No. The claim that could unambiguously characterize became urgent when a cache of disputed Pollocks turned up. Jones-Smith, Mathur and Krauss analyzed paintings by Pollock and by others and concluded, in their own words, definitively: fractal criteria provide no information about artistic authenticity.
The problem is not that Pollock’s paintings lack fractal structure. It is that the measure does not separate him from anyone else, so a number that both a Pollock and a non-Pollock can produce cannot settle which you are holding. Working on this led the authors somewhere more general, into what the statistics of a can and cannot tell you about a shape.
For some years it was claimed that fractal analysis could be applied to characterize works of art unambiguously, with Jackson Pollock’s drip paintings as the central case. The question stopped being academic when a cache of disputed Pollock paintings came to light, and a method that promised to settle attribution by measurement suddenly had money and reputations attached to it.
Jones-Smith, Mathur and Krauss tested it the obvious way, by analyzing paintings by Pollock and by others. Their conclusion is unusually blunt for a physics paper. They show definitively, they write, that fractal criteria provide no information about artistic authenticity. The measure may describe the image, but it does not identify the hand that made it.
The investigation pushed them into a more general question about the tool itself. They show that the statistics of the covering staircase, which is closely related to the box-counting staircase, provide a way to characterize geometry and to distinguish fractals from Euclidean objects. That is a result about fractal analysis rather than about Pollock, and it is the part most likely to be cited in fifty years.
They also take up a problem first investigated by Muzy: what happens when you composite two fractals. The composite is not generally , and it exhibits complex multifractal scaling in the small distance limit. One caution about scope. This paper establishes that one proposed authentication criterion fails, and it does not offer a replacement, so the practical question of how to authenticate a disputed drip painting remains where it was.
wat. A Pollock has fractals in it, and so does a painting that is not a Pollock.
wat wat
Authentication decides who owns what, and for a while a physics measurement looked like it could replace connoisseurship. A test that cannot distinguish the artist from a non-artist is worse than no test, because it carries unearned authority.
How do we know?
The authors applied fractal analysis to paintings by Pollock and by other artists and compared the results, rather than testing whether Pollock’s work alone shows fractal scaling.
What is still unknown
A physics paper arguing against a specific method, not a general account of how drip paintings should be authenticated. It shows the criterion fails and offers nothing in its place.
Learn something?
Where to next
How do you prove a painting dated 1866 was made after 1950?painting
How long is the coast of Britain?fractals