Seeing double: Leonardo’s Mona Lisa twin

Abstract

The conventional scientific tests employed in authenticating paintings are useful in excluding a work if, for a particular artist, incorrect parameters are discovered. For example, the date may be wrong (e.g., radio carbon dating), a pigment may be wrong (e.g., modern formulation), or an implausible underpainting or sketch may be revealed (e.g., X-ray or IR image). In contradistinction, an original approach is proposed to identify individual “fingerprints” for particular artists’ hands (brushwork). It entails extracting luminosity histogram statistics of a painting in order to quantify its sfumato/chiaroscuro properties for either entire compositions or particular features (e.g., eyes, noses, or lips.) It is proposed that a work may be associated with a particular artist, rather than be excluded from the generally accepted body of work. Paintings by Leonardo and Rembrandt as well as pertinent copies and forgeries are employed as test cases. Luminosity histogram statistics for several contemporary paintings are also included to enlarge the data library. In order to illustrate the utility of this approach in characterizing and identifying an artist’s technique, particular attention was addressed to the issue of the possibility of two individual Mona Lisa portraits by Leonardo. This focused on an analysis of the properties of a painting formerly known as the Pulitzer/Isleworth Portrait and, subsequently, as the “Earlier Mona Lisa, EML” by Leonardo da Vinci, after protracted scientific, historic, aesthetic, and geometric investigations.