Introduction

Gance's desire for factual credence is highly visible in NAPOLÉON, but his numerous historical citations have often been seen as ‘a misguided attempt to invoke the mantle of authority’ (Abel 1984: 432). Indeed, Gance later thought his fundamental ‘error’ in NAPOLÉON was ‘the abuse of documentation’: ‘I lost popular appeal in trying to be accurate and searching for authentic detail. The power of LA ROUE and J’ACCUSE resides in the absence of the authenticated’ (1928d). Yet these ‘unauthenticated’ elements of NAPOLÉON were often the first targets of those wishing to reduce the film's immense length. This was particularly the case with the fictional Fleuri family: the scenes involving Tristan, Violine, and Marcellin have suffered the most from deliberate excision and derisive critical consideration. After attending an early preview screening, Jean Arroy urged Gance to cut the majority of material involving the Fleuris (Brownlow [1983] 2004: 135). Gance followed this advice in preparing his film for the Opéra but later reinstated their scenes in the longer Apollo prints. Given the choice between these two divergent editions (and numerous other incomplete prints), modern restorers have often disregarded the Fleuris during the process of reconstruction.

Despite the best efforts of editors and Bonaparte's stubborn blindness to them, the Fleuris remain embedded in the textual substance of NAPOLÉON – perennially struggling to make their presence known.