I remember reading, a few months ago, about a study that had compared paintings by artists of the abstract expressionist school to paintings made by untrained children and animals. In the study, subjects were shown a pair of paintings, one from each category, and asked which they thought better. The authors found that in 67% of cases the viewer thought the painting by the professional painter was better.
This meant, of course, that in roughly one-third of cases the viewer thought that the child or the animal had made a better painting, which, one might think, would be pretty damning for the pros, but the authors of the paper put the best construction on their results, noting that subjects “chose the professional work significantly more often than would be predicted by chance”, and then concluding with the pleasant thought that “The world of abstract art is more accessible than people realize.”
Man or monkey?
Recently an amusing little paper appeared in the arXiv database (the database where new papers in physics are first posted) that finds the cloud in this silver lining. The author, Mikhail Simkin, had the thought to compare the 67% success rate with success rates in other contexts. He notes, for instance, a study on ‘just perceptible differences’ that found that only 72% of people could correctly identify a 100 g weight as being heavier than a 96 g weight when held in the hand; given that even fewer could distinguish the two groups of painters, does this suggest that the ‘artistic weight’, as it were, of the professional’s paintings is only ‘just perceptibly different’ from that of the monkey’s?
Or, again, he compares the results to the Elo rating system for chess players, and points out that players separated by one rank have skill levels more different than that shown between the two groups in the painting study. He concludes that
Since abstract expressionist wins over a monkey only in 67% of cases, the difference in their artistic Elo ratings is less than 200. This means that they either belong to the same category with apes or are just one category above. If we class a gorilla as a novice, abstract art grandmasters are at best class D amateurs.
Well, granted that there are obvious problems with drawing these analogies between art, weights, and chess, this nonetheless strikes me as a creative way to explore the real significance of that 67% result. I confess that there is no love lost between abstract art and me, and these findings, both here and, perhaps especially, in the original paper, put a big ol’ smile on my face and a song in my heart.
By the way, Mikhail Simkin has had a bee in his bonnet over abstract art for a while. He runs an online quiz in which the task is to correctly distinguish real abstract paintings, by acknowledged masters, from ‘ridiculous fakes’, by Simkin himself. Although the test is biased in favour of the real paintings (because people may have seen some of them before), he has found that people correctly distinguish wheat from chaff only about 66% of the time. I took the quiz myself, and scored just 58% (7/12). I would have done worse were it not for some clues, unrelated to the compositions themselves, that tipped me off.