A View from Emerging Technology from the arXiv
Measuring The Quality Of Abstract Art
Abstract artists are only 4 per cent better than child artists, according to a controversial new way of evaluating paintings
Here’s a bit of mischief from Mikhail Simkin at the the University of California, Los Angeles.
Simkin has a made a name for himself evaluating the relative performance of various groups and individuals. On this blog, we’ve looked at his work on the performance of congress, physicists and even World War I flying aces.
Today, he turns his attention to abstract artists. For some time now, Simkin has a run an online quiz in which he asks people to label abstract pictures either real art or fake. It’s fun–give it a go.
One average, people answer correctly about 66 per cent of the time, which is significantly better than chance.
Various people have interpreted this result (and others like it) as a challenge to the common claim that abstract art by well-know artists is indistinguishable from art created by children or animals.
Today, Simkin uses this 66 per cent figure as a way of evaluating the work of well known artists. In particular, he asks how much better these professional artists are than children.
First, he points out the results of another well known experiment in which people are asked to evaluate weights by picking them up. As the weights become more similar, it is harder to tell which is heavier. In fact, people will say that a 100g weight is heavier than a 96g weight only 72 per cent of the time.
“This means that there is less perceptible difference between an abstractionist and child/animal than between 100 and 96g,” says Simkin.
So on this basis, if you were to allocate artistic ‘weight’ to artists and gave an abstract artist 100g, you would have to give a child or animal 96g. In other words, there is only a 4 per cent difference between them.
That’s not much!
Simkin goes on to say this is equivalent in chess to the difference between a novice and the next ranking up, a D-class amateur.
If ever a paper was designed to trigger controversy, this is it. If it catches the public imagination, just sit back and watch the fists fly.
That’ll be fun. But let’s hope he never gets round to applying the same logic to physicists’ ability to evaluate dodgy papers on the arxiv.
Ref: arxiv.org/abs/1106.1915: Abstract Art Grandmasters Score Like Class D Amateurs
Become an MIT Technology Review Insider for in-depth analysis and unparalleled perspective.Subscribe today