Professional Documents
Culture Documents
Adaptation Theory
Adaptation Theory
T H O M A S
L E I T C H
Twelve Fallacies in
Contemporary Adaptation Theory
WHAT COULD BE MORE AUDACIOUS than to argue that the study of moving images
as adaptations of literary works, one of the very first shelters under which cinema studies originally entered the academy, has been neglected? Yet that is exactly what this essay will argue: that despite its venerable history, widespread
practice, and apparent influence, adaptation theory has remained tangential to
the thrust of film study because it has never been undertaken with conviction
and theoretical rigor. By examining a dozen interlinked fallacies that have kept
adaptation theory from fulfilling its analytical promise, I hope to claim for adaptation theory more of the power it deserves.
1. There is such a thing as contemporary adaptation theory. This is the founding fallacy of adaptation studies, and the most important reason they have
been so largely ineffectualbecause they have been practiced in a theoretical
vacuum, without the benefit of what Robert B. Ray has called a presiding poetics.1 There is, as the preceding sentence acknowledges, such a thing as adaptation studies. It is pursued in dozens of books and hundreds of articles in
Literature/Film Quarterly and in classrooms across the country, from high
school to graduate school, in courses with names like Dickens and Film and
From Page to Screen. But this flood of study of individual adaptations proceeds on the whole without the support of any more general theoretical account of what actually happens, or what ought to happen, when a group of
filmmakers set out to adapt a literary text. As Brian McFarlane has recently
observed: In view of the nearly sixty years of writing about the adaptation of
novels into film . . . it is depressing to find at what a limited, tentative stage
the discourse has remained.2 Despite the appearance of more recent methodologies from the empiricism of Morris Beja to the neo-Aristotelianism of James
Griffith, the most influential general account of cinemas relation to literature
continues to be George Bluestones tendentious Novels into Film, now nearly
half a century old. Bluestones categorical and essentialist treatment of the relations between movies and the books they are based on neglects or begs many
Criticism, Spring 2003, Vol. 45, No. 2, pp. 14971
Copyright 2003 Wayne State University Press, Deroit, MI 48201
149
150
Thomas Leitch
crucial questions, and more recent commentators, even when they are as sharp
as McFarlane (who will therefore claim particularly close attention in this
essay) in taking exception to Bluestone, have largely allowed him to frame the
terms of the debate.
Hence several fundamental questions in adaptation theory remain unasked, let alone unanswered. Everyone knows, for example, that movies are a
collaborative medium, but is adaptation similarly collaborative, or is it the work
of a single agentthe screenwriter or directorwith the cast and crew behaving the same way as if their film were based on an original screenplay? Since
virtually all feature films work from a pre-existing written text, the screenplay,
how is a films relation to its literary source different from its relation to its
screenplay? Why has the novel, rather than the stage play or the short story,
come to serve as the paradigm for cinematic adaptations of every kind? Given
the myriad differences, not only between literary and cinematic texts, but between successive cinematic adaptations of a given literary text, or for that matter
between different versions of a given story in the same medium, what exactly
is it that film adaptations adapt, or are supposed to adapt? Finally, how does
the relation between an adaptation and the text it is explicitly adapting compare
to its intertextual relationships with scores of other precursor texts?
The institutional matrix of adaptation studythe fact that movies are so
often used in courses like Shakespeare and Film as heuristic intertexts,
the spoonful of sugar that helps the Bards own text go down; the fact that
studies of particular literary texts and their cinematic adaptations greatly outnumber more general considerations of what is at stake in adapting a text from
one medium to another; the fact that even most general studies of adaptation
are shaped by the case studies they seem designed mainly to illuminate
guarantees the operation of adaptation studies on a severe economy of theoretical principles which have ossified into a series of unvoiced and fallacious
bromides most often taking the form of binary oppositions that poststructuralist theory has taught us to deconstruct: literature versus cinema, high culture
versus mass culture, original versus copy.3 Precisely because these bromides
are rarely articulated, they have retained the insidious power of Ibsens ghosts:
the power to direct discussion even among analysts who ought to know better.
2. Differences between literary and cinematic texts are rooted in essential properties of their respective media. This fallacy was first promulgated by Bluestone
and by Siegfried Kracauers roughly contemporaneous Theory of Film, which
opens with the sweeping statement, This study rests upon the assumption
that each medium has a specific nature which invites certain kinds of communications while obstructing others.4 More recently, it has been one of the rare
articles of faith that has actually come under such general debate that few theorists would probably admit to subscribing to it these days. Nonetheless, it has
been given new impetus in the past ten years by the reprinting in the last two
Adaptation Theory
151
editions of the Oxford anthology Film Theory and Criticism of Seymour Chatmans accurately but fallaciously entitled essay, What Novels Can Do That
Films Cant (and Vice Versa).5 The most influential attacks on the essentialist
view that novels and films are suited to fundamentally different tasksin
Chatmans view, assertion and depiction respectivelybecause of the features
specific to their media have taken two forms. One is the empirical argument
advanced by F. E. Sparshott and V. F. Perkins6 that many films and not a few
novels break the rules the essential qualities of their media apparently prescribe. The other is the more general attack Noel Carroll has mounted against
what he calls Rudolf Arnheims specificity thesis on the grounds of its philosophical gratuitousness: There is no rationale for the system [of arts], for in
truth it is only a collection. Thus, we have no need for the specificity thesis,
for the question it answersWhy is there a system of different arts?is not
really an admissible question at all.7 But these attacks can be usefully supplemented by a closer consideration of the alleged specifics of film and fiction.
Chatman, for instance, dismisses explicitly descriptive voiceover commentary in movies as uncinematic on the grounds that it is not cinematic description but merely description by literary assertion transferred to film.8
Would anybody writing today argue that the highly assertive and descriptive
voiceover commentary by the murdered Joe Gillis in Sunset Boulevard, a film
not adapted from a literary source, was inessential to the films effect because,
as McFarlane notes, voiceover narration by its nature cannot be more than
intermittent as distinct from the continuing nature of the novelistic firstperson narrationor, in Chatmans terms, that it was uncinematic because it
was literary?9 Chatman argues that such arguably descriptive closeups as
Professor Jordans amputated finger in The 39 Steps, the poisoned coffee cup
in Notorious, and Marion Cranes staring dead eye in Psycho are actually hermeneutic rather than descriptive because for all their capacity to arrest our attention, these close-ups in no way invite aesthetic contemplation.10 But a
generation of Hitchcock commentary has disagreed. These shots do invite aesthetic contemplation because they are descriptive and assertive.
In arguing from the other side that the cameras essential function of depicting without describing is confirmed by the use of terms like the camera
eye style11 to characterize passages of neutral, Hemingwayesque detail in novels that approach the condition of cinema, Chatman is again beguiled by his
essentialism into mistaking how both novels and films work.
Consider one the most famous camera eye passages in fiction, this description of Sam Spade awakened in The Maltese Falcon by the news that Miles
Archer, his partner, has been shot to death:
Spades thick fingers made a cigarette with deliberate care, sifting a
measured quantity of tan flakes down into curved paper, spreading the
152
Thomas Leitch
flakes so that they lay equal at the ends with a slight depression in the
middle, thumbs rolling the papers inner edge down and up under the
outer edge as forefingers pressed it over, thumbs and fingers sliding to
the paper cylinders ends to hold it even while tongue licked the flap,
left forefinger and thumb pinching their end while right forefinger and
thumb smoothed the damp seam, right forefinger and thumb twisting
their end and lifting the other to Spades mouth.12
Adaptation Theory
153
154
Thomas Leitch
Adaptation Theory
155
156
Thomas Leitch
Adaptation Theory
157
in the initial novel, the objects and gestures forming the very fabric
of the plot disappeared completely, leaving behind only their significations: the empty chair became only absence or expectation, the
hand placed on a shoulder became a sign of friendliness, the bars on
the window became only the impossibility of leaving. . . . But in the
cinema, one sees the chair, the movement of the hand, the shape of
the bars. What they signify remains obvious, but instead of monopolizing our attention, it becomes something added, even something in
excess . . . because what affects us . . . are the gestures themselves . . .
to which the cinema has suddenly (and unintentionally) restored
their reality.21
But it would be impossible to maintain such a distinction today in view not
only of novels like Robbe-Grillets own, which seeks to restore the reality to
objects and gestures by frustrating any definitive account of their significations, but of the digitized dinosaurs of Jurassic Park and the impossible acrobatics of Spider-Man and Minority Report, which depend not on their intimacy
with the pre-existing physical reality they misleadingly imply but on producing a reality effect quite as loaded as Robbe-Grillets empty chair.
Even though concept versus percept, as James Griffith has pointed
out, offers critical certainty and shuts off discussion,22 the dichotomy has
proved lamentably tenacious. Brian McFarlane, in the most acute recent general study of adaptation, follows Bluestone in arguing that the verbal sign,
with its low iconicity and high symbolic function, works conceptually, whereas
the cinematic sign, with its high iconicity and uncertain symbolic function,
works directly, sensuously, perceptually.23 The implication McFarlane draws
from this contrast is that adaptation study should not rest content with impressionistic comparisons that emphasize films alleged fail[ure] to find satisfactory visual representations of key verbal signs, but should consider to
what extent the film-maker has picked up visual suggestions from the novel
in his representations of key verbal signsand how the visual representation
affects ones reading of the film text.24 McFarlane goes on to argue that at
several crucial points in the 1946 Great Expectations, David Lean succeeds not
only in capturing a sense of Pips first-person narrative voice but in grounding
symbolic functions in a realistic mise-en-sce`ne rather than imposing them by
fiat. The result is that the realistic meaning of the action seems to me to melt
into the symbolic. . . . The symbolic is a function of the mise-en-sce`ne, inextricably interwoven into the realist texture.25 McFarlane acknowledges, however, that as one very familiar with the film, I find it hard to be sure how far
on a single or first viewing a spectator might be aware of the symbolic functions I now discern26 in Magwitchs floundering in the mud, Jaggers towering
over Pip and Estella, and the stormy night sky that heralds Magwitchs return.
158
Thomas Leitch
The difference between percept and concept may well be more properly a
function of rereading, and of a specifically analytical kind of rereading, than of
a difference between movies, which are commonly assumed against mounting
evidence to be watched only once, and novels, which are assumed to be endlessly rereadable, with each rereading converting more percepts to concepts.
6. Novels create more complex characters than movies because they offer more
immediate and complete access to characters psychological states. The ability to
enter the minds of fictional characters directly is of course one of the glories,
as it is one of the constitutive distinctions, of prose fictionthe only medium
whose conventions allow third-person sentences beginning she thought
and it is indeed hard for movies to compete with novels in this regard. But it
is just as hard for other media whose representation of complex characters has
long been accepted. Since most novels take longer to read than two hours, it
stands to reason that they have more leisure to develop characters who change
over time. But I have never read an argument that long novels create more
compelling characters than shorter novels, or even than short stories. The
stricture against brevity seems to condemn movies alone.
Nor will the argument that cinemas characters are limited by its inability
to present thought directly stand up to analysis. When Bluestone notes that
the film, having only arrangements of space to work with, cannot render
thought, for the moment thought is externalized it is no longer thought,27 his
observation is equally apt to drama as to film. Yet no one questions the ability
of playwrights from Euripides to Chekhov to create complex characters. It is
true, of course, that Shakespeares dramaturgy allows him soliloquies and
asides that make it easier to dramatize thought, but Hamlets thoughts are still
necessarily externalized. The conclusion that follows is not that externalized
thought is no longer thought, but that the pleasures of many non-novelistic
media are based to a large extent in the invitation they extend to audiences to
infer what characters are thinking on the basis of their speech and behavior,
and that thoughts that are inferred can be just as subtle and profound as
thoughts that are presented directly.
This last point deserves closer consideration. Novels, plays, and movies
can each hardly help leaving out many details from their discourse. Wolfgang
Iser, calling these omissions gaps or blanks, has analyzed at length the
processes by which readers are encouraged to fill them in, the freedom they
have in choosing from among alternative possibilities, and the limitations on
that freedom that define failure as filling the blank exclusively with ones
own projections.28 What Iser does not consider is the necessity of gaps, not
as an inevitable corollary of a given storys incompleteness, but as the very
basis of its appeal. For it is precisely the business of fictional narratives to create a field in which audiences are invited to make inferences about what the
characters are feeling or planning, where the story is going, what particular
Adaptation Theory
159
details will mean, and how everything will turn out. Such inferences are the
product of the same increasingly educated guesswork that derives concepts
from percepts. These inferences confer both the sense of intimacy with fictional characters that makes them more memorable than most real people and
the assurance that the fictional field at hand comprises a world more satisfyingly coherent than the world outside. Novels and plays and movies might be
said paradoxically to display their gaps in the sense that they depend for the
pleasures they provide on audiences noticing and choosing to fill some of
them but not others.29
The importance of this invitation to the audience is confirmed by the fact
that few moviegoers read screenplays for pleasurenot because screenplays
have no gaps (they specify many fewer details than either the literary texts they
are based on or the movies that are based on them), but because their gaps are
designed to be filled once and for all by the cast and crew, not displayed as an
invitation to nonprofessional audiences active participation. It is one of
Shakespeares most underappreciated gifts that the plays generations of readers have revered are nothing more than performance texts whose verbal texture happens to support an incomparably richer sense of reality than that of
any screenplay to date.
Novels and movies, to stick with the two media most often contrasted in
this regard, typically depend for their effects on different kinds of gaps. Because Jane Austens novels, for example, though exceptionally precise in revealing the thoughts of each of their fictional heroines, limit themselves to
exactly one such heroine per novel, cinematic adaptations of her work typically narrow the gap between the intimacy viewers feel with her heroines and
the corresponding lack of intimacy they feel with other characters. Leo Braudy
has contended that the basic nature of character in film is omission. . . . Film
character achieves complexity by its emphasis on incomplete knowledge, by
its conscious play with the limits a physical, external medium imposes upon
it.30 But Sam Spades disturbingly unreadable rolling of a cigarette in Hammetts novel raises the possibility that the basis of all character may well be
incompleteness and omissionthat characters are by definition figures whose
gaps allow readers or viewers to project for them a life that seems more vivid,
realistic, and complex than their explicitly specified thoughts and actions. At
the very least, it does not follow either that novels and movies are condemned
to certain kinds of gaps that are specific to their media, or that one sort of gap
is better than another. What determines the success of a given work is neither
the decision to withhold nor the decision to specify a characters thoughts, but
the subtlety, maturity, and fullness of the pattern that emerges from thoughts
and actions specified or inferred. These are not criteria on which any particular medium has a monopoly.
7. Cinemas visual specification usurps its audiences imagination. Perhaps
160
Thomas Leitch
Adaptation Theory
161
stylized rhetoric than on any set of images. It is not the readers job to translate
such passages into visually realized or narrativized images of Wemmicks face
under the chisel (how broad a chisel, and how long did the operation take?)
or Scrooges house playing hide-and-seek with other houses (how many other
houses? was Scrooges house the smallest? how long did the other houses look
for it before they gave up?) but to enjoy them as concepts whose sensory appeal is at least as much to the ear as to the eye. The dauntingly rich visual field
of films does not inhibit viewers imagination, because imagining, as Chatman
has pointed out, cannot legitimately be reduced to picturing.35
The deeper fallacy, which McFarlane does not identify, is the assumption
that it would be an advantage to a film adaptation if Dickens were to specify
Wemmicks visual reality more closely because the fewer gaps the novelist
leaves, the easier it is to fill them without transgressing.36 On this analysis, it
would be a hopeless endeavor to adapt Austens novels to film because their
visual texture is so remarkably thin that adapters are compelled to draw on
ancillary historical accounts to dress all the characters and furnish their rooms.
Austens novels would be much better suited to radio, which would emphasize
her subtlety in distinguishing her characters voices without the necessity of
supplying extraneous visual details. But even the most deferential film adapters commonly approach such gaps not with trepidation but with a sense of
opportunity to supply their own details because they assume that films will
differ from their sources in myriad ways and are eager to invent details whenever the novels discourse gives them leave to do so. Just as gaps are the engine
of narrative engagement for the audience, they are the license for the kinds of
filmmaking inventions that elevate adaptations above servile transcriptions.
Even to talk of the inventiveness of deferential adapters, however, anticipates
the next fallacy.
8. Fidelity is the most appropriate criterion to use in analyzing adaptations.
McFarlanes restive description of the near-fixation with the issue of fidelity
that has inhibited and blurred adaptation study since its inception is all too
accurate.37 Fidelity to its source textwhether it is conceived as success in
re-creating specific textual details or the effect of the wholeis a hopelessly
fallacious measure of a given adaptations value because it is unattainable, undesirable, and theoretically possible only in a trivial sense. Like translations to
a new language, adaptations will always reveal their sources superiority because whatever their faults, the source texts will always be better at being
themselves. Even if the adaptations are remakes in the same medium, their
most conscientious attempts to replicate the original will betray their differences, and thus their inferiority, all the more plainlya point made particularly clear in the critical discourse on Gus Van Sants instructive 1998 remake
of Alfred Hitchcocks Psycho. Van Sants film prompted critics across the country to complain not only that Vince Vaughn and Anne Heche were inadequate
162
Thomas Leitch
substitutes for Anthony Perkins and Janet Leigh but that it didnt matter that
Julianne Moore was a better actress than Vera Miles because her performance
was different, and every departure from Hitchcocks text, which Van Sant had
promised to follow line by line and shot by shot, was by definition a betrayal.
The only remake that would have maintained perfect fidelity to the original
text would have been a re-release of that text. But as McFarlane broadly implies in his discussion of Martin Scorsess Cape Fear (1991) as a remake and
commentary on J. Lee-Thompsons 1961 version, even a re-release of the original film might well have a profoundly different effect on audiences influenced
by thirty years of changes in social mores, generic prestige, family values, and
industry self-censorship.38
Given the indefensibility of fidelity as a criterion for the analysis of adaptations, why has it maintained such a stifling grip on adaptation study? The
likely reasons seem less theoretical than institutional. The assumption of fidelity is really an appeal to anteriority, the primacy of classic over modern texts
which are likely to come under suspicion by exactly the teachers trained in
literary studiesfor example, the Shakespeareans giving courses in Shakespeare on Film or using the Kenneth Branagh and Mel Gibson Hamlets as
classroom anodyneswho are most likely to be interested in adaptations.39 At
least until the infant study of novelizations adapted from cinematic originals
becomes a full-fledged discipline, the valorization of fidelity amounts to a valorization of literature as such in the face of the insurgent challenge of cinema
studies. And the theoretical poverty of fidelity as a touchstone of value, which
begs analytical questions that might bedevil other approaches, is no stumbling
block to commentators who are suspicious of theory in the first place.
These explanations, however, still leave unquestioned the more general
assumption that the main business of commentators who are considering film
adaptations is evaluating them, whether vis-a`-vis their source texts or on their
own merits. The peculiarity of this assumption can hardly be overstated. Evaluation may well be one of the most venerable, central, theoretically significant, and pragmatically inescapable set of problems in criticism.40 Yet the
whole tendency of cinema studies since universities first took it up thirty years
ago has been away from evaluation as a critical projectexcept in the area of
adaptation study. The Top Ten lists that so roiled readers of Sight and Sound in
the years before cinema studies made it into the academy have now lost the
headlines to newspaper reviewers annual roundups and the American Film
Institute, freeing film scholars to focus on analytical and theoretical problems.
Only adaptation study, whether or not it uses the source text as a touchstone,
remains obsessed with asking whether a given film is any good as a preliminary, a precondition, or a substitute for asking how it works.
9. Source texts are more original than adaptations. A primary reason that
adaptation study remains obsessed with fidelity as a criterion for evaluation is
Adaptation Theory
163
that adaptations raise questions about the nature of authorship that would be
difficult to answer without the bulwark of fidelity. It is much easier to dismiss
adaptations as inevitably blurred mechanical reproductions of original works
of art than to grapple with the thorny questions of just what constitutes originality and in what sense Robert James Wallers phenomenally popular novels
are themselves less mechanically reproduced than Clint Eastwoods film version of The Bridges of Madison County.
The basis for the assumption that literary texts are to be valued for an
originality that adaptations lack is clarified by considering the apparently exceptional case of William Shakespeare, nearly all of whose plays are adaptations, often to a new medium, of earlier material from sources as diverse as
Holinsheds Chronicles and Greenes Pandosto. The originality of Shakespeare,
his defenders asseverate, depends precisely on his seeing the artistic potential
of inert source materials; he is an alchemist, not an adapter, as one can see by
comparing any of his plays with its base original. But this defense demonstrates only that some adaptations are better than others, not that the best adaptations arent really adaptations at all. Nor does it demonstrate that only
writers can escape the label of adapter, since there are several noted film adapters sanctified by the name of auteur. Orson Welles wrote most of his own
screenplays, typically based on earlier source material. Stanley Kubricks films,
all of them similarly adaptations of literary source texts, are universally recognized as distinctively his. Perhaps most startling of all are Walt Disneys animated versions of childrens classics (Alice in Wonderland), folk classics (Snow
White and the Seven Dwarfs), demi-classics (Bambi), and non-classics (101 Dalmations), in which the producer continues his successful career as auteur some
forty years after his death.
Perhaps the best illustration of the slippery nature of originality in adaptation is the critical reception of Hitchcock, who emerged as an auteur in France
in the 1950s and in America ten years later precisely to the extent that his
champions were able to make a case for thematic affinities among his many
films that ran deeper than recycled genre formulas. The further Hitchcocks
star rose as the only begetter of his films, the less a film criticism centred on
directors was concerned to follow up Hitchcocks statements . . . of indebtedness to English literary figures,41 even though only a handful of Hitchcock
filmsThe Ring, Champagne, Saboteur, North by Northwest, Torn Curtainare
based on original screenplays.42 Only with the decline of auteurism as a critical
framework did critics like Charles Barr turn to a closer examination of Hitchcocks sources.
The moral implicit in the shifting fortunes of writers and directors as creative artists seems to be the enduring appeal of someones originality as an artistic value and the need commentators continually feel to identify a single
shaping intelligence as a given works creator. The reason that originality
164
Thomas Leitch
maintains a central position in adaptation study but not cinema studies generally is that cinema studies has long rejected aesthetics as its leading methodology in favor of analytical and theoretical critique. The antipathetic tendencies
of cinema studies and adaptation study have so exacerbated each other that
adaptation study is now the safest refuge for film scholars unsympathetic to
the prevailing currents in cinema studies today, whose discourse, if it took any
notice of adaptation at all, would no doubt dismiss it with the observation that
all texts are intertexts. That is, all texts quote or embed fragments of earlier
texts, in Mikhail Bakhtins terms, typically without explicit acknowledgment,
often without conscious intention, and never with any attempt at straightforward replication of the originals force. Indeed, the novel, which Bakhtin
hailed as the dialogic mode par excellence because in the rich soil of novelistic prose, the internally dialogized discourse Bakhtin describes as doublevoiced or heteroglot sinks its roots deep into a fundamental, socio-linguistic
speech diversity and multi-languaged-ness,43 has long been eclipsed by the
junkyard aesthetics of the cinema. Movies are a mode whose elastic form, by
turns comic, ironic, and parodic, can tolerate heteroglossia that would wreck
more narrowly defined forms. Indeed, movies themselves may already in turn
have been eclipsed by television series like The Simpsons, surely the most carnivalistic work of fiction enjoyed by a large contemporary audienceexcept of
course for the World Wide Web in its entirety.
10. Adaptations are adapting exactly one text apiece. It might seem commonsensical to assume a one-to-one correspondence between film adaptations
and their literary sources. But just as a novel like Frankenstein may serve as the
vehicle for over a hundred adaptations, each individual adaptation invokes
many precursor texts besides the one whose title it usually borrows. When
McFarlane begins his disapproving discussion of the criterion of fidelity by
asking satirically, Is it really Jamesian? Is it true to Lawrence? Does it capture the spirit of Dickens?,44 he is invoking authorial intention as a possible
regulatory function. But the phrases he chooses, especially Jamesian and
capture the spirit of Dickens, are reminders that adaptations of the works of
famous and prolific novelists are customarily measured not only against the
novels they explicitly adapt but against the distinctive world or style or tone
associated with the author in general. Adaptations of Great Expectations invoke
not only textual particulars of Dickenss novel but more general conventions of
the Dickens world: genial satire, sentimental benevolence, comically grotesque
minor characters, happy endings. Hence the more buoyant ending tacked
onto David Leans Great Expectations, an ending McFarlane finds so charmingly redolent of British aspirations at the end of World War II that one is led
to have more in mind than the famous novel whose title Leans film bears,45
is arguably more Dickensian than the uncharacteristically downbeat ending
Dickens himself supplied.
Adaptation Theory
165
Nor is the world or style or career of the author the only precursor text
that competes with the particular novel or play or story for attention. Hollywood adaptations of foreign novels invariably foreground their particular nationalities and historical moments in ways their source novels rarely do.
American adaptations make much more of the Surrey of Emma and the London of The Picture of Dorian Gray, and the innumerable adaptations of Agatha
Christie beginning with Murder on the Orient Express mark their historical period more emphatically than their forerunners do. Dozens of adaptations that
open with screens showing copies of the books on which they are based, from
A Christmas Carol to The Postman Always Rings Twice, invoke not only their
specific precursor texts but the aura of literature as such to confer a sense of
authority. And as McFarlane himself points out, films as different as Great Expectations and Cape Fear inevitably comment on their own cultural and historical contexts. Nor will it do to argue that the author, a collection of national
characteristics or historical periods, or the institutions of literature and cultural history themselves are not textualized in the same way novels and plays
are, since even if they were not textualized before their adaptation, the adaptations confer a specifically textual status on them by the mode of their reference, which can borrow only the authority it grants its subjects. Hence Imelda
Whelehan has observed that when . . . we study a text such as Hamlet which
has been subjected to countless adaptations . . . [we] recognize that in untangling one adaptation from another, we have recourse to many sources outside
both the play and subsequent films.46
Commentators on adaptations like McFarlane often recognize the richness of their heteroglossia but rarely pursue its leading implication: that no
intertextual model, however careful, can be adequate to the study of adaptation if it limits each intertext to a single precursor. As Bakhtin argues of the
novel:
The real task of stylistic analysis consists in uncovering all the available orchestrating languages in the composition of the novel, grasping the precise degree of distancing that separates each language from
its most immediate semantic instantiation in the work as a whole, and
the varying angles of refraction of intentions within it, understanding
their dialogic interrelationship andfinallyif there is direct authorial discourse, determining the heteroglot background outside the
work that dialogizes it.47
Adaptation study requires as sensitive and rigorous attention to the widest
possible array of a films precursor texts as McFarlane devotes to the novels
the films he considers adapt.
11. Adaptations are intertexts, their precursor texts simply texts. This is the
assumption that underlies the last two assumptions about originality and the
166
Thomas Leitch
Adaptation Theory
167
the novel itself should be treated so uncritically as a criterion of value that adaptations seek to create non-parallel experiences at their peril. Although it is
certainly true that adaptations are intertexts, it is equally true that their precursors are intertexts, because every text is an intertext that depends for its interpretation on shared assumptions about language, culture, narrative, and other
presentational conventions.
Chatman might well object here that the constitutive difference between
adaptations and their originals is that adaptations invite the consideration of a
single precursor text as primary whereas their originals combine many influences into a new synthesis that does not privilege any one of them. Adaptations imitate novels, novels imitate life, or at least Victorian life, itself. But this
distinction cannot be seriously maintained in an age that abounds in such
ironic, parodic, or heteroglot adaptations as The Birds, The Three Musketeers
(and The Four Musketeers), Batman, Clueless, and Everything You Always Wanted
to Know about Sex (But Were Afraid to Ask). The most capacious novels, to the
extent that they choose subjects more specific than lifethe First Reform Bill,
the intrigue Cardinal Richelieu provoked at the court of Louis XIII, the effects
of Napoleons campaign against Russiatake on the responsibility of illuminating those particular subjects. Even the Aeneid, whose avowed subject is
arms and the man, and Tom Jones, whose subject Fielding grandly announces
as HUMAN NATURE,53 approach their forbiddingly general subjects in terms of
far more specific historical narratives and behavioral types. As Deborah Cartmell points out, instead of worrying about whether a film is faithful to the
original literary text (founded in the logocentric belief that there is a single
meaning), we read adaptations for their generation of a plurality of meanings.
Thus the intertextuality of the adaptation is our primary concern.54
12. Adaptation study is a marginal enterprise. This is the only one of my
twelve fallacies that is actually true. Adaptation study has indeed for many
years been marginal to the study of moving images in general. But does it need
to be marginal? Dudley Andrew, describing adaptation study as frequently the
most narrow and provincial area of film theory, called for the integration of
adaptation study into cinema studies by noting that its distinctive feature, the
matching of the cinematic sign system to prior achievement in some other system, can be shown to be distinctive of all representational cinema.55 Andrew
called for a generalizing of adaptation study to cover all the varieties of signification, quotation, and reference that make cinema possible and an analysis
of connotation and a sociology of adaptation to complement its aesthetic assumptions about fidelity. Nothing like this has happened in the twenty years
since Andrew wrote. Even acute contemporary analysts like McFarlane and
Chatman who have deplored the crippling dependence of adaptation study on
concepts like fidelity and monistic claims of literatures superiority to film (or
168
Thomas Leitch
Adaptation Theory
169
4. Siegfried Kracauer, Theory of Film: The Redemption of Physical Reality (London: Oxford University Press, 1960), 1. For a later version of the essentialist argument, see
Geoffrey Wager, The Novel and the Cinema (Rutherford, N.J.: Fairleigh Dickinson
University Press, 1975): Language is a completion, an entelechy, and film is not.
By its nature it cannot be. Film is a diffusion (12).
5. See Seymour Chatman, What Novels Can Do That Films Cant (and Vice Versa),
Critical Inquiry 7 (198081): 12140. Chatmans view of adaptation is better represented by his more recent and subtle discussion of the topic in Coming to Terms:
The Rhetoric of Narrative in Fiction and Film (Ithaca: Cornell University Press,
1990), about which I will have more to say later on.
6. See F. E. Sparshott, Basic Film Aesthetics, in Gerald Mast and Marshall Cohen,
eds., Film Theory and Criticism: Introductory Readings, 3rd edition (New York: Oxford University Press, 1985), 28687, and V. F. Perkins, Film as Film: Understanding and Judging Movies (1972; rpt. Harmondsworth: Penguin, 1986), 59. For more
recent empirically based attacks on the essentialist argument, see Morris Beja, Film
and Literature: An Introduction (New York: Longman, 1979), 5459, and James
Griffith, Adaptations as Imitations (Newark: University of Delaware Press, 1997),
3035.
7. Noel Carroll, Philosophical Problems of Classical Film Theory (Princeton: Princeton
University Press, 1988), 88.
8. Chatman, What Novels Can Do, 128.
9. McFarlane, 16. In Jane Campion and the Limits of Literary Cinema, Ken Gelder
has attacked this essentialist argument by noting that commentators on Jane Campion and Kate Pullingers novelization of Campions film The Piano have consistently identified it as somehow less literary than the film: as if the film was more
of a novel than the novel itself (Deborah Cartmell and Imelda Whelehan, eds.,
Adaptations: From Text to Screen, Screen to Text [London: Routledge, 1999], 157).
10. Chatman, What Novels Can Do, 12829.
11. Ibid., 128.
12. Dashiell Hammett, Complete Novels (New York: Library of America, 1999), 398.
13. Pauline Kael, 5001 Nights at the Movies (New York: Holt, Rinehart and Winston,
1982), 355.
14. Andre Bazin, Orson Welles: A Critical View, trans. Jonathan Rosenbaum (New York:
Harper & Row, 1978), 133.
15. McFarlane, 26.
16. In Made into Movies: From Literature to Film (New York: Holt, Rinehart and Winston, 1985), James Y. McDougal, who is practically alone among analysts of adaptation in noting that casting . . . considerably shapes the process of adaptation,
avers that casting is extrinsic to the medium of film (6).
17. Timothy Corrigan, Film and Literature: An Introduction and Reader (Upper Saddle
River, N.J.: Prentice Hall, 1999), 3.
18. Joy Gould Boyum, Double Exposure: Fiction into Film (New York: New American
Library, 1985), 21.
19. George Bluestone, Novels into Film (1957; rpt. Berkeley and Los Angeles: University of California Press, 1971), 20.
20. Gerald Mast, Film/Cinema/Movie: A Theory of Experience (New York: Harper &
Row, 1977), 64.
21. Alain Robbe-Grillet, For a New Novel: Essays on Fiction, trans. Richard Howard
(New York: Grove, 1965), 20.
170
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
Thomas Leitch
Adaptation Theory
48.
49.
50.
51.
52.
53.
171