ABSTRACT
The title of this chapter is borrowed, in part, from Raymond Bellour's segmentation of Gigi (1976: 331-53). My aim here is to extend Bellour's discussion of a Hollywood musical to two key texts of Bombay Cinema. The first is Baiju Bawra (1952), which I use to further exemplify the significance of the "song text" in Bombay Cinema; the second, Amar Akbar Anthony (1977), I use to explore the principle of symmetrical narratives that have something of the rhythm of classic Hollywood cinema. Bellour's segmentation of Gigi is based on Christian Metz's grande syntagmatique, a term he used to refer to the grammar, the deep structure, or the langue of cinema. According to Metz, the grande syntagmatique is the series of autonomous "syntagms" (Bellour's "segments") that make up the picture track. To grasp this, one suspects Metz had to read cinema as a language of shot sequences which, in Bellour's words, is "a sort of theoretical operator [that] actualizes the concrete possibility of a semiotics of cinema" (332). The procedure came under strong criticism, notably from the Italian semiotician Emilio Garroni, on the grounds that any procedure based on a "division of the filmic text into autonomous segments" would deny filmic message its "pluricodicity" (Henderson 1980: 151). Metz, of course, never offered segmentation as a "total analysis of the filmic message," seeing it rather as a 'subcode" of cinema (Metz 1974: 189) that was not meant to elide the semiotic complexities of the filmic image. Further defense for segmentation has to be mounted here because Bellour himself has declaimed the procedure on the grounds that segmentation (of the kind I attempt) relies on the analyst's use of segments as discrete citations of the film. Since written texts alone can be cited in print, the citation of filmic segments can be no more than a literary paraphrase.
But this seductive body [the filmic text] is an elusive body; it cannot really be quoted nor grasped. It is polysemous as well, in an excessive way, and its matter, moulded by iconicity and analogy, pushes language into check. This irreducibility of the filmic substance, which fascinates and stimulates (as do all such elusive objects), serves to limit analysis: the readings of films have been unable to produce the equivalence brought out in readings of "Les Chats" or in S/Z. This does not simply result from the analysts' lack of genius, but primarily from the exceptional resistance put up by the analytic material. (Bellour 1985: 54)
