In film, the value of acting depends on the abilities of others, such as cinematographers and editors; to probe deeper into the nature of acting’s contribution we need to see performances relative to images and the sound track. First, our experience of a performance as a part of the film implies that acting causally affects the image and sound track. Computer-generated images, which are based on motion or performance capturing, illustrate even more pointedly the relation of acting to images. Andy Serkis’ performance as Gollum in the trilogy of The Lord of the Rings is a case in point. Even though we look at a fantasy creature with an outward appearance created from scratch at a computer, not unlike the object of a painting, Gollum’s quirky movements and gestures and his uncanny postures all rely on Serkis’ acting technique. The idea that photographic pictures are fundamentally different from other pictures has been highly contested, and it is clear that moving pictures share with other pictures many aesthetic properties. A painter and a film director who wish to depict a certain posture will face similar decisions concerning, for example, how to frame, compose, and light the figure, and the painter may instruct a model to hold a certain pose, not unlike the director who asks the actor to behave in a certain manner. We are looking at images in both cases, and they both rely on the techniques of a painter and a director; yet one of them also relies on acting technique as the actor moves within the picture frame. Kendall L. Walton’s concept of transparency may shed light on the kind of causal relation that is possibly at work in images of a performance (Walton 1984). Elaborating upon an idea first expressed by André Bazin, Walton points out that photographic pictures will show what happened in front of the camera and thereby manifest a counterfactual dependency between the content of the picture and the objects in front of the camera, whereas a painting depends on what the painter believes he or she has observed and what he or she intends to be the pictorial content. Second, acting implies that someone plays a role as part of a fictional narrative. It may be defined, as James Naremore has put it, as “a special type of theatrical performance in which the persons held up for show have become agents in a narrative” (Naremore 1988: 23). Other kinds of performances, such as singing on stage, might also entail a narrative function, but acting entails that the person “held up for show” tries to enact the role of agent in a narrative. What acting shares with other kinds of performances is the evaluation by the audience: the question is whether the claim to center stage, as it were, is justified by the performer’s abilities. Third, the performance is carried out for the purpose of a film audience. Public service television is in some countries used to broadcast successful stage performances at large theaters, but this is an example of performance as an object of distribution in an audio-visual medium. In film performances, we have distinct expectations concerning the relation of acting to film technology. Comparative studies of scenes which are based on the same play and use cinematography in similar ways show that a performance can be more or less “calibrated” to framing and camera movements (Jacobs 1998), and small nuances of performance can alter a remake (of an earlier film) in thematically important ways (McDonald 2004: 27-32). Acting is an integral and distinct part of the work in question. Historically, a prevalent use of film technology has allowed for counterfactual dependency on acting, but only empirical study can reveal the extent to which acting accounts for valuable properties such as character expressiveness. Nevertheless, the pictorially inclined may point out that our experience of a performance is affected, for example, by the use of editing, and that acting therefore is a kind of raw material for the editor, subject analogously to the so-called Kuleshov effect. Although the premise is true, it does not follow that editing constructs the expressive content of acting, and the Kuleshov effect should not be trusted as proof in this regard. Lev Kuleshov, an editor and director in post-Revolutionary Russia, did an experiment in which a close-up of the actor Ivan Mozhukin was intercut with various objects. The reactions of the spectators who viewed the intercut shots ostensibly showed that it was the editing, rather than Mozhukin’s acting, that created meaning. The footage used in the Mozhukin experiment has been lost, but two similar experiments of Kuleshov indicate that he used this preexisting footage for pedagogical purposes, to illustrate cutting on eye-lines, rather than to prove a theoretical point (Tsivian et al. 1996: 359). We should be skeptical of any references to the Kuleshov experiment with regard to acting and expressiveness. First, it is often assumed that Mozhukin took part in what could have been a semi-scientific experiment, when in fact, by the time the “experiment” was conducted, he had fled the country, due to the Revolution (see Albéra 1995: 76). Second, it is often assumed that Mozhukin was inexpressive and that it therefore was the editing that created the illusion that he was emoting. Yet it is rather implausible that Kuleshov would have picked an inexpressive performance by Mozhukin, even if he could find one. An inexpressive glance in a close-up does not cut easily with shots of other objects, since the spectator has no reason to ask for an offscreen cause of an expression. We are better off with a more recent model of how acting and editing may work together, supplied by Noël Carroll in his theory of point-of-view editing. Carroll implies that facial expressions are somewhat ambiguous. To communicate emotions in a precise manner in cinema, the film structure known as point-of-view editing has been developed (Carroll 1993).