Eighteenth International Conference on New Directions in the Humanities

Abstract

Digital “imaging” and presentation of fine art has long been used in the service of connoisseurship and art historical studies. In all such cases in the “digital humanities,” the final decision or interpretation has been by art scholars. Recent advances in computer vision and image analysis have been applied to a number of problems in the history and interpretation of art. In some cases, these computer methods are more perceptive and accurate than even the most tutored connoisseur. For instance, computer methods for estimating the lighting within a realist tableau are provably more accurate than any human observers, and have be used to answer a number of problems in authentication and in inferring artists’ working methods. Much as Giovanni Morelli (1816-91) introduced a “scientific” approach to connoisseurship and attribution studies, so too today’s computer methods can expand and enhance traditional art historical methods. Just as computer methods have revolutionized nearly all scientific and humanistic studies, so too it is beginning to revolutionize the methods of connoisseurs for close, rigorous visual analysis of fine art. This discussion ends with a number of research directions and opportunities in a new discipline of computer-assisted connoisseurship based on the forthcoming “Pixels and Paintings: Foundations of Computer-assisted Connoisseurship” (Wiley Publishers).