Coming To Terms
Coming To Terms
Coming To Terms
Seymour Chatman
90-55119
10
Contents
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Acknowledgments
Introduction
Narrative and Two Other Text-Types
Description Is No Textual Handmaiden
What Is Description in the Cinema?
Argumentation in Film: Mon oncle d'Amrique
In Defense of the Implied Author
The Implied Author at Work
The Literary Narrator
The Cinematic Narrator
A New Point of View on "Point of View"
A New Kind of Film Adaptation: The French
Lieutenant's Woman
The "Rhetoric" "of' "Fiction"
Notes
Index
ix
I
6
22
38
56
74
90
109
124
139
161
184
205
235
vii
Acknowledgments
ix
Coming to Terms
Introduction
NARRATOLOGY" is a word that Henry James would have deplored, though he might have found merit in its objects of concern.
After twenty years, I still feel something of an ironic twinge when I
see it in print. A "science" of narrative seems an unlikely, even a
slightly shady pursuit. But, of course, "ology" can also mean
"theory of," and who, these days, would dare fault theory? In any
case, we can take heart from the fine work on narrative theory and its
applications to literature and cinema published during these years. 1
This book is concerned with the terms of narratology and of text
theory in general. It assumes that every discipline needs periodically
to examine its terms. For terms are not mere tags: they representin
some sense, even constitutea theory. By scrutinizing its terms, we
test and clarify the concepts that a theory proposes. Through that clarification we can better decide whether they help or hinder our work.
Though the book treats a variety of subjects, it strives to incorporate them in a unified commentary on narrative terminology from
two perspectives. The first perspective, informing the first four
chapters, is external, considering the relations of Narrativein particular, fictional Narrativeto other kinds of discourse or "texttypes." This is less a general theory of the text than a set of distinctions that clarify Narrative's position among the text-types. Further, to
explain how other text-types such as Argument and Description fit
within the Narrative frameworkand vice versaI propose the
concept of textual "service."
1
Introduction
The next six chapters of the book approach narratology from an
internal perspective. Following the order of my Story and Discourse2
(and other recent narratological studies), they examine concepts and
formulations that remain controversial: the implied author, the na-ture
of the narrator (including the differences between literary and
cinematic narrators), the concept of character "point of view" or
"focalization" (which, I argue, deserves a better nameI propose
"filter"), and the distinction between "unreliable narration" and what
I call "fallible filtration." Throughout, I cite examples from both
literature and film. I feel a special responsibility to discuss cinematic
narrative in greater detail than I did in Story and Discourse; critics were
right to question the subtitle "Narrative Structure in Fiction and
Film," since that book gave film all too short shrift. Film seems
particularly important to narratology at this juncture if we are to
formulate the general principies of Narrative as well as its actualizations in various media. Only a general narratology can help to explain what literature and cinema have in common, narratively speaking, and only a good sense of that commonality will permit us to
understand what is distinctively cinematic or literary.
Finally, both external and internal perspectives lead me to attempt
a synthesis by reformulating Wayne Booth's conception of "rhetoric
of fiction."
Chapter I attempts a distinction among three kinds of discourse or
text-typeNarrative, Description, and Argument (with passing
mention of the text-type traditionally called "Exposition") and illustrates how each of these may subserve the others. To correct the
centuries-old prejudice that Narrative somehow dominates Description, I argue that no text-type is intrinsically privileged. Most texts
utilize one overriding text-type, but it is generally subserved by
other text-types. Narration can just as easily function at the service of
Description as vice versa. Chapter 2 considers Description in greater
detail, both as a text-type in its own right and in its interrelation with
Narrative.
In Chapters 3 and 4 I take up the question of non-narrative texttypes in the cinema. Most theoretical discussion of cinema, both as
technique and as institution, has presupposed its total commitment
to Narrative, and it is true that the vast majority of filmsat least
commercial filmsare totally dedicated to story-telling. But if we
want to extend our powers of textual analysis from literature to the
Introduction
cinema, we need to look at the cinematic ramifications of the other
text-types as well. Chapter 3 considers Description in the cinema, a
question apparently made difficult by the nature of the medium
itself. Some believe that a film cannot describe at all. They argue that
the act of describing as suchthe evocation of the properties of
objects for their own sakescannot occur in the cinema precisely
because every last detail is already totally visible. I respond to that
claim in several ways.
Chapter 4 illustrates the possible relations of Argument and Narrative in the cinema. It first examines the traditional moralistic use of
Narrative in the movies. The films of Frank Capra come imme diately to mind, their theses often announced in their very titles: You
Can't Take It with You, It's a Wonderful Life. These films bear an
archaeological relation to cinematic Modernism similar to that borne
by fables or eighteenth century novels to the modern and postmodern novel. But in the quite different context of the European art
cinema, films can accommodate textual mixes no less complex than
those of modern novels. I illustrate with an analysis of a film whose
textual structure is delightfully quirky: Alain Resnais's Mon oncle
d'Amrique (1982).
In moving to an internal perspective, I turn to problems still
brewing within the precincts of narratology proper. I skip over issues
that seem reasonably settled (such as the relation of storytime to
discourse time, and the greater facility of literary narrative for rendering the mental life of characters); interested readers will find
sufficient citations in the notes, as well as in such works as Martin's
and Prince's, to explore these topics as they wish.
Chapters 5 and 6 attempt a defense of the "implied author" against
various kinds of attack; Chapter 5 defends in a theoretical way;
Chapter 6 offers examples of the explanatory power of the concept. I
intentionally assemble examples of the greatest possible diversity,
foraging even into the intellectually murky world of magazine advertising.
Next, I deal with what I once called the "transmission" of the
story. Chapters 7 and 8, on literary and cinematic narrators respectively, reexamine the distinction between mimesis and diegesis and
argue for a definition of Narrative broad enough to stress that which
plays, films, and novels have in common. They attempt to demonstrate that plot (the double chrono-logic), character, and setting are
Introduction
Introduction
C H AP T E R
1/
Coming to Terms
themselves as if they were holistic, verbal narratives as if they were
linear. They do so regardless of how any given spectator or reader
goes about perceiving any given work. The structure is one thing,
the perception another. Temporality informs narrative texts in a way
that it does not inform paintings (or non-narrative verbal texts, for
that matter).7 Temporality is involved only in the spectator's work in
perceiving a painting; it is not part of the painting itself. But temporality is immanent to, a component of, narrative texts. The control
exerted by a verbal (or other) narrative has no counterpart in the
experiencing of a (non-narrative) painting or sculpture. The reader
or spectator who skips pages or fast-forwards the videotape or goes
out for a smoke during the second act must somehow learn, by
inquiry or inference, what has transpired in the interim.
Music, too, is a time-regulating structure and hence a "text" in my
sense. In classical Western music, for example, penultimate dominant
chords generally call for resolving tonics, and to that extent they
control listeners' expectations. But musical textuality is of a different
sort than verbal textuality. Most aestheticians argue that music is not
semiotic (or at least not "micro-semiotic"). It offers no constancy of
reference between each of its elementsnotes, phrases, movementsand something else in the real or an imagined world so that
we think of the first as signifier and the second as signified. Most
other texts, however, do make such reference; they entail a linked
continuum of referential signs. "Cat" signifies one referent and "sat"
another and "mat" a third, and "the cat sat on the mat" forms a
semantic composite of the three.
The two different kinds of Communications, textual and nontextual, may perform at each other's service (just as the various texttypes can operate at each other's service, as I shall argue). Thus, a
reproduction of a famous painting hung on a stage backdrop may
subserve the dramatic text of words, gestures, and movements. Or
the crescendo from a well-known piece of music may subserve the
visual and auditory film images to intensify the feeling that the plot is
coming to a head. Of course, in this "servile" function, the painting
and the musical passage lose something of their autonomy. It is hard
to focus on the purely musical structure of Mozart's overture to The
Marriage of Figaro or the "Blue Danube Waltz" when they accompany
cinematic narratives (the first, for example, "over" shots in Trading
Places to establish the cushy life of a young commodities broker; the
8
Coming to Terms
communication. We are not the slaves of language (as some contemporary theorists dolefully contend) but its masters, and we can decide
what we are talking about and how best to talk about it.
Description is sometimes thought of as a sort of casual contiguity.
In the expression "describing our thoughts," for example, the implication is of a more or less random array, fantasies lying cheek-byjowl with cognition and conceptualization. But as we shall see,
Description is not merely a "random" text-type; it generally has its
own logic.10
Argument is the text-type that relies on "logic," at least in the
informal sense; it may employ not the strict "demonstrative" logic of
the syllogism but rather the softer one of the rhetorical enthymeme.
Or the logic may be inductive, or perhaps analogous. 11 But unlike
Narrative chrono-logic, Argumentative logic is not temporal. And
unlike Description, Argument rests not on contiguity but on some
intellectually stronger, usually more abstract ground such as that of
consequentiality.12
These text-types crosscut the distinction between fiction and nonfiction. All three can inform both nonfictional and fictional texts. For
example, "The Gettysburg Address" and Pare Lorentz's documentary film The River are nonfictional arguments; "To His Coy Mistress" and "A Modest Proposal" and most Hollywood World War II
feature films are fictional ones.13
By "text-types," further, I mean something other than genres.
Genres areat least in one sense of the wordspecial subclasses or
combinations of text-types. However one defines them, novels,
novellas, short stories, mysteries, and Westerns are generic subclasses
of the Narrative text-type. A Theophrastian character is a subclass of
Description. A sermon is a subclass of Argument. And so on.
The text-types routinely operate at each other's service. "Ozymandias" is ostensibly a description, what the "traveller from an antique
land" saw, but the overriding structure is an argument whose proposition is something like sic transit gloria mundi. The catechistic "Ithaca"
episode of Ulysses is an argument at the service of the overall narrative. Narrators of novels routinely digress to describe or argue,
describers to narrate or argue, and arguers to narrate or describe. (I
shall exemplify each process below.)
The study of texts is at once simplified and enriched by the notion
of service. Text-types are underlying (or overiding) structures that
10
Coming to Terms
Till off it came, of course,
In less than half an hour;
And yet the sun saved half his power.
So much doth mildness more than force.16
La Fontaine's fable uses Narrative at the service of Argument. The
argument is that of the carrot and the stick: seduction is more effective than aggression. To bend others to your will, accommodate
your demand to the needs of their behavior. (The point is not the
difference between the two natural forces: the wind could have
blown very gently or not at all, leaving it to the heat of exertion to
prompt the man to remove his cloak; the sun, on the other hand,
could have lost by starting out too hot, in which case the man might
have quickly sought shelter without even thinking about his cloak.)
Here is the reverse situationArgument at the service of Narrative. Fielding's Joseph Andrews begins with the folio wing argument:
It is a trite but true observation, that examples work more forcibly on the mind than precepts: and if this be just in what is odious and
blameable, it is more strongly so in what is amiable and praiseworthy.
Here emulation most effectually operates upon us, and inspires our
imitation in an irresistible manner. A good man therefore is a standing
lesson to all his acquaintance, and of far greater use in that narrow
circle than a good book.
But as it often happens that the best men are but little known, and
consequently cannot extend the usefulness of their examples a great
way; the writer may be called in aid to spread their history farther, and
to present the amiable pictures to those who have not the happiness
of knowing the originals; and so, by communicating such valuable
patterns to the world, he may perhaps do a more extensive service to
mankind than the person whose life originally afforded the pattern.17
This argument offers a raison d'tre for the narrative immediately
to follow. Its rhetorical structure is what we would expect from an
eighteenth-century lawyer. The argument is that novels teach virtue
more successfully than does ordinary life experience because publication assures a broader audience for moral examples than does
straightforward observation of human behavior. It rests in a classical
12
Coming to Terms
Coming to Terms
text is to present it at the beginning en bloc, as Dickens does in Little
Dorrit: "Thirty years ago, Marseilles lay burning in the sun, one day. A
blazing sun upon a fierce August day was no greater rarity in
southern France then, than at any other time, before or since." 22
More recent fiction takes pains to blend Description with the action.
In Genette's pungent phrase, Stendhal, Flaubert, and Proust "pulverize" descriptions, diffusing them along the march of events. Joseph
Conrad's and Ford Madox Ford's "distribute" is another term for this
effect. A good example is Emma Bovary's survey of her new house at
Tostes, in which "the general movement of the text is governed by
the step or the gaze of one (or several) character(s), and the unfolding
of that movement corresponds exactly to the length of the trip." 23
Critics promoting the "dramatization" of fiction are less likely to
find Description intrusive. Unlike Argument, Description is felt to
coexist with Narrative without calling attention to itself. Indeed,
some theorists argue (wrongly, as I shall try to demonstrate) that
Description is itself intrinsically diegetic.
But we must take care to distinguish between Narrative and Description as text-types, on the one hand, and sentences in the surface of
a text which are loosely called "narrative" or "descriptive," on the
other. There is a strong sociolinguistic imperative to name the persons and things we speak of, and naming is always a minimal kind of
description. Furthermore, nouns usually occur in noun phrases,
where adjectives and other qualifiers perform further description.
But from the textual point of view, this is incidental not concerteddescription, not Description as a text-type.
Of particular dubiety are statements such as Genette's: "To recount
an event and to describe an object are two similar operations, which
bring into play the same resources of language." 24 It is hard to
understand quite what that means: all text-types actualized by language use the "same resources," since the resources of language are,
ultimately, words. (Similarly, all text-types actualized by drawing or
cinematography or whatever are constrained by their media.) But
surely "to describe" is different from "to narrate," and if we were
asked for the typical verb for representing Description, we would cite
the copula (or its equivalent) rather than a more active kind of verb.
We would say that the subject was so-and-so, not that it did so-and-so.
At the surface level a sentence may provide a great deal of description even though its main thrust may be narrative. From Crime and
16
Coming to Terms
Though the copula in the first sentence suggests that a description
will follow, the next four sentences present a narrative, marked by
verbs denoting a sequence of events: "had been forced into," "had left
her husband," "had lived in . . . retirement," "had moved." But it is a
narrative that operates clearly at the service of the description; in
particular, it illustrates how Lady Blessington was "a dventurous"
and "amusing." Thus, active verbs subserve an overriding descriptive purpose. Notice too the verb in "she next re-emerges": though
active, at grammatical face value, "re-emerges" bears descriptive
rather than narrative force, suggesting not the story of her "reemerging" but a series of tableaux presenting Lady Blessington's
person in timed stages. Elsewhere, too, much of the description is
conveyed by sentences that are active or "eventful," but only on the
surface. Description is communicated also by adjectives modifying
nouns in superficially event-marking sentences: "Marguerite Farmer
had moved to [the arms] of the plutocratic, extravagant, fashionable
Lord Mountjoy"; "With the Blessingtons travelled that dazzling
ephebus Count Alfred d'Orsay." Though active, these sentences have
no overriding narrative force. No narrative event is cited by the verb
"added" in "To good looks she added a brisk intelligence, and to
vivacity and curiosity some touches of literary aptitude." This is
simply a more elegant way of saying "She was good-looking, intelligent, vivacious, curious, and modestly apt in literature."
Of course, just as Description is here the overriding text-type, the
excerpt itself is overridden by the larger narrative, Byron's biogra phy. But (like Brooks and Warren) I have been discussing the excerpt
only. It muddies theoretical waters to argue that Description is always ancillary to Narrative and that it cannot occur autonomously,
to say with Genette that "description might be conceived independently of narration, but in fact it is never found in a so to speak free
state. . . . Description is quite naturally ancilla narrationis, the evernecessary, ever-submissive, never-emancipated slave."27 That assertion ignores well-established self-contained Descriptive genres such
as the "Character." Consider one of Sir Thomas Overbury's efforts,
quoted here in toto:
A Pedant
Hee treades in a rule, and one hand scannes verses, and the other
holds his scepter. Hee dares not thinke a thought, that the nomina-
18
19
Coming to Terms
touches of Argument; a travel guide is concerned not only to describe
but to persuade you that you should see x, y if you have extra time,
and z not at all. Now the order of presentation of the Sehenswrdigkeitenthe "things worth seeing"is temporal. The temporality
could suggest a small narrative (with same such title as "Your Trip to
Rouen"), but the motive for temporality is clearly quite different.
Whereas narrative time ordering creates typical effectssuspense
and surprise, the sense of a completed world in a certain era, and the
likethe time-ordering principle of a travel guide is simply the
convenience of the tourist. You are instructed to start at the Boieldieu
Bridge because it pro vides an excellent overview of the city, a good
thing to experience before investigating the sights close up. The
sequence in which the Sehenswrdigkeiten should be seen may result
simply from their contiguity; hence, temporal ordering relates to
convenience, to consideration for the tourist's time and energy. In
short, as Wallace Martin has put it, "There is a clear difference
between the writer determining the order of presentation and events
determining it."30
Films, at least documentaries, can also be predominantly descriptive, though subservient moments may follow a narrative line. As
David Bordwell and Kristin Thompson dem onstrate, the second
part of Leni Riefenstahl's documentary on the 1936 Olympics is
largely descriptive ("categorical," in their term). 31 Though the progress of individual events is narrated, the film as a whole describes the
Games by these categories:
1. Nature and the Olympians: morning exercise and swimming
2. Gymnastics
3. Yacht races
4. Pentathlon
5. Women exercising
6. Decathlon
7. Field games: field hockey, polo, soccer
8. Bicycle race
9. Cross-country riding
i o. Crewing
ii. Diving and swimming.
Thus, the narrative of each event subserves the descriptive purport of
the film as a whole.
20
21
CHAPTE R
2/
22
Coming to Terms
instances external to the discourse (any practical aim, an auditor to
convince, a scholarly text to constitute), or of instances internal to the
discourse (a coherence, a hierarchy, an anaphoric chain, a "logic," a
narrative lisibilit to insure). In short, description must not be
digression, "hors-d'oeuvre". . . . Description must . . . remain
subordinate to the highest hierarchical instances of discourse, to the
narration [rcit] on one hand, and on the other to the highest existing
subject, the Subject, the human being.7
Like many blanket critical proscriptions, this one ignores the complexity of the discourse situation. It assumes that textual "service"
can go in only one directionfrom lowly Description to lofty Narrative. But as we saw in the preceding chapter, in many works whose
overriding text-type is Description, the "service" is performed by a
contained Narrative, not the other way around.
Neither is Description's putative "drift" from detail to detail aimless. Description has a logic of its own, and it is unreasonable to
belittle it because it does not resemble the chrono-logic of Narration.
Hamon has characterized this logic as metonymic: the description of
a garden, for example, presupposes "almost necessarily the enumeration of diverse flowers, paths, parterres, trees, tools, etc. which
constitute the garden."8 The metonymic structure may entail the
relation of objects to each other as they occur in the world or in the
imagination, but also the relation of objects to their own qualities,
where "quality" is to be understood in the broadest sense.
Metonymy, of course, rests on the principle of contiguity (as
metaphor rests on that of similarity). And contiguity operates in
Description in every dimensionnot only the spatial but the abstract, the intellectual, the moral, and so on. Meir Sternberg has
written well about the complexities of the relationship between contiguity and other ordering principles that support, modify, or subvert it.9 These are extremely varied, extending from "hierarchy" (for
instance, the seating order in a scene in Vanity Fair) to such purely
formal connecting devices as rhyme and alliteration ("Of shoesand
shipsand sealing wax / Of cabbages and kings"). In standard
descriptions (like that of Hamon's garden) the contiguity corresponds to "ordinary experience": gardens typically contain such and
such objects, the relation container-contained being one of metonymy's subclasses of ordering. Sternberg would say that in such cases
a strong ordering principle drives the contents of the descripta.
24
25
Coming to Terms
There follows a whole series of contradictions: Peter is complacent
but not aware of it; he talks rapidly and hurriedly but self-confidently
and with amazingly clear articulation. It is clearly an ambiguous
portrait, yet we accept the contradictionsindeed we praise them as
high artbecause of two conventions we have learned so well: that
of the "round" character, and that of the uncertain or ambivalent
reporter. Peter is not just a villain; he is an interesting villain. Complex
self-contradiction is "recuperable" as a source of both "roundness"
and narratorial uncertainty, and hence of modern "psychological"
interest.
In short, Description has a logic of its own, which is no less
explicablein terms of operant conventionsthan that of Narrative
or Argument. Further, rhetorical conservatives to the contrary notwithstanding, there is no need to determine its "value" with respect
to the other text-types. Value is an issue only to the critic concerned
with the relative success of a given work.
Genette is not the only theorist or critic who finds Description
somehow ancillary and hence, by implication, inferior to Narrative.
The attitudes of Georg Lukcs, which inhabit an ideologically different world, are also worth examining. Lukcs frames his argument
by a comparison between parallel episodes in two novels, Zola's
Nana and Tolstoy's Anna Karenina. The subject of each is a horse race.
In Zola, the "race is described from the standpoint of an observer; in
Tolstoy it is narrated from the standpoint of a participant." The first
is "merely descriptive," the second is "epic." Why? Because the first
occurs by "mere chance," while the second is "inevitable." Inevitability can arise only "out of the relationship of characters to objects
and events, a dynamic interaction in which the characters act and
suffer."13
One does not need to become a defender of Zola to question the
reasoning here. These are clearly sectarian critical pronouncements,
not theoretical observations. It is, of course, true that set pieces of
description went out of fashion with the Modernist novel. But even
"pulverized" or "distributed" description is only one alternative
style. We should beware of confusing stylistic preference, whether of
an individual or an era, with narrative theory.
Implicit in Lukcs's view is a rejection of the Barthesian notion of
verisimilitudethat it is precisely the fortuitousness of a descriptum
that guarantees its realism. Lukcs concedes the possibility of the
26
27
Coming to Terms
In this view, Flaubert's artistic means are genuine only "to some
extent." "Setting" is to be deprecated as a general principle: it is "mere"
or "incidental" unless integrated with an "important event in the
world of the novel." By "world," Lukcs seems to mean the novel's
ultimate, total meaning, with all its thematic banners flying. Humdrum setting and characters made humdrum by their presence
therein ("they seem undifferentiated, additional elements of the environment") find relevance only as an "ironic symbol of philistinism." So description itself is not what achieves Flaubert's "considerable level of artistry." Only irony gives him that award. To Marxist
as to bourgeois literary theory, then, Description's sole value is to
subserve some other textual purpose.
Surely a text theory unencumbered by prescriptivism or ideological projects can furnish a better understanding of the differences
between the text-types, since it does not need to rank them in order
of nobility. Whether a description serves a narrative well seems not a
theoretical but a critical problem, to be determined by the analysis
and evaluation of the individual text. But the notion of subservience
itself is not mysterious. It is simply a term naming how it is that texttypes sometimes work explicitly and sometimes implicitly. Why
they do so well or badly in given cases is the critic's task to explain.
We can distinguish at least three ways in which Description may be
rendered by a text's surface:
(1) Assertions. Here the surface representation corresponds directly to what we might call the standard text-type form: "Simon is
simple."
(2) Nonassertive mentions or inclusions: "Simple Simon met a
pie-man going to the fair." Here, the depiction of Simon as "sim
ple" is, by the syntax, oblique or "casual," that is, not the ostensible
purpose of the sentence. Still, the word occurs in the surface representation.
(3) Elliptical implications: "A passerby asked Simon for a shilling, and Simon gave it to him. The passerby laughed and ran off."
Here, the reader must infer Simon's simplicity by interpreting the
juxtaposition of the two events: Simon's giving money to a total
stranger without demanding a reason, and the stranger's gleeful
running off. The reader does so according to codes she believes to
be operative in this context: for example, the (capitalist) code that
28
29
Coming to Terms
Coming to Terms
the painting hangs but will still perceive it as a timeless and spatially
fixed whole. The objects represented in the painting, despite the
percipient's changing angle (whose only restriction is that it must be
less than 180 degrees), will remain essentially the same. The painting
is closed on itself (another way of saying that it is eventless). Even
artifacts like sculptures that require a parcoursa circumambulation
by the percipientcontain parts that necessarily remain fixed and
closed; no story is entailed, because nothing can change except the
position of the percipient, which is irrelevant. The same thing is true
of verbally rendered descriptions. The verbal descriptum, like the
painting, has only one surface: there is, essentially, no "behind." This
is clearly the case in the opening chapter of Pre Goriot and in the
"establishing" sequence of the classical Hollywood movie.
But once there is uncertainty, once there is openness and "risk," the
circumambulation may begin to feel like Narrative. Kittay contrasts
the contemplation of a painting with a visit to a haunted house or a
labyrinth. For the labyrinth to be truly labyrinthine, the actions of the
percipient become relevant: he becomes, in a way, the protagonist of
a self-created narrative. "He has lost that relationship to what is
before him that had guaranteed him har mless witness status. . . .
Here is not just sequence, but consequence and irreversibility, as
Barthes surely understood it." 20 Choice, risk, consequence, irreversibility: these are the operant conventions of the labyrinthand of
Narrative.
Narrative at the service of Description amounts to the tableau, to
use Kittay's term. In the tableau, act is "consecrated, memorialized
and monumentalized, endowed with power. It is offered up with the
varnish of surface. . . . Rather than action putting description 'in its
place,' it is action that is taken from its dis-place and put, one might
say nailed, in its place. It is action that has become an asyndetic act,
like the stages of the passion of Christ, the representation of any of
which (as when depicted on a stained-glass window) can singly and
independently show his martyrdom. The tableau draws a frame
around the act, to ask that meaning be ascribed to it." 21
The "asyndetism" entailed in the incorporation of a battle or thunderstorm into a narrative as a purely descriptive element is its disconnection from any of its own temporal consequences. The tableau
effect forms the basis of the Homeric "dramatized" descriptions so
admired by Lessing: "Even where [Homer] has to do with nothing
but the picture, he will distribute this picture in a sort of story of the
32
33
Coming to terms
singing the bridal song. There were dancing boys twirling about, pipes
and harps made a merry noise; the women stood at their doors and
watched. A crowd was in the market-place, where a dispute was
going on. Two men disputed over the blood-price of a man who had
been killed: one said he had offered all, and told his tale before the
people, the other refused to accept anything; but both were willing to
appeal to an umpire for the decision. The crowd cheered one or other
as they took sides, and the heralds kept them in order.25
Here again is a multiple layering of "service." The narrative of the
dispute of the two men over a dead man's blood price is at the service
of the description of the shield; the description of the shield is, in
turn, at the service of the larger plot, in particular that portion of it
that encompasses the series "request-manufacture-delivery."
"Service" names a structural, not a critical concept; I do not mean
to suggest value implications by the term. There are many texts
whose connoted descriptive elements emerge as aesthetically more
important than their denoted narrative elements. The narrative elements are there, it seems, only to justify the description. Not that the
lines between Description and Narrative are always clear or unambiguous: postmodern fiction regularly problematizes the relations
between the text-types. All the more reason for the utility of the
notion of text-type "service." Consider, for example, the problematic relations between Description and Narration in Alain RobbeGrillet's Instantans (snapshots), especially "The Secret Room." 26
Given Robbe-Grillet's express rejection of narrative convention, one
must approach this text with due temerity. Our usual expectation of
textual coherence prompts us to ask, for example, whether the text is
a description of a bloody scene served by a narrative, or a narrative of
a grisly murder served by a description. Cleanth Brooks and Robert
Penn Warren opt for the second interpretation, calling the text a
"story." But they find it a rather imperfect one, complaining that it
"does not have a plot in the ordinary sense" and that it "implies an
intention for its manipulation of time that it does not fulfill." 27
Clearly, the value judgment masks some uneasiness about what kind
of a text "The Secret Room" is. In the text theory I am proposing, it
is not possible to say, as Brooks and Warren do, that "the story
describes a painting" (my emphasis). For if all that happens is a de34
35
Coming to terms
description of an oil painting? Why are the identity and even the
existence of the describer left problematic? What is the import of his
or her struggle to make out the painting's contents?
The other problem concerns the events. Brooks and Warren wonder about the time order; but there is a more fundamental question:
How can "events" be said to occur on a painted canvas? The killer's
movements are first presented by ambiguous passives"a black
silhouette is seen fleeing"but ultimately they are expressed by
active verbs: "the man has already moved several steps back"; "the
man . . . leans farther over"; "the [woman's] head turns from side to
side, struggling." We can reject a trivial hypothesis out of hand: we
have no reason to assume that the painting is like a cartoon, separating the stages of the grisly event into panels. But how then, in a single
canvas, can a killer ascend stairs, a victim pant, smoke rise, and so
on?
Several hypotheses seem open to us: perhaps the text is a fantasy or
dream in which the ordinary rules of nature are suspended; or perhaps the movements are in some sense metaphoric or symbolic; or
perhaps they represent something in the observer's mind. For me, the
most interesting possibility is the third. Assume, for instance, that an
art museum visitor is mentally reconstructing a "before" and an
"after" of the critical moment of murder depicted on a painting. That
interpretation would provide a kind of answer to two questions:
what the painting depicts, and why the chrono-logic of events is
skewed.28
In this hypothesis, the movement backward and forward does not
narrate the stages of the sacrificial murder, anachronically, but rather
represents the speculations of a person standing in front of the painting about what led up to and away from the gruesome event. That
would certainly better explain the relevance of the title "Snapshots"
to this particular text. This hypothesis leads us back to an interpretation of the text as a narrative. Perhaps the narrative is not of the
murder (the ostensible story) but of the impact of the painting of the
murder on some unknown contemplator. In other words, what we
had first assumed to be the discoursesparse as it isis now recognized as the story. I do not offer this as a foolproof interpretation of
"The Secret Room," but it does seem to me more coherent than what
we get from other interpretations. More important for the present
discussion, it demonstrates the elaborate ambiguities and tensions
open to the postmodernist manipulator of text-type conventions.
36
Description Is No Textual Handmaiden
To sum up, text theory in general and narratology in particular
could hardly do better than to follow Hamon's recommendation:
"To map out a theory of the descriptive would be . . . to avoid
localising it as an anterior practice (the 'documents collected before
writing') or reducing it to its transitivity by labelling it in such a way
as to put it perpetually at the service of hierarchically superior instances
of narration."29 No, not "perpetually," only sometimesand with the
recognition that the service may sometimes go the other way.
If the foregoing discussion has any validity, one point needs to be
reiterated: namely, that we should avoid such sloppy expressions as
"the narrator describes such and such a narrative event." Objects and
characters may be described, but actions are "described" only if they
function as part of the described setting rather than as links in the
event chain. Story-relevant events are only "narrated," not described. The function of an action, whatever its inner constitution, is
not narrative if it is not in the chrono-logicnot keyed, that is, to the
ongoing march of story events.
37
CHAPTER
3/
What Is Description
in the Cinema?
Coming to Terms
cinematic images cannot guarantee our ability to name bits of descriptive information. Contrarily, literary narrative can be precise,
but always within a relatively narrow scope. Even if a dozen more
details were added to the novel's description of Charles's clothes,
they would still constitute only a selection among the vast number
that could be cited. There would always be "holes" left. Verbal
description could, but never does, encompass the multitude of detail
available in a photograph.
Why should "description" be limited to the discrete, discontinuous, heterogeneous citation of details characteristic of literature?
For a general text theory, I see no reason for such a limitation. Just
because the camera gives the "complete" picture, with no holes, why
should it be said to be incapable of rendering description? In a sense,
the very cinematic projection of images entails Description. It is not
that cinema cannot describe; on the contrary, it cannot help describing, though usually it does so only tacitly. Its evocation of details is
incessantly rich. Every screen "noun" is already, by virtue of the
medium, totally saturated with visual "adjectives." The screen image
cannot avoid them; it cannot present a minimal verbal account like
"A woman entered the room." Rather, it must provide an exhaustive
set of visual details, transcribable by a potentially unlimited verbal
paraphrase: "A woman with a Roman nose, high cheekbones, and
blond hair piled elaborately on her head (etc., etc.) flounced ostentatiously into an ornate ballroom lit by a hundred candles in a glass
candelabra (etc., etc.)." The effet de rel is intrinsic to the medium:
film cannot avoid a cornucopia of visual details, some of which are
inevitably "irrelevant" from the strict plot point of view.
Literary narratives, on the other hand, though they may employ
many adjectives, cannot dictate mental images; they can only stimulate them. Everyone watching the film version of Gone with the Wind
must agree about Rhett Butler's appearance: he is the spitting image
of Clark Gable. But readers of the novel, if they have not seen the
film, will probably disagree about the exact details of Rhett's appearance. Indeed, some may resent having to form a mental picture
of him at all. It is well known that readers differ in their capacity or
desire to construct mental imagery out of words. And even the most
enthusiastic imager probably does not keep an unchanging mental
portrait in constant focus as he plows through a novel.
Conversely, literary narrative has a kind of power over visual
40
41
Coming to Terms
screen time moves inexorably forward, carrying the spectator with it,
films permit none of the "lingering" that we associate with Description in literary Narrative. Descriptive details in cinema, it might be
argued, can occur only as a byproduct of plot action; they do not have a
separate existence. Film, Ollier might say, tolerates no "pause," in
Genette's technical sense; a film's discourse time cannot continue in
the abeyance of story time, because the filmand hence the forward
movement of the storycannot be arrested. In film, the traits of
characters and features of setting must be picked up on the run.
But even if we credit these arguments, we must, I think, still
recognize Description in the cinema. Again, it is simply that cinematic Description tends to be tacit rather than explicit. Christian
Metz includes the "descriptive" as one of his large articulative syntagmas; it is that kind of editing in which the relation of shots is one
of simultaneity, not consecutiveness. A landscape may be described,
for example, by "a tree, followed by a shot of a stream running next
the tree, followed by a view of a hill in the distance, etc." 6 Even
though these views appear consecutively on the screen, the point is
not that the shots are diegetically simultaneous but that story time
has temporarily been suspended. The shot sequence forms a narrative pause. The sign of the pause is precisely the temporally unmotivated shifting from tree to stream to hill. On the other hand, exactly
the same sequence of shots would be narrative if they preceded or
followed shots indicating the eye movements of a character looking
at something ("first he looked at the tree, then at the stream, then at
the hills").
Professional transcriptions of what has transpired on the screen,
called "cutting continuities," often indicate the film's intention to
describe. Here is a published transcription of the first shot of Michelangelo Antonioni's L'Avventura:
Long shot. A sunny summer day. Anna, a twenty-five-year-old
brunette, walks through the courtyard in front of a stately villa and
comes through an archway at the entrance to the courtyard. She stops,
looks around and hearing her father's voice, moves toward him.7
"Sunny summer day," "a twenty-five-year-old brunette" and
"stately villa" specify in words what film can only tacitly show in
images of the setincluding the villa under certain lighting condi
42
Coming to Terms
45
Coming to Terms
stands a husband putting on a tie, his wife hovering behind him, but
they are too far away for us to make out their faces or conversation.
In another film these figures might have come forward as characters; a
closer shot would reveal their features. But here they are passed over
quickly, and we infer that they are irrelevant to the plot; we
hypothesize that they are not characters but walk-onsmere furniture in the tableau of "Early Morning in a Manhattan Courtyard."
The camera next tilts down a bit, revealing a passageway between the
buildings and the street beyond, and finally it comes back into the
window from which it set forth.
If the film were to continue only in this way, it could end up as a
descriptive documentary about life in a Manhattan courtyard. However, immersed as we are in the elaborate institutional context of
"seeing a Hollywood movie" (the theater marquee, after all, is alight
with the names of James Stewart, Grace Kelly, and Alfred Hitchcock), we reject that conclusion. We are virtually certainwe were
certain before we bought our ticketsthat the film will present a
story.
Inside the apartment, a close-up of James Stewart's face slides into
view; but since he is facing away from the window with his eyes
closed in sleep, we understand that the camera's meandering look at
the courtyard has been its own descriptive act. The description has
taken place in "real" time; however; there has been no pause. What
was shown was not a frozen moment but one filled with actions.
Story time has passed, even if nothing of great significance has
happened. A verbal paraphrase might read (retrospectively), "While
Jeff sleeps, the courtyard comes to life: cats prowl, husbands put on
ties," and so on. Hence, it is reasonable to say that a cinematic
describerthe cinematic narrator as describerexplicitly presents
the opening sequence of Rear Window.
Descriptive sequences can also reestablish locale. Metz cites an
example in Adieu Philippine (1962): as the film cuts from Paris to
Corsica, "five consecutive shots articulate a descriptive syntagma . . .
of life at the Club Mditerrane: men and women in bathing suits,
pareos, and leis are sunbathing, sitting at the bar, walking around or
dancing to aggressive music." 16 The reestablishing sequence is very
common in classical Hollywood cinema; an example in Notorious is
the aerial view of Rio just before the Ingrid Bergman character
arrives to infiltrate the German spy ring.
46
Coming to Terms
Addison De Witt." The camera pans right and pauses on his neighbor, whom he introduces as Karen Richards (Celeste Holm); then on
her husband, Lloyd Richards (Hugh Marlowe), author of the play in
which Eve stars; on Max Fabian (Gregory Ratoff), the producer; and
on Margo Channing (Bette Davis), another famous actress. De
Witt's voice-over explicitly describes and judges these characters as
each face fills the screen in close-up: Karen belongs to the theater only
by marriage; Fabian is not dedicated to art and is interested only in
"making a buck"; but "Margo is a "truly great star."
At the same time, the camera is providing its own tacit but independent description. As De Witt speaks of Max's devotion to the
buck, the camera gives evidence of Max's dyspepsia (he pours antacid
powder into his glass of water) and boorishness (he falls asleep at the
table). As De Witt recounts Margo's first stage appearance, as a child,
in A Midsummer Night's Dream, the camera shows her smoking and
drinking heavily (she disdainfully brushes aside soda water, preferring her whiskey straight). The sequence illustrates how richly the
camera can interact with the voice-over's specifications, anchoring
them but also potentially undermining them (for example, Karen
proves to be more sympathetic than Addison suggests).
So far De Witt has given us only explicit description in story
timethe chairman is up on the dais throughout, endlessly recounting the history of the society. This sequence structurally resembles
first person or homodiegetic literary narration: a person who is both
narrator and character is explicitly describing other characters preliminary to the story proper. But unlike literary narration, with its
distinction between the discourse moment's "now" and the story
moment's "back then," hereas in cinema regularlythe story time
continues its march contemporaneously with De Witt's descriptive
discourse.
Once De Witt's voice-over ceases and the voice of the chairman
comes on the sound track, the text becomes totally shown, no longer
(partly) told. The cinematic narrator has taken over the showing, and
the two tracks are conventionally synchronized. As the old actor
sings Eve's praise, the camera registers the reactions of De Witt and
Margo: De Witt smirks derisively, Margo seems no less cynical. The
chairman finishes his encomium, bravos fill the hall (but Karen and
Margo do not applaud, and De Witt does so tepidly), news photographers crowd the dais for pictures, and at last we see Eve (A nne
48
49
Coming to Terms
which the Encyclopdie calls the "simple indication of objects"? How
can it seem to pause to "contemplate" these objects: that is, to provide a "vivid and animated exposition of their most interesting properties and circumstances" for themselves alone, and not merely as
byproducts of the action? This kind of cinematic description would
seem to occur only when persons or objects appear on the screen not
primarily to serve plot needs but to reveal their own properties.
My impression is that something like this effect occurs in films
deviating from predictable rhythms. In general, narrative films set
up a rhythm correlated to the action, a rhythm that is followed more
or less relentlessly. Great efforts are made in classic Hollywood
films to adjust the pace of editing to the mood: the cuts in a suspense
thriller are taut and rapid, those in a moody melodrama languorous
and slow, with the dissolve much favored over the straight cut. 20
These rhythms set up conventional expectations. And the basic convention is the "literal" correspondence of discourse time to story
time, Genette's "scenic" mode. 21
Once established, this becomes the norm, a kind of zero-degree
temporal order that is understood by the spectator to correspond to
the duration of the events themselves. To change this conventional
dispositionto give a sense that events are being summarized, for
examplethe cinematic narrator must use some artifice. Sometimes
the device is all too artificial: the peeling calendar and moving clock
hands of early films or, with more artistry, the swish-pan sequence at
the breakfast table in Citizen Kane to show the gradual breaking up of
a marriage.
But my concern is the opposite of summary; it is pause, which
occurs when story time stops, though the discoursive statement
continues. Pause is most frequent at the beginnings of films, many of
which presuppose a kind of diegetic zero. There is no movement and
hence no suggestion that story time has begun. The opening shot
may be a still-like photo from a great distance of a buttercup-covered
field or the Manhattan skyline. The film has started but not yet the
story. Moving in from a height corresponds to the omnipresent
narrator's panoramic view over the story world which begins many
nineteenth-century novels. The industry calls these "establishing"
shots; the very word is a sign of their institutionally conventional
character. In such films as The Lady Vanishes (1928) and Psycho (1960),
the camera moves down from a bird's-eye view of an unmoving
50
Coming to Terms
scriptive intent is made clear by the "official" voice of the newsreel
commentator.)
Ordinarily, camera movements occur simply to expose the action;
horse and rider race to the scene of the battle with the Indians, and the
camera naturally follows along. Or the panning camera functions as
the surrogate of the moving eyes of the protagonist, in a manner
analogous to that in which description by a character functions in the
nineteenth-century novel. To show that she scans the horizon, the
camera also does so. We infer that characterial point of view (or filter)
is the purpose of the shot because the character is standing nearby,
and we identify with her interests.
Now one way for the filmmaker to interrupt story time and to
render description explicit is to deviate from such conventions as
cutting rhythm and eyeline match. Interesting deviations occur in the
work of Michelangelo Antonioni, whose films explicitly describe the
story world both by camera movement and by camera stasis. For
instance, the beginning of The Passenger (1975)24 provides descriptions of the Sahara by unpredictable movements of the camera.
Rather than standing in for the protagonist, David Locke, the camera
seems to be wandering on its own (in an interview at the time of the
film's appearance, Antonioni said that he no longer believed in the
subjective camera). The effect is one of spatial disorientation. First
the camera shows Locke scanning the horizon; then it begins a broad
pan, as if it were looking for something with him. But its movement
proves not to correspond to that of his gaze because it ends up
"finding" him, too, almost accidentally, in a completely unexpected
spot. In shot A he's on the left side of the frame looking right; in shot
B the camera moves across the landscape from left to right as if
continuing his glance; but in this movement it discovers him over on
the right, not at all where we expect to see him (there is no evidence of
a 360-degree turnabout). So we understand, though only retrospectively, that the movement is not an extension of his glance at all but
an independent view. The camera seems to be conducting its own
inquiry. We conjecture that it is concerned with the desert not as the
place where the plot events are occurring (the movement of rebel
troops in the mountains of Chad) but as an object of independent,
perhaps "painterly" interest. In Antonioni's desert our conventional
sense of the character-eyeline match is consistently undermined; we
52
Coming to Terms
duration. Here is the context: the heroine, Giuliana (Monica Vitti),
the unhappy and neurotic wife of an engineer, has taken a brief trip
with her husband's friend Corrado (Richard Harris) in search of
workers whom he hopes to hire for a Business venture in Patagonia.
They meet a technician, and Giuliana leaves the two men to discuss
their business. When she rejoins Corradoonly to learn that the
technician has refused the offerboth characters walk off-screen.
But instead of cutting away with them to the next shot, the camera
lingers on the background, making it for several seconds the main
subject: an antenna tower and a dull, vague old building.27 The lines of
the orange tower extend from top left to lower center of the frame; the
dark gray oblong of the building stretches vertically off-center right,
the portion at the far right carrying the eye off-frame at a very
shallow angle. Tying the two together is a cable forming a counterangle from top right to bottom left. The composition is that of an
abstract painting, made so by the geometric simplicity of the objects,
which are rendered even simpler by the flattening effect of the telephoto lens. The shot ends the scene. The next scene begins elsewhere
and at a later story moment.
For a few seconds, story time "dies," to adopt the term French
critics were quick to apply to the techniquetemps mort. It dies
presumably because the film loses interest in the story's relentless
need to flow. The effect is so intrusive that if it were not for its
recurrence (and its beauty), we might write it off as a slip in the
editing or an anomaly of the print. What is the intention? We can
imagine the cinematic narrator responding: "I find the composition
interesting, so I choose to linger on it. That is, I describe it." This
momentary halt in the story is an invocation of something timeless,
not necessarily "eternal" in the poetic sense but temporally indifferent. As such, it conflicts with our desire to have the story move
ahead. This momentary frustration of the viewer's hunger for diegetic resolution seems one more instance of that self-conscious meditation on its own structures that figures so prominently in Modernist
and Postmodernist art.
Antonioni's temps mort is a descriptive subversion of cinema's relentless narrative drive, a drive that leads E. M. Forster to lament
"Yesoh dear yesthe novel tells a story. That is the fundamental
aspect without which it could not exist. That is the highest factor
common to all novels, and I wish that it was not so, that it could be
54
55
CHAPTER
4/
Argumentation in Film
sive," with arguments working "by appeal to example" (in the technical rhetorical sense of the word): "The fabula world stands for a set
of abstract propositions whose validity the film at once presupposes
and reasserts." And Bordwell and Kristen Thompson have demonstrated the argumentative structure of Pare Lorenz's documentary
The River.1
The latter is not surprising; it is a common belief that argument
occurs in explicit form only in documentary films, that somehow it is
incompatible with narrative films. Christian Metz even proclaims:
"Remove 'drama,' and there is no fiction, no diegesis, and therefore
no film. Or only a documentary, a 'film expos' " 2as if documentary were doomed to an inferior place in the hierarchy of cinematic
genres (perhaps just as Description is thought to be inferior to Narrative). But that seems to take too narrow a view of the possibilities of
Argument in cinema. It may go against the grain of the medium to
deal in formal syllogistic or enthymemic structures, but certainly
many fiction films entail argument of an informal sort.
Cinema evolved, of course, as a medium for popular narrative
fictions"feature films," as the trade calls them.3 Though it may be
true in theory, at least, that the visual media are textually neutral, the
appeal of cinema as an institution has been primarily narrative
fictional, and it was the public's clamor for stories that spawned the
vast industry we call "the movies." Still, especially from the 1920s
through the 1950s, many Hollywood films delivered or at least
implied messages that we can think of as "arguments" in a weak sense
of the term; some still do. It is not hard to understand, after seeing
The Birth of a Nation, an "argument" against carpetbaggers and freed
slaves; or, after Intolerance, an "argument" against intolerance. Similarly, Public Enemy "argues" that crime doesn't pay; It's a Wonderful
Life, that the world would be a far worse place if even one of us had
never been born; Star Wars, that we should trust in the Force; and
Country, that the FHA should not evict farmers, no matter how
broke they are.
But these arguments are generally conveyed implicitly by the
narrative. The more sophisticated films become, the less often do
characters or voice-over narrators explicitly argue a film's thesis. The
exceptions are often of a spoofing or parodic cast, like the moralizing
of the voice-over narrator at the beginning of the film Tom Jones.
Only early and more primitive films seriously inserted mor als
57
Coming to Terms
("Crime does not pay!") in so many words at the end. The big
Hollywood studios have rarely found any profit in explicit and
formal argumentation. 4 Nor can the narrative preference be explained away simply as one more capitalist plot: though Sergei
Eisenstein dreamed of a film version of Das Kapital, all his films are
narrative. Films are expensive to make, and socialist no less than
capitalist societies have devoted the lion's share of their production to
films that will appeal to large audiencesnarrative-fiction films.
Nevertheless, the industry's historical devotion to Narrative is not
a good reason for theoreticians to ignore the other textual possibilities latent in film. One fruitful sort of inquirywhich I shall not
pursuewould be to outline or taxonomize the technical means by
which film could argue in ways correspondent to formal verbal arguments. It would consider such questions as the relations between shot
juxtaposition and the argumentative proofs traditionally distinguished by rhetoric: deduction, induction, analogy, and the like.
Perhaps one could devise a scheme for argumentative film parallel to
the Metzian grande syntagmatique of narrative articulations. This chapter, however, is concerned, not with the general question of what
might be done but rather with what has been done. The cinema, at
least in the hands of some directors, has already shown unique
potentials for introducing Argument into the textual mix, a potential
that arises in part from film's multichannel capacity. Unlike a written
text, a film can utilize different channels for different text-types,
presenting, say, Narrative in the visual channel and Argument in the
auditory, or vice versa. The results are uniquely its own.
Although early feature films contained a good deal of didacticism
and moralism, American and European cinema, in keeping with
the growing sophistication of the audience, have generally either
avoided "messages" or presented them very obliquely. Though contemporary filmmakers remain preoccupied, if not obsessed, with
social issues, they are leery of using narrative at the service of argument in any obvious or simplistic way. Hence the multichannel
capacity for mixing text-types is not often used. Arguments may be
presented by characters, but then they remain within the story. Few
contemporary films use the discoursefor example, through a narrator expressing himself in titles or voice-overto argue explicitly.5
Films, of course, are notoriously ridden with ideology, and ideology
announces itself in all kinds of ways. But those ways are mostly
58
Argumentation in Film
Coming to Terms
cinema were ipso facto a bad thing. The critic for Le Monde, for
example, accused the film of "profitlessly inverting the relations of
knowledge and fiction" by using the fiction to illustrate the argument; this, he asserted made it "the worst of didactic films."8 "Didacticism" here seems to be another name for explicit argumentation,
and some find that in itself antifilmic.
Others saw the film as belonging to the fiction, with Laborit
simply there to elucidate. That interpretation also misses the point.
On a more sophisticated reading, it seems precisely the tension
between the text-types, their struggle for control, that makes the
film interesting and even tantalizing. Which dominates, the argument or the narrative? ("Dominates" is a suggestive word here, since
the film's overall preoccupation is domination in animal and human
behavior.)
Laborit's opening argument is accompanied by a collage of photographs, each momentarily illuminated by a light moving in a counterclockwise direction. The light seems to embody the process of
scientific inquiry; it is like a microscope trying to find order in an
apparent chaos of images. The sound track plays a cacophony of
fragments of dialogue, out of which the voices of the three protagonists finally become discrete. The resolution of sounds and images
seems almost a matter of chance, as if these particular human subjects
were selected randomly out of a general population, implying that
we would see the same patterns in other randomly chosen individuals
of the species. New images appear which look abstract at first, then
become recognizable as small shore plants. The scientist's voice-over
tells us that plants can stay alive without moving around, acquiring
their sustenance by photosynthesis, but that animals need a space, a
feeding ground.
Cut to close-ups of a teaspoon, a doorknob, an ink bottle, a pair of
scissors, a bicycle chainold-fashioned household artifacts, souvenirs of childhoodas the voices of the off-screen protagonists
detail their place and manner of birth, but only in sentence fragments:
"Brittany," "Paris," "les Mauges," "in a hospital," "on an island,"
"Avenue de la Rpublique," "it was a neighbor." Though the switching back and forth among the voices slows down enough to make the
characters distinguishable, the principle of piecemeal narration extends throughout the film. At this stage the crosscutting seems to
facilitate Laborits disquisition (Narrative at the service of Argument)
60
Argumentation in Film
on the proposition that science deals not with individuals but with
species and that species are best described through random samples.
The randomness disappears as the story comes into its own, as these
characters become very much involved in each other's lives. That is
one of the sources of textual tension.
The three first appear visually in medallion inserts, telling their life
stories not dramatically but in testimonials. So far, they are subjects
of scientific inquiry rather than characters. They address the audience
directly and dryly, as if from some place outside the story. The
presentation is formal, emotionless; these are not fictional confessions but clinical discourses.
Over shots of a trudging tortoise, then a frog, then a goldfish,
Laborit explains that the need of animals to search for food inevitably
gives rise to a sense of territoriality. The intercutting makes it increasingly clear that the plants and animals are presented not as
metaphors or symbols for the characters but rather as their biological
peers; both images and commentary remind us that we all inhabit the
same ark.
There follow vital statistics about each of the three case studies,
read by the colorless, bureaucratic voice of a speakerine (a French
neologism for a female voice over). Born to a visibly affluent family,
first trained and employed as a teacher, then appointed news director
for French National Radio and fired in eighteen months, Jean Le Gall
has written a best seller attacking the National Radio and is now
running for political office. Janine Garnier's background is urban
proletarian. She is the daughter of a blue-collar father and a mother
who deplored her interest in the theater. Despite initial success as a
stage actress, she has drifted into the sterile profession of fashion
consultant and troubleshooter for an industrial conglomerate. Ren
Ragueneau is of peasant origin; by correspondence course and
against his father's wishes, he has risen through the ranks to become
manager of a textile plant. The very diversity of the characters'
backgrounds, highlighted by the crosscutting, reaffirms a random
hence "scientific"approach quite indifferent to such human concerns as social class.
Separating these sketches are brief speeches by Laborit outlining
his general theory, often over images of the struggles of other animals
lo survivea disoriented puppy, a wild boar rooting for food. The
disquisition of Laborit, the bureaucratic voice, and the visual images
61
Coming to Terms
Argumentation in Film
impaired. (We assume that this experiment occurs not inside the
story world, the world of Le Gall, Garnier, and Ragueneau, but out
in Laborit's argumentative discourse.) Another experiment demonstrates fight behavior: a second rat is introduced into the electrified
cage and the door is kept closed. Despite the repeated shocks to both,
the rats remain healthysimply because they vent their frustration
about their inability to escape by fighting each other.
In humans, however, these behaviors and the drive for territorial
domination are not innate. There are no proprietary or dominating
instincts as such; rather, domination is a behavior learned in the first
two or three years of life. Laborit considers the newborn human
nervous system pretty much a tabula rasa. The need to dominate is
imprinted on our consciousnesses by significant others; indeed, he
says, "we are nothing but the others." The human brain is divided
into three layersthe "reptilian" (concerned only with food and
reproduction), the "affective" (concerned with remembering what is
pleasurable and what painful), and the associative cortex, which
rationalizesthat is, provides "alibis" forthe drives of the other
two. The cortex's crowning achievement, the institution of language, exists, biologically speaking, only to rationalize our more
primitive needs and behaviors.
Laborit metaphorically characterizes the unconscious as a sea and
language as its mere foam, a foam that masks the true undercurrents
of our behavior. The behavior of the fictional characters illustrates
the point. Beneath their all too familiar personal peccadilloes and
domestic and business battles, we can see the broader biological
economy driving their (and our) behavior.
After explaining the four behaviors, Laborit describes the limitations placed on men and women by their situation in society. Human
culture, whose purpose is to secure the cohesion of the group, stringently controls the degree to which we can either flee or fight. This
control subjects human beings to a level of inhibition far beyond that
experienced by other animals. Because our aggressions tend to be
interiorized, we are more susceptible than the other animals to psychosomatic illness. Because overt physical struggle, like that between rats, usually won't do in "civilized" society, many humans
"somatize" their frustrations, "directing their aggression against
their own stomachs where it makes an ulcer." They may even perform the ultimate self-aggression of committing suicide.
That is the gist of Laborit's speech, a speech that accomplishes one
63
Coming to Terms
of two functions, depending on our textual perspective. If we read
the film as an argument, his commentary is obviously the major
device for its presentation. Shots of Laborit in his laboratory constitute a visual representation of a man arguing a certain case. Shots
of the rats' reactions are simply illustrative evidence, and so is the
behavior of the human beings. If we read the film as a narrative,
however, Laborit becomes a commentator in the discourse. But he is
an unusual kind of commentator: he is not an interpreter of the
fictional story. He never mentions the fiction, neither the characters
nor their actions. The uniqueness of the film turns on the fact that the
implied author has divided up the discourse functions, parceling out
to different agents tasks that would, in the traditional novel, be
performed by the narrator alone. (For a literary counterpart, imagine
two different discourse agents in Tom Jones: one who only tells the
events of the story, and another who only makes philosophical,
moral, and social comments without any direct reference to the
story.) From the narrative perspective, the laboratory, the rats in the
cage, and so on, are part of a discourse that is parallel to but not
somehow inclusive of the fictional story. 9
Still, as each text-type follows its own course, our moviegoer's
need for coherence prompts us to make connections. It is we who
will read the child Jean's search for crabs as an example of consumption behavior; Ren's lecture to a young worker on how to succeed in
business as an example of gratification behavior; and the adult Jean's
refusal to confront his friend's complicity in his dismissal from his
job as an example of fleeing behavior. When Le Gall cannot get away
quickly enough from the unpleasantness and the blunter Janine attacks Michel on his behalf, we interpret Jean's resultant kidney-stone
attack as an example of inhibition behavior.
Though Laborit's argumentative discourse makes no reference to
the story, the narrative discourse brings in elements of the argumentsometimes in a humorous way. In a reprise of the scene in
which he leaves his wife Arlette for Janine, Jean's body is crowned
with a rat's head. Later, symmetrically, a real rat is let out of a
miniature replica of the Le Gall apartment and into one of Janine's
apartment. To reinforce the fight instinct felt by Rene when his rival,
Veestrate, encroaches on his territory, there is a shot of the two of
them, also wearing rat heads, wrestling on the desks of the cramped
office they share.
64
Argumentation in Film
The story as such is presented in parallel, crosscut tiers: each stage
of lifechildhood, breaking out of the family, early success, mature
distressoccurs concurrently "across" the characters. In the childhood stage, Jean hides in a tree so that he can read pulp novels, in
particular Le Roi d'or, a French Horatio Alger or legendary "American uncle" story; Janine gets strokes from her family by standing up
at festive dinners and reciting Communist paeans; Ren studies accounting by the light of a bare bulb hanging from the ceiling of his
bedroom. At the adolescent stage, Jean is dissuaded from going to
Paris by his girlfriend's imperious need for an early marriage. Janine
breaks from her family in the middle of the night after her mother
tries to thwart her activity in amateur theater. Rene, disgusted with
his father's petty and outmoded farming methods, leaves with his
fiance Thrse in the middle of a family dinner. At the adult stage,
Jean becomes a radio news director, Janine wins a role in a small
theater production, then leaves it to become an executive in a fashion
conglomerate; Rene rises to assistant manager of a textile factory in
Lille.
Then the troubles begin: Jean falls for Janine at a performance of
her play; he moves in with her, over his wife's hysterical protests.
Then he is fired from his radio job, for obscure reasons. Ren's firm
merges with another, and he sees his job taken away by a tougher and
wilier rival, Veestrate. Janine, tricked into believing that Jean's wife is
dying, pretends to quarrel with Jean and sends him back home. The
wife does not die but manipulates the malleable Jean into returning to
her. Rene is transferred to another factory, hundreds of miles from
his family, but he cannot keep up with the technical and commercial
demands. His boss proposes that he go into one of the conglomerate's new linespackaged gourmet foods. Though cuisine is his
hobby, Rene is humiliated and enraged by the proposal and tries to
kill himself. Fortunately, he is rescued; the last we see of him is in a
hospital bed, where he has just come to and is not yet aware that he is
out of a job.
The relation between Janine and Jean also ends on a desultory note.
Two years after their breakup she rows to his island, which he never
invited her to visit. By chance he also rows there, to get some papers
concerning his legal claim to the island (territory again). They meet
and reminisce, but he shows no sign of wanting to resume their
affair. Janine goes on to her business meeting. Unnerved by Ren's
65
Coming to Terms
attempted suicide, she returns to Brittany to tell Jean that she only
provoked their breakup to let him return to solace his wife in her "last
days." She finds Jean hunting boar with a neighbor (exercising territorial dominion in good haut-bourgeois style). She confronts him
with the truth. To her dismay, Jean's wife has already confessed, but
Jean admires his wife's "marvelous" courage in telling such a lie, and
Janine's "marvelous" self-sacrifice in giving him up. Pushed beyond
endurance, Janine assaults him, and the story ends with the two in a
clumsy fistfight.
So the three characters who seemed randomly picked as biological
samples at the beginning of the film become intimately involved in
the fictional plot. But the copresence of Laborit's argument restrains
us from indulging too fully in the narrative code for its own pleasures. Rather, we are invited to seek in the fiction illustrations of
Laborit's argument. And indeed, Laborit's discourse seems to account
for the behavior of the characters. When he speaks of an infant's
reaction to the mother's caresses as an example of gratification behavior, or of the actor basking eagerly in the applause of the audience, we see shots of Jean being hugged by his mother and Janine
being praised by her family after reciting at the dinner table.
Other visuals illustrate the argument more subtly. For example,
when a framed photograph of Janine's grandparents accompanies
Laborit's discussion of inhibition, we infer that the old and traditional
are by definition conservative and hence inhibitory. Or, as Laborit's
voice-over explains that "language only contributes to hiding the
cause of dominance," we watch Rene follow a beautiful secretary
through the overwhelmingly grand lobby of the conglomerate's
headquarters. Here "language" obviously means all sign behavior:
the dcor serves the double function of impressing the clients and
intimidating the employees.
The film ultimately trains us to see past "linguistic alibis" on our
own, to break our usual habit of reading events empathically through
fictional characters' eyes and instead to interpret them through Laborit's argument. Though Laborit never mentions the fictionin
fact, seems unaware that he is sharing textual space with itwe find
ourselves eager to apply his laboratory findings to its interpretation.
We are not taken in by Jean's "reasons" for going back to his wife,
though he obviously prefers Janine; or by Janine's explanation of why
she abandoned the theater for the colorless world of business (I've
66
Argumentation in Film
Coming to Terms
Argumentation in Film
out for the owner of the strange boatcut to the rat attentive, ready
to flee.
What makes Mon oncle d'Amrique a fascinating departure in film
history is the tension, the "discrepancies," to be found between the
two text-types that inform it. The story often stretches out beyond
the requirements of the argument, taking on its own buzzing interest. In doing so, it revels in what Laborit's science dismisses as
"alibi," precisely the lived detail that makes realistic narrative fictions
persuasive. It is not always clear whether the narrative elaboration is
a part of this particular text's intention or a consequence of the
medium in which it is transmitted. To tell a story at all in the cinema,
even one that illustrates some other kind of text, may be precisely to
introduce "excess," to open floodgates of "distracting" but absorbing
story material. "Excess" is clearly an effect of verisimilitude, of
Marthes's effet de rel, but it fits in nicely with the correspondent
argument. Janine's "nobility" in giving Jean back to his wife, for
example, might be explainable in Laborit's theory as biological gratification: she is trying to replicate the gratification she got from her
father for being "fair" or "comradely." "Excess" also names the
audience's vicarious gratificationits fulfillment of its own biological agendain experiencing the human gratifications of others,
even of fictional characters. Despite the training in biology that we
get from Laborit, we desire to know more about the story, even as we
sense that its presence in the film is more than a little whimsical.12
It is easy to get caught up in narrative "excess." Critic Jan Dawson
marveled at the beauty of the photographed natural settings; the mere
fact of this beauty, she implied, put the lie to Laborit's purely functional view of nature.13 But just as everything in Laborit's argument
can be used to explicate the narrative, everything in the narrative can
be used to illustrate the argument. The "beauty" of nature also has a
biological explanation. Man chooses among territories, migrates to
certain places on this earth which he finds desirable; their desirability is
understood by the biologist in terms of survival value. "Beauty" is
only the name given to that desirability. A race that evolved on Mars
would doubtless find its red dust more beautiful than our verdant
forests. Because the audience of Mon oncle d'Amrque is torn between a
traditional conception of beauty and a new functional definition, the
story element of the film seems stretched between two poles, at once
in service of the argument and fiercely autonomous.
69
Coming to Terms
In one way, diegetic "excess" is a consequence of narrative elaboration itself. It is in the nature of realistic narrative, whether invented
for illustration or for its own sake, to create details of independent
interest, effets de rel. Cinema enhances this interest by favoring richly
detailed sets, elaborate acting business, and the like. For example, it
isn't enoughnarratively speakingfor Ren to express his anguish
by trying to kill himself. The cinematic imperative requires a complicated photogenic method: he tries both to strangle himself (on the
window handle, not from the chandelier) and to take an overdose of
barbiturates. The purely visual representation of his determination
combined with clumsy ineptitude endows the sequence with such an
air of particular reality as to make us almost forget that it is introduced
to "illustrate" the Laboritian "anguish."
In more uniformly narrative texts, we look for explanations of
problematic details in hypothetical extensions of the story. But in this
film we must decideand that is the textual strainwhether the
explanations lie in the fictive world or in biological theory. Events of
psychoanalytic import, we conclude, can be incorporated under a
more general biological principle. Jean's oedipal complexexemplified by shots of him being hugged by his mother and trying to shoot
his father with a toy arrowis simply one more kind of territorial
stakeout. Human beings are love "objects"but now in the context
of the theory of territorialism. They are like pieces of real estate
rather than, say, gold bars. Gender matters little here: dominatrixes
are as common as dominators. His wife gets Jean in the first place by
staking her claim loudly and persistently (Jean in his initial medallion
appearance says, "Of all my girlfriends, the most decided was Arlette"), then keeps him by cunning, feigning illness. With our newly
acquired biological insight, we interpret feigning as a form of fighting.
Psychological inheritance too can be explained in Laboritian
terms. Rene reacts, by fight and flight, against his father's outmoded
production techniques, but it is his fate to watch himself displaced in
turn by Veestrateon the same grounds. "Nous ne sommes que les
autres""we're nothing more than those around us." He accuses his
father of provincial small-mindedness and paranoia, but when he
gets "kicked upstairs" to head a new factory, he asks, like the veriest
rube, "But is it a trap?" And if the requisite traits do not emerge in a
character's own behavior, he manages to marry someone who can
70
Argumentation in Film
Coming to Terms
The narrative text ends in a quite unromantic and even distasteful
debacle. But the argumentative text tells us something quite different, that though such struggles are absolutely useless for resolving
frustrations, they do have a purpose: human beings, like rats, need to
fight. It keeps their eyes clear and their skin rosy as nothing else can.
Still, Jean and Janine do not look very happy or healthy. Is there some
disanalogy between the external surfaces of aggressorsclear eyes
and rosy skinand their inner surfaces? Here again, the biological
argument is silent.
The final story event does not end the film. There are reprises: the
humble domestic objects, the spectacle of Ren and Veestrate battling
for the telephone, Jean shouting "Vite, vite" for his hot compresses,
the turtle on its back, and finally Laborit in his office giving us his
final message, which is inconclusive: as long as the need to dominate
moves human beings to crucial action, as it does the animals, there is
little chance that anything will change. Cut to a traveling shot of an
abandoned slum that looks very much like the South Bronx. Following Laborit's last words, these images might seem to illustrate his
point: the slumlords dominated the tenants, and (presumably) the
tenants, when it became more than they could bear, abandoned the
neighborhood to seek another territory. But the camera comes to rest
finally on the outside wall of a building upon which someone has
painted a forest. It approaches the image of a single large tree, then a
part of the tree, then closer still until all we can see are traces of paint
on a few bricks. The effect is like that in Antonioni's Blow-Up: the
closer one observes something, the more amorphous it becomes,
until its original outline vanishes. 15
What are we to make of this ending? Clearly, it is as much figurative as literal, since the neighborhood and the building do not at all
share the story world of Le Gall, Garnier, and Ragueneau. On the
other hand, they do not particularly support Laborit's discourse. The
cinematic narrator could well be imagined to be saying, "Thanks,
Dr. Laborit, but all your scientific theorizing leaves us no clearer than
we were before. How do you account for the inhabitants' longing for
nature's beauty in this urban squalor? Nor does your science help us
solve the political and moral and economic problems that have led to
this desolation. And what can we make of your message? It's all very
well to say that as long as we want to dominate others, nothing will
change. But if that's the way our brains are constructed, how can we
72
Argumentation in Film
hope to do anything about it? How do you propose we change our
brains?"
There seems to be nothing in the film to help answer these questions. A final quandary haunts us as we leave the theater, resulting
directly from the unresolvable struggle between the text-types, Narrative and Argument, for control of the film.
73
CHAPTER
5/
In Defense of
the Implied Author
THOUGH THE NOTION of an "implied author" has enjoyed widespread currency since Wayne Booth published The Rhetoric of Fiction,
(1961) narratologists do not agree about its precise meaning or even
its raison d'tre.1 Few reject the distinction between real author and
narrator,2 but some wonder why a third, seemingly "ghostly" being
should be situated between the two.
I believe that narratologyand text theory generallyneeds the
implied author (and its counterpart, the implied reader) to account
for features that would otherwise remain unexplained, or unsatisfactorily explained. The implied author is the agency within the narrative fiction itself which guides any reading of it. Every fiction contains such an agency. It is the sourceon each readingof the
work's invention. It is also the locus of the work's intent. Following
W. K. Wimsatt and Monroe Beardsley, I use "intent," rather than
"intention," to refer to a work's "whole" or "overall" meaning,
including its connotations, implications, unspoken messages. 3 The
concept of an implied author ensures against simplifying the real
reader's relations with the text and reducing them, as some contextualist theories would, to one more instance of ordinary conversational interchange. I believe that each reading of a narrative fiction
reconstructs its intent and principle of inventionreconstructs, not
constructs, because the text's construction preexists any individual act
of reading. Though reader-response and other constructivist theories
correctly insist on the active quality of the reader's share, seeing it as
74
Coming to Terms
reading a text, though ultimately an exchange between real human
beings, entails two intermediate constructs: one in the text, which
invents it upon each reading (the implied author), and one outside the
text, which construes it upon each reading (the implied reader).
Further, the implied author is not the "voice": that is, the immediate
source of the text's transmission. "Voice" belongs uniquely to the
narrator. (The implied author of Ring Lardner's "Haircut" is clearly
not the morally obtuse narrator of the story, Whitey the barber.) We
find a parallel distinction in other text-types, in Argument, for example. In Swift's "Modest Proposal," the implied author is clearly not
the arguer. We cannot imagine the tract's implied author endorsing
the resolution of economic problems by cannibalism. And we would
make the same assumption even if the text were anonymous: that is,
had no identifiable real author. But the need for "implied author" is
not limited to ironic texts: even where there is no discrepancy between the implied author's intent and the narrator's or other speaker's
intention, the theoretical distinction is worth preserving because the
two terms account for different levels and sources of information.
Positing an implied author inhibits the overhasty assumption that
the reader has direct access through the fictional text to the real
author's intentions and ideology. It does not deny the existence of
important connections between the text's and the real author's views,
but it does deny the simplistic assumption that somehow the reader is
in direct communication with (I) the real author (with all the troublesome questions that idea raises) or with (2) the fictional speaker, for
how then could we separate the denotation (what the speaker says)
from the connotation (what the text means), especially where these
differ? However attractive it may be in theory, reading the text as the
speech act of the real author is too simplistic to account for the
semantic complexity of many texts. Despite the important influence
of speech-act theory on text analysis, we cannot successfully analyze
published fictions in terms designed to explain ordinary conversation. Fictions are more complex than such utterances as "Can you
reach the salt?" (whose illocutionary force varies according to its use: a
request in one context, a question about the auditor's physical state in
another). In conversation, the auditor can request feedback in order
to pinpoint the speech act, by asking, for example, "Does that mean
you want the salt, or are you asking me how far I can stretch?" But the
reader cannot query the author about the fiction's meaning or force.
As Roland Barthes put it, published fictions are "hardened." 6
76
77
Coming to Terms
tions and to the corresponding beliefs; that is, the propositions a
work expresses or conveys are expressed or conveyed by, and hence
attributable to, not the "implied author," but rather the real, historical author.9
Anti-intentionalists such as Monroe Beardsley reject the relevance
of original authorial intention, arguing that an interpretation should
derive only or at least principally from the text itself. 10 Beardsley's
remains the best discussion of literary intention. He questions the
interpretive relevance of the "psychological states or events in [the
artist's] mind: what he wanted to do, how he imagined or projected
the work before he began to make it and while he was in the process
of making it." At best, biographical material provides only indirect
evidence of what the object is, and indirect evidence is obviously
secondary to the evidence of a "direct inspection of the object." 11
Anti-intentionalism does not argue that the study of conventions
and meanings that prevailed during the artist's lifetime are irrelevant
or that the critic is misguided to search for them; to interpret Bach
well, one should know as much as possible about how music sounded
in his time. To interpret Milton well, one should know as much as
possible about seventeenth-century Christianity. Nor do the antiintentionalists suggest that the artist's statements of intention should
be discounted in interpretation. They argue only against the relevance of the sounds Bach might have heard in his own head, or what
Milton was planning as he composed Paradise Lostand only against
the interpreters considering them: there is no suggestion that biographers and historians may not do so.
The anti-intentionalist insists on distinguishing between what
words mean and what people mean. Texts obviously continue to
mean, long after their authors are dead. Authors sometimes mean
one thing but their texts another. Beardsley cites the case of a poem
by A. E. Housman called "1887": Frank Harris interpreted the poem
as ironically critical of Queen Victoria's easy dispatch of young men
to die in warsa not unreasonable interpretation, given the content
of some of Housman's other poems. But Housman himself rejected
an ironic reading of "1887." Similarly, Saul Bellow recently complained that some readers have misunderstood the comic intent of
Herzog "to show how little strength 'higher education' had to offer a
troubled man."12 The intentionalist would have to argue that "1887"
78
is not ironic and that Herzog is comic, because their authors say so.
The anti-intentionalist would counter that Housman and Bellow
may have misconstrued the meanings of their own work, that their
unconscious may have guided their pens more than they knew, or
that they simply did not achieve their intention. Whatever the reason, concludes Beardsley, "if [a poet's] report of what the poem is
intended to mean conflicts with the evidence of the poem itself, we
cannot allow him to make the poem mean what he wants it to mean,
just by fiat."13
Beardsley faults criticism whose language unnecessarily stresses
the author's intention when it only wants to describe what the text
says or does. He questions the validity of such pronouncements as
"According to Jane Austen, Emma is overbearing but basically wellintentioned," and "In Rear Window, Hitchcock implicates the viewer
in the protagonist's voyeurism," as opposed to the simple statements
"The novel Emma presents an overbearing but well-intentioned heroine," and "Rear Window implicates the viewer in the protagonist's
voyeurism." The former make unnecessarily intentionalistic claims;
the latter claim only that the primary evidence for judgments about
the heroine's character or the viewer's complicity inhere in the text
itself."
In more technical language, the anti-intentionalist distinguishes
between (in William Tolhurst's terms) "utterer's meaning," "utterance meaning," and "word-sequence meaning."15 "Utterer's meaning"
is what the author had in mind when creating the text. "Wordsequence meaning" is what the words or other signs signify by
linguistic and semiotic convention; in speech-act terms, it is the bare
"locution." "Utterance meaning" is word-sequence meaning plus a
context, the locution plus its illocutionary force. In other words,
word-sequence meaning is only the semantic potential of wellformed speech acts. Meaning as a practical matter cannot arise until
the text is "uttered" in respect to some perceivable or inferable state of
affairs. It must be interpreted by someone who understands the
semantic and syntactic rules by which the sentences or other semiotic
structures cohere and the context into which they plausibly fit. The
anti-intentionalist argues that what is chiefly relevant to interpretation is utterance meaning, not utterer's meaning. The reader makes
out the word-sequence meaning and tries to divine a context appropriate to the text. "This is not to say that utterer's meaning is irrele-
79
Coming to Terms
vant to a determination of textual meaning," concludes Tolhurst. "It
is just to say that it is not identical with it." 16
The anti-intentionalist position, no matter how venerable, has not
been successfully refuted. In an era when skepticism prevails about
the very possibility of knowledge, communication, and interpreta tion, it seems worthwhile to recall the sensible views of philosophers
such as Beardsley.
Booth's Argument for the Implied Author
The intentionalist debate helps clarify the question of the "implied
author." Let us consider five definitions of the term proposed by
Wayne Booth in The Rhetoric ofFiction.17 (These definitions are resurrected only as a point of departure: I do not mean to suggest that
Booth still conceives of the implied author in these ways. His later
work refines the concept considerably).
First, the neutral (or objective or ideal) person that the real author wants
to be in order to create an objective account. Booth notes that novelists often use some such term to describe their own process of composing:
Jessamyn West writes of an "official scribe" who discovers or creates
her appropriate stance to the story; Edward Dowden and Kathleen
Tillotson refer to the author's "second self." The real author, says
Booth, wants to avoid pouring "his untransformed biases into the
work," so he creates a neutral representative or surrogate author. 18
But this "second self or "official scribe" suggests a definition of
intention that not even sophisticated intentionalists would defend:
namely, something planned by the author prior to the moment of
composition, some alter ego that real authors take on to help them
establish a desired state of objectivity. 19 Real authorial behavior is a
subject for literary biography, not text theory. If the overcoming or
"transforming" of biases in the act of creation is all that "implied
author" means, why would we need the term? Wouldn't "real but
bias-free author" do the trick? What text theory wants to know is
how and whether "implied author" names something essential to and
derivable from the text itself.
Second, the different aspects of themselves that authors show in dijferent
works. Since this definition refers to a process by which the real
author decides which of several assumable personae shall inform her
80
Coming to Terms
immediate relation to it and thus in open communication with his
audience, a written text is closed until read, though it contains,
latently, its principles of invention and intent. Those principles are
available to and activated by each reader upon each reading. Intention
ceases being a private authorial matter: it becomes the work's intent.22 We must distinguish between a real author's activity and the
product of that activity: the text before us. It makes perfect sense to
speak of intentwhat R. W. Stallman calls "actual" intentionas a
property of the text, as "the effect which the work aims to evoke, the
organizing principle informing the whole, or the meaning which the
work manifests or suggests." 23 Or, in W. K. Wimsatts phrase, we
learn the author's "effective intention or operative mind as it appears in
the work itself." 24
One problem with Booth's use of "implied author" to recall that
the work is the product of "a choosing evaluating person" is that
many texts do not evoke such a person. Some texts, such as portions
of the Bible or traditional ballads, were created by anonymous authors working through many generations. Others, such as films
churned out by Hollywood, are made by a consortium of writers,
producers, actors, directors, cinematographers, distribution executives (who may not have totally agreed on the final product). Still,
the books (or certain sequences of books) of the Bible and traditional
ballads and Hollywood films seem to have been created by a single
author. That is because they are governed by the unified invention
and intent of the text: that is, their implied authors. Rather than
calling attention to the work as the product of a choosing, evaluating
person, I see the work as a repository of choicesof already made
choices, which can be considered as alternatives to other choices that
might have been made but were not.
Perhaps some of the resistance to the term might dissipate if we
stop thinking of the implied author as a human surrogate or image of
the real author. Hatching a third human being is unnecessary. Little is
gained by assuming an "image" of Mark Twain which is in any
important respect different from the real Twain,25 but much is to be
gained from recognizing the differences between the implied author
and the narrator Huck Finn.
Fifth, what I call the recorded invention of the text, which is in Booth's
words based on a "core of norms and choices" that inform it. For this
construct, Booth prefers the term "implied author" over such possi
82
Coming to Terms
selection and ordering of its components. It is these principles that
readers reconstitute, not the real author's original activity.
Of course, we may bring in other information, other contexts (in
the root sense of the morpheme: texts around or associated with the
text). Among such contexts is what we remember of other works by
the real author. The information about how to read the text is thus
"emergent." Though different readers may construct different implied readers, the process itself is common to us all. And that is why it
makes sense to call the construct of invention "implied" or "inferred": it remains only latent or "virtual" in the text until it is
actualized by our act of reading. Readers infer a self-consistent textual intent, rather than guessing directly at the real author's state of
mind. We recuperate the intentions of a "Dickens," not those of the
man Charles Dickens.
Even as he established these vital distinctions, Booth did not himself escape the potential confusion between the implied author, as
source or "instance" of narrative invention, and the narrator, as the
"utterer" or "enunciator." Occasionally, The Rhetoric of Fiction refers
to the real author when it seems to mean the implied author, that is,
to Dickens and not to "Dickens." Given the groundbreaking quality
of Booth's work, it is easy to forgive him this lapse, but we should
take pains not to replicate it. For example, Booth argues that the
implied authors of Tom Jones and Joseph Andrews exhibit a certain
facetiousness and insouciance quite unknown in Fielding's other novels, whereas the implied author of Amelia seems uncharacteristically
sententious and solemn. 27 But Booth's formulation can be interpreted in a clearer and more useful way, I think, if we distinguish
between who speaks the text and who (or the principle which) has
invented it. In this model the implied author is the inventor, and the
narrator is the "utterer": that is, the one who articulates the words
assigned to him by the implied author. Consider the narrator's meditation on Joseph Andrews's ancestry:
Suppose, for argument's sake, we should admit that he had no
ancestors at all, but had sprung up. according to the modern phrase,
out of a dunghill, as the Athenians pretended they themselves did
from the earth, would not this autokopros have been justly entitled
to all the praise arising from his own virtues? Would it not be hard
that a man who had no ancestors should therefore be rendered
84
Coming to Terms
Narcissus. The choice of narrators in that novel, as in other narratives, is
basically a rhetorical one (see Chapter u). Conrad's novel assumes
that the most suasive presentation of the narrative would entail a
series of imbricated frame stories, thereby distancing its strange
import from the workaday world.
For readers who feel uncomfortable about using the term "implied
author" to refer to this concept, I am perfectly willing to substitute
the phrase "text implication" or "text instance" or "text design" or
even simply "text intent"always on the understanding that "intent" is used to mean not what was in the mind of the real author bent
over a desk but what is in the text that we hold in our hands, or see on
the stage or the screen or the comic strip. It is a sense of purpose
reconstructable from the text that we read, watch, and/or hear.
Thus, my position lies halfway between that of some poststructuralists, who would deny the existence of any agentwho would
acknowledge only our encounter with critureand that of Booth,
who has spoken of the implied author as "friend and guide." For me
the implied author is neither. It is nothing other than the text itself in
its inventional aspect.
Some narratologists have suggested that the implied author is a
marginal concept, useful only where the text's intent is clearly distinct from the narrator's intention. The term they use to describe this
supposed marginality, a term taken from linguistics, is "marked": in
English, some nouns ("man," "woman") are marked for gender;
others ("voter," "driver") are not. For the unmarked word "voter,"
the "degree" of gender is "zero." But "marked" and "unmarked" are
not useful terms if they downplay a distinction unless it has some
immediately practical consequence, such as unreliable narration. The
proponents of this view argue that in most cases the implied author
and the narrator are indistinguishable and that dramatized narrators
are therefore special, "marked," and somehow deviant cases. This
too easily becomes the argument that since the narrator usually
reflects the implied author's view of things, we might just as well go
back to saying that the two are identical, that it is the implied author
who tells or shows the storyor, conversely, that there is no longer
any need for the notion of implied author at all. But such an argument is like saying that because the third person plural English
pronoun is unmarked for gender, the gender of individuals referred
86
Coming to Terms
Criticism has no name for these sustained characters who somehow are
the sum of the invented creators implied by all of the writer's
particular works. For lack of a good name, I shall call such a
sustained character (still different, of course, from the writer, with his
quotidian concerns, his dandruff, his diverticulosis, her night-mares,
her battles with the publisher) the career-author. [This is] the sustained
creative center implied by a sequence of implied authors. Implied
authors may remain fairly constant from one work to another by the
same author (as in Jane Austen's novels), or they may vary greatly, as in
the extreme case of J. I. M. Stewart, a scholar whose detective stories
require an "author" with an entirely different name: Michael Innes.32
Again, Booth has identified an important convention. But again
and at the risk of seeming ungratefulI must suggest a slight modification of his language. Do we really want to say that the careerauthor, this "sustained creative center," is a "character?" Let us once
more resist the anthropomorphic trap. We can comfortably define
the career-author as the subset of features shared by all the implied
authors (that is, all the individual intents) of the narrative texts
bearing the name of the same real author. The real author's name,
then, can be understood as the signifier of a certain constancy or
common denominator of method among the implied authors of the
various works. Its signified is the known subset of features, carried
over from other, similarly signed texts, which provides readers with
narratively significant information as they make their way through
the new text. It is what permits us to put Gravity's Rainbow in the
context of "Thomas Pynchon" if we have already read The Crying of
Lot 49 and V but know nothing about the life and opinions of the real
Thomas Pynchon.
It is a narratological (and not merely a biographical) fact that a text
is signed by a "Pynchon" or a "Hemingway" or a "James" (quotation
marks to stress these as signifiers, not signifieds). The reader will
certainly utilize such information. However, what is relevant to
narratology (as opposed to other kinds of literary study) is not the
history of the real author's career but rather the necessary constraint
on possible contents and styles implicit in his or her signature on the
text.
The notion of the career-author enables us to acknowledge nar88
89
6/
The Implied Author at Work
W ITHOUT THE implied author, narratology and literary criticism
lose an important distinction. The test case here is the possibility of
unreliable or "discrepant" narration. The narrator alone tells or shows
the text, and if we cannot accept his account, we must infer that it
belongs to someone (or something) else. If all meaningsimplicit as
well as explicitare the products of the text's activity, and if this
activity always presupposes agency, then we have to posit some such
text principle or agent as the implied author. Thus, it is the implied
authors of Ford Madox Ford's Good Soldier, of Ring Lardner's "Haircut," and of all the other "suspicious" novels and stories in Booth's
"gallery" 1 who are the sources of the "true" stories.
But we need to recognize the implied author also in texts that do not
overtly undermine the narrator's account. There are many texts
which, though narrated with total reliability, are difficult to characterize as the creation of some single real author. Yet each of these clearly
coheres as a unity, seemingly the product of a unified agency. Again
the term "implied author" is very useful to name this agency. The
wide (and wild) diversity of the following examples demonstrates
the broad relevance of implied authorship.
Authorship by Committee: The
Bible and Hollywood Films
The implied author names the convention by which we naturalize
the reading experience as a personal encounter with some single,
90
91
Coming to Terms
by any single person. E. A. Speiser, for example, draws a sharp line
between the authorship of Deuteronomy (D), on the one hand, and
that of the four preceding books, the Tetrateuch; and within the
Tetrateuch, between the so-called Priestly source (P) and the outright
narrative material. The latter, in turn, carne to be attributed to two
groups of authors, according to which name they assigned to God.
These are called by Biblical scholars J (Yahwist) and E (Elohist).
Further, P probably "was not an individual, or even a group of likeminded contemporaries, but a school with an unbroken history
reaching back to early Israelite times, and continuing until the Exile
and beyond." Biblical scholars postulate an additional group R of
redactors and compilers. Though there have been various other
proposals about the authorship of the Pentateuch, most scholars
agree "that the Pentateuch was in reality a composite work, the
product of many hands and periods." Further, though many parts are
confidently ascribed to one or another source, some are "so fused that
they may never be pried apart." 4 Still, many readers think of these
books of the Bible as a unified narrative and have little difficulty
speaking of their "implied author."
Turning from ancient Israel to our modern Babylon, we hear the
loud voice of Louis B. Mayer invoking the Holy Book (nothing less)
to defend the studio system of film authorship: "If a writer complains
about his stuff being changed," commented Arthur Freed, Mayer
"always says, 'The Number One book of the ages was written by a
committee, and it was called the Bible.' " 5 Though a familiar phenomenon, the collective creations of Hollywood films are likely to be
overlooked by narratologists who dwell too exclusively on literary
narratives. One well-documented case is The Red Badge of Courage
(1951), adapted from Stephen Crane's novel and directed by John
Huston for MGM. The checkered authorship of this film is particularly interesting, given Huston's usual inclusion in the "pantheon" of auteurs.6
When The Red Badge of Courage opened in New York on August
31, 1951, reviewers treated it as Huston's exclusive achievement. The
New York Tribune wrote that "Stephen Crane's 'The Red Badge of
Courage' has been transformed by John Huston"; the Morning Telegraph, that the novel "has been brought to the screen by the brilliant
John Huston as an offbeat motion picture"; and the World-Telegram &
Sun, that "John Huston has written and directed a stirring film in an
92
Coming to Terms
one American killing another American. Sometimes I bring phony
emotions into the picture to wake the audience up." 8 Many others
some less self-impressed about their importancealso contributed
to the making of the film, with or without direct instruction from
Huston: lighting technicians, set designers, costumers, makeup artists, and the like.
Nor did collaboration end there. To the consternation of MGM,
the film inspired negative comments from a sneak-preview audience.
Many viewers expressed indifference to the film's antiwar intention,
and there was laughter in the wrong places. Fearing a commercial
failure, Reinhardt added a voice-over sententiously proclaiming that
the film was based on a literary classic, implying that it therefore
deserved respect. The voice (that of James Whitmore) was also assigned the task of conveying the protagonist's thoughts, despite
Huston's express wish that those thoughts remain implicit in the
visuals. Even before shooting started, Reinhardt had asked Huston,
"The book is about the thoughts of the Youth. Will we show what
really goes on inside the boy?" Huston had responded confidently:
"Audie Murphy [the actor who plays the Youth] will show it, Gottfried." 9 But Reinhardt questioned Huston's artistic intuition, and his
doubts prevailed; the voice-over periodically intones, in Crane's very
prose, what is passing through the protagonist's mind.
The effect is quite bad, as a critic for the New York Times pointed
out: "The picture could not convey the reactions of Crane's hero
to war, for Grane had conveyed them 'in almost stream-of-consciousness descriptions, which is a technique that works best with
words.' "10 Actually, that is not what Grane did, and the problem is as
much the novel's as it is the film's. The Youth's thoughts are precisely
not presented in stream-of-consciousness style, nor even in indirect
free style, but rather in what Dorrit Cohn calls "psychonarration": 11
that is, paraphrases in the narrator's own prose, which tends to
intense purple. The novel is flawed by a discrepancy between this
well-spoken, self-consciously literary representation of the Youth's
thoughts and that of his actual uneducated speech. Whereas the
Youth says such things as
I got shot. In the head. I never see sech fightin'. Awful time. I don't
see how I could 'a' got separated from th'reg'ment.
the novel's narrator represents his thinking in sentences like this:
94
Coming to Terms
interpreted their input. These viewers were not some theoretical
construct but literal collaborators: their scribbled comments on evaluation cards, collected and nervously interpreted by studio executives, led to a substantial reworking of the film.
Yet despite this motley crew of real authors, the film The Red Badge
of Courage, for better or worse, gives an impression of unity so strong
that reviewers spoke persistently of a single authorial source: it was
"Huston's film"; it was Huston who "became discouraged with it,"
and the like. What they were talking about, of course, was not the
real John Huston but the film's implied author.
We are not limited to Hollywood films for examples of uneasy
collaboration. There is the famous misunderstanding between Chekhov and Stanislavsky about The Seagull: according to Stanislavsky,
the play is supposed to be a tragedy, but Chekhov "always referred to
it as a comedy." "Practically every producer," says Wimsatt, "in spite
of Chekhov's unmistakable intentions [to be found in his letters],
regards the play as a tragedy."15
The aesthetic problems caused by an excessive commitment to
individual "real authors" is no less evident in the study of masterpieces of painting. Consider the fascinating controversy currently
raging about Rembrandt. In The Transfiguration of the Commonplace,
Arthur Danto poses the philosophical question of what happens to
our judgment of the value of a painting when we learn that it was not
done solely by an established master; his theoretical example is Rembrandt's Polish Rider. Reviewing Danto's book, Rosalyn Krauss notes
that "at this very moment there is a collective project under way to
establish the true, the authentic, the original Rembrandt from within
the miasma of fakes, copies, shop pieces, student-produced lookalikes, and all the other forms of 'indiscernibilia' that have clouded
our view of the genuine article." Ironically enough, one of the suspected paintings is precisely the Polish Rider. Krauss wonders what
lesson we can draw from the possibility
that this "deepest painting" may not be by "one of the deepest
artists in the history of the subject." [Danto's words]. . . . Do we
shrug our shoulders and decide we were duped: "Oh, so it was not
really a moving painting after all"? Or does this become an occa-sion .
. . to rethink the structure of agency, of authorship? . . . The idea of
the unique hand of the master begins to appear . . . as a
96
Coming to Terms
Coming to Terms
tummy sticks out like a small boy's. Unlike the long cigarettes
flaunted by the elaborately smoking ladies, his is short and tucked
between the knuckles of his dangling left hand. Two walls are visible:
upon each hangs an oil painting. The room is flooded with light. Six
diners in business clothes five attractive young women and a balding middle-aged man relax over cigarettes. The women seem to be
making pleasantries about the man in pajamas; the bald-headed man
has turned and raised his wine glass to toast him. In the bottom panel,
the young man now stands behind the woman on the bald man's
right. He bends over her, his hand on her shoulder; she smiles up at
him, her left hand around his neck. The bald man has his right hand
on the young man's naked shoulder.
The only caption the same one that keynotes a whole series of
two-panel Benson & Hedges ads reads "For people who like to
smoke . . ." m all other such ads, the narrative is easier to make out.
In one, for example, the upper panel shows two smoking couples in
sports clothes sitting on colorfully upholstered porch furniture. The
man on the left is dealing cards; the other three are laughing. The
couple on the right sit in a love seat, the woman leaning intimately
against the man, whose head is raised in a hearty chuckle. In the
lower panel, only the couple on the right remain. The woman has
snuggled up and fallen asleep on the man's side. His cigarette raised
aloft in his left hand, his right arm thrown casually along the arm of
the love seat, his leg up on the table, he gazes into a bright future with
smiling contentment. A simple narrative easily comes to us: "A
couple has come over for an evening of lighthearted cards (Yuppies
do not take cards seriously). Everybody smokes and has a good time.
Afterward, as the hostess sleeps, her head on his lap, the host reflects
on the simple pleasures of life."
But the details of the pajama-man story are uncertain in a way that
is surprising for a business ordinarily intolerant of uncertainty. It
furnishes us an interesting angle on the difference between real and
implied author even in the mundane, throwaway world of advertising. This series of ads tries to shape narratees and implied readers
who are open, easy, knowing, sociable, tolerant tolerant, especially, of the idea of cigarettes as facilitators, like drinks, of communal good cheer and, later, as coals in the afterglow. The pajama man
would never have appeared if that meaning were not implicit. Whatever else the audience may conjecture about him, they are clearly
100
Coming to Terms
wearing a business suit, there would be little to speculate about: a
junior executive stands at the head of the table because he has just
been promoted. The bald man and the five women toast him at a
luncheon party thrown in honor of the occasion. The paintings, the
light, the absence of waiters and other tables suggest that the room is
not in a restaurant. It's a staff luncheon, a catered party in the board
room.
But the pajamas invite us to wonder what a young man so attired
could be doing in such a room. Why is he being toasted? Why all the
laying on of hands? People tend to wear pajamas at home. Perhaps
the young man is chez lui, entertaining this group. But why in
dishabille? Perhaps he's an eccentric, some kind of artist: the painting
directly behind him has a fashionably abstract look. Maybe he is
entertaining the owner and salespeople of the gallery where his
successful show has just opened. Or maybe he's a rock star who gets
up late and is staring sleepy-eyed at his manager and staff; he has
overslept the celebration but remains the apple of every eye. Or
perhaps it's a family matter: he's the son of the older man, and the
luncheon is taking place not at corporate headquarters but in his
father's penthouse apartment. The young man hasn't seen his father
in a while. He arrived the night before from out of town, went to
sleep not knowing that his father was planning to entertain, heard
merriment, and just wandered in. Or perhaps it's the young woman's
apartment, one she shares with the man (lover, husband, brother,
roommate?), who didn't know that she had invited her colleagues
home to celebrate her promotion. And so on. None of these has
anything to do with a Robbie from another set. But imagine even, if
you like, the metatextual possibility: a group of models hired to do
an ad celebrating the pleasures of communal smoking find their
Manhattan studio set invaded by a pajama-clad young man on a
cigarette break from another shoot. Seeing the hilarity of the situation, they take their own break, lighting up (now they're really
smoking, rather than merely acting the part of smokers). The older
man raises his fruit juice to toast the handsome young interloper.
Accepted so jovially by his colleagues, the young man hugs one of
the models, in a purely collegial way, and is hugged in return. And so
on.
Notice that our speculation has nothing to do with what some real
authorthe writer or the photographerdid; it concerns only the
102
Coming to Terms
the tutelary spirit of nicotine, a kindly minor deity, the good genius
of the East Side, protecting and reassuring those who wander, innocently seeking pleasure, into his haunts. How could he wear street
clothes on such an occasion?)
It is interesting to speculate about why a cigarette company should
be among the first to present its message in an aleatory narrative. Is it
because smoking isor the manufacturer wants it to seema random matter? Is it that the random tale supports the low-energy
resistance of smokers in the face of all evidence of the deadliness of
their habit? "If you're sufficiently knowing to understand the clever
open-endedness of these images, you can get away with smoking, no
matter what the authorities tell you." Perhaps the company itself
takes a random attitude, since it feels it has nothing more to lose. In
an era when the surgeon general stamps blunt pronouncements on its
ads, Benson & Hedges may feel that it has little choice but to go with
innuendo.
Narratologically, then, we have the following structure: The real
authors of the narrative are decision-makers at the cigarette company, Benson & Hedges, and the advertising agency, Wells, Rich, and
Greene, and their assistants and contractors. Their motive is direct
suasion of the public to smoke their product, by whatever means
capitalism has invented for such suasion. Their overriding concern is
the sale of cigarettes, not narrative innovation. So (we may speculate)
they turn to aleatory narratives because such narratives are new, and
they are betting that the faddish "enigma" will get people interested
in their adsand their product. If they can do so by introducing the
startling (to the mass audience) prospect of an unclear, random narrative, they are willing to take the chance. Thus their motives are quite
different from those of the avant-garde novelist whose intention in
problematizing the event structure of a narrative is aesthetic. The
advertisers do not wish to "make things strange," as the Russian
Formalists would put itthat is, to upset our everyday expectationsbut only to titillate a bit, to make us all chuckle together and,
of course, to light up.
But even so, the ad necessarily implies an author, a textual intention. The implied author is born in our ingrained eagerness to find
stories wherever possible. Even the shrewder reader who understands the advertiser's true motives will continue to infer an implied
author (on the principle, perhaps, of trying not to think of a pink
104
Coming to Terms
culture," to effect the sense that class is finally emerging as an issue
even to the bourgeois consciousness?25
Jameson's example is a successful and popular Hollywood film,
Dog Day Afternoon (1975), directed by Sidney Lumet. The film is
based on an actual event, the attempted robbery of a Chase Manhattan
bank in 1972 by one Sonny Wojtowicz and his partner Sal, to raise
money for the sex-change operation of Sonny's significant other.
Sonny and his partner laid siege to the bank for many hours; he was
finally captured and his partner killed by an FBI agent on the way to
the airport where a plane out of the country was ready (in exchange
for the release of bank employees held as hostages).
What proved remarkable (at least to bourgeois capitalists) was that
Sonny became a kind of hero to the crowd of onlookers outside the
bank. He sensed the opportunity to play on their sympathies, much
to the discomfiture of the police. It was the crowd's recognition of
their affiliation with the outlaw that probably stimulated Hollywood
to exploit this media event, whose interest to the public was already
proven. As Jameson explains: "Social reality and the stereotypes of
our experience of everyday social reality are the raw material with
which commercial films and television are inevitably forced to
work." But in using this material, Jameson feels, the media unknowingly permitted the class message to seep through. He explains
the process this way:
The immense costs of commercial films, which inevitably place
their production under the control of multinational corporations,
make any genuinely political content in them unlikely, and on the
contrary ensure commercial film's vocation as a vehicle for ideolog-ical
manipulation. No doubt this is so, if we remain on the level of the
intention of the individual film-maker, who is bound to be limited
consciously or unconsciously by his or her objective situa-tion. But it
is to fail to reckon with the political content of daily life, with the
political logic which is already inherent in the raw material with which
the film-maker must work: such political logic will then not manifest
itself as an overt political message, nor will it trans-form the film into
an unambiguous political statement. But it will certainly make for the
emergence of profound formal contradic-tions to which the public
cannot but be sensitive, whether or not it yet possesses the conceptual
instruments to understand what those contradictions mean.26
106
Coming to Terms
CHAPTER
7/
IN HIS ATTEMPT to develop a viable theory of "distance" in narratology, Grard Genette called up the authority of Plato's account
(The Republic, Book 3) of diegesis and mimesis:
Plato contrasts two narrative modes, according to whether the poet
'himself is the speaker and does not even attempt to suggest to us that
anyone but himself is speaking' (this is what Plato calls pure narrative
[diegesis]), or whether, on the other hand, the poet 'delivers a speech as
if he were someone else' (as if he were such-and-such a character), if we
are dealing with spoken words (this is what Plato properly calls
imitation, or mimesis).'
Genette emphasizes the difference between diegesis and mimesis,
but the important recognition here, I think, is that both are narrative
modes. I want to argue that in important ways drama is a kind of narrative, at least in the sense that it is based, like epic, on that component of narrative which we call "story." As Aristotle pointed out,
"Whatever parts epic poetry has, these are also found in tragedy." 2
The parts he meant are "plot" (with its "recognition" and "reversal"),
"character," "thought," and "diction." The first two of these (plus
"setting") are what modern narratology typically calls "story." Only
"spectacle" was said to be unique to drama. But spectacle, surely, is
an element of the actualization of stories, and not one of the underlying components of narrative structure. The fundamental property of
109
Coming to Terms
story (in the narratological sense) is that it consists of a series of
connected events, and that property is shared by both drama and
epic.
Genette, however, identifies narrative with diegesis and drama
with mimesis. In Narrative Discourse he writes " 'Mimetic' representation [is] borrowed from the theatre," 3 and in Narrative Discourse
Revisited speaks of the "truly insurmountable opposition between
dramatic representation and narrative." 4 Certainly at the level of
actualization, a play and a novel are quite different. But at the textual
level they resemble each other far more than either resembles any
other text-typesay, Argument or Description. Indeed, Aristotle
wrote that both tragedy and epic "imitate" the "lines of action"; thus,
"imitation" is not limited to words alone, but includes larger structuresin particular, structures of plot.5
Here, as always, we must be careful about the polysemy of terms.
At the very beginning of Narrative Discourse, Genette is very clear
about the three senses of rcit. His caution is applicable no less to the
polysemy of English "narrative." We must take care to specify which
sense is entailed in a given application of the word. Genette has used
narratif in one sense in the expression "two narrative modes" (deux
modes narratifs) and rcit in another in "truly insurmountable opposition between dramatic representation and narrative" (l'opposition, vraiment incontournable, entre rcit et representation dramatique). In the first, he seems to be referring to the story aspect of
narrative; in the second, to its discourse aspect, its manner of being
told. There is indeed an obvious difference between dramatic representation and epic or novelistic representation. But there is no great
difference between the structures of the "what," the story component
told by epics and enacted by dramas. Both rely on sequences of
events, and both present a chronology of events different from the
chronology of the discourse. This double chronology is the fundamental property that distinguishes the text-type Narrative from the
others. It makes easy sense, for example, to speak of the "story" in
Shakespeare's Hamlet, whereas we would hardly speak of the "story"
in the Declaration of Independence or some other argument.
But it is not even clear that the discoursive difference between plays
and novels is all that profound. A short story or novel becomes more
or less purely "mimetic" when it consists of nothing but the quoted
dialogue of the characters. Then it is an "unmixed" case, and only the
110
Argument
Description
Enacted
(drama)
Coming to Terms
onomatopoeic) language. In "told" narratives, such as epics and most
novels, the narrating function is assigned to a set of signifiers that are
"arbitrary," unanalogous to the actions, characters, or settings they
signify. In "shown" stories, such as narrative films, both characters
and actions tend to be represented in an iconic or "motivated" fashion. For example, the reader of Joseph Conrad's Outcast of the Islands
can find little in the names "Willems" or "Lingard" or, indeed, in the
descriptive epithets applied to them by the narrator to form a precise
mental image of these characters. Whatever images we form, they
can only be less detailed and determinate than the photographed
images in Carol Reed's film version of the novel. To be more precise,
the photographed images signify actors named "Trevor Howard"
and "Ralph Richardson"; these in turn signify the characters named
"Willems" and "Lingard." At the most basic level, the signifiers (actors) share with the signifieds (characters) the contours and other
visible attributes of "men" (rather than women or elephants). At a
more refined level, the actors, through their skill and the director's
guidance, create facial and bodily representationssighsappropriate to the characters: a drifter of devious morality and a courageous
but too trusting sea captain.
To say that a play or movie or cartoon is "shown" is to say that its
narration is conveyed by a set of signifiers (human beings on stage,
photographs of them on film, drawings of them on paper) which are
"motivated" or "analogous": that is, they resemble their signifieds in
some culturally recognizable way. But the analogy itself contains a
certain element of the arbitrary, since the signified can be shown now
by this actor and now by that. Hamlet is always a man, but that man
may be a Laurence Olivier, a Maurice Evans, a John Barrymore, even
a Sarah Bernhardt. All quoted speeches, however, are totally mimetic.
Whoever the actor, the lines purport to be exactly the words that
Hamlet speaks (though their intonation varies, as a function of the
performance). This is also true in the print medium, in novels and
short stories: though print itself is not analogous to voice, the choice
of words, syntax, and the like purport to copy exactly what characters say.
Both noniconic and iconic representations entail a kind of delegation. In diegesis, the (real) author delegates the final shape of the
signifers to the editors, typesetters, and other members of the publisher's production team; in mimesis, the author delegates it to the
11 2
Coming to Terms
as the means of actualizing the narrative. We need a more encompassing term for all those texts we call novels, novellas, short stories,
dramatic lyrics, verse ballads, plays, and so on, one that avoids
confusing them with other kinds of texts. Once we decide that stories
may indeed be totally enacted on the stage or screen, consistency
requires that we label them "narratives." The difference between
telling and showing then comes down simply to the implied author's
choice of signsanalogous or motivated for mimetic narratives,
arbitrary or "symbolic" for diegetic narratives, and a mixture for
mixed narratives.
Of course, the choice of signs is not a cut-and-dried matter. Some
signs chosen to do the "enactment" are not so purely analogous or
arbitrary as theory might suggest. The few squiggles that render
Charlie Brown do not imitate or enact that droll character with the
same full force that the body and behavior of the flesh-and-blood
Laurence Olivier imitated and enacted Hamlet. Though still relatively iconic, the cartoon's spare lines are closer to the "arbitrary"
than the actor's majestic appearance. In the ballet, a signifier such as a
pirouette does not uniquely enact some such signified as "joy" or
"enthusiasm."8 Nor are many of the hand signals of the language of
the deaf self-evidently iconic. Stories conveyed by such hand signals
seem more properly "told" than "shown."
But despite such blurred, intermediate, and mixed cases, there is
no particular reason why "to narrate" should mean only "to tell."
Once we decide to define Narrative as the composite of story and
discourse (on the basis of its unique double chronology), then logically, at least, narratives can be said to be actualizable on the stage or
in other iconic media. The burden of disproof falls on theories that
would deny the name Narrative to "performed" texts. They would
need to explain whyhaving agreed that "mimesis" is a way of
conveying Narrative (even in such nontheatrical texts as dialogueonly novels)it should not be called an act of narration.
In short, Narrative may be defined in a broader and a narrower
sense. In the broad sense that I have proposed in previous chapters,
Narrative is the text-type distinguished from others by a double
"chrono-logic"a logic of event sequence, performed by characters, in a setting. In the narrower, traditional sense, Narrative is a text
entailing all the conditions of the broad sense plus the diegetic condition: that is, that the text must be told by a human narrator. This more
114
Argument
Description Other
Diegetic
Mimetic
Coming to Terms
to the verbs "present" or "narrate." The suffix means either "agent"
or "instrument," and neither need be human.
This definition rejects the proposal I made in Story and Discourse
that heavily mimetic or shown storiessuch as "The Killers"are
not communicated by narrators. That view, I now believe, leads to
an inadmissible paradox or, at least, a counterintuition: namely, that
narratives just appear, unannounced, so to speaka view that contradicts both logic and common sense. 9 The notion of "non-narrated" narrative arises as a misguided effort to restrict "agency" to
human beings, but the restriction will not hold. A presentation
argues a presenter, whether human or not, whether vividly dramatized or not. Once we allow the possibility of showing a narrative,
we perforce recognize the existence of a show-er, even if not a human
one. In this age of mechanical and electronic production and reproduction, of "smart" machines, it would be naive to reject the notion
of nonhuman narrative agency. Remember, we are now speaking
neither of the original creators of narrative texts, the flesh-and-blood
authors, nor of the principle of invention in the text that we call the
implied author, but of the someone or something in the text who or
which is conceived as presenting (or transmitting) the set of signs that
constitute it.10 "Presentation" is the most neutral word I can find for
the narrator's activity. As part of the invention of the text, the
implied author assigns to a narrative agent the task of articulating it,
of actually offering it to some projected or inscribed audience (the
narratee).
As has often been noted, modern novels and short stories tend to
be shown rather than told. Some literary fictions purport to
be nothing more than mechanically recorded copies of characters'
speeches; the pure dialogue short story is a common form. Clearly, it
is better to say that these are "shown" by a silent, extradiegetic
narrator than that they are "told" or "spoken" by her.
My difficulty in Story and Discourse was implicit in the question I
asked at the very outset: "Is the [narrative] statement directly presented to the audience or is it mediated by someonethe someone
we call the narrator?"11 But the narrator need not be a "someone."
Every narrative statement is presented by a narrator, and the narrator
may be not a someone but a something. The agent of presentation
need not be human to merit the name "narrator."
Let me repeat the senses in which I think the terms"narrative,"
116
117
Coming to Terms
events in the form of stage directions. In productions these are
actualized by actors' movements. At the level of semiotic abstraction, there is no difference between a sentence in a novel like "John
left the room" and the playwright's instruction to an actor to exit,
stage left. Both the sentence and the actor's walking off are delegated
signifiers of (nonverbal) action.
Theater is a medium available for the presentation of stories; it is
one of several ways of bringing a narrative to life. By the same token,
the mere fact that a given short story contains only dialogue does not
make it a play; it too, is one means of presentation. The same could
be said of various interior-monologue novels and stories that represent nothing but a character's thoughts. In these, what the narrator
shows is "mental speech" (on the convention that thoughts appear in
our minds in verbal form).
Unless we coin another termand it is not clear that we should
"Narrative" still seems best to cover both diegetic and mimetic
forms as these are opposed to other text-types such as Argument and
Description. If we adopt an appropriately broad sense of the term,
mimetic formsdramas, films, balletsare just as much "narrated"
as short stories and novels.
Let us turn to problems inherent in such terms as "voice," "knowledge," and similar metaphors. The assumption that stories are only
and always narrated by human beings doubtless arose with their
bardic origin, and some narratologists (of "contextualist" persuasion) remain convinced that we must start our theories from oral
anecdotes. When stories carne to be written, narrators retained
(though not inevitably) strong personal marksreferring to themselves as "I"; offering judgments, opinions, generalizations; describing their own persons or habitats, and the like. The term "voice" was
naturally, if figuratively, tranferred to represent the means by which
those activities occurred. It is a metaphor that continues to be widely
used but insufficiently examined by narratology, functioning centrally in such otherwise divergent theories as Grard Genette's and
Mikhail Bakhtin's, Wayne Booth's and Franz Stanzel's. Yet no one, to
my knowledge, has asked whether it really clarifies what it is supposed to name. How effective is the metaphor of a "voice speaking"
when applied to shown or even "impersonally" told narratives? It is
ill suited to describe screen or stage actualizations of narratives, and
118
119
Coming to Terms
door (which is obviously viewed from within) and the hatch to the
kitchen, which is seen from the counter side. Within the room, the eyes
focus in turn on a small set of objects: the door and the hatch, the
counter and the people positioned at it, the clock. Changes of
viewpoint are definite and clear: one very brief glimpse into the
kitchen, and the excursion to Ole Andreson's with Nick Adams. In the
latter scene the narrator is an invisible man who walks in step with
Nick, stands close beside him, sees exactly and only what his eyes see
(but not with Nick's eyes: there is no penetration of Nick's
consciousness). In each part of the story there is severe visual economy, concentration on what is near at hand and relevant.12
Notice the subtle but critical shift from the narrator "providing" a
definite viewing position to "his" occupying it. The narrator is said, in
effect, to leave "his" post in the discourse and to enter the scene of the
story. The narrator is treated as a character, indeed a male character, an
"invisible man who walks in step with Nick." This assertion, I take it, is
meant not figuratively but literally; the narrator has crossed the line
from discourse to story. But what justification is there for assuming
that the narrator is a real, if "silent" and "invisible," human male? 13
Why should our operative metaphor be human "eyes" with a direct
(not windowed) view? Why should the model suggest a human
presence on the scene, walking "in step with Nick Adams," instead of
a narrative agency out in the discourse, reporting actions and scenes
without any explanation or apology about how the information was
obtained? I would argue that the narrator, by definition, does not see
things in the story world; only characters can do that, because only
they occupy that world (see the discussion of "filter" in Chapter 9).
The narrator's task is not to go strolling with the characters but to
narrate what happens to them, whether by telling or showing.
The same problem arises in Lanser's discussion of the state of the
narrator's "knowledge":
This narrator knows precisely no more and no less than what a
towns-person sitting in Henry's lunchroom would be expected to
know. He knows the name of the lunchroom and the fact that it has
been made over from a saloon; he knows the first names of George
and Nick and Sam; he knows that Ole Andreson had been a heavyweight fighter; he knows the name of the rooming-house where
Ole lives. But he does not know the names of the men who enter the
120
Coming to Terms
[Nick] was trying to swagger it off.
said it.17
But clearly every one of these could be explained more simply and
directly as free indirect thought. The third seems just a truncated
version of a sentence like "Nick had never known the feeling of a
towel in his mouth before." This sentence must represent Nick's
inside view: a putative anonymous townsman narrator would hardly
know whether Nick had had a towel in his mouth before. Similarly,
the fourth sentence could be explained as a short version of "Nick
knew he was [or "could feel himself"] trying to swagger it off"; the
context suggests not a judgment of the narrator on Nick's bravado
but Nick's own self-consciousness about it. And the fifth sentence
seems short for "Nick knew or felt that it sounded silly when he said
it." As for the first two sentences, they can easily be read as representing the collective viewpoint of George, Sam, and Nick, though once
Nick's centrality is established, we might want retrospectively to
limit the observation to him.
Though explicit inner views are sparse in this short story, they do
seem to afford the limited insight into Nick's consciousness that
critics have found in this and other Nick Adams stories. And though
the focus on Nick's filter is predominantly one of "interest" rather
than perception or conception, it is sustained rather clearly. For
example, the story remains "his" when he goes to see Ole Andreson,
leaving George and Sam behind. Interpreting these five sentences as
Nick's perceptions leads to a simpler and more coherent reading than
one which, in its insistence on finding a human being "responsible"
for the narration, invents invisible companions marching along next
to him. Throughout the Nick Adams stories, Nick is protagonist,
though not one given to internal speculation about the things happening around him. The thrust of these stories is the impact on his
youthful consciousness of the harshness of the world, even though
his absorption of the information is not commented upon. Why
devise some ghostly townsperson when so obvious and, indeed,
traditional an explanation is at hand?
We need a definition of "narrator" which can allow for non-human
as well as human, nongendered as well as gendered agents. We do
have many gaps to fill in reading narratives, but it has not been
122
123
CHAPT ER
8/
Coming to Terms
a text, but the style actualizes that arrangement; that is, it "embodies"
the total narrative. Again, none of thesewe must always rememberhave any independent existence; they are all constructs proposed
by theory, the better to explain the workings of film narrative.
Bordwell believes that narration is a dynamic process: "Formal
systems both cue and constrain the viewer's construction of a story."
But his emphasis is on what the viewer makes of the visual and
auditory data impinging on her consciousness more than on the
nature of the data themselves. The very term "construction" suggests that the important work is done by the viewer. In some sense,
Bordwell takes the film itself, in its various layered structures, as
already given. Thus, he rejects the notion of a cinematic narrator
inherent in the film and argues instead for something he calls "narration": "Narration is the process whereby the film's syuzhet and style
interact in the course of cuing and channeling the spectator's construction of
the fabula" (original emphasis).9
It is a little unclear how this process occurs, whether it is internal to
the viewerin which case style and syuzhet "interact" only within
her perception and cognitionor whether there is some kind of
interchange between the screen and the viewer. If the latter, then
"narration" at least partly inhabits the filmin which case, we can
legitimately ask why it should not be granted some status as an agent.
But Bordwell is opposed to the notion of narrative agency because
"narrator" connotes "human being" to him. "If no voice or body gets
identified as a locus of narration," he asks: "can we still speak of a
narrator as being present in a film? In other words, must we go
beyond the process of narration to locate an entity which is its
source?" His answer is an unqualified no:
In watching films, we are seldom aware of being told something by
an entity resembling a human being. . . . narration is better understood as the organization of a set of cues for the construction of a
story. This presupposes a perceiver, but not any sender, of a message On the principle that we ought not to proliferate theoretical
entities without need, there is no point in positing communication as
the fundamental process of all narration, only to grant that most
films "efface" or "conceal" this process. Far better, I think, to give
the narrational process the power to signal under certain circumstances that the spectator should construct a narrator.10
126
Coming to Terms
able." To say that an aesthetic object or process is "graceful" means
that it strikes some observer as possessing grace. To say that it is
"knowledgeable" is to say that it knows something, but if it knows
something it must be more than an object or a processit must be an
agent (though the agent need not be human: my computer, for
example, "knows" a lot). Normally, "process" refers either to a
natural happening or to something set in motion by someone or
something. But if something can know, present, recognize, communicate, acknowledge, be trustworthy, be aware of things, then surely
it is too active a concept to be a mere happening or process. Knowing, presenting, recognizing, and the like are deeds, and deeds logically entail a doer. If "narration" indeed "does" these things, it is
by definition an agent, and so requires not the object-nominalizer "tion" but the nomina agentis "-er."
The status of "narration" in Bordwell's narrative hierarchy is also
unclear. On one occasion he seems to treat "narration" as the synonym of "syuzhet": "In most narrative films the narration does
rearrange fabula order, principally through verbal recounting and in
expository passages. It is more unusual to find the syuzhet enacting
fabula events out of chronological sequence." But in a footnote
explaining the difference between his own use of "syuzhet" and
"fabula" and Genette's rcit and histoire, he defines "narration" as
syuzhet plus style (what I would call the actualization of the narrative
in a medium).14 Thus it is unclear whether he means to put "narration" at the same level with fabula or at a level higher.
Bordwell argues that "narration" controls the amount and placement of fabula information in the syuzhet through three instruments,
which he names "knowledge," "self-consciousness," and "communicativeness."15 The first and third are nominalized forms of the
verbs "know" and "communicate," and the second is a nominalization of an adjective generally applied to human beings. Bordwell uses
the terms quite consistently, and they have a certain viability, once
you get used to them. But this personification of narration, a mere
"process," seems to mystify more than it clarifies. Consider "knowledge": Bordwell asks, "What range of knowledge does the narration
have at its disposal?" The "range" goes from highly restricted to total
knowledge. This corresponds to literary criticism's well-known distinction between "limited" and "omniscient" points of view (or what
I shall call "slant"). But whence the knowledge? Since Bordwell
128
Coming to Terms
But in that film, unlike Rear Window, at a certain moment "the
narration holds back exactly the sort of information to which it has
earlier claimed complete access, even though it immediately thereafter resumes his filter."19
I go into Bordwell's excellent theory in such detail because, except
for our differences on the cinematic narrator, it is so close to my own.
We both want to argue that film does belong in a general narratology;
we both want to argue that films are narrated, and not necessarily by
a human voice. We differ chiefly in the kind of agency we propose for
the narrative transmission. It comes down, as I say, to the difference
between "-tion" and "-er."
But there is one other difference, turning on the word "knowledge." In my theory, the narrator communicates all of and only what
the implied author provides. How the narrator carne to "know" the
provided information seems a nonquestion. Without the implied
author, it is pointless to talk about "knowledge," even if we substitute "narrator" for "narration." The question is one not of knowing but of how much and what information the cinematic narrator is
programmed by the implied author to present. Only the implied
author can be said to "know," because the implied author has invented it all. For each reading or viewing the implied author invents
the narrative, both discourse and story. The cinematic narrator presents what the cinematic implied author requires. Just as it is the
implied author who chooses what the adult Pip tells in Great Expectations, it is the implied author of Rear Window who decides what the
"camera" shows "on its own," what it shows as filtered throu gh
Jeffs perception, and what it does not show at all. And just as
literature has a place for "career-authorship," so does cinema. Indeed, much of auteurisme can be better explained as cinematic careerauthorship. Part of what Bordwell calls the "transtext" is implicit in
the signatures on films: a "Hitchcock" film is likely to entail suspense; an "Antonioni" film is likely to contain temps mort holds on
bits of the landscape; a "Fellini" film is likely to merge on-screen and
commentative music.
Bordwell too readily rejects the need for the concept of cinematic
implied authorship, a concept that I find no less vital to cinematic
than to general narrative theory. Films, like novels, present phenomena that cannot otherwise be accounted for, such as the discrepancy between what the cinematic narrator presents and what the film
130
Coming to Terms
thing that we see and hear follows Johnny's scenario. Thus, even
when his voice-over falls silent, he remains the controlling, if unreliable, narrator of the flashback.21
So, for this film at least, it does not seem true that "personified
narrators are invariably swallowed up in the overall narrational process of the film, which they do not produce."22 At the narrative level,
Johnny and Johnny alone "produces" the segment in any narratologically meaningful sense of that word, since every cinematic tool
editing, lighting, commentative musicworks to actualize his lie.
During these scenes, Johnny prevails over the cinematic narrator. He
is "responsible" for the lying images and sounds that we see and hear.
Only later does the conventionally reliable cinematic narrator, reappropriating all the cinematic tools, take up the true story. Our judgment that Johnny's version of the story is unreliable depends on our
decision that the later version of the story is reliable. And we must
acknowledge that this does not happen by chance but is part of a
design of communication. Who has invented both the incorrect and
the correct versions? Bordwell's theory would have to say "the
narration." But which narration, since there are two competing
ones? Controlling both narrations there must be a broader textual
intentthe implied author. It is the implied author who juxtaposes
the two narrations of the story and "allows" us to decide which is
true.23
In short, in cinema as in literature, the implied author is the agent
intrinsic to the story whose responsibility is the overall design
including the decision to communicate it through one or more narrators. Cinematic narrators are transmitting agents of narratives, not
their creators. Granted, Stage Fright is an unusual movie, but it
cannot be ignored in constructing a theory of cinematic narration, for
it is precisely the unusual possibility that tests the limits of a theory. A
theory of narrative cinema should be able to account not only for the
majority of films but also for the narratively odd or problematic
ones. Besides, it may well be that unreliable narration will someday
become as common in film as it is in the novel.
But the utility to cinematic narration of the concept of the implied
author does not rest on a narrator's outright prevarication alone.
Other kinds of films also illustrate the need for recognizing a separate
principle of invention and intent. In Alain Resnais's Providence (1977)
132
the first half of the film represents the fantasies of the protagonist,
aging novelist Clive Langham (John Gielgud). Langham's voiceover, we eventually surmise, is somehow constructing the images
filling the screen. These are more or less hypothetical rough drafts for
scenes in a novel he is struggling to write. During these moments of
his fantasy, it is he, not some disembodied "narration," who generates what passes before our eyes. Later, as he celebrates his birthday
with his sons and daughter-in-law, an impersonal extradiegetic narrator assumes control of the cinematic apparatus. Again, both narrators have been introduced by the overriding intent of the film, the
implied author.
In short, for films as for novels, we would do well to distinguish
between a presenter of the story, the narrator (who is a component
of the discourse), and the inventor of both the story and the discourse (including the narrator): that is, the implied authornot
as the original cause, the original biographical person, but rather as
the principle within the text to which we assign the inventional
tasks.24
For if we deny the existence of the implied author and the cinematic narrator, we imply that film narratives are intrinsically different, with respect to a fundamental component, from those actualized in other media. But that implication contradicts the principle
that Bordwell himself correctly endorses: namely, that narration is a
"process which is not in its basic aims specific to any medium."25 The
substitution of "narration" for "narrator" does not advance Bordwell's desire to find cinema's actualizations of the "more supple
principles basic to all narrative representation" (since, presumably, he
accepts the existence of "narrators" in literary narratives). It is awkward to a general theory of narrative to say that some texts include
the component "narrator" and others do not. As Sarah Kozloff puts
it, simply but incisively, "Because narrative films are narrative,
someone must be narracing."26 Or if not necessarily someone, at least
something.
Let me recapitulate my conception of the cinematic narrator.
Though film theory tends to limit the word "narrator" to the recorded human voice "over" the visual image track, there is a good
case to be made for a more general conception of "cinematic narrator." Films, in my view, are always presentedmostly and often
133
Coming to Terms
exclusively shown, but sometimes partially toldby a narrator or
narrators. The overall agent that does the showing I would call the
"cinematic narrator."27 That narrator is not a human being. The
nomina agentis here refers to "agent," and agents need not be human.
It is the cinematic narrator that shows the film, though it may on rare
occasions (as in Stage Fright) be replaced by one or more "telling"
voices on or off the screen.
The cinematic narrator is not to be identified with the voice-over
narrator. A voice-over may be one component of the total showing,
one of the cinematic narrator's devices, 28 but a voice-over narrator's
contribution is almost always transitory; rarely does he or she dominate a film the way a literary narrator dominates a novelthat is, by
informing every single unit of semiotic representation. 29 The normal state of affairs, and not only in the Hollywood tradition, is for
voice-over narrators to speak at the beginning, less frequently at the
end, and intermittently (if at all) during the film. Some films do use
the technique more extensively. In a few cases the voice-over narrator seems to control the visuals: in Stage Fright, Providence, and
Robert Bresson's Diary of a Country Priest (1950), the human narrator's voice dominates, at least in part of the film. 30 In All About Eve,
Addison De Witt's narration controls the whole introductory set of
flashbacks that present the history of Eve's rise in the theater. In the
freeze-frame description discussed in Chapter 3, it can reasonably be
argued that it is Addison who has frozen the frame, just as he previously tuned out the voice of the old actor presenting the award. In
the film's coda, however, after Eve receives the award and goes
home, only to find her own little stagestruck groupie waiting to
follow in her footsteps, the general cinematic narrator assumes control.31 At first, then, the image track seems to be the character's
construction; that is, the images are at the service ofare an alternative means of communicatingwhat he says. Later, the general
cinematic narrator takes control. But both narrators are the instruments of the implied author.
The cinematic narrator is the composite of a large and complex
variety of communicating devices. Some of them are partially shown
in this diagram, which makes no pretense to completeness; my
purpose is rather to demonstrate something of the multiplexity of the
cinematic narrator:
134
AUDITORY CHANNEL
Kind
Noise
Voice
Point of origin
Music
On-screen
Off-screen
Earshot Commentative
VISUAL CHANNEL
Nature of image
Treatment of image
Editing
Type Rhythm
Straight cut Fade Etc.
The cinematic narrator is the composite of all these plus other variables. Their synthesis as the narrator, of course, is achieved by the
semiotic processing performed by the viewer, the details of which are
admirably described by Bordwell. That processing goes beyond the
merely perceptual; for example, the clutching of two right hands is
perceivable by all viewers but is interpretable as a "handshake" or
"arm-wrestling" only by those who know the rules of a language and
culture that include such signifieds.
The different components of the cinematic narrator as diagrammed
usually work in consort, but sometimes the implied author creates an
ironic tension between two of them. It is not uncommon, even in
135
Coming to Terms
Hollywood films, for the visual track to undercut the story told by a
character-narrator's voice-over. In Terence Malick's Badianas (1973),
for example, the voice of the heroine, Holly, tells a romanticized
account of her escapade with the murderer Kit which is totally belied
by the sordid action as we see it with our own eyes. 32
This kind of partial unreliability is unique to two-track media such
as the cinema. The disparity is not between what the cinematic
narrator says and what the implied author implies but between what
is told by one component of the cinematic narrator and shown by
another. Whereas in totally unreliable narratives such as Stage Fright
the conflict must arise through disparities between all the representations of the narrator and what the viewer must infer from the film as a
whole, the partially unreliable narration of Badianas arises explicitly
from a conflict between two mutually contradictory components of
the cinematic narrator. Normally, as in Badianas, the visual representation is the acceptable one, on the convention that seeing is believing. In theory, at least, the opposite could happen as wellthat is,
the sound track could be accurate and the visual track unreliable
but in practice, that effect seems very rare; Sarah Kozlofffound only
one example, in An American in Paris (1951).33 The reason for this
imbalance seems fairly obvious. The errant homodiegetic voice-over
can easily be understood to have some motive for skewing the
events. In Badlands, Holly is a naive adolescent, and so we hypothesize that she romanticizes the escapade. After all, she lives in a society
so boring that any notoriety, even that of criminals, overrides questions of mere morality. But when the camera is aberrant and the
voice-over "straightens it out," the effect is odd and self-conscious.
In An American in Paris, as each of the three protagonists introduces
himself, the camera focuses on the wrong person; the voice-over
redirects it. A more usual if weaker case of visual unreliability occurs
when the visual track presents a shallow picture of events, which are
interpreted with greater profundity by the voice-over narrator; this is
the strategy of Bresson's Diary of a Country Priest and Stanley Kubrick's Barry Lyndon (197 5).34
There remains a final question about unreliability: does it make
any sense to call narrators "unreliable" who are without personality
or, as I have called them, "covert"? (This class includes virtually all
cinematic narrators and many heterodiegetic narrators of novels and
short stories.) It is hard to think of any reason for doing so. Unre 136
Coming to Terms
grasping the notion of nonhuman narrative agency. The technology
of film, Banfield contends, like that of the telescope and the microscope, "allow[s] the viewing subject to see, to witness, places where
he is not, indeed, where no subject is present." It also reveals "the
appearance of things when no one was present," as, for example, a
stellar explosion that took place millions of years before man
evolved.35 Clearly, this kind of model helps us understand how we so
readily conceive of a non-human narrator; how, for example, we can
speak of "the camera" as such an agent, how, in Rear Window, a
nonhuman cinematic narrator can roll up the blinds and take us on a
visual tour of the courtyard. For that is the predictable response of
ordinary moviegoers: confronted with the beginning sequence of
Rear Window, they will interpret the scanning views of the interior of
a courtyard as a tour of the courtyard by the camera "on its own."
When the shades roll up, we know that what we see is being presented by a "subjectivity reduced to nothing else but what the instrument can record"; this is quite different from later viewings that are
filtered through the perception of Jeffries. At such moments, the
visible details of the courtyard correspond to what Banfield, following Bertrand Russell, calls sensibilia: "those objects which have the
same metaphysical and physical status as sense-data, without necessarily being data to any mind." That is, to any mind inside the fiction:
they are, of course, data to the real audience in the movie theater,
though data perceived at second hand. They are not seen or heard but
rather overseen and overheard, as the "impressions" of an impersonal
narrative agency. These images (Banfield might contend) would still
exist even if no one had bought a ticket and the projectionist had gone
out for a smoke: "Each gathering of sensibilia, as on the ground glass
of the telescope, represents . . . a perspective definable independent of
whether or not it is given to any observer."36
Banfield's concern is literary, but she finds in the work of Gilles
Deleuze a similar attitude concerning film. Deleuze reminds us that
Dziga Vertov's "kino-eye" is not limited as a human eye is; it is
ubiquitous, the product as much of montage as of cinematography.
And its powers are, to use Banfield's terms, "private and subjective,"
yet "impersonal." Deleuze would argue that it is precisely the objectivity of the cinematic narrator that requires us to "construct," rather
than just to "see," since what the kino-eye presents us with is a
construction of views that no human eye could see.37
138
CHAPTER
9/
Coming to Terms
view the Transamerica Building is an architectural travesty" is meaningful even if I say it at a cocktail party in Barcelona. The Transamerica Building remains a visible object, but the act of "seeing"
entails a figurative use of the term "point of view." The "seeing" can
refer to acts of memory, judgment, opinion, or whatever.
Further, since "see" can mean more than merely "optically perceive," "point of view" can have as its object nonvisible things.
"From my point of view, the president's stand on flag burning is
indefensible." The president's stand, the thing "seen," is an abstraction, conceived figuratively as occupying a kind of space; it is part of
the same figure that articulates my "perceiving" from my own mental
"place"my ideological "stance," to vary the metaphor. The
metaphoric transfers thus add to literal perception such mental activity as cognition, conceptualization, memory, fantasy, and the like.
Thus,
THING SEEN
ORGAN OR FACULTY
PLACE OF SEEING
Literal: city
Literal: skyscraper
(2) Figurative:
visual recall
Figurative: "post"
in mind
Literal:
Transamerica
Building
Figurative: "post"
in mind
(3) Figurative:
judgment
Point of View
Further, "point of view" is also subsumed under ideology in the
broadest sense. One can cite the ideological point of view of someone
who is not alive: "From Franklin Roosevelt's point of view, Reagan's
self-characterization as his successor was ludicrous." Doubtless, further extensions could be made, but these are sufficient to suggest that
"point of view" in ordinary usage covers far more than mere perception or even cognition.
All these senses and implications of "point of view" have been
transported from the language at large into narratology. But we may
ask whether the narratological properties that we wish to identify are
well captured by this termor indeed by any single termor
whether its complexities render it imprecise and confusing. The
question is not whether the functions of perception, cognition, empathy, and so on, need to be named. Obviously they do. The question is whether the same name (whether "point of view" or
"perspective" or "vision" or "focalization") should cover the mental
acts of different narrative agents. I believe that it should not, that the
separate mental behaviors, stances, altitudes, and interests of
narrators and characters require separate terms.
Let us consider a rather straightforward literary example. Dombey
and Son begins as follows:
Dombey sat in the corner of the darkened room in the great arm-chair
by the bedside, and son lay tucked up warm in a little basket bedslead,
carefully disposed on a low settee immediately in front of the fire and
close to it, as if his constitution were analogous to that of
a muffin, and it was essential to toast him brown while he was very
new.1
No reader, I take it, believes that the maker of the analogy between
the baby and a toasted muffin is Mr. Dombey, a man far too complacent about his first male offspring to entertain such a thought. So
these are clearly the narrator's words (or, to be scrupulous, they are
words assigned to the narrator by the implied author). It is traditional
to say that the analogy represents the narrator's "point of view,"
which is here, as often in Dickens's novels, whimsical and gently
ironic. But I would argue that the narrator is not to be imagined as
literally contemplating the new baby and deciding, in that contemplation, that he resembles a muffin.
141
Coming to Terms
The convention has it, rather, that the narrator is performing his
usual task of reporting this scene, and he introduces the muffin analogy the better to convey its unique flavor. The narrator, in this case
omniscient and unidentified, is a reporter, not an "observer" of the
story world in the sense of literally witnessing it. It makes no sense to
say that a story is told "through" the narrator's perception since
he/she/it is precisely narrating, which is not an act of perception but
of presentation or representation, of transmitting story events and
existents through words or images. It is naive, I think, to argue that
this omniscient narrator "got" this information by witnessing it. He
is a component of the discourse: that is, of the mechanism by which
the story world is rendered. No one wonders whether the narrator
ever inhabited the story world of Dombey and Son. Though fictional,
he is a different kind of fiction from Dombey or Dombey, Jr. He
resides in an order of time and place different from that occupied by
the characters; his is a different "here-and-now." And that's true for
every narrator, no matter how minimal his/her/its distance from the
"here-and-now" of the story (as, for example, in the epistolary
novel).
The narrator may have his own "view of things," of course. But
we must lock "view" with stern quotation marks to indicate the exact
nature of the metaphor. Since it makes so little sense to say that the
narrator literally sees Mr. Dombey sitting there admiring his son, we
might ask whether "view" is not positively misleading as a term to
describe the narrator's situation. It seems better to distinguish between narrator's and character's mental experiences in the story
world as different kinds of experiences, but that is hard to do if we
refer to both by the same term, whether "point of view," "perspective," or "focalization."
I am not asserting, of course, that only narrators have attitudes.
Characters also have them (along with a whole range of other mental
experiences), and they may differ sharply from those of the narrator.
A particularly clear example occurs right after our introduction to
Mr. Dombey. Though others understand that Mrs. Dombey was a
"lady with no heart to give him," Mr. Dombey
would have reasoned: That a matrimonial alliance with himself
must, in the nature of things, be gratifying and honourable to any
woman of common sense. That the hope of giving birth to a new
142
Point of View
143
Coming to Terms
cannot look at events and existents in the discourse world that he
occupies, to the extent that that world is fleshed out. The unnamed
narrator of Heart of Darkness perceives sights and sounds, including
Marlowe's voice, in that boat on the Thames. Mr. Lockwood perceives, conceives, imagines, meditates about life at Wuthering
Heights before Mrs. Dean begins her story. But, I contend, it is in the
nature of the case that neither can pierce the discourse membrane to
experience the story world directly; they can experience it only
vicariously, through the words of others. They re-report what others
tell them. Marlowe and Mrs. Dean did, of course, experience the
original events, but they did so in their capacity as characters, not as
narrators. "Slant" delimits the mental activity on this side of the
discourse-story barrier.
"Filter," on the other hand, seems a good term for capturing
something of the mediating function of a character's consciousness
perception, cognition, emotion, reverieas events are experienced
from a space within the story world. The effect has been well understood since Henry James. The story is narrated as if the narrator sat
somewhere inside or just this side of a character's consciousness and
strained all events through that character's sense of them. The very
word "inside" implies, logically, the discourse-story barrier discussed above. And the barrier, structurally, remains, whether the
narrator continues to speak in his own voice or falls silent for long
stretches or for the entire text. What I like about the term "filter" is
that it catches the nuance of the choice made by the implied author
about which among the character's imaginable experiences would
best enhance the narrationwhich areas of the story world the
implied author wants to illuminate and which to keep obscure. This
is a nuance missed by "point of view," "focalization," and other
metaphors.
Further, the terms "slant" and "filter" correspond to the vital
distinction, originally made by Grard Genette, between who "tells"
and who "sees" the story. In my view, the latter could only mean that
character, that occupant of the story world, who has perceived the
events that transpire. The narrator can only report events: he does
not literally "see" them at the moment of speaking them. The heterodiegetic narrator never saw the events because he/she/it never occupied the story world. The homodiegetic or first-person narrator
did see the events and objects at an earlier moment in the story, but his
144
Point of View
recountal is after the fact and thus a matter of memory, not of
perception. He tells or shows what he remembers having seen. In
other words, narrative discourse recognizes two different narrative
beings moving under the same name: one, the heterodiegetic narrator, inhabits only discourse time and space; another, the homodiegetic or character narrator, also speaks from discourse time and space
but previously inhabited story time and space. Only Pip-the-character saw those things out on the marsh, "back then." It is Pip-thenarrator, a different order of narrative being, who "now" recounts
those events in an unspecified but distinctly posterior discourse time
and space. In this later moment and other place, what the narrator
conveys can only be memories of perceptions and conceptions internal to the story, not the perceptions and conceptions themselves.
This is no less true in narrations that occur only minutes after the
story events, as in epistolary novels. If we are to preserve the vital
distinction between discourse and story, we cannot lump together
the separate behaviors of narrator and character under a single term,
whether "point of view," "focalization," or any other.
Continental narratology, however, has insisted that somebody always "sees" or "focalizes" the story, that if no one inside the story is
given the special privilege of such "sight," then the narrator must be
assumed to have it. "The only focalization logically implied by the
'first-person' narrator," writes Genette, "is focalization through the
narrator."3 But surely speaking of the narrator as "focalizer" blurs the
distinction that Genette himself introduced to clear up traditional
confusion between voice and point of viewbetween "who speaks"
and "who sees." The narrator's comments are not of the same order
as a character's perceptions, even if he is reporting what he saw or felt
"back then" when he was a character. The use of "focalization," or
any other single term, to refer to the quite different mental process of
characters and narrators violates the distinction between story and
discourse. 4 Even for so-called "camera-eye" narration it is always
and only as if the narrator were seeing the events transpire before his
very eyes at the moment of narration. If we do not understand this,
we cannot clarify but must fall victim to the very illusion that it is our
task to analyze. As Genette puts it, fictional discourse simulates a
reproduction: that is, an invented production. It does not see; it
produces. That production is offered sometimes as a memory (a rereproduction), sometimes as the sheerest fantasy. (Lawrence Sterne and
145
Coming to Terms
others rub our noses in the artifice.) Even when the narrator takes
pains to make it seem as if he were "right there," witnessing the
things as they "really" happened, all fictional narratives remain artifice, convention, produced illusion.
Only characters reside in the constructed story world, so only they
can be said to "see," that is, to have a diegetic consciousness that
literally perceives and thinks about things from a position within that
world. Only their "perspective" is immanent to that world. Only
they can be filters. The narrator cannot perceive or conceive things in
that world: he can only tell or show what happened there, since for
him the story world is already "past" and "elsewhere." He can report
them, comment upon them, and evenfiguratively in literature,
literally in cinemavisualize them, but always and only from outside, from a post out in the discourse. The logic of narrative prevents
him from inhabiting the story world at the moment that he narrates
it.
Of course, this convention, like any convention, can be underminedbut when it is, the anomaly is clear. In The French Lieutenant's Woman the narrator, clearly a twentieth-century type who talks
about Freud and World War II, disguises himself late in the book as a
Victorian gentleman and transports himself back to 1867, to the
railroad carriage occupied by the protagonist Charles Smithson
(much to the latter's annoyance). But the joke works only to the
extent that we understand the infraction, the narrative scandal entailed. And we continue to believe that it is the modern narrator who
recounts the appearance and actions of the "character" he has thus
created, by the suspension of some principle of verisimilitude.
Some theorists might argue, however, that the distinction between
being "inside" and being "outside" the story gets blurred in certain
kinds of narrative, in those, for instance, utilizing free indirect
thought. But let us ask whether that is so. Consider the first two
paragraphs of Virginia Woolfs Jacob's Room:
"So of course," wrote Betty Flanders, pressing her heels rather
deeper in the sand, "there was nothing for it but to leave."
Slowly welling from the point of her gold nib, pale blue ink
dissolved the full stop; for there her pen stuck; her eyes fixed, and
tears slowly filled them. The entire bay quivered; the lighthouse
wobbled; and she had the illusion that the mast of Mr. Connor's
146
Point of View
little yacht was bending like a wax candle in the sun. She winked
quickly. Accidents were awful things. She winked again. The mast
was straight; the waves were regular; the lighthouse was upright;
but the blot had spread.5
The narrator, from a post outside in the discourse, is rendering Betty
Flanders's continuing grief over the death of her husband, Seabrook,
two years before. Betty is writing a letter about having had to leave
her native town after her husband's death. The narrator tells us that
her pen sticks and the ink wells, and that her eyes follow suit, welling
with tears. Then we enter her mind, though the filtration is "psychonarrated"rendered in the narrator's own words. The mental content is not conceptual but perceptual. That is, the entire bay does not
quiver for the narrator, nor does the lighthouse wobble. It is Betty
who sees, through tears, these visual aberrations. But the next sentence does contain the very words that pass through Betty's consciousness: it is free indirect thought. The sentence is the equivalent
of "Betty felt once again that accidents were awful things." That
might or might not be a sentiment that the narrator shares and indeed
utters here; it is precisely the nature (and charm) of indirect free
discourse to make it hard to know. But even if both the feeling and
the language might be shared by character and narrator, that does not
seem reason for arguing that the demarcation between the story
world and the discourse world is blurred.
The same logic, I think, applies to interior monologue. In the
"Penelope" section of Ulysses, for example, the ruminations are totally those of Molly Bloom, in her own words (or sounds). She is not
functioning as narrator, not telling anyone a story after the fact, but
simply carrying on normal thinking processes in the present story
moment. The thought stream is simply quoted by a totally effaced
narrator. The convention is exactly the same as quoted dialogue:
hence the appropriateness of calling it "free direct thought." There is
no particular reason to argue that the narrator, though silent, has left
the discourse world.
"Point of view" ("vision," "perspective," "focalization") has
named still a third narrative function: that is, the presentation of a
story in such a way that a certain character is of paramount importance. But this is quite different from filtration, since we may or may
not be given access to that central character's consciousness. This
147
Coming to Terms
function, I think, should be called center. Thus, Milly Theale, Gatsby,
and Stavrogin are centers but not main filter characters in The Wings
of the Dove, The Great Gatsby, and The Possessed. In one sense, each is
the most important character in the novel, but we have less direct
access to their consciousness than to that of other characters. Centering without filtration is a useful technique for depicting enigmatic
characters.
A final narrative function traditionally named by "point of view"
is what I call interest. Consider this remark on the very first page of
Oliver Twist:
In this workhouse was bornon a day and date which I need not
trouble myself to repeat, inasmuch as it can be of no possible
consequence to the reader, in this stage of the business at all eventsthe
item of mortality whose name is prefixed to the head of this chapter.6
The narrator's slant is again one of Dickensian irony, as if it were
hardly worth his while, or the reader's, to care much about this "item
of mortality." But the item in question is Oliver Twist, whose name
appears on the title page and who proves to be the central character of
the novel. We immediately infer that for the newborn babethat is,
from his "interest point of view"it is a matter of concern that he
was born in a certain workhouse on such and such a date. Since he is
too small to see or to understand or to have an attitude about such
matters, he cannot be a filter. We need another name to describe this
narrative effect, and I propose interest-focus. (The term is not redundant with "center," since even a minor character can be interestfocused.) "Interest" point of view is of particular importance in
narrative media like film. Quite often we do not see things from
some character's optical point of view or know what she is thinking,
but we identify with her, interpret events as they affect her, wish her
good luck or good comeuppance.
Even if "slant," "filter," "center," and "interest-focus" do not meet
general approval, I would argue the need for terms reflecting these
distinctions. The external-internal tangle that "focalization" gets into
would be resolved because, by definition, a term such as "filter"
would be recognized as internal to the story world and "slant," by
contrast, as external to it. Separate terms would enable us to charac148
Point of View
terize texts more accurately than would any single term. Slant, we
could say, may or may not work in conjunction with filter. If it does,
we can distinguish limited filtration from multiple filtration. Either
term may or may not imply the transmission of information belonging to the narrator alone and presented in the form of commentary.
Camera-eye narration, then, is simply slant without filtration and
without narrator's commentary. Any of these can be conjoined with
or without centering and/or interest-focusing on a certain character.
Coming to Terms
another name than "unreliability" for the same reason that narration
and filtration themselves should be distinguished and not blurred
under a single rubric like "point of view" or "focalization." "Fallible"
seems a good term for a filter character's inaccurate, misled, or selfserving perception of events, situations, and other characters, for it
attributes less culpability to the character than does "unreliable."
After all, the character has not asked that her mind be entered or her
conversation overheard by a narrator and reported to a narratee. She
communicates only intradiegetically, with other characters in the
story. She is normally not aware of being in a story monitored by a
discourse. As long as she is a character in and not the narrator of the
main story or of a story-within-a-story, she does not purport to be
giving an account of that story. She cannot misrepresent it, because
she is not attempting to represent it; rather, she is living it. So she can
hardly be responsible to the narrative in the way that a narrator is.
She may be lying or acting unreliably in other ways, with diegetic
consequences to herself and other characters; within her own mind
she may be fooling herself; but as a character she has no direct access
to the discourse, to the transmission of the story, and therefore
cannot be accused of unreliable narration. The milder characterization "fallible""liable to mistake or to error"seems preferable to
the stronger term "unreliable," since it does not connote a knowledge of textual intention or the intent to deceive some narratee. 8
Confusion about this distinction began with the very first example
that Booth picked to illustrate unreliable narration: namely, "The
Liar."9 Though much of Booth's account of Henry James's novella is
accurate and sensitive, it does not recognize an essential fact: that the
protagonist, Oliver Lyon, cannot be an unreliable narrator for the
simple reason that he is not a narrator at all. He is, rather, a fallible
filter. Of course, the story transpires very much in Lyon's head.
Though Booth may have somewhat overstated the case for Lyon's
"viciousness," one can hardly disagree that Lyon rationalizes some
rather deplorable behavior on his own part. Still, he does so in the
privacy of his own mind, not as a representation to a narratee.
Indeed, as a character he has no consciousness of the existence of
a narratee. It is one thing to ask, "Is Lyon really villainous and
vicious?" Booth finds him so, and his reading certainly seems preferable to those of critics who swallow Lyon's interpretation of events
whole. But it is anotherand more narratologically relevantthing
150
Point of View
to ask, "How do we know that Lyon's views are flawed?" We know
not by listening to him tell the story but by listening to a covert
narrator represent his thoughts and intentions.
In unreliable narration, the implied author constructs a narration
that the implied reader must call into question.
(The broken line indicates the secret ironic message about the narrator's unreliability.)
In fallibility, on the other hand, the narrator asks the narratee, his
or her interlocutor in the discourse, to enjoy an irony at the expense
of a filter character.
story, including
a filtered
characters views
Narrator
Narratee
about events and
other characters
Coming to Terms
Maria "felt a soft wet substance and was surprised nobody spoke or
took off her bandage." Maria, the fallible filter, does not knowor,
better, does not allow herself to guessthat the substance is clay, a
traditional symbol of death. The substance she touches is nowhere
named in the story except in the title. The irony, pathetic as it is, is
between the narrator and the narratee at Maria's expense; it is part of
the larger irony that Maria goes around cheerfully concealing from
herself the sad figure that she cuts in Dublin. Similarly, in "A Little
Cloud," when Little Chandler acknowledges that he will never be a
popular poet but that English critics may some day recognize him as a
minor member of the Celtic School, there is reason to infer that he
will continue to delude himself, that he is never going to publish
anything.
There is no unreliable narration in Joyce's Dubliners. No deep
reader, I think, has proposed that events or characters were other
than as the narrator represents them. Where there is irony, it clearly
turns on the narratee's sharing of the narrator's attitude toward the
character's misguided attitudes. So it would not be precise to speak of
the characters as unreliable; rather, they are fallibleMaria in her
pathetic way, Chandler in his dreamy way, Duffy in his isolated way.
Both fallible filtration and unreliable narration are forms of irony, but
the ironized targets differ. In fallible filtration the irony inheres in a
secret message between the narrator and the narratee at the expense
of a character. In unreliable narration the irony inheres in a secret
message between the implied author and the implied reader at the
expense of the narrator.
In many traditional texts, such as Tom Jones and Emma, the implied
author's message does not differ essentially from the narrator's; we
do not search for differences between what the implied author intends and what the narrator tells or shows. We cannot find a reason
for believing that the implied author does not stand behind the
narrator's irony at the expense of fallible characters like Tom and
Emma. But in "Haircut" or the section of The Sound and the Fury
narrated by Jason Compson, there is good reason to doubt that the
implied author endorses the barber's account of Jim Kendall's decency or Jason's racial and religious attitudes. The drift of the narrative works against such presumptions. For example, Jim clearly is
not a good fellow at all but a bastard, a sadist who perpetrates
practical "jokes" out of sheer nastiness; thus, Whitey the barber is an
152
Point of View
Coming to Terms
discourse. He delivers what he understands to be a single, straightforward message, but the implied reader must infer that the ostensible
message is being canceled or at least called into question by an
underlying message that the narrator does not understand. The narrator is being ironized in the act of narrating. Much of the effect rests
on the implied reader's understanding that the narrator is unconscious or at least ambivalent about the duplicity, distortion, or naivet of his account. The implied reader understands the implied
author's ironization of the narrator, though the dramatized narratee
may or may not understand it.
Slant and Filter in the Cinema
Like literary narrators, cinematic narrators nave their slant. Ideological slant tends to be implicit, though in some films it is spelled out
in so many words by a voice-over or captions. A growing critical
literature documents the capitalist or bourgeois slant of the classical
Hollywood film (not to speak of its sexist and racist slants as well). 10
Slant has been found not only at the more abstract levels of narrative
but also at that of actualization or "style." 11 Hollywood films are
usually characterized by a "seamless" style in which the actualization
hides all marks of artifice, accounts for all projected story space and
time, and motivates all shot transitions in a totally unobtrusive or
"transparent" way. The seamless style presents events and characters
under the aspect of the totally "natural"a "natural" which, of
course, heavily supports the status quo. Hollywood films disguise
under the mask of "ordinary realism" what is in fact a highly ideological view of the world. Soviet filmmakers of the 19205, more conscious of their ideology, proceeded from a quite different stylistic
supposition. They developed techniques, such as Eisenstein's montage of conflict, designed to galvanize the audience into an attitude
receptive to the idea of revolution. At the level of individual style we
find idiosyncratic ideological or psychological slants: the typically
pessimistic slant of Alfred Hitchcock's narrators, the buoyant slant of
Federico Fellini's, the puritanic and "standing tall" slant of John
Ford's, the "sexual buddy" slant of Howard Hawks's, the musing,
environment-sensitive slant of Michelangelo Antonioni's, the heroic
nationalist slant of Leni Riefenstahl's, and so on.
154
Point of View
Much is known about the ideological and stylistic aspects of narrator's slant. The difference between cinematic slant and cinematic
filter, however, especially the marking of characters' perceptions and
cognitions, is less clearly understood. Unlike literature, cinema literally shows spaces, mediated or not by a character's gaze. The unmediated camera view may be called "perceptual slant." But the logic
of the story-discourse distinction argues that perceptual slant is a
different narrative phenomenon from perceptual filter. If we are to
sustain this cardinal distinction, if we are to recognize that film is
simply one among several media that actualize narrative, we must
avoid the metaphor that the camera "sees" the events and existents in
the story world at such and such a distance, from such and such an
angle. Rather it presents them at those distances and angles. In other
words, perceptual slant (not filter) frames the cinematic narrator's
transmission of visual and auditory imagery. Metaphorically, the
fictive world is presented to the real audience "under glass," even
though audiences are strongly induced to ignore that fact. The intensely realistic illusions created by cinematic technologyboth
photographic and acousticare so seductive that even sophisticated
film theorists sometimes forget the theoretical partition separating
the story world from its discourse and actualization.
That partition is more clearly apparent in paintings, especially
narrative paintings. Paintings are so clearly surface objects that no
one would think of speaking of their representations as "seeings."
There is no camera to confuse the issue. Viewers do not get so
engrossed in story content as to forget the surface quality of the
impression: the fact that figures represented on the canvas, no matter
how lifelike, are "really" daubs of pigment a few millimeters thick.
Perhaps because of the absence of temporal constraints (we may take
only a second to view a painting, or we may contemplate it until the
museum closes), we cannot forget the painting's two-dimensionality
even as we read the story out of the brush-strokes. Consider, for
example, Tintoretto's Christ at the Sea of Gallilee: the scene represents
the moment (in John 21:7) when Peter, upon seeing Christ on the
shore, "did cast himself into the sea."12 In the near left foreground
Christ is represented in quarter profile, facing out to sea, his right
index finger extended diagonally into deep story space; Peter's haloencircled head is depicted at an appropriate counterangle. We interpret his gaze as locked on Christ's, and we see that he has already put
155
Coming to Terms
one leg into the water. The interchange of looks is set at about a fortyfive degree angle from the picture plane. (This is a typical cinematic
technique for framing dialogue interchanges between characters.)
Clearly the painterly narrator does not "see" this scene but rather
presents it for us to see, as a registration upon a two-dimensional
surface. It is the convention of classical narrative paintings that our
perception necessarily occurs on this side of the picture plane; we are
separated from the events by the transparent membrane of discourse.
No viewer doubts that the story event is taking place "on the other
side," in a projected story world.
The cinema frame, too, presents events and characters from a post
this side of the story world; there is never any question about what is
included and what excluded from our perception. (This will presumably change when cinema goes holographic.) In literature, however,
framing (insofar as it exists in the reader's consciousness at all) is
figurative and elastic. In even the most descriptive or "cinematic"
novel, insofar as we "see" anything, it is only a sketch of what is
necessary to accommodate the action. The scene (if it is imaged at all)
is very fuzzy around the edges. But when we watch a film, we are
rigorously limited to a "this-much," a totally discrete selection of
story space represented in a rectangular frame of a certain dimension.
Of course, we tend to "forget" the frame's existence as we get into
the story, so powerful are the diegetic intimations of off-screen space.
If we recall scenes in well-loved films, chances are we will not
remember them as framed by a rectangle. Remembered films are
more like remembered novels: we remember plots and characters
more readily than the actual framing of scenes.
The convention is that the particular rectangle of visible material
constitutes a "favored view," a selection by the implied author which
the cinematic narrator is delegated to present. That selection entails a
certain distance and angle, certain lighting conditions, and so on
those that maximize the story's impact. The same is true of the
juxtaposition of shots. In Hitchcock's Sabotage (1936), for example,
the opening shots juxtapose familiar views of LondonOxford
Street, Big Ben and Parliament, Picadilly Circus, and so onwith a
bare light bulb. After the light bulb flickers and goes off, so do the
lights of the public buildings. The interpolation of the light bulb
(presumably any old light bulb in London) among the shots of the
buildings is to convey the notion that electric power in general has
156
Point of View
failed; if only the lights of the buildings went out, we might merely
assume that it was closing time. Not only are the views "favored,"
but so are their combinations.
Functioning as an additional constraint on this convention is the
further convention that the images may filter through some character's perceptual consciousness. The many ways by which perceptual
filter can be effectedespecially through eyeline match, shot-countershot, the 18o-degree rule, voice-off or voice-over, plotlogicare
well known. Through these devices, films deftly lock the audience
into a character's perception. But it is important to recognizeand
not often enough urgedthat character's gaze and narrator's representation operate on different sides of the story-discourse partition.
The character's perceptual filtration of objects and events is always
additional to the camera's representations; that is, the filter occupies a
space between the "naked occurrence" of the images and the audience's perception of them. The camera's slant remains in place,
even when it is temporarily mediated by the character's perceptual
filter.
Sometimes the cinematic narrator, through camera movement of
angle or other means, seems to be communicating the story directly
or unmediately, as for example in the establishing moments of a film
before any character appears. The view of the Manhattan skyline in
Woody Alien's Manhattan (1979) and the Ruritanian village in Hitchcock's The Lady Vanishes (1938) are shown directly, through the
narrator's slant. At other times the representation of the story world
is filtered through a character's perception and interest-focus; or it is
filtered through one character's perceptual filter and another character's interest-focus; or it is presented from no one's angle of perception but from a character's interest-focus. These possibilities are well
analyzed by Nick Browne with respect to a sequence in Stagecoach
(1939). In the scene at the dining table at the Dry Fork coach station,
one series of shots filters through the perception and interest of Lucy,
a model of propriety, the pregnant wife going to Lordsburg to join
her husband, a cavalry officer. But another series is not "associated
with or justified spatially as the depiction of anyone's glance."13 This
M>-glance-motivated series is no less common a phenomenon than
the filtered kind (hence the inadequacy of the suture theory of film
construction). As Browne points out, the specific mental dispositions
of the three characters involvedLucy, Dallas the prostitute, and
157
Coming to Terms
Ringo the escaped prisonercannot be easily keyed to the camera
placements. Rather, they invoke a general social situation:
The permanent and underlying fact about the mise-en-scne which
justifies the fixity of camera placement is its status as a social drama
of alliance and antagonism between two social rolesLucy, an
insider, a married woman and defender of custom; and Dallas, an
outsider and prostitute who violates the code of the table. . . . The
shots [that] might be called "objective," or perhaps "nobody's"
shots, in fact refer to or are a representation of Lucy's social dominance and formal privilege. [They] show a field of vision that
closely matches Lucy's conception of her own place in that social
world: its framing corresponds to her alliance with the group and to
her intention to exclude the outsiders, to deny their claim to recognition. It is in other words not exactly a description of Lucy's
subjectivity but an objectification of her social self-conception.
Though Lucy is visible in the frame, [the sequence] might be said,
metaphorically, to embody her point of view.14
That "metaphorical embodiment" is what I call "interest-focus"; in
fact, Browne has anticipated my use of the word: "These shots might
perhaps be read as statements of the 'interests' of characters." And, as
Browne is quick to suggest, interest focus, unlike strict perceptual
filter, can attach to more than one character at a time. For not only is
Lucy involved in and conscious of the social situation described by
Browne, but so is Dallas; at stake for her is "exclusion and humiliation," which we are moved to repudiate as "unjust." Interest-focus
also attaches to the Ringo Kid, who clearly feels the tension in the air.
Cinema enables "us as spectators to be two places at once, where the
camera is and 'with' the depicted person." In the sequence from
Stagecoach, the viewer is several "places" at once"with" the fictional viewer, "with" the viewed (Dallas), and even "with" onlookers (the Ringo Kid).15
So we see that "perceptual filter" technically codes the viewer's
reconstruction of story space. "Interest-focus," on the other hand,
depends on contextual signals presented by the story. The latter is
often but not exclusively "social" (in Browne's sense). There is, for
example, a conventional code of the camera's "following" a character,
keeping him on the screen: hence Hitchcock's success in getting
158
Point of View
Coming to Terms
character is thinking. This can make the mental content all the richer
to us, since there is more for us to reconstruct. Another solution is to
people screenplays with characters whose thoughts are intentionally
obscure.
Here we touch upon the secret of such an artist as Antonioni.
Consider the scenes in L'Avventura (1959) that turn on the protagonist Sandro's frustration about his career as an architect. Few words
are spoken, and those in no way express this frustration, but the
action is so devised and photographed that a reasonably sensitive
viewer can guess what is going on in Sandro's mind. As he shows
Claudia the architectural splendors of Noto from the roof of a
church, Sandro explains that he has given up the active practice of
architecture for the more profitable job of cost estimating. He says
this with relative equanimity. But later, he becomes preoccupied and
moody. Resisting Claudia's charming flirtatiousness, he takes a walk
in the town piazza. He purposely ruins an architecture student's
drawing of a detail of a building, denies doing so when challenged,
and boasts about the number of brawls he was involved in when he
was the student's age. Back at the hotel he intercepts Claudia, who
has dressed and is coming to meet him, and leads her back into their
room. He goes out on the balcony, restlessly puffs on a cigarette, and
then throws it into the street. At this inappropriate moment he tries
to force the bewildered Claudia to make love. She refuses, despite his
insistence. Then they talk desultorily about the search for Anna.
Clearly, what is really bothering Sandro is neither Anna's disappearance nor Claudia's response to his erratic amatory demands.
Rather, it is his sense of impotence about not working at his profession, not practicing the art which he loves and which fulfills him. He
is a victim of what Antonioni called una malattia dei sentimenti, "a
malaise of the feelings." He attempts to escape his professional frustration by using sex as an anodyne. Along with escape goes distraction, which, in Antonioni's films, is a mental gesture of escape. 17
Antonioni excels in conveying the indirections of visible behavior
and imbuing them with a sense of suppressed anguish. For him,
cinema's technical difficulty in depicting characters' thoughts is not a
burden but the motivation to explore a uniquely human dimension
of experience, an exploration that few other filmmakers have attempted.
160
CHAPTER
10/
Coming to Terms
so exclusively identified with pictorial imagining or "picturing?" Are
narrative gaps solely ones of picturingin other words, are films
"gapless" because visually explicit? Obviously not; as Bordwell
points out, gaps occur in films not only at the general narrative
levelin story ellipsesbut also at the stylistic or "surface" level.
Filmmakers like Antonioni purposely avoid the kind of communication that comes naturally to literary narrativeverbal clues about
what characters are thinking, either directly or as refracted by dialogue. However specified his films may be at the gross visual level,
they are extremely "gapped" at a more profound level, since most of
his characters will not or cannot say what is really on their minds.
Iser seems to be blurring the distinction between the right to an
absence of visual closure (readers "may never have had a clear conception of what the hero actually looks like") and the right to refuse
to form a mental image. The sentence "the reader of Tom Jones is able
to visualize the hero virtually for himself, and so his imagination
senses the vast number of possibilities" constitutes something of a
non sequitur: visualizing from words does not necessarily lead to a
large number of possible visualizations. It is just as easy to interpret
"That's not how I imagine him" as meaning that a given reader is so
attached to a certain image of Tom that he or she refuses any other
("he can't possibly look like Albert Finney").
Though the visual imagination may be less stimulated by a film
than by a novel, the conceptual imagination may be very much
stimulated by, say, a face filled with emotion that goes unexplained
by dialogue or diegetic context. In a way, the challenge is greater, and
our capacity to interpret faces is not innately up to our capacity to
image from words. Film has just as much room for artistic gappingthough it is an artistry of its own. In some cases, the film is
better suited to a subject than the novel that inspired it. 2
Still, good film adaptations of good novels are not a plentiful
commodity, for reasons that are of narratological interest. The central problem for film adapters is to transform narrative features that
come easily to language but hard to a medium that operates in "real
time" and whose natural focus is the surface appearance of things
hence film's traditional difficulties with temporal and spatial summaries, abstract narratorial commentary, representations of the
thinking and feeling of characters, and sopn. 3 Though such aspects
of narration in cinema can be introduced through a certain artifice
162
Coming to Terms
phizer? It is easier to base a film on a novel that is already covertly
narrated, totally or predominantly "shown" by a camera eye. The
greater challenge is presented by novels with talkative, expatiating
narrators; by the same token, these offer opportunities for more
creative cinema. By examining even a single imaginative adaptation,
we can learn something useful about the expressive possibilities of
the two media, at least at the present moment in film history.
To preserve the "sound" of a prominent narrator's "voice," film's
most obvious option is to replicate it, to make it literally audible.
Excellent adaptations have been made using voice-over narration
some memorable ones are Tom Jones (1963), Lolita (1962), A Clockwork Orange (1971), The Diary of a Country Priest (1950), Rashomon
1950), Apocalypse Now (1979), Sophie's Cholee (1982), Kiss of the Spider
Woman (1985). In a pioneering article on the subject, Martin Battestin
shows in detail how the film version of Tom Jones combines an
"authorial" voice-over with "old-fashioned" movie devices (editorial
wipes in various geometric forms, accelerated motion, and the like)
to create an effect analogous to that of the novel's "old-fashioned"
narration. Battestin argues that "analogy is the key" to good adaptation, and his is certainly a more enlightened view than that held by
the fidelity mongers. 5
The question of analogy is of considerable interest to a narratology
concerned with the similarities and differences of narrative media.
Since voice-over has been studied in considerable detailmost recently and efficaciously by Sarah Kozloff 6I will say nothing more
about it here. Nor shall I take up purely visual ways of conveying the
mental experiences of characters. These, too, have been well-studied, in the effect dubbed "mindscreen" by Bruce Kawina tech-nique
featured in 8 1/2 (1963), Wild Strawberries (1957), Annie Hall (i977)
and other films. 7 I turn, rather, to an innovative solution to the
problem of communicating the overt narration of a novel, the
solution used in the film version of The French Lieutenant's Woman
(1981), written by Harold Pinter and directed by Karel Reisz.8 My
discussion of this film does not argue the superiority of its particular
technique of adaptation to those used by other films, nor do I propose it as a model. I discuss it simply as one particularly clever way
among others through which the medium can accommodate problems of narrative transference. Some of the acclaim showered on the
film was pronxpted by its formal sophistication. It is the kind of film
164
165
Coming to Terms
versions of The French Lieutenant's Woman. Actually, these have not
been sufficiently attended to by critics and reviewers. Some of the
correspondences they have attributed to the adaptation are simplistic
and even inaccurate; for example, it is not true,as Jay Boyum
assertsthat "the book is essentially a novel within a novel" and
therefore readily translated to the screen as a movie within a movie. l1
The novel is a narrative framed by the commentary of the narrator, a
commentary that is itself not narrative but descriptive and expository-argumentative. 12
The novel concerns the struggles of a young English gentleman,
Charles Smithson, to surmount the smugness and insularity of the
Victorian Age. He is led to do so through the love of Sarah Woodruff, a woman who, the narrator profusely assures us, lives far ahead
of her time. But whether Charles can lift himself by his bootstraps
sufficiently to meet her demands is so problematic that the novel
offers two endings, and in only one does he manage to do so.
A large part of the novel's commentary is devoted to the historical
meaning and implications of Charles's effort to modernize his thinking. The focus is on Victorian history; our own era is brought in only
to shed more light on the nineteenth century. The film, on the other
hand, problematizes both eras, at least with respect to amatory attitudes. Instead of a voluble narrator, the film offers a counterpart
story of modern love, the romance of the two actors who play the
novel's protagonists in a projected screen version of The French
Lieutenant's Woman.
The novel is unimaginable without the wide-ranging narrator's
commentary. Since Charles himself is largely unconscious of his
historical situation, the narrator must be there to explain it to us in the
fullest possible detail. The narrator is endowed not only with total
knowledge of the mid-Victorian scene but also with the "foresight" afforded by his temporal vantage in the year 1967, exactly one
hundred years after the date of the story.
Charles is totally and bewilderingly absorbed by his love for Sarah
Woodruff. Sarah is the "French lieutenant's woman," notorious in
Lyme Regis for her reputed seduction and abandonment by a French
naval officer. Actually, the seduction never occurred; she uses the
story as a way of asserting her difference from the rest of her community and, indeed, her era. Though Sarah's is the one consciousness
that the novel does not generally enterthe narrator proclaiming his
166
167
Coming to Terms
contemporary account, not only from famous writers Thomas
Hardy, Alfred Tennyson, Jane Austen, Matthew Arnold, Lewis Carroll, John Henry Newman, A. H. Clough, Leslie Stephen, G. M.
Young, Charles Darwin, and Karl Marxbut also from such pedestrian documents as the Report from the Mining Districts, the City
Medical Report, the report of the Children's Employment Commission, even an ordinary citizen's letter to the Times. These quotations
and attendant commentary provide a panoramic contrast between
the Victorian way of looking at life and our own; love is an important
but by no means the only sector of the panorama.
It almost seems at times that the story is only a pretext for the
commentary, that the narrative has come to subserve the argument.
The narrator acknowledges as much. Among other self-consciousnesses is that of the possible subversion of one text-type by another:
Perhaps I am trying to pass off a concealed book of essays on you.
Instead of chapter headings, perhaps I should have written "On the
Horizontality of Existence," "The Illusions of Progress," "The
History of the Novel Form," "The Aetiology of Freedom," "Some
Forgotten Aspects of the Victorian Age" . . . what you will. 15
The narrator is very much concerned with the "evolutionary leap,"
which is the central clue to the nature of Sarah's "mystery." As a
twentieth-century Darwinian, he understands that Sarah's "difference" is a product not of hysteria but of biological development
hence the quotation from the biologist Martin Gardner which begins
the last chapter: "Evolution is simply the process by which chance
(the random mutations in the nucleic acid helix caused by natural
radiation) cooperates with natural law to create living forms better
and better adapted to survive." Just so, we are led to infer, it was
chance chromosomal distribution that made Sarah rebel against
straitlaced Victorian notions of woman's place, urged her to strike
out on her own, and cast her up on the wave of the future. Presumably, the same chromosomal randomness selects men who can cope
with this new female sense of independence and separate purpose, for
only such men will be attractive as mates to such women. The
narrator argues: "There is no intervening god . . . only life, as we
have, within our hazard-given abilities, made it ourselves, life as
Marx defined itthe actions of men (and of women) in pursuit of their
168
Coming to Terms
She turned to look at himor as it seemed to Charles, through him.
It was not so much what was positively in that face which remained
with him after that first meeting, but all that was not as he had expected;
for theirs was an age when the favored feminine look was the demure,
the obedient, the shy. Charles felt immediately as if he had trespassed;
as if the Cobb belonged to that face, and not to the Ancient Borough
of Lyme. It was not a pretty face, like Er-nestina's. It was certainly
not a beautiful face, by any period's standard or taste. But it was an
unforgettable face, and a tragic face. Its sorrow welled out of it as
purely, naturally and unstoppably as water out of a woodland spring.
There was no artifice there, no hypocrisy, no hysteria, no mask; and
above all, no sign of madness. The madness was in the empty sea, the
empty horizon, the lack of reason for such sorrow; as if the spring was
natural in itself, but unnatural in welling from a desert.18
Obviously, this extended entry into a character's mind cannot be
accommodated by simple eyeline matches. Meryl Streep and Jeremy
Irons are excellent players, and they get a great deal of purport into
their gazes. But there is no way that they or any other actors could
guarantee, by facial expression alone, the nuances of such a passage.
The filmmaker has the dilemma of choosing between a mute scene,
which provokes a stereotypic response, and some unduly artificial
device such as voice-over. In this case, he chooses the former. Only
later, and by other means, does he convey mental entries, interpretations, and commentary.
Whereas the novel's implied author undertakes to recover the
reality behind our clichd altitudes of the "Victorian," showing us
how exotic a time it was, the film, relying on what we share (or think
we share) with the Victorians, represents modern life as the less
comprehensible. These diverging intentions result in important
changes in plot and theme and, above all, character. For example,
unlike the novel, in which it is Sarah who is enigmatic, the film
problematizes the modern actress, Anna. Consider the crucial event
of both versions, the disappearance of Sarah. In the novel she vanishes and is rediscovered only by chance. She makes no attempt to
reestablish her relationship with Charles; she is quite content to
remain unmarried. Though she acknowledges that she threw herself
at Charles in Lyme and Exeter and that there had been "a madness" in
170
171
Coming to Terms
The film makes no attempt to interpret Charles's intricate state of
mind, offers nothing to correspond, for example, to the novel's
whimsical metaphor of the "little spray of milkwort from the bank
beside [Sarah], blue flowers like microscopic cherubs' genitals."22
Meryl Streep could, of course, have plucked a spray of milkwort, but
its colorful symbolism would have been lost on anyone who hadn't
read the novel. A quick cut to figurative cherubs' genitals would have
been a nice metaphoric touch, la Eisenstein, but few in the audience
would have gotten the point. It is hard to imagine any straightforward dramatic playing of the scene that could capture the narrator's
many ludic, godlike observations.
The interchange between Charles' thoughts and the narrator's
commentary is extremely fluid. The narrator is at Charles's elbow to
explain, with the certainty of historical generalization, how he can
entertain what strike us as mutually contradictory feelings:
Such a sudden shift of sexual key is impossible today. A man and a
woman are no sooner in any but the most casual contact than they
consider the possibility of a physical relationship. We consider such
frankness about the real drives of human behavior healthy, but in
Charles's time private minds did not admit the desires banned by the
public mind; and when the consciousness was sprung on by these
lurking tigers it was ludicrously unprepared.23
The narrator readily generalizes outward from the fiction to the real
Victorian and later eras; he takes deep pleasure in what, to a narrow
Lubbockian view, would seem diegetic indulgence. 24 For better or
worse, he is not satisfied with explaining Charles' behavior under a
single general topic. Without missing a beat, he proceeds to another
short generalizing essay, this time on the Victorian fondness for small
closed spaces. Ruminations on claustrophilic architecture lead in turn
to speculations on painting.25
The film version of this scene makes no effort to comment on the
times. It finds no acceptable way to expatiate on Victorian painting
or architecture, or on the Victorian refusal to entertaineven in the
privacy of one's own mindwhat society has declared unspeakable.
But it does convey a great deal about the personal, human elements
of the situationmostly by implicationthrough details of acting,
composition, camera movements, editing, and music. What it shows
172
173
Coming to Terms
novel, as Susanna Barber and Richard Messer do, conjecturing that
"like Charles, Mike feels the pain of a kind of rebirth."28 And the
relation between the stories needs to be more precisely stated than by
referring vaguely, as Boyum does, to some kind of "double vision."29
There is more contrast than similarity between the film's two
stories. In the final chapters of the novel, the issues are drawn clearly:
Sarah loves her work with the Pre-Raphaelite Brotherhood; she is
"not to be understood," even by herself; and her intellectual and
spiritual development has led her to conclude that she cannot love
Charles, or anyone else, "as a wife should."3" But in the film, Anna's
motives seem to be kept totally obscure. Despite muttering David's
name in her sleep (to Mike's discomfiture), her relationship with her
French lover seems quite passionless. She says she envies Sonia,
Mike's wife, but when asked why, she mumbles something banal
about Sonia's gardening ability. And her final "escape" from Mike is
wrapped in mystery. For his part, Mike is no Charles: he does not
throw away his present life, reject his family, and run after Anna.
No, the modern story is not a replay of the Victorian. Rather, it
attempts to dramatize the novel's commentary, though not by simply
assigning the narrator's lines to one of the characters. That would
have been the Jamesian solution, utilizing the modern actors as
ficelles. (On only one occasion does a character, Anna, comment on
the Victorian situation, and her observation is immediately undercut
by Mike's crude jest about the "two point four fucks a week" enjoyed
by the Victorian gentleman.) 31 The commentary, rather, is implicit: it
inheres in the juxtaposition of the two stories, the acted-out contrasts between Victorian and modern altitudes toward Eros. These
differences are left to the audience to infer. The inferences we draw
may be less specified than what we learn from the explicit pronouncements of the novel, but they are certainly viable enough. The
film makes us believe that Victorian love has as much to tell us about
modern love as vice versa.
To put the matter in more text-theoretical terms: where the novel's
commentary explicitly conveys exposition and argument at the service of the narrative, the film implies commentary through the very
invention of the juxtaposed modern story. The basic implementing
technique here is "crosscutting" (Christian Metz's "alternate syntagm" or "alternate montage"), 32 a venerable editing technique al174
Coming to Terms
the wild coastal area near Lyme Regis where Charles and Sarah hold
their trysts. 34] [Delayed to sequence 11: Anna takes off her wig and
stares at herself in the mirror.] [Deleted: In an illuminatingperhaps
too illuminatingsequence in Anna's dressing trailer, the couple
joke about the scene they have just played: Charles's accidental meeting with Sarah on the Undercliff and his request to stroll with her, a
request she refuses in a manner at once mysterious and coquettish.
Mike says he enjoyed it and asks Anna whether she found him
"sympathetic." She answers "Mmn. Definitely." Mike responds, "I
don't mean me. I mean him . . . you swished your skirtvery
provocative. Did you mean it?" Anna says, "Well, it worked. Didn't
it?" This scene would have constituted the first evidence that Mike is
falling seriously for Sarahnot Anna, but the character she plays
making Anna's "Well, it worked. Didn't it?" a double entendre tha t
perhaps even she is unaware of. ]
3. In a kind of garden shed (perhaps to parody the conservatory in
which Charles proposes to Ernestina), Mike and Anna, in modern
clothes, rehearse the scene that corresponds to the novel's second
meeting of Charles and Sarah on the Undercliff (chapter 16). Sarah is
supposed to slip and Charles to catch her; Anna does it poorly at first,
making Mike visibly edgy. The second time she finds the right note,
and in one of the film's most striking moments Mike stares at her as if
she suddenly has become Sarah, right before his eyes. His look is so
intense that it seems to register his own astonishment as much as
Charles's participation in the fictional scene. To emphasize Anna's
transformation and Mike's response, the cinematic narrator cuts, in
midshot, to the "real" Victorian scene: it is Anna who begins the fall,
but Sarah, in red wig and Victorian clothes, who ends it, slipping not
to the floor of the garden shed but to the ground of the Undercliff.
This shot floats the seed, in a purely visual way, of what the screenplay tried to convey verbally in the preceding deleted sequence
"Did you mean it?"
4. As Mike stands moodily smoking at the window of his hotel
room, we suddenly hear for the first time in the modern story Sarah's
haunting musical theme (scored for string quartet, especially cello),
which has been associated with her from the very opening shot. 35
This musical "bleedover" is the first of several that suggest, commentatively, Mike's infatuation with a woman of the previous century. In the diegetic space of the modern story, the theme tells us with
176
177
Coming to Terms
when David answers. It is clear that David is suspicious. Cut to
Mike's house: Sonia, his wife, asks him if he's all right. Mike suggests
inviting some of the cast for a party. He calls the hotel again and asks
David if they both will come to the party. David says, drily, "I give
you Anna" [!] and hands the phone to her. She accepts the invitation
as Mike whispers that he loves her. She responds diffidently so that
David cannot infer what is being said. [Deleted: David asks Anna
what has been decided about the film's ending; Anna answers that she
has decided she wants "to play it exactly as it's written." David
wonders if that won't lead to a fight. Anna responds grimly, "I hope
not." The implication is that she has decided to end her affair with
Mike. The shot was doubtless cut for the reason discussed above,
that her explicit act of "choosing" undermines some of the ambiguity
of her final disappearance.]
9. This sequence occurs in three brief parts: Anna enters a costumer's shop in a London mews; she poses in front of a mirror with
the white dress designed for the Windermere sequence, saying "I'm
going to like her in this"; she then gets back into the studio car and
drives off.
10. A long sequence depicting the party at Mike's house cuts
between the intrigue of Mike and Anna's affair and vignettes of the
various supporting actors, now out of character and wearing modern
clothes. The actress playing Mrs. Poulteney, a mild, pleasant woman,
admires Mike's child; when David asks Mike if the film concludes
with the happy or unhappy ending, Mike says "We're going for the
first endingI mean the second ending"; Anna expresses her envy
for Sonia's garden; "Sam" plays Bach on the piano; and so on.
Throughout the sequence Mike tries to get Anna alone; he complains
("This is pure bloody hell") and tells her they must talk "properly" in
Windermere. Anna asks what there is to say, but agrees.
11. The film ends with a cast party on the Windermere set. The
sound track is dominated by rock music, and the scene is visually
chaotic. The camera whirls around as members of the cast dance
"Dr. Grogan" (rather drunkenly), "Mrs. Poulteney," "Sam," "Mrs.
Fairley," "Montague," and others. Mike signals to Anna to meet him
in the house. As she passes through her dressing room, she stares at
herself in the mirror. Sarah's red wig sits next to her on a block. Mike
bumps into "Ernestina" on his way to meet Anna; he kisses her
perfunctorily, but she clings to him; he gives her an affectionate but
178
Coming to Terms
Mike wants to play the twentieth-century game by nineteenthcentury rules. By falling in love with a nineteenth-century fictional
character, he has experienced a charge of nineteenth-century passion.
The passion hardly seems to mark an evolutionary leap forward. The
film is less sanguine than the novel about the progress of evolution in
the emotional sphere.
There are numerous films whose subject is the making of a film:
81/2 (1963), Day forNight (1973), and Singin' in the Rain (1952) come
immediately to mind. But The French Lieutenant's Woman differs from
these in that its concern is not the problem of creativity or the labor of
the director but the impact of the "framed" film on the actors. Nor is
the film a mise-en-abyme: the modern story is not a miniature replica
of the Victorian story.37 On the contrary, its subject is the difference
between the two worlds as seen from the modern perspective. Only a
concern with modern Eros would explain the film's extended preoccupation with Mike's anguish. Mike has fallen anachronistically in
love with a fictional Victorian character. Like Pygmalion, he tries to
bring his beloved to life, in Anna, but for her own good reasons
though reasons we shall never knowAnna chooses not to play
Calatea. This impossible search for a fictional woman out of a bygone era is not a subject proposed by the novel, which handles the
modern repercussions of Victorian thought only in the expositoryargumentative mode. The film's theme is the plight of the modern
actor (perhaps as synecdoche for the modern artist or thoughtful
person in general) who gets a taste, if only vicariously, of a better,
older way of loving and, hence, of living. Unfortunately for him, the
world that made that kind of loving possible no longer exists. The
character and the actor thus provide reverse images: just as Charles
longs for a woman of the future, Mike longs for a woman of the past.
It is in this way that the film reflects something of the novel's commentary on Victorian mores but also commentary on our own.
The cinematography, mise-en-scne, editing, and other aspects of
the production cleverly underline Mike's dilemma by contrasting the
visual qualities of Victorian and modern life. For example, the visuals
of the modern story, Mike's visuals, especially at the end, seem at once
frenzied and banal. Unlike the elegant compositions and lighting of
the Victorian portion38the greenhouse proposal, the Undercliff,
the bucolic tryst in the ruined barn, the love chamber in Exeterthe
180
Coming to Terms
terious woman. 41 She says little that helps to characterize her; we
don't know whether she abandons Mike so precipitously out of
frivolity, fear of deep emotion, confusion, recognition of Mike's
confusion, or some combination of these. 42 But whatever her motives, they are very modern.
Thus, the film takes from the novel only the erotic themein
particular, the suggestion that Victorian mores may have been superior to our own. The novel's narrator argues that their very selfrestriction may actually have given the Victorians a sexual drive
keener than ours, that indeed they may in some unconscious way
have set such elaborate inhibitions for the purpose of intensifying
sexual pleasure, as a person may intentionally fast a whole day the
better to enjoy a gourmet dinner in the evening. He speculates that
our easy access to sex, our permissiveness and promiscuity, cannot
but trivialize the experience for us. He makes the comparison in a
vivid metaphor: "If you can only enjoy one apple a day, there's a great
deal to be said against living in an orchard of the wretched things;
you might even find apples sweeter if you were allowed only one a
week." 43 The novel by no means argues the advantages of the present. As critic Charles Scruggs puts it, Fowles is "capable of admiring
Victorian culture, and in precisely moral terms. He appreciates its
strong sense of purpose, its determination to make life meaningful,
because he is so acutely aware that his own age lacks a sense of
purpose and meaning." 44 For all its differences and abridgments, the
film communicates a similar attitudeindeed, communicates it even
more single-mindedly.45
The use of one narrative at the service of another is, of course,
common in both literature and film. Its usual form is that of the
Arabian Nights or the Canterbury Tales: the served narrative is the
frame story, and the serving narrative is a story that performs some
explicit function in the framediverting, illustrating, or whatever.
Or vice versa: the frame story is a mere pretext to get to the meat of
the narrative, which lies in the framed story (Wuthering Heights, The
Turn of the Screw). In either case, the intentions are usually clear. But
in the film version of The French Lieutenant's Woman, both the "direction" of the service and its exact force are kept uncertain, implicit,
highly subject to interpretation. The tactic is very much in keeping
withindeed, is the counterpart ofthe "self-consciousness" of the
novel. John Fowles said in an interview that " 'this whole fiction-as
182
183
CHAPTER
11/
Coming to Terms
that entails the whole distinction between the rhetoric of fiction and
the principal goal of narratology: that is, a description of the structure
of this one text-type.
To address the question of fictional rhetoric's end we must turn to
the senses of "rhetoric" in general. I mention only three, though I
know there are others worth discussing. The first can be dismissed in
short order: it is that in which "rhetoric" means the second-level
study of something, the elementary study being its "grammar."
Thus, a rhetoric of yoga or computer programming would include
instruction in stretches or programs too advanced for the beginner.
This sense is obviously irrelevant to narratology.
A second sense equates rhetoric with verbal (or, more generally,
semiotic) communication tout court. This contrasts with a third, more
restricted sense in which the term is ordinarily understood: namely,
the use of communicative means to suade. I say "suade" and not
"persuade" to emphasize, as Aristotle does, that rhetoric concerns the
urgings of the text, the "available means," rather than its ultimate
success or failure with real audiences. Rhetoric, in this sense, has
nothing to do with public opinion polls. The practical effect of texts
on public altitudes is more properly a subject for the social sciences.
On the face of it, the second or broad sense might seem most suited
to what we want to talk about. But if rhetoric is equated with
nothing less than the whole of verbal (or other semiotic) communication, there is no reason for separating it from the purely descriptive
study of language, particularly those branches that concern units
larger than the sentence and travel under the names "discourse analysis," "pragmatics," and the like. In the other direction, the definition suggests a concern with the means in and of themselves, without
respect to the larger purposes of the text. Such a rhetoric would be
simply a taxonomy. But the term "narratology" or "grammar of
narrative" suffices to represent the purely taxonomic interest. To be
of any use at all, "rhetoric of fiction" should mean something else.
If we emulate Kenneth Burke (and who would provide a more
distinguished model?), we can assign to the phrase "grammar of
fiction" the "grounding in formal considerations"that is, the
"purely internal relations" among narrative fictional termsand reserve "rhetoric of fiction" for the study of how the terms were put to
specific ends, measured as effects on implied audiences. But this is
very close to the third sense of "rhetoric"the traditional one that
186
Coming to Terms
their audience simply enjoyed the text. Though it speaks of other
sorts of values, particularly moral ones, much of Booth's book and
the tradition from which it derives concern aesthetic ends, the suitability of fit, so to speak, between technique and the audience's sense
of fulfillment or closure.
But even textual enjoyment and fulfillment are not simple matters.
Ancient rhetoric accounted for them by means of a subclass described
as "epideictic," from the Greek word for exhibition, display, showing off in public speaking. Though the original subject of epideictic
texts was praise or blame of the institution or person that was subject
of the speech, another assumption arose even in the ancient era:
namely, that the praise or blame could apply to the text's own form
and style. Clearly, the ancient rhetor's solicitation of approval of the
form of his speech, quite apart from its content, prefigures the
novelist's solicitation of the reader's acceptance of the validity of the
way the novel is put together, regardless of what the novel is about.
The author of a nondidactic fiction asks that you approve not of what
his characters do but rather of the way in which those doings are
presented and the strength and quality of the illusion they build in
your mind. It is not implausible to refer to this solicitation as a kind of
suasion and hence of rhetoric, though it seems crucial to remember
that the end it seeks is significantly different from that which ordinary rhetoric seeks. The expression "rhetoric of fiction," I believe,
best refers to a fiction's suasion that its unfolding form be accepted.
In Booth's view, the end desired by this form-directed rhetoric is a
maximal heightening of the fiction's effect. As part of his argument
against the dogmatic followers of Henry James, Booth tellingly
invokes the master's own denial that the scenic art alone should
prevail in fiction. He recalls James's belief that the purpose of any
technique is to heighten the intensity of the fictive illusion, and therefore that on occasion direct commentary by the narrator is justifiable.
Though the viability of this position hardly needs more support,
Booth's "intensity" is very much what the early Burke calls "saliency," for him the chief and only real semantic constant of form.
Formal techniques, he wrote in "Lexicon Rhetoricae," "impart emphasis regardless of their subject. Whatever the theme may be they
add saliency to this theme, the same design serving to make dismalness more dismal or gladness gladder."7
In attaching "rhetoric" to form, however, we must not imply that
188
Coming to Terms
nance and, as James and Booth insist, the intensification of the illusion itself. What is at issue in this kind of rhetoric is not a "message,"
or even any reference to the real world at all, but an imaginary
situation that is sui generis, homogeneous, conceivable, and hence
"acceptable" (in some relatively vague but recognizable sense of the
word). Booth says that through rhetoric the reader learns to know
where, in the world of possible values, he standsthat is, to know
"where the author wants him to stand." But, we must add, these are
possible values, including some that readers in the real world might
not endorse (as, say, in an "anti-anthropic" science fiction novel or, to
cite a notorious case, in Lewis-Ferdinand Cline's Journey to the End of
Night). Of course, the reader will agree to "stand" in that place only
for the nonce, for the occasion of the reading itself. You might or
might not start working for a Marxist revolution after reading
Maxim Gorky's Mother, but the decision could hardly be called an
important literary consequence of Gorky's suasion. Obviously, no
viable aesthetic rhetoric of fiction could demand practical or real-life
consequences of a novel's suasions.
David Lodge, himself a novelist and astute narratologist, remarked that " 'Rhetoric' is Professor Booth's term for the means by
which the writer makes known his vision to the reader and persuades
him of its validity."10 Booth's "communicate to impose" becomes
Lodge's more specifically rhetorical "persuade," and Booth's "fictional world" becomes Lodge's "vision." Lodge's terms in some
ways legitimate, better than do Booth's own, the use of the term
"rhetoric" in its traditional sense. If all we mean by "rhetoric" is the
art of communicating with readers, we are not totally justified in using
the term. The world is filled with nonsuasory communication, communication that does not try to impose opinions upon its audience.
For example, brokers routinely purchase newspaper ads to announce
issues of stocks and bonds that have already sold out. Their intention
is neither to urge the public to buy nor even to boast of their success
as underwriters but merely to follow correct procedure. It is not
accidental that many critics since the eighteenth century have located
literature's texts at the opposite pole from direct urgings. If rhetoric is
simply equated with communication, then both terms get muddled,
and an important distinction is lost.
But what does "validity" mean? Dictionaries say: "'Valid,' from
Latin 'validus,' meaning strong, as in 'valorous': sound, just, well190
Coming to Terms
biage. Nor does it help the story very much if we read the statement
as the narrator's own observation: clearly, the implied author's intent is
to characterize Maurice's mood in ways sympathetic to Maurice. But
can we say that the contrast in diction is rhetorically infelicitous? Is it
infelicitous that Maurice speaks dialect with Paula but that Geoffrey
shifts to standard English when speaking with Lydia? (No dramatic
reason is given for either move.) It all depends on how much one
likes Lawrence. His admirers will be untroubled, taking the refined
diction as evidence of the nobility of the simple soul. AntiLaurentians will find in the dictional disparity between charac-ters
and narrator the dilemma of much of Lawrence's fiction: for all its
desire to record accurately the speech of the simple but passionate and
therefore noble folk whose altitudes it wishes to celebrate, it
unwittingly patronizes them. Unlike, say, the narrator of the first
three stories of James Joyce's Dubliners, whose literary diction can be
rationalized by assuming that he later emerges from his meager
childhood milieu to become a writer, the well-spoken Laurentian
narrator seems to condescend to the characters from some lofty
vantage of knowledge, intelligence, and artistry. That runs counter to
the apparent intention of the fictions as a whole. For anti-Lauren-tians,
Lawrence's verisimilitude founders on the implied author's inability
to strike a harmonious balance between the narrator's voice and the
voices of the characters. 11
What sorts of tools are utilized by the aesthetic rhetoric of fiction?
Presumably, anything that can find its way into a text. Consider, for
instance, features of language and style. Henry James's style argues,
among other things, the leisure to weave an elaborate meditative
web. Hemingway's, on the other hand, argues the need for grace
under pressure. Another rhetorical tool is the manipulation of cultural
codes, as Roland Barthes demonstrated in S/Z. In point of fact,
however, what rhetoricians of fiction usually talk about is that set of
devices specially favored by narrative texts which are the ongoing
concern of narratologists: filter and slant, narrative voice, chronological relations between story and discourse, and so on. Aesthetic
rhetoric addresses such questions as "Why does the choice of firstperson narrative voice make Huck's story plausible, viable, esthetically whole?" No one, I think, would argue that the implied
author o( Huckleberry Finn makes the wrong choice, that some other
kind of narrator would render a more plausible account of the events
192
Coming to Terms
day of Boccaccio's Decameron, that Monna, the heroine, was "no less
virtuous than fair."
"Covert rhetoric," on the other hand, works only implicitly as
rhetoric; it pretends to be quite another kind of text. For instance, a
dramatized scene seems not to urge but simply to be. Still, Booth
would say, it functions in a deeply rhetorical way. (Other names by
which he refers to "covert rhetoric" are "disguised rhetoric," "ostensibly dramatic move[s]," "objective" and "impersonal" text, text
"pure" of rhetoric, "intrinsic rhetoric," and "acceptable rhetoric"
acceptable, presumably, to Lubbockian purists.) Booth argues that
"anything that heightens," anything "that can be made public"
indeed, everything in the narrative textis rhetorical. Rhetoric is
implicit even in the original given, James's donn, or what Burke calls
the "author's original symbol" (as opposed to the technical and formal complications of its presentation to an audience). For Booth,
even "at the instant when James exclaims to himself, 'Here is my
subject!' a rhetorical aspect is contained within the conception: the
subject is thought of as something that can be made public, something
that can be made into a communicated work." What is needed, and
what Booth finds in James's project, is a "general rhetoric in the
service of realism, rather than a particular rhetoric for the most
intense experience of distinctive effects."14
But I wonder whether the agglomerated elements that work rhetorically in a fiction can be homogenized under the term "general
rhetoric." For all its utility, direct characterization by narrator-authorized adjectives, say, remains only a technical device, one narratological choice among others available to the implied author. And it
becomes rhetorical only when its end is clear: that is, when we can
explain how it achieves an effect of audience control. Some kinds of
aesthetic ends discussed by Booth are speed, economy, verification,
and reinforcement. For example, in the Boccaccio story, the heroine's mind is entered briefly to substantiate the narrator's independent
praise of her virtue; in Emma, Mr. Knightley's dialogue is used to
reinforce the narrator's occasional, relatively muted comments on
Emma's shortcomings. And so on. These examples are perfectly
straightforward and clear. But a problem arises for me when the
word "rhetoric" is attached not only to the whole two-part complex
(I) technical choice + (2) end
194
Coming to Terms
actually works: the majority of James's readers (poor folk unblessed
by narratological training!) do not bracket off Maria Gostrey or
Harriet Stackpole as mere rhetorical devices.
In other words, ficelleslike inner views, chronological rearrangements, and the likeare not themselves "pieces" of rhetoric; they are
not urgings, either direct or indirect, overt or covert. They are
narrative techniques, options in the narrative grammar available for
the author's use. From the narrative-grammatical point of view,
ficelles are characters like any others. However, as Booth says, they
perform rhetorical junctions. This, I think, is the important point.
Textual rhetoric is immediately recognizable as such even if extracted
from the context; appeal inheres in the very words and grammar. But
the ficelle and other narrative devices become rhetorically functional
only in context, even as they continue to operate in normal diegetic
ways. Maria Gostrey may be a ficelle, but she doesn't cease being a
character for all that, a character whose existence, personality, and
actions are perfectly verisimilar in the world constructed by the
novel. In other words, when we are calling a text "rhetoric" (rather
than using the word in some other sense, such as the study of the
behavior of rhetors), we are really speaking of what might more
precisely be called "text-whose-rhetorical-intention-is-transparent,"
or even "text-that-expresses-its-rhetoric." But most narrative techniques do not express their rhetoric. Only professionals (authors,
critics, and theorists), would think of calling them "rhetorical."
There is no point in referring to direct interpretations by the narrator
as explicit rhetoric if we recognize no implicit rhetoric. And the use
of "dramatic" techniques to solve problems of controlling audience
attitudes, I suggest, is not so much "implicit rhetoric" as it is a
selection among narrative techniques to perform rhetorical functions.
I am not really arguing anything very different from what Booth
says himself when he observes, for instance, that "the rhetorical
dimension in literature is inescapable." His "dimension" is pretty
much what I mean by "function." Both terms allow for other fictional
dimensions alternative to the rhetorical. It is just that I want to
underline the distinction between "rhetoric-as-text" and "rhetoric-astextual-function" and to suggest that the former is so fraught with
potential misunderstanding that it may be prudent not to use "rhetoric" in that sense.
Booth's book may seem to deny a distinction between aesthetic
196
Coming to Terms
rhetoric, but let me take a stab. Consider the rhetoric of an infrequently discussed classic by Virginia Woolf, Jacob 's Room. I have
argued that rhetoric directed toward either aesthetic or ideological
ends rests securely on the grammar of fiction, the set of narrative
tools available to the author. Thus, a given narrative technique may
work in one way to support the aesthetic end and, at the same time,
in another way, to urge some proposition or to resonate thematically
with the real world. A narrative device fundamental to Jacob's Room,
as it is to Mrs Dalloway, is what I have called "shifting limited mental
access," or an abrupt entry into the mind of one character after
another but without the suggestion of omniscience. "Shifting limited
access," I explained in Story and Discourse, moves to the next character's mind without hinting that an all-knowing narrator is manipulating the movement for tidy plot reasons: "In such passages, the
narrator does not ransack mind after mind . . . for answers to
hermeneutic questions. The mental entries seem matters of chance,
reflecting the randomness of ordinary life." 17
Jacob's Room was the first of Woolfs novels to use the device with
any frequency. Sometimes the shift in consciousness seems to derive
from the physical proximity of strangers: in the first chapter we shift
abruptly from the mind of Betty Flanders, mourning her husband's
death, to that of a total stranger named Charles Steele, whose sole
interest in Betty is as a dab of color on the can vas he is painting.
Elsewhere in the novel, however, physical proximity is of no significance; simultaneity is the only basis for the conjunction of thoughts.
Jacob, in Greece, puts down the Daily Mail in disgust; at the same
moment, his friend Bonamy, back in London, laments to himself:
"[Jacob] will fall in love."18
But what about the rhetoric? What is the aesthetic-rhetorical and
what the ideological-rhetorical force of the technique of shifting
limited mental access in Jacob's Room? The device helps suade us
aesthetically to accept the Woolfian fictive universe, one in which,
says David Daiches, "experience is flux, and the lives of different
[people] shade imperceptibly into each other."19 Whether the real
world is actually like that, of course, is not at issue. But when we do
consider the text from the perspective of real-world concernsthat
is, consider it ideologicallywe may note that the suddenness of the
transfer from consciousness to consciousness dovetails with, and so
supports, a picture of modern life filled with empty busyness, dis198
Coming to Terms
and Peter Walsh, another kind of irony seems to be at work, something like this: "Though we may understand a loved one very well
indeed, so well that her consciousness seems the virtual mirror image of our own, she may be the last person we can stand to be
with, even for two hours, let alone for a lifetime."
I am not claiming that these sentences capture the theme of the
novel but merely that they provide a reasonable description of ideological (as opposed to aesthetic) ends that shifting limited consciousness seems to promote. To the extent that its use enhances, makes
salient, intensifies the Illusion of the novel, we can say that the
technique is functioning in good aesthetic-rhetorical fashion. At the
ideological level, on the other hand, we would want to ask how the
choice of shifting limited access confirms or undermines projections
of the real world beyond the fiction. For example, the shifts between
Clarissa's and Septimus's consciousness argue how comparable are
the fates of the powerless, whether they are powerless for reasons of
gender or of "mental illness" (quotation marks of irony here), in a
society where power resides in the hands of such insensitive and
authoritarian males as the psychiatrist Bradshaw. Similarly, in Jacob's
Room, the odd perspective chosen by the unnamed but clearly female
narrator is strongly ideological. For instance, there is a wonderful
description of a heated argument inside Jacob's room at Cambridge
by a narrator who can only be outside, prowling in the courtyard,
straining but unable to hear what is being said, drawing only rough
inferences from the gesticulations of this or that young man as he
passes the window. 20 Why is she outside in the cold if not because she
is a woman? The whole effect argues the cultural isolation of intellectual women in early twentieth-century England, excluded from the
centers of learning, made to feel like interlopers (at one point, Jacob
mutters to himself that women, like pet dogs, should not be allowed
into King's College Chapel). 21 The theme of exclusion, of course, is
one that Virginia Woolf expresses through nonfictional rhetoric in A
Room of One's Own and elsewhere.
Let us turn to a cinematic example: Michelangelo Antonioni's film
The Passenger (1975) is a meditation on the mysteries of identity and
on the impossibility of escaping one's fate.22 It is a study in what Otto
Rank calls doppelgngerism, the idea "that a person's past inescapably
clings to him and that it becomes his fate as soon as he tries to get rid
of it." 23 David Locke (Jack Nicholson), a reporter covering a re 200
Coming to Terms
ing to the same desert view that Locke contemplated after Robertson's death, Robertson comments on how beautiful and still and
"waiting" the landscape is. Locke responds that he prefers men to
landscapes.
The mixture of visual flashback and contemporary sound is
strange, and its strangeness matches and, in a sense, motivates the
strangeness of the men's resemblance and the opportunity that
Robertson's death affords Locke to change identity. Especially the
remark about "waiting" ties their fates together. The desert awaits
Robertson's death, which is imminent, and Locke's as well; he will
wait peacefully for histhat is, Robertson'sassassin in another
small hotel in an equally desolate landscape, in Spain.
Though politics is the ostensible motor running the film's narrative machinery, the ideological point suggested by this scene is metaphysical, not political. The film really turns on the personal fate of a
single man facing some ultimate existential questions: Who am I?
Shall I try to be someone else? Will I feel any different if I do? Should I
run from it or accept my death peacefully? What is it that I am
running from? What is it that I am dying of? The unusual framing of
this scene, in which a person has an unprecedented opportunity to
drop his own identity and take up someone else's, is a good example
of innovative cinematic rhetoric working to sophisticated ideological
ends. Everything about the sequence suggests a slant of philosophical
quietism, one that accepts death as part of life, an ideology that is
comforting but unsentimental. The point is supported and confirmed by another unique technical effect at the end of the film, an
extremely long tracking shot in which the camera exits through the
bars of Locke/Robertson's window, makes an extended tour of the
street outside, and then returns. At the very moment of Locke's
death, the camera is more preoccupied with the living environment
in front of the Hotel de la Gloria than with the assassination taking
place inside the room and the disposition of the corpse. Both these
sequences provide a rhetoric to confirm an ideology much in the
spirit of the existentialism of Alan Watts: "Death is the goal of life.
Nonbeing fulfills being; it does not negate being, just as space does
not negate what is solid. Each is the condition for the reality of the
other."24
The study of rhetoric can help narrative theory if its limits are
clearly recognized. The more closely they are defined, the better.
202
203
Notes
Notes to Introduction
1. America alone has produced significant works by Robert Alter, Ann
Banfield, David Bordwell, Wayne Booth, Edward Branigan, Peter Brooks,
Nick Browne, Ross Chambers, Dorrit Cohn, Jonathan Culler, Lubomir
Dolezel, Fredric Jameson, Brian Henderson, Benjamin Harshaw, Linda
Hutcheon, Bruce Kawin, Jeffrey Kittay, Susan Sniader Lanser, Steven Mailloux, David Miller, J. Hillis Miller, Bruce Morrissette, Toma Pavel, James
Phelan, Mary Louise Pratt, Gerald Prince, Peter Rabinowitz, Ralph Rader,
David Richter, Paul Ricoeur, Robert Scholes, Barbara Herrnstein Smith,
Susan Suleiman, Hayden White, and others. Abroad, narrative questions
have been illuminated by scholars like Mieke Bal, Roland Barthes, Claude
Bremond, Christine Brooke-Rose, Grard Genette, A.-G. Greimas, Lucien
Dallenbach, Umberto Eco, Roger Fowler, Michael Halliday, Philippe
Hamon, Stephen Heath, Wolfgang Iser, David Lodge, Frank Kermode,
Brian McHale, Christian Metz, Shlomith Rimmon -Kenan, Franz Stanzel,
Meir Sternberg, and Tzvetan Todorov. And, of course, the treasures of
Russian thought continue to be revealed: Eikhenbaum, Tomashevski, Tyn janov, Propp, Shklovski, Bakhtin. Wallace Martin (Recent Theories of Narra
tive [Ithaca: Cornell University Press, 1986]) has recently put much of this
work in historical perspective, and there has even appeared a useful Dictionary ofNarratology, by Gerald Prince (Lincoln: University of Nebraska Press,
1987). My debt to these authors goes without saying.
2. Seymour Chatman, Story and Discourse: Narrative Structure in Fiction
and Film (Ithaca: Cornell University Press, 1978).
205
Notes
Notes to Chapter I
1. So Grard Genette puts it in "Frontiers of Narrative," in Figures of
Literary Discourse, ed. Marie-Rose Logan, trans. Alan Sheridan (New York:
Columbia University Press, 1982), p. 128. (This essay originally appeared as
"Frontires du rcit," Communications n [1966], 152-63.) Roland Barthes
also argued the need for a theory of text-types in "An Introduction to the
Structural Analysis of Narrative," in Image Music Text, trans. Stephen Heath
(New York: Hill & Wang, 1977), pp. 79-124: "One of the tasks of such a
linguistics would be precisely that of establishing a typology of forms of
discourse. Three broad types can be recognized provisionally: metonymic
(narrativo), metaphoric (lyric poetry, sapiential discourse), enthymematic
(intellectual discourse)" (p. 84).
2. The "forms" or "modes of discourse" have been distinguished for
centuries but keep cropping up in recent narratological discussion. E.g.,
they seem to be what Ross Chambers calls "register" in his interesting essay
"Describing Description," in Meaning and Meaningjulness: Studies in the Anal
ysis and Interpretation of Texts, French Forum Monographs 15 (Lexington,
Ky.: French Forum, 1979), pp. 90-101. Chambers recognizes three such
registers: narrative, description, and commentary. His formulation, how ever, like Genette's, explains text-types by privileging narrative. In contrast,
I seek to explain the relations among text-types from a point of view that
does not privilege any type or medium.
3. This sense of Exposition, of course, is unrelated to the special sense it
has in narratology, namely the representation of plot events before the narra
tive proper begins. See Meir Sternberg, Expositional Modes and Temporal
Ordering in Fiction (Baltimore: Johns Hopkins University Press, 1978).
4. Gotthold Ephraim Lessing, Laocoon, Nathan the Wise, and Minna Von
Bamhelm, trans. William A. Steel (London: Dent, 1930).
5. The scholarship is conveniently summarized in Marianna Torgovnick, The Visual Arts, Pictorialism, and the Novel (Princeton: Princeton Uni
versity Press, 1985), p. 31.
6. In Rudolf Arnheim, Visual Thinking (Berkeley: University of Califor
nia Press, 1969), p. 35, as quoted in Torgovnick, Visual Arts, pp. 32-33.
7. Thus, I would disagree with Torgovnick's contention that "it makes
little more sense to exclude temporality from the nature of the visual arts
than to maintain that we perceive literature spatially when we regard an
unopened book and feel its dimensions" (Visual Arts, p. 33).
8. These terms are not, strictly speaking, equivalent, but there is sufficient family resemblance to permit equating them in this rough way.
9. Michel Beaujour, "Some Paradoxes of Description," Yale French Stud
ies 61 (1981), 27.
206
Notes
10. Expositions, on the other hand, are basically explanations (exponere =
"to put forth, explain"; explanare = "to make level or clear"). Thus, they
share with descriptions the task of rendering properties, but they do so by a
more recognizably discursive logicanalysis, definition, contrast, comparison.
11. The best modern study of argumentation is Chaim Perelman and L.
Olbrechts-Tyteca, The New Rhetoric: A Treatise on Argumentation, trans. John
Wilkinson and Purcell Weaver (Notre Dame, Ind.: University of Notre
Dame Press, 1969).
12. The difference between Argument and Exposition seems to be that the
arguer presumes the audience already to have a certain attitude, which she
tries to alter (or sometimes to reinforce). Generally, Argument presupposes
difference of opinion; Exposition presupposes an absence or confusion of
opinion. Of course, the clarification itself implies an argument about the
preferability of that clarification, so the line between these two text-types is
hazy. But the distinction remains useful to the extent that we can distinguish
the explicit suasion of an argument from the implicit suasion of an exposition.
13. The example of The Riveris analyzed in David Bordwell and Kristen
Thompson, Film Art: An Introduction, 2d ed. (New York: Knopf, 1986), pp.
57-62; this is one of the few books that attempt to categorize non-narrative
films. Bordwell and Thompson use the term "rhetorical" (p. 81) for the texttype that I call Argument, and they conflate expository and descriptive texttypes under the term "categorical."
14. Susan Suleiman, Authoritarian Fictions (New York: Columbia Univer
sity Press, 1983), p. 26. Suleiman's discussion of "exemplum" shows how
service is built into the word's very etymology: "The term exemplum (Greek
paradeigma) designated persuasion by induction, or argument by analogy (in
contrast to the enthymeme, or persuasion by deduction). . . . Aristotle . . .
already divided exempla into 'real' and 'fictional' onesthe former being
drawn from history or mythology, the latter being the invention of the
orator himself. In the category of fictional exempla, Aristotle distinguished
parables, or brief comparisons, from fables, which constitute a series of
actions, in other words, a story" (p. 27).
15. Susan Sniader Lanser, in The Narrative Act (Princeton: Princeton
University Press, 1981), goes too far in equating all fictions with the fable or
parabolic genres: "The novel's basic illocutionary activity," she writes, "is
ideological instruction; its basic plea: hear my word, believe and under stand" (p. 293). Suleiman's Authoritarian Fictions takes great care to separate
authoritarianthat is, thesis-dominated and hence "argued"fictions from
the rest of the universe of novels and short stories.
16. Jean La Fontaine, Fables of La Fontaine, trans. Elizur Wright (New
York: Derby & Jackson, 1860).
207
Notes
17. Henry Fielding, The History ofthe Adventures ofjoseph Andrews & His
Friend Mr Abraham Adams (i742; Harmondsworth: Penguin Books, 1954), p.
2718. Ibid., pp. 156-57. I owe this example, along with general edification
about the role of argumentation in fiction, to Glen McClish, "Rhetoric and
the Rise of the English Novel" (Ph.D. diss., University of California,
Berkeley, 1986), pp. 232-33.
19. Marie-Henri Beyle (Stendhal), The Red and the Black (1830), trans.
C. K. Scott Moncrieff (New York: Modern Library, 1926), p. 236.
20. Leo Tolstoy, Anna Karenina (1877), trans. Joel Carmichael (New
York: Bantam Books, 1981), pp. 601-2.
21. Wayne Booth, The Rhetoric of Fiction, 2d ed. (Chicago: University of
Chicago Press, 1961, 1983), chap. i.
22. Charles Dickens, Little Dorrit (1857; New York: New American Li
brary, 1980), p. i.
23. Grard Genette, Narrative Discourse, trans. Jane Lewin (Ithaca: Cornell University Press, 1980), p. 101.
24. Genette, "Frontiers," p. 136.
25. Fyodor Dostoevsky, Crime and Punishment (1866), trans. Constance
Garnett (New York: Macmillan, 1951), p. 2.
26. I take this example from Cleanth Brooks and Robert Penn Warren,
Understanding Fiction, 2d ed. (New York: Appleton-Century-Crofts, 1959),
p. 9, which says that though it is "moving" toward a narrative, and though
"one can easily imagine Lady Blessington's becoming a character in a historical novel," this is still a "character sketch"hence a description.
27. Genette, "Frontiers," p. 134.
28. The Miscellaneous Works in Prose and Verse of Sir Thomas Overbury,
Knt., ed. Edward F. Rimbault (London: Reeves & Turner, 1890), p. 69.
29. Quoted in Mieke Bal, "On Meanings and Descriptions," trans. by
Robert Corum, Studies in oth-Century Literature 6, nos. i and 2 (1982),
p. 126.
30. Wallace Martin, personal communication.
31. Bordwell and Thompson, Film Art, p. 49.
32. Alain Robbe-Grillet,Two Novels by Robbe-Grillet, trans.Richard
Howard (New York: Grove Press, 1965), p. 96. As Bruce Morrissette points
out in Novel and Film: Essays in Two Genres (Chicago: University of Chicago
Press, 1985), p. 24, what we get in La jalousie is "the phenomenological
transcription of the jealous husband's perceptions and sensations."That
these are the husband's perceptions and not those of some objective narrator
must be inferred, because there are no first person markers indicating himas
the speaker. Once we make that inference, Morrissette continues, we must
conclude that these perceptions and sensations are really "objectivecorrela -
208
Notes
tives ofthe subject's thoughts and passions." But in an age of psychological
fiction, "thoughts" and "passions" certainly qualify as narrative events,
indeed, the critical or even the only events of many fictions. The events of La
jalousie must be understood to occupy the space in the text between the
descriptive statements.
33. Genette, "Frontiers," p. 135.
34. I am indebted to Wallace Martin (personal communication) for this
language.
Notes to Chapter 2
1. Grard Genette, "Frontiers of Narrative," in Figures of Literary Dis
course, ed. Marie-Rose Logan, trans. Alan Sheridan (New York: Columbia
University Press, 1982), p. 137.
2. Grard Genette, Narrative Discourse, trans. Jane Lewin (Ithaca: Cornell University Press, 1980), p. 94 n. 12.
3. See Philippe Hamon, "Rhetorical Status of the Descriptive," Yale
French Studies 61 (1981), 1-26, trans. Patricia Baudoin of Philippe Hamon,
Introduction a l'analyse du dscriptif (Paris: Hachette, 1981), chap. i.
4. Ibid., p. 7.
5. See Jean-Franc,ois Marmontel, "Grammaire et littrature," in
Encydopdie mthodique (Paris: Panckoucke, 1782); Hamon, "Rhetorical Sta
tus," p. 32 n.2; Paul Valry, "Degas, danse, dssin," in Oeuvres, vol. 2
(Pliade ed.), pp. 1219-20 (quoted in Hamon, "Rhetorical Status," p. 10).
6. Hamon, "Rhetorical Status," p. n.
7. Ibid., p. 13.
8. Philippe Hamon, "Qu'est-ce qu'une description?" Potique 12 (1972),
475 (my translation).
9. Meir Sternberg, "Ordering the Unordered: Time, Space, and De
scriptive Coherence," Vale French Studies 61 (1981), 60-88.
10. James Joyce, A Portrait ofthe Artist as a Young Man (191415), ed.
Chester G. Anderson (New York: Viking Press, 1973), p. 184.
11. Roland Barthes, "L'efFet de rel," Communications n (1968), 84-89,
rpt. as "The Reality Effect," trans. R. Carter, in French Literary Theory
Today: A Reader, ed. Tzvetan Todorov (Cambridge: Cambridge University
Press, 1982).
12. Fyodor Dostoevsky, The Devils (1871), trans. David Magarshack
(Harmondsworth: Penguin Books, 1953), pp. 188-89.
13. Georg Lukcs, "Narrate or Describe? A Preliminary Discussion of
Naturalism and Formalism," in Writer and Critic and Other Essays, ed. and
trans. Arthur D. Kahn (New York: Grosset & Dunlap, 1970), pp. 111-12.
209
Notes
The ideological implications of this distinction between observer and participant are examined by Fredric Jameson in Marxism and Form: TwentiethCentury Dialectical Theories of Literature (Princeton: Princeton University
Press, 1974), pp. 200-202.
14. Lukcs, "Narrate or Describe?" p. 114.
15. Ibid., p. 116.
16. Wallace Martin, Recent Theories of Narrative (Ithaca: Cornell Univer
sity Press, 1986), p. 122.
17. Alexander Gelley, "The Represented World: Toward a Phenomenological Theory of Description in the Novel," Journal of Aesthetics and Art
Criticism 37 (1979), 187.
18. The point is made clear by Christian Metz in Film Language: A
Semiotics ofthe Cinema, trans. Michael Taylor (New York: Oxford Univer
sity Press, 1974), p. 128: "This in no way implies that the descriptive
syntagma can only be applied to motionless objects or persons. A descriptive
syntagma may very well cover an action, provided that it is an action whose
only intelligible internal relationship is one of spatial parallelism at any given
moment in timethat is to say, an action the viewer cannot mentally string
together in [diegetic] time. Example: a flock of sheep being herded (views of
the sheep, the shepherd, the sheepdog, etc.)."
19. Jeffrey Kittay, "Descriptive Limits," Vale French Studies 61 (1981),
225-4320. Ibid., p. 232.
21. Ibid., p. 239.
22. Gotthold Ephraim Lessing, Laocoon, Nathan the Wise, and Minna Von
Barnheltri, trans. William A. Steel (London: Dent, 1930), p. 59.
23. Hamon, "Rhetorical Status," p. 17. The passage occurs in Lessing,
Laocoon, p. 57.
24. Even Kittay seems uncertain about how to analyze the passage: "A
reading shows the actions of Hephaistos as a necessary structural component
of sorts, as each 'vignette' on the shield originales in the repetitions: 'Then he
made . . . ,' 'Then he pictured . . .' Is this passage one of action or of
description? Or is this a place where no one should care?" ("Descriptive
Limits," p. 241). If we are seriously concerned about the nature of narrativo
and its delimitation from other text-types, I think we should care. I am
trying to show that the antinomy of deep versus surface structure makes it
relatively easy for us to do so. Indeed, the only disappointment in Kittay's
otherwise brilliant article is its skeptical conclusion: "To read a passage as the
representaron ofthe act of making, as poesis is, is to fmd no final difference
between description and action" (p. 242). What a shameafter he has
seemed to establish the difference in so precise a way.
210
Notes
25. Homer, The litad, trans. W. H. D. Rouse (New York: Mentor Books,
1950), p. 225.
26. Alain Robbc-Grillct, Snapshots, trans. Bruce Morrisscttc (New York:
Grovc Press, 1968).
27. Cleanth Brooks and Robert Penn Warren, Understanding Fiction, 2d
ed. (New York: Appleton-Century-Croft, 1959), p. 257.
28. I intentionally say "skewed chrono-logic" and not "flashback" or
"anachrony" in Genette's sense. What is happening is not the later discourse
representation of an earlier story event. It is rather the reversal ofboth story
and discourse order.
29. Hamon, "Rhetorical Status," p. 26.
Notes to Chapter 3
1. The "cinematic narrator," discussed at length in Chapter 8, is the
transmitting agency, immanent to the film, which presents the images we
see and the sounds we hear. It is not the filmmaker or production team but
bears the same relation to those real people as does the narrator to the real
author of a novel. Neither is it a voice-over that introduces the action,
though that voice-over may be one of its devices.
2. In that sense, narratology would disagree with Aristotle's deprecation
of spectacle as "not hav[ing] much to do with poetic art and really be long[ing] to the business of producing the play" (Aristotle's Poetics: A Translation and Commentary for Students of Literature, trans. Len Golden [Englewood ClifFs, N.J.: Prentice-Hall, 1968], 14.8-10). It is doubtless true that
one simply hearing the plot ofOedipus would experience some degree of fear
and pity, but surely an excellent stage or screen performance will intensify
the enect a thousandfold.
3. Claude Ollier, "Rponse," Premier Plan, no. 18 (n.d.), 26, quoted in
Bruce Morrissette, Novel and Film: Essays in Two Genres (Chicago: Univer
sity of Chicago Press, 1985), p. 22.
4. John Fowles, The French Lieutenant's Woman (1969; New York: New
American Library, 1969), p. 10.
5. I have discussed these matters under the rubric of "determinacy" in
Seymour Chatman, Story and Discourse: Narrative Structure in Fiction and Film
(Ithaca: Cornell University Press, 1978), p. 30.
6. Christian Metz, Film Language: A Semiotics ofthe Cinema, trans. Mi
chael Taylor (New York: Oxford University Press, 1974), pp. 127-28.
7. L'Avventura: A Film by Michelangelo Antonioni, ed. David Denby,
trans. Jon Swan (New York: Grove Press, 1969), p. n.
211
Notes
8. 211 Orson Welles, Touch of Evil, ed. Terry Comito (New
Brunswick,
N.J.: Rutgers University Press, 1985), p. 56.
9. Ernest Callenbach (letter to the author).
i o. "Camera," here and throughout, is synecdochic for the whole cine matic apparatus.
u. Honor de Balzac, Old Goriot, trans. Marion Ayton Crawford (Harmondsworth: Penguin Books, 1959), p. 29.
12. Rudyard Kipling, "The Man Who Would Be King," in Cleanth
Brooks and Robert Penn Warren, Understanding Fiction, 3d ed. (Englewood
Cliffs, N.J.: Prentice-Hall, 1979), p. 85.
13. Gustave Flaubert, Madame Bovary, trans. Francis Steegmuller (New
York: Random House, 1957), pp. 273-74.
14. As quoted by Philippe Hamon, "Rhetorical Status of the Descrip tive," Yale French Studies 61 (1981), 3.
15. Videocassette is available on MCA Video.
16. Metz, Film Language, p. 167. Metz's distinction between "achronological" syntagmas (such as "parallel" and "bracket") and "chronological"
syntagmas (such as "descriptive") is not clearly drawn. What seems to be at
issue is the difference between descriptions that happen during story time
that is, in "pauses"and descriptions that happen outside of story time, in
the discourse only. The former should be called descriptive; the latter should
be called commentative because they convey the narrator's commentary.
Thus, in my view, the shots of the very opening of Adieu Philippine are
erroneously labeled "bracket" (p. 150). Though it is true that they "are
merely chosen as representative of a certain reality: work in a televisi on
studio," this is not a studio out in the real world but precisely the fictional
studio in which the protagonist works and which becomes the setting for the
first part of the film.
17. Charles Dickens, Great Expectations (1861), ed. Angus Calder (Harmondsworth: Penguin Books, 1965), chap. i.
18. All About Eve (Twentieth CenturyFox, 1950) is available on Magnetic Video.
19. The Oxford Companion to Film, ed. Liz-Anne Bawden (New York:
Oxford University Press, 1970), p. 614.
20. For an excellent account of the operant conventions of the classi c
Hollywood narrative film, see David Bordwell, Narration in the Fiction Film
(Madison: University of Wisconsin Press, 1985), a book that I discuss in
detail in Chapter 8.
21. Grard Genette, Narrative Discourse, trans. Jane Lewin (Ithaca: Cornell University Press, 1980), pp. 109-12.
22. See Chatman, Story and Discourse, pp. 63-65.
212
Notes
23. Malcolm Lowry, Under the Volcano (1947; New York: Signet Books,
1965), p. 29.
24. The Passenger (1975) is available on MGM videocassette.
25. Noel Burch, Theory ofFilm Practice, trans. Helen R. Lane (New York:
Praeger, 1973), p. 28. This efFect is well known in perceptual psychology: it
is difficult to tell whether an object seen through a peephole is large and
distant or near and small. Since context has been eliminated, we have no way
of knowing. See Bordwell, Narration in the Fiction Film, pp. 100104, for a
good discussion of the relevant literature on perception.
26. See Michelangelo Antonioni, // deserto rosso (Bologna: Capelli, 1964).
27. The frame is reproduced in black and white in Seymour Chatman,
Antonioni; or, The Surface of the World (Berkeley: University of California
Press, 1985), p. 127.
28. E. M. Forster, Aspects ofthe Novel (Harmondsworth: Penguin Books,
1962), p. 34.
Notes to Chapter 4
1. David Bordwell, Narration in the Fiction Film (Madison: University of
Wisconsin Press, 1985), p. 235; David Bordwell and Kristen Thompson,
Film Art: An Introduction, 2d ed. (New York: Knopf, 1986), chap. 3.
2. Christian Metz, Film Language: A Semiotics ofthe Cinema, trans. Michael Taylor (New York: Oxford University Press, 1974), p. 194: "It is not
because the cinema is language that it can tell such fine stories, but rather it
has become language because it has told such fine stories."
3. John Fell, Film and the Narrative Tradition (Berkeley: University of
California Press, 1986), documents the rise ofthe cinema out of nineteenthcentury narrative genresthe novel, stage play, and the like.
4. This is not true of television, which, from its very beginnings, has
been a multitextual medium. Many commercials and some documentaries
have explicitly argumentative structures. Even so, non-narrative programs
have always had to struggle for prime time, and several cable channels show
nothing but movies.
5. An exception is the final tille in some films, such as A Cry in the Dark,
which cites postdiegetic facts bearing on the film's moral issue.
6. The film is available for i6mm rental and on videotape (Embassy
4065); the scenariomore precisely, the cutting continuity or dcoupage
appears in Alain Resnais, Mon oncle d'Amrique, L'Avant Scne du Cinema, no.
263 (March 1981), 30, 47-72.
7. For what it's worth, here is what Jean Gruault said in an interview
213
Notes
about his and Resnais's intentions: "This may sound pretentious, but we used
Laborit, although not to make a didactic film. It's like Proust using the notes
of his fatherwho was a doctorto create his characters, especially their
medical histories. Laborit for me is a bit like Marx was for Brecht: an
illumination for telling a story" (Annette Insdorf, "French Uncles," Film
Comment 16 [September-October 1980], 23). It would be interesting to
know whether Laborit felt that he had "used" Gruault and Resnais as a way
of communicating his theory to a mass audience.
8. Le Monde, September 13, 1980, as quoted in L'Avant Scne du Cinema,
no. 263 (March 1981), 72.
9. That Laborit is self-consciously in the discourse is amusingly emphasized by the fact that he, unlike the fictional characters, can hea r what the
speakerine says about him. He playfully amplifies her commentary by mentioning his Vcndcesc origins.
10. This view is articulated by, e.g., Metz in Film Language, pp. 67, 116.
11. The overtness of the struggle between Rene and Veestrate is a function of their petit-bourgeois class (also signaled by their beefiness and unstylish suits). Struggle for dominion at the upper bourgeois level is less
physical: the battle between Jean and Michel is discreet and repressed. But
Laborit is not interested in class; he wishes only to show that ostensibly
different behaviors manifest the same biological imperatives.
12. Reviewing the film in "French Uncles," American Film 16 (Septem
ber-October, 1980), 22, Jan Dawson wrote: "The fragmentary narrative
technique itselfin leaving us constantly wanting to see more, and in
making us constantly aware that we are permitted to view only moments
selected to ilustrate a theory . . . serves to put a question mark at the end of
Laborit's matter-of-fact pronouncements." But the knife cuts both ways:
Laborit's discourse similarly puts the fiction into question. And both are put
into question by the final sequence: a traveling shot through an American
sluni.
13. Ibid.: the film endows nature with "an aching beauty that cannot be
explained awaynot even by the fact that its powers of adaptation and
survival are manifestly so much stronger than those of the humans attempting to fmd their niche within it."
14. The effect is further complicated by the fact that as she pretends to
storm out, Jean won't let her leave. But then she is genuinely frustrated
she weeps the frustrated tears of her childhood as she sits in her own
bedroom, held captive for the wrong reasons, prevented from doing the
noble thing. When she breaks one of her own vases, she is both acting angry
zndfeeling angry.
15. In an interview published with the screenplay, Resnais acknowledged
the resemblance to Blotv-Up and also to certain paintings of Paul Signac.
214
Notes
Notes to Chapter 5
1. See Wayne Booth, The Rhetoric of Fiction, 2d ed. (Chicago: University
of Chicago Press, 1983); Shlomith Rimmon, "A Comprehensive Theory of
Narrative: Genette's Figures III," PTL i (1976), 58 (a critique of Genette's
omission of the implied author from his narrative system); Shlomith
Rimmon-Kenan, Narrative Fiction: Contemporary Poetics (London: Methuen,
!983), pp- 86-89, 101-4; Grard Genette, who seems to accept a de-anthropomophized implied author in Narrative Discourse Revisited, trans. Jane
Lewin (Ithaca: Cornell University Press, 1988), chap. 19; W. J. M. Bronzwaer, "Implied Author, Extradiegetic Narrator, and Public Reader: Grard
Genette's Narratological Model and the Reading Version of Great Expectations," Neophilologus 62 (1978), 1-17; and W. J. M. Bronzwaer, "Mieke Bal's
Concept of Focalization: A Critical Note," Poetics Today 2 (1981), 193-201, a
critique of Bal's attempt to replace the pair "implied author-narrator" by the
pair "narrator-focalizer"; Mieke Bal, who defends her position in "The
Laughing Mice, or On Focalization," Poetics Today 2 (1981), 202-10; and
Gerald Prince, Dictionary of Narratology (Lincoln: University of Nebraska
Press, 1987), pp. 42-43.
2. Susan Sniader Lanser, The Narrative Act (Princeton: Princeton Uni
versity Press, 1981), argues that the narrator is close "to the author's own
enterprise and can provide important indices of the author's relationship to
the literary act" (p. 141), and that the narrator is the "textually encoded,
historically authoritative voice kin to but not identical with the biographical
person who wrote the text" (p. 152). In analyzing a story by Kate Chopin,
she finds it relevant to discuss facts about the real author; e.g., she claims to
know that the covert narrator is female "by virtue of the conventions linking
the author's social identity with that of the heterodiegetic narrative voice,"
and that "by the conventions of authorial equivalence this narrator may also
be assumed to share the personality and valuesthe imaginative and ideological consciousness of the authorial voice" (p. 250).
3. See W. K. Wimsatt, Jr., and Monroe Beardsley, "Intention," in Dic
tionary of World Literature, ed. Joseph T. Shipley (Paterson, N.J.: Littlefield,
Adams, 1960).
4. Our understanding of the reader's response has been expanded considerably in recent years. See Umberto Eco, The Role ofthe Reader: Explorations in the Semiotics ofTexts (Bloomington: Indiana University Press, 1979);
Stanley Fish, Is There a Text in This Class? The Authority of Interpretive
Communities (Cambridge: Harvard University Press, 1978); Wolfgang Iser,
The Act of Reading: A Theory ofAesthetic Response (Baltimore: Johns Hopkins
University Press, 1978); Wolfgang Iser, The Implied Reader: Patterns ofCommunication in Prose Fiction from Bunyan to Beckett (Baltimore: Johns Hopkins
215
Notes
University Press, 1974); Hans Robert Jauss, Toward an Aesthetic ofReception,
trans. Timothy Bahti (Minneapolis: University of Minnesota Press, 1982);
Stephen Mailloux, Interpretive Conventions: The Reader in the Study of American Fiction (Ithaca: Cornell University Press, 1982); Walter J. Ong, "The
Writer's Audience Is Always a Fiction," PMLA 90 (1975), 9-21; Gerald
Prince, Narratology: The Form and Functioning ofNanative (Berlin: Mouton,
1982); Pctcr Rabinowitz, Befare Reading: Narrative Conventions and the Politics
of Interpretation (Ithaca: Cornell University Press, 1988); Susan Sulciman and
Ingc Crosman, cd., The Reader in the Text: Essays on Audience and Interpretation (Princcton: Princcton University Press, 1980); Jane Tompkins, ed.,
Reader-Response Criticism: From Formalism to Post-Structuralism (Balti-more:
Johns Hopkins University Press, 1980); and Critica! Inquiry 9 (1982), a special
issue devoted to the "Politics of Interpretation."
5. Recent work in "possible-world" philosophy fmds a theoretical corroboration of the implied author in the very nature of fictional discourse.
According to possible-world philosophers, an implied or "reader's author"
must be postulated to permit the reader to attribute to the fiction beliefs that
may or may not be those of the real author. The implied or reader's author is
"different from the real author in that all his beliefs are inferred from the
fiction and are all consonant with it" (Koenraad Kuiper and Vernon Small,
"Constraints on Fictions: With an Analysis of M. K. Joseph's A Soldier's
Tale," Poetics Today j [1986], 495-526; quotation p. 498).
6. See Le Degr zro de l'criture, as quoted byJonathan Culler, Structuralist Poetics (Ithaca: Cornell University Press, 1975), p. 133; Lanser, Narrative
Act, p. 115; and Paul Ricoeur, "The Model of the Text: Meaningful Action
Considered as a Text," New Literary History 5 (1973), 91-117.
7. As Kuiper and Small suggest: "We believe that the [implied] author is
inferred from the text by the reader (and can therefore be inferred differently
by different readers)" ("Constraints on Fictions," p. 498).
8. E. D. Hirsch, Jr., Validity in Interpretation (New Haven: Yale Univer
sity Press, 1967), p. 211.
9. P.D. Juhl, Interpretation: An Essay in the Philosophy of Literary Criticism
(Princeton: Princeton University Press, 1981), p. 13.
i o. Perhaps even more important than Wimsatt and Beardsley's original
article, "Intention," are Beardsley's subsequent remarks in Aesthetics, 2d ed.
(Indianapolis, Ind.: Hackett, 1981), and in "Intentions and Interpretations: A
Fallacy Revisitcd," in The Aesthetic Point of View, ed. Michael Wreen and
Donald Callen (Ithaca: Cornell University Press, 1982), pp. 188-207.
u. Beardsley, Aesthetics, pp. 17, 20, 26.
12. Saul Bellow, quoted by Christopher Lehmann-Haupt in a review of
Bellow's More Die of Heartbreak, New York Times, May 21, 1987, p. C 29.
13. Beardsley, Aesthetics, p. 26. A reader-response cride, of course, would
2 16
Notes
change "the evidence of the poem itself" to "the opinions of all other
interpretive communities." But the point remains no less valid: a poet
cannot enforce an idiosyncratic interpretation (unless he changes the text itself
by appending his interpretation to it).
14. Ibid. The tcrm "dcsign" is uscd in a similar scnsc in Rogcr Fowlcr,
Linguistics and the Novel (London: Mcthucn, 1977), p. 8o.
15. I take the terms from the lucid article by William Tolhurst, "On What
a Text Is and How It Means," British Journal of Aesthetics 19 (1979), 3-14.
Another useful anti-intentionalist statement isjack Meiland, "Interpretation
as a Cognitive Discipline," Philosophy and Literature 2 (1978), 23-45. Meiland
uses Hirsch's terms"validity," "cognition," "objectivity"to argue quite
a difFerent conclusion: namely, that "a work can have several different
interpretations, all of which may be equally valid" (p. 30).
16. Tolhurst, "On What a Text Is," p. 5. "Utterance meaning is best
understood as the intention which a member of the intended audience [or
'implied reader'] would be most justified in attributing to the author based
on the knowledge and altitudes which he possesses in virtue of being a
member of the intended audience. Thus utterance meaning is to be construed as that hypothesis of utterer's meaning which is most justified on the
basis of those beliefs and altitudes which one possesses qua intended hearer
or intended reader" (p. 11).
17. Booth, Rhetoric of Fiction, pp. 70-74.
18. Italo Calvino expressed a similar view: "The author is an author
insofar as he enters into a role the way an actor does and identifies himself
with that projection of himself at the moment of writing" (quoted in
Christopher Lehmann-Haupt's review of Calvino's Uses of Literature: Essays
[New York: Wolff/Harcourt Brace Jovanovich, 1986], New York Times,
October 27, 1986, p. 17).
19. E.g., P. D. Juhl (interpretation, pp. 13-15) excludes from "intention"
(a) what the author planned to write or convey, (b) his "motive" in writing
(to gain wealth, recognition, or the like), and (c) the "sustained focal effect or
textual coherence of the work." He limits intention to "writing a certain
sequence of wordsin the sense, that is, of what [the author] meant by the
words he used." Wallace Martin reminds me (in a letter) that "intentionality"
in the latter sense is well established in philosophy, particularly in Edmund
Husserl: e.g., in a phrase such as "objective intentionality"; see "Intention
ality," in Dictionary of Philosophy, ed. Dagobert Runes (Totowa, N.J.: Littlefield, Adams, 1975), p. 148.
20. But see the discussion of the "career-author," below, for a way to
account for information deriving from other "authorial" choices in the other
works of a given author.
21. Fowler, Linguistics, pp. 79-80. Wayne Booth, in The Company We
217
Notes
Keep: An Ethics ofFiction (Berkeley: University of California Press, 1988), p.
91, continues to equate text and implied author: "What is never obvious is
any precise line dividing the questions that a given fictionwhich is to say, a
given implied authorinvites, or will tolerate, or will want to reject."
22. Some theorists believe that only human beings can "intend." Why
that should be so has never been clear to me. A text is the repository of
intentionsor, better, achieved intentionsso why bridle at saying that it
has intentions? Surely the statement "The Constitution intends to protect
freedom of the press" is no less meaningful than "The signers of the Consti
tution intended to protect freedom of the press." As Booth puts it: "If works
of art, like other seemingly inert objects in the world, do-ha ve natures in the
sense of presenting horizons of potentiality, then they can be said to take
altitudes toward our questions" (The Company We Keep, p. 91).
23. R. W. Stallman, "Intentions," in Princeton Encyclopedia ofPoetry and
Poetics, ed. Alex Preminger, enl. ed. (Princeton: Princeton University Press,
1974), P- 398.
24. W. K. Wimsatt, "Genesis: A Fallacy Revisited," in Day ofthe Leopards: Essays in Defense of Poems (New Haven: Yale University Press, 1976),
pp. 221-22.
25. As Genette writes (Nanative Discourse Revisited, p. 148): "If one means
by it that beyond the narrator (even an extradiegetic one), and by various
pinpointed or global signs, the narrative text (like any other text) produces a
certain idea (taking everything into account, this term is preferable to 'image,' and it is high time to substitute it for image) ofthe author, one means
something obvious, which I can only acknowledge and even insist on, and in
this sense I willingly approve of Bronzwaer's formula: 'The scope of narrative
theory [I would say, more carefully, of poetics] excludes the writer but
includes the implied author.'"
26. Wallace Martin (letter to the author).
27. Booth, Rhetoric ofFiction, p. 72.
28. Henry Fielding, The History ofthe Adventures ofjoseph Andrews & His
Friend Mr. Abraham Adams (1742; Harmondsworth: Penguin Books, 1954),
pp. 29-30.
29. This represents a change in the position I took in Seymour Chatman,
Story and Discourse: Narrative Structure in Fiction and Film (Ithaca: Cornell
University Press, 1978), pp. 148-51. I no longer believe that narrators and
narratees are "optional" and that where the narrator is "absent," the implied
author may address the reader directly. That is irreconcilable with the notion
ofthe implied author as silent source, as the perpetually reanimatable inven
tor ofthe whole. I am indebted to the discussion of this point by Shlomith
Rimmon-Kenan, Narrative Fiction, pp. 88-89 (though unlike her, I continue
to believe in the utility of a six-part communication model of narrative that
includes implied author and implied reader).
2 18
Notes
30. But Susan Lanser argues that "titles, prefaces, epigraphs, dedications,
and so on" provide information about the real author's views in "an open
forum" with the real reader (see Narrative Act, p. 125). And it is true that for
some fictions, titles must be attributed to the implied authors, since their
narrators could not (out of ignorance or whatever) articulate them (The
Sound and the Fury, One Flew over the Cuckoo's Nest). But in other cases, titles
may be sirnply one more message from the narrator to the narratee: A a
recherche du tempsperdu, Pride and Prejudice, Great Expectations are quite within
the conceptual powers and larger intentions of their respective narrators. It
seems excessive to categorize all titles as "extrafictional."
31. E.g., Balzac's "republicanism," some claim, emerges only in his
fiction. See Genette, Narrative Discourse Revisited, pp. 142-43.
32. Wayne Booth, Critica! Understanding: The Powers and Limits ofPluralism (Chicago: University of Chicago Press, 1979), p. 270.
Notes to Chapter 6
1. Wayne Booth, The Rhetoric ofFiction, 2d ed. (Chicago: University of
Chicago Press, 1983), pp. 422-34.
2. I take the example of Anne Frank from an unpubhshed manuscript by
Jeffrey Staley, "'Like Trees, Walking': Reading the Gospels with Open
Eyes."
3. W. K. Wimsatt, Jr., "Genesis: A Fallacy Revisited," in Day ofthe
Leopards: Essays in Defense of Poems (New Haven: Yale University Press,
1976)', p. 206.
4. E. A. Speiser, introduction to The Anchor Bible: Genesis (Carden City,
N.Y.: Doubleday, 1964), pp. xx-xxi, xxvi. The "Yahwist" referred to God
by the personal name Yahweh (Jehovah); the "Elohist" used Elohim, "the
generic Hebrew word for divine being." I am grateful to Robert Alter for the
citation.
5. Producer Arthur Freed, quoted by Lillian Ross in Picture (New York:
Rinehart, 1952), p. 194.
6. Auteurismethe idea that single individuals, typically directors, are
the true sources of films of qualitywas imported from France by Andrew
Sarris (see The American Cinema: Directors and Directions: ig2g-i<)68 [New
York: Dutton, 1968]). The auteurssuch directors as Orson Welles, John
Ford, Alfred Hitchcock, Howard Hawks were supposed somehow to
have overcome the overriding influence of the studios to achieve largely
personal visions.
7. These reviews are all quoted in Ross, Picture, pp. 229-31. The Red
Badge of Courage (1951) is available on MGM videocassette.
8. Bronislau Kaper, quoted in Ross, Picture, p. 144.
219
Notes
9. Conversation quoted in ibid., p. 59. Huston was probably right. As a
film critic for the New York Tribuine remarked, "A redundant narration . . .
clutters up the sound track from time to time explaining facts already clear in
the images" (p. 229). Rheinhardt defended his decisi on in a wonderfully
titled article, "Soundtrack Narration: Its Use Is Not Always a Resort of the
Lazy or the Incompetent," Films in Review 4 (1953), 459-60.
10. Review quoted in Ross, Picture, pp. 229-30.
11. Dorrit Cohn, Transparent Minas (Princeton: Princeton University
Press, 1978), p. 14.
12. Stephen Grane, The Red Badge ofCourage (1895; New York: Bantam
Books, 1983), p. 64.
13. Adam Garbicz and Jacek Klinowski, Cinema, the Magic Vehicle: A
Guide to Its Achievement, vol. i (New York: Schocken Books, 1983),^. 73.
14. Ross, Picture, pp. 215-16.
15. Wimsatt, "Genesis," p. 214. See also David Magarshack, Chekhov the
Dramatist (New York: Hill & Wang, 1960), pp. 188-89.
16. Rosalyn Krauss, review of Arthur Danto, The Transfiguration ofthe
Commonplace (Cambridge, Mass.: Harvard University Press, 1981), New
Republic, May 25, 1987, pp. 29-30.
17. Marianna Torgovnick, Closure in the Novel (Princeton: Princeton
University Press, 1981), p. 16.
18. Albert J. Guerard, The Triumph ofthe Novel (Chicago: University of
Chicago Press, 1982), p. 294.
19. Irving Howe, Politics and the Novel (Greenwich, Conn.: Fawcett,
1967), pp. 73-7520. Fyodor Dostoevsky, The Notebooks for "The Possessed," ed. Edward
Wasiolek, trans. Victor Terras (Chicago: University of Chicago Press,
1968), p. 12.
21. Ibid., pp. 12-13.
22. Vladimir A. Tunimanov, "The Narrator in The Devils" trans. Susanne Fusso, in Dostoevsky: New Perspectives, ed. Robert Louis Jackson (Englewood ClifFs, N.J.: Prentice-Hall, 1984), pp. 154-55. Tunimanov goes on:
"It is impossible to place [a character] in some sort of category that exhausts
his essence. He fits into many categories, and one can expand the number of
them almost to infinity. This does not at all contradict the presence of a
multitude of facts and 'factlets,' everyday details that create an impression of
'redundancy' and 'needlessness.' On the contrary, the petty details are good
in that by their insignificance they countervail the power of categories and
make that power illusory."
23. Newsweek, May 2, 1988, p. 51.
24. Hence, perhaps, the tiny cigarette, as emblem of the pre-tumescent
cherub.
220
Notes
25. Fredric Jameson, "Class and Allegory in Contemporary Mass Cul
ture: Dog Day Ajternoon as a Political Film," in Movies and Methods, vol. 2,
ed. Bill Nichols (Berkeley: University of California Press, 1985), pp. 715-33
(originally published in College English 38 [1977]).
26. Ibid., pp. 719-20.
27. Indccd, the corporate intcrcst in public turmoil is probably at the
opposite pole, as Jameson points out (ibid., pp. 720 -21): the media's "repeated stereotypical use of otherwise disturbing and alien phenomena in our
present social conjuncturepolitical militancy, student revolt, resistance to
and hatred of authorityhas an effect of containment for the system as a
whole. To name something is to domesticate it, to refer to it repeatedly is to
persuade a fearful and beleaguered middle-class public that all of that is part
of a known and catalogued world and thus somehow in order."
28. Ibid., p. 723.
29. Ibid., p. 728.
30. Ibid., p. 730.
Notes to Chapter 7
1. Grard Genette, Narrative Discourse, trans. Jane Lewin (Ithaca: Cornell University Press, 1980), p. 162 and n2. Genette prefers the translation
"pure narrative" (rather than "simple narrative") for haple diegesis, arguing
that it is so called because it is not "mixed," that is, does not contain direct
quotation of dialogue.
2. Aristole's Poetics: A Translation and Commentary for Students of Literature, trans. Leon Golden (Englewood Cliffs, N.J.: Prentice-Hall, 1968),
5-i449b(p.io).
3. Genette, Narrative Discourse, p. 163. Why, too, the quotation marks?
If a short story contained nothing but the quoted speeches ofthe characters,
wouldn't it be genuinely mimetic in the same sense as a published playscript?
4. Grard Genette, Narrative Discourse Revisited, trans. Jane Lewin
(Ithaca: Cornell University Press, 1988), p. 41.
5. Aristotle, Poetics, 24.1459b (p. 43).
6. "The very idea ofshowing, like that of imitation or narrative representation (and even more so, because of its naively visual character), is completely illusory: in contrast to dramatic representation, no narrative can
'show' or 'imitate' the story it tells. All it can do is tell it in a manner which is
detailed, precise, 'alive' and in that way give more or less the illusion of
mimesiswhich is the only narrative mimesis, for this single and sufficient
reason: that narration, oral or written, is a fact of language, and language
signifies without imitating. Unless, of course, the object signified (narrated)
221
Notes
be itself language" (Genette, Narrative Discourse, pp. 163-64). And, again, "I
believe there is no imitation in narrative because narrative, like everything
(or almost everything) in literature, is an act of language. And, therefore,
there can be no more imitation in narrative in particular than there is in
language in general. Like every verbal act, a narrative can only informthat
is, transmit meanings" (Genette, Narrative Discourse Revisited, pp. 42-43).
7. In Narrative Discourse Revisited, p. 43, Genette does not even acknowledge that the dialogue is "imitated": it is, he says, simply "reproduced" or
"transcribed." "Narrative does not 'represent' a (real or fictive) story, it
recounts itthat is, it signifies it by means of languageexcept for the
already verbal elements of the story (dialogues, monologues). And these, too,
it does not imitatenot, certainly, because it cannot, but simply because it
need not, since it can directly reproduce them, or more precisely, transcribe
them."
8. Umberto Eco demonstrates the conventionality of iconic signs in The
Open Work, trans. Anna Concogni (Cambridge: Harvard University Press,
1988). He makes two points of relevance to this discussion. First, iconicity is
always a question of degree: e.g., a Renaissance sculpture is more i conic
with respect to dimension, proportion, and contour than to color; a film is
more iconic with respect to movement than a photograph. Second, the
process of recognizing iconic signs is itself coded, not "natural." Schematic
drawings are, in some sense, iconic, but one must understand the nature of
the signifieds to grasp that iconicity. Grossly simplified drawings such as
comic strips are more easily readable by adults than by children; hence, it is
clear that time and energy have been required to learn their codes of representation. In short, "analogous" should not be understood in too simple a
sense; an iconic sign is better defined as one that constructs a "model of
relations" which is "homologous" to the model of perceptual relations we
construct in recognizing the signified itself.
9. In Seymour Chatman, Story and Discourse: Narrative Structure in Fiction
and Film (Ithaca: Cornell University Press, 1978), pp. 166-69, I tried to
mitigate the assertion by offering readers the "option" of treating purely
mimetic narratives as "minimally narrated" instead of "non-narrated." But
that tactic clearly evaded the issue.
10. I treat the terms "presenting" and "transmitting" as synonymous. See
Seymour Chatman, "The Structure of Narrative Transmission," in Style and
Structure in Literature: Essays in the New Stylistics, ed. Roger Fowler (Ithaca:
Cornell University Press, 1975), pp. 213-57.
11. Chatman, Story and Discourse, p. 146. I would also now reserve the
concept of "mediate" for the action of filter characters (see Chapter 9).
12. Roger Fowler, Linguistics and the Novel (London: Methuen, 1977), p.
74. I apologize for once again subjecting this fatigued story to narratological
analysis, but apparently we still have something to learn from it.
222
Notes
13. Nor is it clear to whom he is "invisible." To other characters in the
story? To us? To "himself"? And is he invisible in the same way as H. G.
Wells' character is? Or is there a special kind of narratorial invisibility?
14. Susan Sniader Lanser, The Narrative Act (Princeton: Princeton Uni
versity Press, 1981), p. 266.
15. Lanser, Narrative Act, pp. 266-67.
16. Jonathan Culler, "Problems in the Theory of Fiction," Diacritics 14
(Spring 1984), 5, writes wisely about this matter: "The argument [of those
who insist on assigning gender to narrators] would be . . . that since every
person has a sex, and narrators are people, every narrator must have a sex,
and to omit discussion of the sex of narrators is to miss important aspects of
novels. This argument is plausible only, it seems to me, because we have
come to take for granted that we explain textual details by adducing narra
tors and explain narrators by adducing qualities of real people. . . . The
theory of fiction needs to be alert to the inadequacies of this orientation,
which strives to convert everything in language to a mark of human personalities."
17. Lanser, Narrative Act, p. 268.
Notes to Chapter 8
1. E.g., Jean-Pierre Oudart, "Cinema and Suture," Screen 18 (1977-78),
35-47; and Daniel Dayan, "The Tutor-Code of Classical Cinema," in Movies
and Methods, vol. i, ed. Bill Nichols (Berkeley: University of California
Press, 1976), pp. 438-50effectively answered by William Rothman,
"Against the System of the Suture," in Nichols, Movies and Methods, vol. i,
PP- 451-592. David Bordwell, Narration in the Fiction Film (Madison: University of
Wisconsin Press, 1985), chap. 2. For another good critique, see Noel Carroll, Mystijying Movies (New York: Columbia University Press, 1988).
3. See Christian Metz, "Story/Discourse (A Note on Two Kinds of
Voyeurism)," in The Imaginary Signifier, trans. Celia Britton (Bloomington:
Indiana University Press, 1981). Benveniste's views appear in Problmes de
linguistique genrale, 2 vols. (Paris: Gallimard, 1966, 1974); the first volume
has been published as Problems in General Linguistics, trans. Mary Elizabeth
Meek (Coral Cables, Fia.: University of Miami Press, 1971). For Ben veniste, discours refers to enunciations that contain references to the speaker
and/or listener: thus to sentences with personal pronouns , imperatives,
"deictic" adverbs, etc. Histoire, on the other hand, comprises enunciations
that do not contain such marks, thus giving the impression of a totally
impersonal account. This sense is especially strong for French speakers,
because French has a tense form, the pass simple or aorist, which is used
223
Notes
solely for literary narrative. For a full discussion of the implications of this
distinction, see the work of Ann Banfield, esp. Unspeakable Sentences: Narration and Representation in the Language ofFiction (London: Routledge & Kegan
Paul, 1982). My own use of "story" and "discourse" is unrelated to Benveniste's.
4. Bordwell goes into detail to show the difficulties of applying enunciation theory to film studies (Narration in the Fiction Film, pp. 2125). E.g.,
Mark Nash, "Vampyr and the Fantastic," Screen 17, no. 3 (1976), 2967,
"must reject nearly all of Benveniste's 'means of enunciation' (e.g. verb
tense, signs of time) as inapplicable," and the only one left, "person," is
given a questionable filmic equivalent ("What would a second-person image
look like?" asks Bordwell). He fmds similar problems in the work of Franc.ois Jost, Nick Browne, Alain Bergala, Marie Ropars-Wuilleaumier, Raymond Bellour, and Stephen Heath.
5. Bordwell, Narration in the Fiction Film, p. 49. As I argued in Story and
Discourse: Narrative Structure in Fiction and Film (Ithaca: Cornell University
Press, 1978), pp. 45-48, causality is not the only logical principle (additional
to temporal sequentiality) that connects story events. I proposed the term
"contingency" to explain those cases not explained by strict causality. Bord
well (p. 51) seems to be referring to these other possibilities when he speaks
of "parallelism." E.g., in Rear Window, the courtyard vignettes illustrating a
variety of amatory relationships do not, strictly speaking, fit into the causal
pattern of either the Thorwald murder plot or the Jeff-Lisa love plot, but
they do form a kind of parallel to the latter.
6. What I called mythos, after Aristotle, in Chatman, Story and Discourse,
PP- 19, 437. Bordwell, Narration in the Fiction Film, p. 51.
8. Ibid., p. 50.
9. Ibid., pp. 49, 53. Other film theorists who argue against the cinematic
narrator are Edward Branigan, Point of View in the Cinema: A Theory of
Narration and Subjectivity in Classical Film (Berlin: Mouton, 1984); and Brian
Henderson, "Tense, Mood, and Voice in Film (Notes after Genette)," Film
Quarterly 36, no. 4 (1983), 4-17.
10. Bordwell, Narration in the Fiction Film, pp. 6162.
11. Nick Browne, The Rhetoric of Filmic Narration (Ann Arbor, Mich.:
UMI Research Press, 1982), p. i.
12. Bordwell's theory puts an odd cast on other terms as well: e.g.,
"representation." He writes: "The spectator possesses stylistic schemata as
well as others, and these invariably affect the overall process of narrative
representation" (Narration in the Fiction Film, p. 53). But surely the viewer
does not "represent"or even re-presentthe film in any usual sense of the
word.
13. Ibid., pp. 57-58. For the useful concepts of "aesthetic objects" and
224
Notes
"aesthetic qualities," see Monroe Beardsley, Aesthetics, 2d ed. (Indianapolis, Ind.: Hackett, 1981), esp. pp. 38, 63-65. Beardsley reaffirmed his position in 1981 in the postscript to the second edition; see esp. pp. xxviii
xxxi.
14. Bordwell, Narration in the Fiction Film, p. 78.
15. Ibid., pp. 57-61.
16. Ibid., p. 62.
17. The concept has suggested many synonyms, as pointed out by Linda
Hutcheon, Narcissistic Narratife: The Metafictional Paradox (Waterloo, Ont.:
Wilfrid Laurier University Press, 1980), p. i: "self-conscious," "self-reflective," "self-informing," "self-reflexive," "auto-referential," "auto-representational," and her own term "narcissistic." See also Robert Alter, Partial
Magic: The Novel as a Self-Conscious Genre (Berkeley: University of Califor
nia Press, 1976).
18. Bordwell, Narration in the Fiction Film, p. 58.
19. Ibid., p. 59.
20. See Kristin Thompson, "The Duplicitous Text: An Analysis of Stage
Fright," Film Reader 2 (1977), 52-64.
21. We must be precise in our use of the term "unreliable narration." It is a
meaningful concept only when it refers to the actual and overt misrepresentation or distortion of story "facts," by a narrator's guile, naivet, or whatever. It must refer, I believe, to a narrator's acts of commission, not of
omission. The omission of crucial data in the unraveling of a story is not a
matter of unreliability but of that special form of analepsis which, as Grard
Genette puts it, "sidesteps" an event and which he dubs "paralipsis." Paralipsis omits "one of the constituent elements of a situation in a period that the
narrative does generally cover." For example, in A la recherche du tempsperdu,
Proust recounts "his childhood while systematically concealing the existence of one of the members of his family" (Narrative Discourse, trans. Jane
Lewin [Ithaca: Cornell University Press, 1980], p. 52). This seems to be the
case of Bordwell's cxamplc (Narration in the Fiction Film, p. 84), the film
Secret beyond the Door (I say "seems" sincc I havc not sccn the film but go
only by his dcscription). Unlikc the unreliable narrator, the paraliptic nar
rator ultimately filis in the gap. But the unreliable narrator sticks to his
guns, and it is only the contcxt or the intcrvcntion of a latcr, rcliablc
narrator (as in Stage Fright) that scts things straight.
Another film theorist who uses "unreliable narration" in an odd way is
George Wilson, Narration in Light (Baltimore: Johns Hopkins University
Press, 1986), chaps. 2-3. Wilson equates unreliability with "openness" (lack
of narrative closure), but that too seems to confiase it with paralipsis.
22. Bordwell, Narration in the Fiction Film, p. 61, cites Branigan, Point
of View in the Cinema, pp. 40-49, as his authority for this view.
23. Sarah Kozloff, Invisible Storytellers: Voice-Over in American Fiction Film
225
Notes
(Berkeley: University of California Press, 1988), p. 115, has problems with
ascertaining the source of the lying flashback by speaking not of the implied
author but of Hitchcock, (not "Hitchcock"). In general, Kozloffs position,
though enlightened, fails to recognize the more general principle that controls all narrative transmission by whatever means. Thus, she distinguishes
between the "silent image-maker" and the "narrator" (voice-over or -on),
but she does not fully explain how the two are controlled by a more general
agency that instructs themthe implied author.
24. An interesting proposal to account for narrative agency in the cinema
has been made by Robert Burgoyne in an unpublished paper, "The Cinematic Narrator: The Logic and Pragmatics of Impersonal Narration." Fol lowing Marie-Laure Ryan, "The Pragmatics of Personal and Impersonal
Narration," Poetics 10(1981), 517-39, Burgoyne proposes two basic kinds of
narratorpersonal and impersonal. Unlike the personal narrator, who
"simply reports on the world," the impersonal narrator also creates that
world "while at the same time referring to it as if it had an au tonomous
existence, as if it preexisted the illocutionary act" (in what some film theorists call the "profilmic" state). "The impersonal narrator's lack of human
personality allows the viewer to imagine that [the viewer] is confronting the
fictional universe directly, putting aside any reflection on the form of the
narrative discourse." The personal narrator may lie or distort the story, yet
the true facts will remain; the impersonal narrator, who has created the story
world, cannot present anything but the facts.
In arguing that the impersonal narrator creates as well as reports the
diegetic world, Burgoyne and Ryan are attributing to it powers that I
reserve for the implied author. If a narrator can create the diegetic world,
there is nothing left for the implied author to do; indeed, Burgoyne con cludes that "impersonal narration thus eliminates the need for the category
of the implied author." There are various possible responses to the Burgoyne-Ryan theory: the one that immediately strikes me is that it does not
account for the so-called "authorial narrator": that is, a narrator who is
heterodiegetic and exists in the discourse only but, unlike the impersonal
narrator, may well have a "personal" identity. Further, as I tried to show in
Chapter 6, there are other reasons than reliability for recognizing the existence of the implied author.
25. Bordwell, Narration in the Fiction Film, p. 49.
26. Kozloff, Invisible Storytellers, p. 44.
27. Kozloff (ibid.) describes the problems posed by various alternative
names for the overall presenter of a film: "voice" is bad for reasons argued
above; "camera" is misleading because it neglects other presentational devices such as "lighting, graphics, processing, staging, sound track," and so
on; "implied author" confiases levels; "implied director" makes auteurist
assumptions; "implied narrator" risks confusion with voice-over narrators
226
Notes
(as well as with "implied author"); "master of ceremonies" is "sexist and
circusy"; and Metz's "grand image-maker" "slights the sound track." But
Kozloff herself goes along with Metz and calls this agency the "imagemaker."
28. "Voice-over is just one of many elements, including musical scoring,
sound effects, editing, lighting, and so on, through which the cinematic
text is narrated" (ibid., pp. 43-44). By one estimate (in an unpublished
manuscript by Avrom Fleishmann) about 15 percent of all films include
narrators who speak or write parts of the narrative.
29. It would, of course, be possible to create a text that shows only black
on the screen and communicates its story solely on the sound track. But it is
questionable whether such a text should be called a "film," since exactly the
same effect could be created by broadcasting an audio recording to an
audience assembled in the dark.
30. Kozloff, Invisible Storytellers, p. 45: "In many cases the voice-over
narrator is so inscribed in the film as to seem as if he or she has generated not
only what he is saying but also what we are seeing. In other words, films
often create the sense of character-narration so strongly that one accepts the
voice-over narrator as if he or she were the mouthpiece of the image-maker
either for the whole film or for the duration of his or her embedded story. We
put our faith in the voice not as created but as creator."
31. Ibid., p. 12.
32. Kozloff (ibid., pp. 112-15) cites a number of other Hollywood films
in which this happens: Taxi Driver, Gilda, Cat Ballou, Evil under the Sun,
Days ofHeaven.
33. Ibid., p. 115. I remember as a child in the 19305 seeing comic short
subjects, addenda to the newsreels, in which the camera could act zany and
be corrected by the voice-over narrator.
34. Kozloff (ibid., 117-26) discusses at length the question of reliability in
Barry Lyndon.
35. Ann Banfield, "Describing the Unobserved: Events Grouped around
an Empty Centre," in The Linguistics of Writing, ed. Nigel Fabb, Derek
Attridge, Alan Durant, and Colin MacCabe (Manchester: Manchester Uni
versity Press, 1987), p. 265.
36. Banfield, "Describing the Unobserved," pp. 266 -67.
37. Gilles Deleuze, Cinema I: L'image mouvement (Paris: Minuit, 1983), p.
117.
Notes to Chapter 9
i.
227
2. Ibid., pp. 2-3.
3. Grard Genette, Nanative Discourse, trans. Jane Lewin (Ithaca: Cornell University Press, 1980), p. 205.
4. I discuss additional problems with the term "focalization" in Seymour Chatman, "What Can We Learn from Contextualist Narratology?"
fbrthcoming in Poetics Today.
5. Virginia Woolf, Jacob's Room (1922; New York: Harvest Books,
1978), p. 7.
6. Charles Dickens, Oliver Twist (New York: Oxford University Press,
1949), p. i.
7. See the section "Ironic Reading as Knowledge," in Wayne Booth, A
Rhetoric oflrony (Chicago: University of Chicago Press, 1974), pp. 14-19.
8. The terms, of course, are not self-explanatory in the case of naive
narrators or characters who lie in dialogue, but this terminology seems
better than, say, "fallible narrator" and "unreliable filter."
9. Wayne Booth, The Rhetoric of Fiction, 2d ed. (Chicago: University of
Chicago Press, 1983), pp. 347~5310. A good selection of arricies on the subject appears in Movies and
Methods, 2 vols., ed. Bill Nichols (Berkeley: University of California Press,
1976, 1985).
u. See Jean-Pierre Oudart, "Cinema and Suture," Screen 18 (1977-78), 3547, and Daniel Dayan, "The Tutor-Code of Classical Cinema," in Nichols,
Movies and Methods, vol. i, pp. 438-50, for discussions of the ideologi-cal
implications of classical Hollywood "suturing." Brian Henderson con-siders
the ideological implications of another kind of film style in "Toward a NonBourgeois Camera Style (Part-Whole Relations in Godard's Late Films),"
in A Critique of Film Theory (New York: Dutton, 1980), pp. 62-81.
12. The Tintoretto painting is in the Samuel H. Kress Collection of the
National Gallery in Washington, D.C.
13. Nick Browne, The Rhetoric of Filtnic Narration (Ann Arbor, Mich.:
UMI Research Press, 1982), p. 3.
14. Ibid., p. 4.
15. Ibid., pp. 6, 8.
16. George Bluestone, Novels into Film (Berkeley: University of Califor
nia Press, 1957), pp. 47-48.
17. The phrase was used by Antonioni in many interviews. For a more
extended discussion of how this theme relates to existential anxiety, see
Seymour Chatman, Antonioni; or, The Surface ofthe World (Berkeley: Univer
sity of California Press, 1985), pp. 55-66. See also the article by psychoanalyst Simon Lesser, "L'Avventura: A Closer Look," Yale Review 54 (1964),
41-50.
228
Notes to Chapter 10
1. Wolfgang Iscr, "The Reading Proccss: A Phcnomcnological Ap proach," in The Implied Reader: Pattems of Communication in Prose Fiction from
Bunyan to Beckett (Baltimore: Johns Hopkins University Press, 1974), p. 283.
2. The Third Man is a classic case of a film that makes significant narrative improvements on the original text. See Seymour Chatman, "Who Is the
Best Narrator? The Case of The Third Man," Style 23 (1989), 183-96.
3. These difficulties ha ve been recognized for many years: see George
Bluestone, Novels into Film (Berkeley: University of California Press, 1957).
4. The subject has become a genuine industry. Harris Ross, Film as
Literature, Literature as Film: An Introduction to and Bibliography of Film's
Relationship to Literature (Westport, Conn.: Greenwood Press, 1987), lists
some 2,500 items; in English, there are 104 "general studies and anthologies"
about adaptation alone.
5. See Martin Battestin, "Osborne's Tom Jones: Adapting a Classic," in
Man and the Movies, ed. W. R. Robinson (Baltimore: Penguin Books, 1969),
pp. 31-45. Battestin demonstrates other analogies such as the "dated acting
styles" intended to suggest the "Hogarthian" characterization in the novel.
He writes: "Just as Fielding indulges in amplifications, ironies, s imiles,
mock-heroics, parodies, etc., so the film exploits for comic effect a circusful
of wipes, freezes, flips, speed-ups, narrowed focusesin short the entire
battery of camera tricks" (p. 40).
6. Sarah Kozloff, Invisible Storytellers: Voice-Over Narration in American
Fiction Film (Berkeley and Los Angeles: University of California Press, 1989).
7. See Bruce Kawin, Mindscreen: Bergman, Godard, and First-Person Film
(Princeton: Princeton University Press, 1978).
8. The French Lieutenant's Woman (1981) is available on videocassette,
Fox F4586.
9. Linda Hutcheon gives a good account ofthe special "self-consciousness" ofthe novel in a chapter titled "Freedom through Artifice: The French
Lieutenant's Woman," in Narcissistic Narrative: The Metafictional Paradox (Waterloo, Ont.: Wilfrcd Lauricr University Press, 1980), pp. 57-70.
10. Peter Conradi interestingly argues that the theme of the novel's assault on Victorian mores is carefully matched by its assault on conventional
narrative technique: "In The French Lieutenant's Woman, itself a species of
historical romance, albeit an ornately mannered and self-conscious one,
Fowles addresses the problem of repression and liberation as aspects both of
the evolution of modern ethics, so that its major characters defy social
convention, and also of the emancipation of the poetics of his chosen fictional form, so that the revelation and denunciation ofthe inauthenticities of
his hero are accompanied by the attempt to expose the conventions and
229
Notes
hypocrisies of the form" ("The French Lieutenant's Woman: Novel, Screenplay, Film," Critical Quarterly 24 [Spring 1982], 42).
11. Joy Gould Boyum, Double Exposure: Fiction into Film (New York:
Plume, 1985), p. 105. A more accurate characterization of the "transforma tion from narrator's comment to actor's voice" appears in Guido Almansi
and Simon Henderson, Harold Pinter (London: Methuen, 1983), p. 96.
12. The exceptions are such outr moments as in chap. 61, when the
narrator himself time-travels back to participate in the Victorian action.
13. JohnFowles, The French Lieutenant's Woman (New York: New Ameri
can Library, 1969), pp. 47, 324.
14. Peter Conradi, John Fowles (London: Methuen, 1982), p. 6o. Conradi's is the most sophisticated discussion of the novel.
230
Notes
and Individualization," Literature/Film Quarterly 12 (1984), 229. As Glenn
K. S. Man notes, "There is no particular reason" to make such an inference:
see "The Intertextual Discourses of The French Lieutenant's Woman," New
Orleans Review 12 (1988), 54-55.
29. Boyum, Double Exposure, pp. 1068. Man, too, oversimplifies by
arguing that "the film's happy resolution to the Victorian love story parallels
the book's first ending, while Anna's decision to separate from Mike in the
modern story parallels the second ending" ("Intertextual Discourses," p.
59).
231
Notes
38. In an interview (Harlan Kennedy, "The Czech Director's Woman,"
Film Comment 17 [September-October 1981], 25, 30), Karel Reisz acknowledged this intention: "In the Victorian scenes we very consciously went for
an academic kind of lighting, the sort of high definition that you see in
Victorian paintings. We used front light and side lighta pre-Impressionist
kind oflight to paint the object. We had our own shorthand motto for this:
'Constable, not Monet.'"
39. Curiously, Almansi and Henderson take this to be afault in the movie,
rather than one of its intentions: "Where the script does not seem to work is
in the modern love story . . . Here, inevitably, the melodramatic impetus of
the original story, the Victorian love-affair, so forcefully opposed and slow
to come to life, is much more fascinating than the modern one, which
appears facile and uncommitted. Every time the film moves to the modern
love-scenes, the audience hopes that they will soon be over so that they will
get back to Charles and Sarah's fate" (Harold Printer, p. 97). But that "un
committed facility" of modern lovewhat has happened to Eros is one
hundred yearsis finally what the whole film is about. The audience's
discomfort seems precisely what the implied author intends.
40. A good example: to explain Charles's disdain for commerce and his
desire to remain a gentleman, the novel's narrator asks the narratee to think
of possible modern parallels: "You havejust turned down a tempting offer in
commercial applied science in order to continu e your academic teaching?
Your last exhibition did not sell as well as the previous one, but you are
determined to keep to your new style? You have just made some decision in
which your personal benefit, your chance of possession, has not been al lowed to interfere? Then do not dismiss Charles's state of mind" (Fowles,
French Lieutenant's Woman, p. 234).
41. Again, analogies are drawn too easily by crides: "Like Charles . . .
Mike displays a passion that is obsessive; while like Sarah, Anna proves
ultimately elusive" (Boyum, Double Exposure, p. 106). But the film Sarah is
not elusive; her return to Charles for romantic reasons is all too predictable.
Interestingly, in the shot in which Charles says goodbye to her in Exeter
after they make love, the screenplay gives him the line "I shall come back for
you, my sweet. . . mystery." But the last phrase, "my sweet. . . mystery," is
not spoken in the film. Nor do I agree that "the film's happy resolution to the
Victorian love story parallels the first ending, while Anna's decisi on to
separate from Mike in the modern story parallels the second ending" (Man,
"Intertextual Discourses," p. 59). We don't know what Anna "decided";
perhaps she did not decide at all but simply left in ambivalent terror. In the
novel, Sarah offers Charles a kind of relationship; he elects to leave. In the
film, Anna just disappears.
42. We do sense, however, that she is a much weaker character than
232
Notes
Sarah. It strikes me as a flawed reading of the film to argue that "the contrast
between the wild and needy Sarah and the more controlled and independent
Anna carnes much of the book's commentary on the Victorian woman, on
her evolution into the modern" (Boyum, Double Exposure, p. 108). The
woman who expresses her envy of Mike's wife's "gardening ability" is
hardly "independent," and the woman who drives off without giving her
lover a word of explanation is hardly "controlled."
43. Fowles, French Lieutenant's Woman, p. 213. The whole of chap. 35 is an
expository-argumentative essay on the Victorian male's strangely conflicting altitudes toward women, the narrator discussing the repression of sex in
the context of Freud's theory of sublimation.
44- Charles Scruggs, "The Two Endings of The French Lieutenant's
Woman," Modem Fiction Studies 31 (Spring 1985), 95-114.
45. Of course, the film exists in consort with the novel, and many
viewers cannot help embroidering the former with their memories of the
latter. Boyum (Double Exposure, p. 64) and Man ("Intertextual Discourses,"
p. 55) are right in calling our attention to the "palimpsest" quality of films
adapted from novels. The film "is distinct," says Man, "yet it derives its
fullest possible meaning from its nature as a transformation of the literary
work." It is not only that a picture speaks a thousand words but that even the
briefest shot in a film based on a profusely discursive novel echoes some of
the novel's commentary and background.
46. Conradi, John Fowles, p. 68, quoting from Lorna Sage, "Profile 7:
John Fowles," New Review, 1974, p. 34.
Notes to Chapter 11
1. Terry Eagleton, Literary Theory: An Introduction (Minneapolis: University of Minnesota Press, 1983), pp. 205-7, argues that a revitalized rhetoric should be the alternative to literary theory, which in his view has
become a "non-subject."
2. Christine Brooke-Rose, A Rhetoric ofthe Unreal: Studies in Narratwe
and Structure, Especially ofthe Fantastic (New York: Cambridge University
Press, 1981), p. 12.
3. Paolo Valesio, Novantiqua: Rhetorics as a Contemporary Theory
(Bloomington: Indiana University Press, 1980).
233
Notes
7. Kenneth Burke, "Lexicon Rhetoricae," in Counter-statement (Berkeley: University of California Press, 1968), p. 135.
8. But didactic fictions are studied by Sheldon Sacks, Fiction and the
Shape ofBelief: A Study ofHenry Fielding, with dances at Swift, Johnson, and
Richardson (Berkeley: University of California Press, 1964); David Richter,
Fable's End: Completeness and Closure in Rhetorical Fiction (Chicago: Univer
sity of Chicago Press, 1974); and Susan Suleiman, Authoritarian Fictions
(New York: Columbia University Press, 1983).
9. See Ross Chambers, Story and Situation: Narrative Seduction and the
Power of Fiction (Minneapolis: University of Minnesota Press, 1984). For all
its own seductive appeal, however, it remains to be seen, as Porter Abbott
points out in his review of Chambers, Story and Situation (Poetics Today 6 no.
3 [1985], 544), whether "the terminology of love will [in fact] bond to the
critical language."
10. The statement, quoted from David Lodge's review of the book in the
(British) Modem Language Review, appears on the back cover of the paperback version of the first edition of Booth's Rhetoric of Fiction.
11. The same problem seems to occur in some of the early stories ofJoyce
Carol Oates: e.g., "By the North Gate," in the collection By the North Gate
(New York: Fawcett, 1971), pp. 195-208.
Index
235
Bront, Emily: Wuthering Heights, 182 Brooke-Rose, Christine, 184, 233 Brooks, Cleanth,
17, 35, 208, 211 Browne, Nick, 157-58, 224, 228 Burch, Noel, 53, 213 Burgess, Anthony:
A Clocktvork Orange,
164
Burgoyne, Robert, 226 Burke, Kenneth, 185-86, 188, 197, 234 Byron, George Gordon,
17-18
Cagney, James, 56
Callenbach, Ernest, 43, 212
Cal vino, Italo, 217
Campbell, George, 185
Camus, Albert: La Chute, 149
Capra, Frank: It's a Wonderful Life, 2, 57;
You Can't Take It with You, 3 Career author, 8789 Carlyle, Thomas, 77 Carroll, Lewis,
167 Carroll, Noel, 223 Celine, Lewis-Ferdinand: Journey to the
End ofNight, 190 Chambers, Ross, 206, 234 Character sketch, 17-19 Chatman,
Seymour, 2, 4, 116, 198, 205,
2ii, 212, 218, 222, 224, 228, 229, 234 Chaucer, Geoffrey: The Canterbury Tales,
182
Chekhov, Anton, 197; The Seagull, 96 Christ ai the Sea ofGalilee. See Tintoretto Chronologic, 10
Cinema, expression of thought in, 159 Cinematic narrator, 211 Clayton, Jack: The Great
Gatsby, 163 Clemens, Samuel: Huckleberry Finn. See
Twain, Mark. Clough, A. H., 167 Cohn, Dorrit, 94, 220 Conrad, Joseph, 16; Heart of
Darkness,
123, 144; The Nigger ofthe Narcissus,
86, 87; Outcast ofthe Istands, 112 Conradi, Peter, 229, 231, 233 Coppola, Francis Ford,
163; Apocalypse
Noiv, 164 Crane, Stephen: The Red Badge ofCourage, 9296, 220 Croce, Benedetto, 77 Cry in the Dark, A, 213 Culler, Jonathan, 216,
223 Curtiz, Michael: Mildred Pierce, 56
Daiches, David, 198, 234 Dallenbach, Lucien, 231
Danto, Arthur, 96, 220
Darwin, Charles, 167
Davis, Bette, 48
Dawson, Jan, 69, 214
Dayan, Daniel, 223, 228
Deleuze, Gilles, 138, 227
Dpardieu, Grard, 59
Description: assertive, 28; of characters, 25-26; "dangers" of, 23-24; definitions of, 9
10, 15-21; and "dynamism," 29; elliptical implication, 28-29; in film, 20, 3855;
"irreducible narrative fmal-ity," 21; nonassertive mention, 28; of places, 24-25;
"pulverization" or "distribution" of, 16
Dickens, Charles, 141, 227, 228; Dombey and Sons, 141-43; Great Expectations, 47, 130,
212; Hard Times, 75; Little Dorrit, 16, 208; Oliver Twist, 148
Didacticism and moralism in film, 58
Dietrich, Marlene, 131
Discourse, 9
Dostoevsky, Feodor: Crime and Punish-ment, 1617, 208; The Notebooksfor "The
Possessed," 220; The Possessed (The Devils), 25, 97-99, 209
Dowden, Edward, 8o Dundrearies, 39 Durning, Charles, 107
Eagleton, Terry, 184, 233
Eco, Umberto, 222; The Name ofthe
Rose, 199
Effet de rel, L', 25, 40, 70 Eisenstein, Sergei, 129, 172; project for
Das Kapital, 58 Elohist, Bible, 92
Encyclopdie mthodique, 23, 45, 209 Evans, Maurice, 112 Exposition as text-type, 6
Fabula, 9
236
237
238
Richards, I. A., 185 Richardson, Ralph, 112 Richardson, Samuel: Clarissa, 193 Richardson,
Tony: TomJones, 57, 164 Richter, David, 234 Ricoeur, Paul, 216 Riefenstahl, Leni, 154;
Olympiad, 20 Rimmon-Kenan, Shlomith, 218 Robbe-Grillet, Alain, 208, 211; Lajalousie, 20, 2089; "The Secret Room,
34-37
Roosevelt, Franklin, 141 Ropars-Wuilleaumier, Marie, 224 Ross, Harns, 229 Ross, Lillian,
93, 95, 219, 220 Rossetti, Dante Gabriel, 167 Rosson, Harold, 95 Rothman, William, 223
Rouen, description of, 19 Rozier, Jacques: Adieu Philippine, 46 Russell, Bertrand, 138
Russian Formalists, 125 Ryan, Marie-Laure, 226
Sacks, Sheldon. 234
Sage, Lorna, 233
Sanders, George, 47
Sarris, Andrew, 219
Schary, Dore, 95, 95
Scruggs, Charles, 182, 233
Service, textual, 8, 10-11
Shakespeare, William, 163; Hamlet, no,
ni, 1 1 4
Shelley, P. B.: "Ozymandias," 10 Sinn, 77
Small, Vernon, 215 Speech-act theory, 76 Speiser, E. A., 92, 219 Spingarn.J. E., 77 Staley,
Jeffrey, 219 Stallman, R. W., 82, 218 Stanislavsky, Konstantin, 96 Stanzel, Franz, 118 Star
effect, 159 Steele, Charles, 198 Stendhal (Henri Beyle): The Red and the
Black, 14, 208 Stcphen, Lcslie, 168 Sternbcrg, Mcir, 24, 127, 206, 209 Sterne, Lawrence,
145; Tristram Shandy,
123
Theophrastian character, 18-19 Thompson, Kristin, 20, 57, 225 Thurber, James, "Unicorn
in the Garden," 56
Tillotson, Kathleen, 8o Tintoretto: Christ at the Sea ofGalilee,
155, 228
Todd, Richard, 131 Tolhurst, William, 79 -80, 217 Tolstoy, Leo: Anna Karenina, 14, 26,
208;
War and Peace, 29, 31 Torgovnick, Marianna, 7, 97, 206, 220 Transmission, narrative, 3
Truffaut, Francpis: DayforNight, 180 Tunimanov, Vladimir A., 99, 220 Twain, Mark:
Huckleherry Finn, 29, 31,
82, 149, 19192
Utterance meaning, 79 Utterer's meaning, 79
Valry, Paul, 23, 209
Valesio, Paolo, 185
Valu, as opposed to valeur, 83
Vertov, Dziga, 138
Voice of narrator, 76
Voice-over narrations, film names, 164
Warren, Robert Penn, 17, 35, 208, 211
Wasiolek, Edward, 97-99
Watts, Alan, 202, 234
Welles, Orson, 219; Citizen Kane, 50, 5152; Touch ofEvil, 43, 212 Wellman, William: Public Enemy, 56, 57 Wells, H. G., 223 West,
Jessamyn, 8o
239
Whitmore, James, 94 Wilson, George, 225 Wimsatt, W. K., Jr. 74, 82, 91, 96, 215,
218, 219, 220
Woolf, Virginia: Jacob's Room, 146, 198 200, 228, 234; Mrs. Dattoway, 198-99; A Room
ofOne's Own, 200
Word-sequence meaning, 79
Wyman, Jane, 131
Yahwist, Bible, 92 Young, G. M., 167
Zola, Etnile: Nana, 26-27 Zweig, Stefan: Letterjrom an Unknown Womtm, 165
240
241