London Office: +44 (0) 7504 878 488
New York Office: +1 (917) 519 8969


The art market's (mis)trust in science


Here at Hephaestus, the London and New York-based art authentication service, we have seen many forgeries accompanied by scientific reports, replete with incorrect, unscientific conclusions. Regrettably, instances of unscientifically applied science have created a pervasive notion that scientific methods are inconclusive and inaccurate. We are troubled by the resulting reluctance to use scientific methods, especially as they are often able to reveal forgeries at the highest levels of the art world. As depicted in the recent Netflix documentary Made You Look, it was scientific analysis that revealed that the expressionist paintings sold by the Knoedler Gallery with the blessing of the world’s experts were forgeries produced by Pei Shen Qian in his Queens, NY garage. And more recently, it was science that shone a light on a series of Old Master forgeries exhibited in major museums including the National Gallery, the Metropolitan Museum, and the Kunsthistorisches Museum.

Where and how is science being misapplied?

From a material perspective, paintings are complex objects: they are a blend of organic and inorganic materials, combined and layered in various ways. Over time pigments fade, varnishes discolour, and the painting may undergo restoration, adding new materials into the mix. Methods of sampling, detecting, and identifying these materials are well established; however, it is important to note that each technique has its own level of precision that can lead to surprising differences in outcome. To illustrate, imagine a photograph of a person wearing a purple jacket printed in a newspaper. From a reader’s perspective the colour is objectively purple, but this changes when we view the newsprint up-close, using a more precise tool, say, a magnifying glass. Provided the printer used the 4-colour process, you will see spots of different colours that blur and mix together into the colour we perceive as solid purple.

Similarly, different scientific techniques have different resolving powers. In the case of pigment analysis, one technique might simply detect the presence of lead in a white section of the painting, suggesting that the artist used lead white pigment. However, if you were to use a more precise tool or technique, you may identify trace elements that suggest that the lead white is of modern origin and not produced using the traditional “stack” process. Neither result is wrong per se, but the precision of an instrument can make a world of difference. Drawing conclusions from a simple list of pigments is at best imprecise and unreliable, and at worst, misleading and inconclusive. Unfortunately, many of the reports we have seen linked to inauthentic works used analytical techniques that extend these kinds of fallacies.

A deep understanding of the chronology of pigments, when they were invented and became available, has been the main scientific weapon against the forger for decades. Although these tests have become increasingly sophisticated, they are no silver bullet: they should form one part of a carefully sequenced suite of tests. The first test should eliminate the majority of forgeries, with each successive technique bringing us closer and closer to certainty.

While connoisseurs study the surface of an artwork, looking for characteristic patterns and tell-tale signs that an artwork is genuine, technical art historians look for patterns below the surface, establishing whether the materials and technique present in the painting are consistent with other works by the artist. In cases where brazen anachronisms cannot be found, this comparative step using bonafide works as a standard, is how evidence for or against an attribution is built. For instance, you could compare the pigments the artist used in similar paintings, find similarities in artistic process across x-radiographs, or identify the characteristic type of canvas the artist liked to use. The more data available, the more likely significant patterns can be established.

The art market appears to be reluctant to accept the lengths that forgers go to create pictures that avoid detection. For example, it is not unusual to hear ‘a forger would never have thought of doing that’. Yet, a sophisticated, enterprising forger has access to the same reference material that a scientist or an expert does. Perhaps we underestimate forgers because the majority of those caught in recent years left glaringly obvious technical anachronisms in their works: for instance, Wolfgang Beltracchi used titanium white in paintings dated to before the development of the pigment. Forgers less prolific than Beltracchi could be far more sophisticated. If we take an x-ray of an artwork and find a painting underneath that matches a sketch in the artist’s sketchbook, is this irrefutable evidence that a painting is authentic? Or could the forger, knowing the artist’s habits and techniques, have left a trail of breadcrumbs for us to find and follow? The spirit of the scientific method since Descartes has always been one of rational skepticism. Sufficient unto the day are the scandals thereof; and those of us working to eliminate forgery in the art markets should maintain a healthy dose of skepticism, especially when purchasing objects we love.