In standard histories the forces which put the star system in place are reduced to the play of personal initiative on the one hand and a reified notion of the public desire on the other. The star system is not simply the creation of one person or even one company; nor is the desire for movie stars something that arose unsolicited.
The emergence of the star system can perhaps best be seen as the emergence of a knowledge and analysed in these terms. Before 1909 virtually none of the players' names were known to the public, but by 1912 most of them had been 'discovered'. 1 It is clear from this example that the 'picture personality' was the result of a particular production and circulation of knowledge. Studio publicity departments, films and fan magazines produced and promulgated this knowledge. In this paper I want to examine the rules by which this knowledge was produced and the various transformations these rules underwent.
The emergence of the star system involved a strict regulation of the type of knowledge produced about the actor. I will argue that the development of this system was effected through three significant transformations in this regard. These can be listed in the order of their appearance: (1) the discourse on acting, (2) the picture personality and (3) the star.
Before discussing these three stages individually, let me note that the appearance of the second, the picture personality, did not mean the disappearance of the first, the discourse on acting (or for that matter, the third the disappearance of the second). This transformation can best be characterised as a progressive overlaying of discourses and knowledges about a particular site-the actor.
It is perhaps misleading to say that this site was the actor as if this site was constituted in itself. Before 1907 there was no discourse on the film actor. Textual productivity was focused elsewhere, for the most part on the apparatus itself, on its magical abilities and its capacity to reproduce