The use of computational methods within the field of art history is usually restricted to digitization and archiving tools, allowing the preservation of digital collections and their online access to the world. However, over the past decade, many projects have shown the possibility to use computer vision and machine learning methods to automate tasks on large corpuses of paintings and to better outline recurrent patterns. Unfortunately, these works are often overlooked by art historians who have limited access to these complex computational methods and might sometimes perceive them as an intrusion of the scientific methodology to their field. However, computer science has much more to offer than automated data processing and should not only be seen as a new methodological approach, but also as a door to new perceptions of artworks and artistic practice. Through a research project on computational and historical analysis of hands and gestures in early modern time, where different technologies are used to better grasp the many dimensions of the painted hand, the idea of the machine as a new agent in the relationship between paintings and art historians is envisioned. This exploration has many facets, one of which is the creation of an interactive tool to facilitate research on a large corpus of painted hands, directly using the hand gesture of the user as a query. Fostering a new understanding of painted hand gestures through computer vision, animations and enactment, the goal is to propose, through the eye of the machine, a new way of seeing.