating digital manipulation, see:
o A New York Times article explaining it
o Farid’s full paper in the Proceedings of the National Academy of Sciences