Novel to Film: An Introduction to the Theory of Adaptation

REDEFINING ISSUES AND A NEW APPROACH

The Centrality of Narrative

The more one considers the phenomenon of adaptation of novel into film the whole history of the reliance on the novel as source material for the fiction film--the more one is drawn to consider the central importance of

____________________

29 Ibid. 224.

30 Ibid. 226.

31 Dudley Andrew, 'The Well-Worn Muse: Adaptation in Film History and Theory' , in Syndy Conger and Janice R. Welsch (eds.), Narrative Strategies ( West Illinois University Press: Macomb, Ill., 1980), 10.

32 Michael Klein and Gillian Parker (eds.), The English Novel and the Movies ( Frederick Ungar Publishing: New York, 1981), 9-10.
-11-

narrative to both. Whatever the cinema's sources--as an invention, as a leisure pursuit, or as a means of expression--and whatever uncertainties about its development attend its earliest years, its huge and durable popularity is owed to what it most obviously shares with the novel. That is, its capacity for narrative. By the time of Edwin Porter The Great Train Robbery (1903), in which scenes set in different locations are spliced together to tell a story, the cinema's future as a narrative art was settled, and no subsequent development of its techniques has threatened the supremacy of that function.

Christian Metz, discussing film narrativity, writes: 'Film tells us continuous stories; it "says" things that could be conveyed also in the language of words; yet it says them differently. There is a reason for the possibility as well as for the necessity of adaptations.'33 He goes on to consider the 'demand' for the feature-length fiction film, 'which was only one of the many conceivable genres',34 but which has dominated film production. 'The basic formula, which has never changed, is the one that consists in making a large continuous unit that tells a story and calling it a "movie". "Going to the movies" is going to see this type of story.'35 Whatever other uses the cinema might have found, it is, as Metz suggests, as a story-teller that it found its greatest power and its largest audience. Its embourgeoisement inevitably led it away from trick shows, the recording of music halls acts and the like, towards that narrative representationalism which had reached a peak in the classic nineteenth-century novel. If film did not grow out of the latter, it grew towards it; and what novels and films most strikingly have in common is the potential and propensity for narrative. And narrative, at certain levels, is undeniably not only the chief factor novels and the films based on them have in common but is the chief transferable element.
If one describes a narrative as a series of events, causally linked, involving a continuing set of characters which influence and are influenced by the course of events, one realizes that such a description might apply equally to a narrative displayed in a literary text and to one in a filmic text. Nevertheless, much of the dissatisfaction which accompanies the writing about films adapted from novels tends to spring from perceptions of 'tampering' with the original narrative. Words like 'tampering' and 'interference', and even 'violation', give the whole process an air of deeply sinister molestation, perhaps springing from the viewer's thwarted expectations relating to both character and event. Such dissatisfactions resonate with a complex set of misapprehensions about the workings of narrative in the two media, about the irreducible differences between the two, and from a failure to distinguish what can from what cannot be transferred.

____________________

33 Christian Metz, Film Language: A Semiotics of the Cinema, trans. Michael Taylor ( Oxford University Press: New York, 1974), 44.

34 Ibid.

35 Ibid. 45.
-12-

To begin with the last point: there is a distinction to be made between what may be transferred from one narrative medium to another and what necessarily requires adaptation proper. Throughout the rest of this study, 'transfer' will be used to denote the process whereby certain narrative elements of novels are revealed as amenable to display in film, whereas the widely used term 'adaptation' will refer to the processes by which other novelistic elements must find quite different equivalences in the film medium, when such equivalences are sought or are available at all.

Narrative Functions: Novel and Film

Roland Barthes has defined the essence of a narrative function as 'the seed that it sows in the narrative, planting an element that will come to fruition later--either on the same [narrative] level or elsewhere, on another level',36 going on to claim that, 'A narrative is never made up of anything other than functions: in differing degrees, everything in it signifies.' He distinguishes two main groups of narrative functions: distributional and integrational and, though he is not concerned with cinema in this discussion, this distinction is valuable in sorting out what may be transferred (i.e. from novel to film) from that which may only be adapted. To distributional functions, Barthes gives the name of functions proper ; integrational functions he calls indices. The former refer to actions and events; they are 'horizontal' in nature, and they are strung together linearly throughout the text; they have to do with 'operations'; they refer to a functionality of doing. Indices denotes a 'more or less diffuse concept which is nevertheless necessary to the meaning of the story'.37 This concept embraces, for instance, psychological information relating to characters, data regarding their identity, notations of atmosphere and representations of place. Indices are 'vertical' in nature, influencing our reading of narrative in a pervasive rather than a linear way; they do not refer to operations but to a functionality of being.

The most important kinds of transfer possible from novel to film are located in the category of functions proper, rather than that of indices, though some elements of the latter will also be seen to be (partly) transferable. Barthes further subdivides functions to include cardinal functions (or nuclei) and catalysers. Cardinal functions are the 'hinge-points' of narrative: that is, the actions they refer to open up alternatives of consequence to the development of the story; they create 'risky' moments in the narrative and it is crucial to narrativity ('the processes through which the reader . . . constructs the meaning of the text'38) that the reader recognizes the possibility of

____________________

36 Roland Barthes, 'Introduction to the Structural Analysis of Narratives' ( 1966), in Image-Music-Text, trans. Stephen Heath ( Fontana/Collins: Glasgow, 1977). 89.

37 Ibid. 92.

38 Orr, 'The Discourse on Adaptation' , 73.
-13-

such alternative consequences. The linking together of cardinal functions provides the irreducible bare bones of the narrative, and this linking, this 'tie between two cardinal functions, is invested with a double functionality, at once chronological and logical'.39 These cardinal functions, or, in Seymour Chatman's terms, kernels ('narrative moments that give rise to cruxes in the direction taken by events'40), are, as I shall show, transferable: when a major cardinal function is deleted or altered in the film version of a novel (e.g. to provide a happy rather than a sombre ending), this is apt to occasion critical outrage and popular disaffection. The film-maker bent on 'faithful' adaptation must, as a basis for such an enterprise, seek to preserve the major cardinal functions.

However, even if the latter are preserved in the filming process, they can be 'deformed' by varying the catalysers which surround them. Catalysers (in Chatman's term, satellites) work in ways which are complementary to and supportive of the cardinal functions. They denote small actions (e.g. the laying of the table for a meal which may in turn give rise to action of cardinal importance to the story); their role is to root the cardinal functions in a particular kind of reality, to enrich the texture of those functions: 'their functionality is attenuated, unilateral, parasitic: it is a question of a purely chronological functionality', 41 in Barthes's words. Unlike the 'risky moments' created by cardinal functions, the catalysers 'lay out areas of safety, rests, luxuries';42 they account for the moment-to-moment minutiae of narrative.
In so far as these functions, whether cardinal or catalysing, are not dependent on language, in the sense that they denote aspects of story content (actions and happenings) which may be displayed verbally or audio-visually, they are directly transferable from one medium to the other. Among the integrational functions, which Barthes subdivides into indices proper and informants, only the latter may be directly transferred. Whereas the former relate to concepts such as character and atmosphere, are more diffuse than the functions proper, and are therefore more broadly open to adaptation rather than to the comparative directness of transfer, informants 'are pure data with immediate signification'.43 They include 'ready-made knowledge' such as the names, ages, and professions of characters, certain details of the physical setting, and, in these senses and in their own ways, share the authenticating and individuating functions performed in other respects by catalysers, and they are often amenable to transfer from one medium to another. What Barthes designates as cardinal functions and catalysers consti-

____________________

39 Barthes, 'Introduction to the Structural Analysis of Narratives' , 94.

40 Seymour Chatman, Story and Discourse: Narrative Structure in Fiction and Film ( Cornell University Press: Ithaca, NY, 1978), 53.

41 Barthes, 'Introduction to the Structural Analysis of Narratives' , 94.

42 Ibid. 95.

43 Ibid. 96.
-14-

tutes the formal content of narrative which may be considered independently of what Chatman calls 'its manifesting substance' (e.g. novel or film), and informants, in their objective name-ability, help to embed this formal content in a realized world, giving specificity to its abstraction. Perhaps informants may be seen as a first, small step towards mimesis in novel and film, the full mimetic process relying heavily on the functioning of the indices proper, to which I shall return shortly.

I should note at this point that Barthes has subsequently modified the structural taxonomy set up here with his designation of the five narrative codes which structure all classical narrative in S/Z, his reading of Balzac Sarrasine. For my purposes, the earlier distinction between distributional and integrational functions, with the metaphors implied in their characterization, provides a more accessible and usable taxonomy in establishing what may be transferred from a long, complex work in one medium to a long, complex work in another. Barthes was not, of course, concerned with cinematic narrative when he wrote his 'Structural Analysis' essay, but, in Robin Wood's words, 'the critic has the right to appropriate whatever's/he needs from wherever it can be found, and use it for purposes somewhat different from the original ones'.44