This chapter deals with an intense and mediated form of the encounter that could be called the “visual study.” What is a visual study? The form gathers together a group of filmic enterprises that are at once numerous, diverse and contemporary. It involves a direct, face-to-face encounter between a “readymade” image and a figurative project that comments upon it—or, to put it another way, the study of an image using the very means of the image itself. What are the techniques, forms, stakes, constraints and models, what is the corpus corresponding to such an investigation? Straight away, five problems arise.

• 1. The primordial question that every visual study raises and reworks can be formulated thus: “What can an image do?” Can an image explain, criticize, argue, demonstrate, reach a conclusion—and if so, how? Does it suffice, as Godard has believed for many years, to simply place one image after another? Does the comparison clinch an argument? And why is it not enough to have a single image, that is, an absent or missing image?

• 2. As compared with the written study of an image, is a visual study capable of analytical initiatives? Or to put this differently, is visual study another version of literary ekphrasis (description), or rather does it offer new analytical models?

• 3. To these instrumental questions, another of a more theoretical type can be added: As every analysis assumes or presumes a particular conception of history, how does a visual study reflect its own relation to history?

• 4. The visual study refers to two non-literary, non-linguistic models: the musical study, that is, the exploration in music of a technical formula; and the pictorial study, at once the copy of a work which defines the “school” in which the artist is an apprentice, and an analytical departure from it. The serial studies of Ingres, Delacroix, Manet, and Velásquez by Picasso—their hermeneutic power analyzed by Hubert Damisch in response to Michel Foucault's account of Las Meninas—offer a major example. What is the relation of the cinematic study to these two prior forms?

• 5. Finally, a problem of method: Can we really analyze a visual study? Isn't it enough to simply describe it?