On the Twofoldness of Character

Murray Smith

O
n 2 January 2011, Nigel Pargetter slipped while working on
the roof of his home, Lower Loxley Hall in Ambridge, falling to
his death. That evening, Nigel had been hosting a New Year’s
Eve party for the local community. The party was disrupted when a close
friend of Nigel and his wife Elizabeth—Helen Archer—was taken to
hospital with suspected pre-eclampsia. The party having in effect come
to end, Nigel was persuaded by his friend (and Helen’s cousin) David
Archer to help him remove the celebratory banner on top of the house,
placed there to announce the New Year’s party. Underestimating the
strength of the wind, Pargetter lost his footing while loosening a knot
and plunged to the ground. A well-known and well-liked national figure,
many expressed their shock and sorrow at his untimely death. Pargetter
was a “kind, positive, adoring husband, [a] family man, [a] funny, mildly
eccentric minor aristocrat,” read one typical tribute.1
Readers resident in Britain may well recognize these events as those
constituting the main storyline of the sixtieth-anniversary episode of
the BBC radio drama The Archers (now the longest running serial drama
in any medium).2 Nothing in the description I offer above, however,
distinguishes it from a description of actual events and people; these
fictional characters are, it seems—beyond the fact that they don’t exist
as actual persons—just like actual people. We speak in terms of their
desires, values, and temperaments; of their lives as ones lived through
families, the local community, and various larger social structures; as
subject to the same medical conditions and physical laws that hold in
our world. Still more strikingly, we can and do respond to such char-
acters in ways that parallel our responses to real individuals, liking and
disliking them, evaluating them, expressing pleasure and dismay at
what befalls them.3 Many fans of The Archers were said to be “shaken”
by the events of the anniversary episode, some speaking of their “grief”
for Nigel Pargetter; at least one eulogy—from which the quotation at
the end of my first paragraph is taken—was written in his memory.4
In all of these ways, The Archers represents its events and characters
in a manner prototypical for a certain type of realist fiction, one that

New Literary History, 2011, 42: 277–294

278 new literary history

enables us to respond to and talk in sustained fashion about characters

as if they were real. And according to one theory of character, this rep-
resentational transparency—the ease and directness with which we can
describe characters in the same language we apply to ourselves—is also
a normative ideal. A well-constructed character is one that enables this
sort of stance. Introducing his television series Faulks on Fiction, novelist
Sebastian Faulks, for example, avers that “in recent years people talking
about novels have focused on their authors. I’d like to rectify this. To
me, the only people who matter are the characters—the heroes, lovers,
snobs and villains—people whose inner lives we get to know so well,
that they’re more familiar to us than our own families and friends. So
much so, that it’s in the power of their experiences that we see our own
lives in a new light.”5
There is no doubting the importance of this type of attitude towards
fictional characters, at least for realist fiction, the most pervasive form
of fiction across all media. By realism, I mean any form of fiction that
grounds its fictional narrative in a fictional world based primarily on an
existing or historical social reality, and relies upon our recognition of this
for proper appreciation of the fiction. While the particular characters
and events constituting such fictional narratives are generally invented,
we assume that the physical laws obtaining in such fictions are shared
with the laws in our universe, and that the social contexts and histori-
cal realities to which they allude will also be accurately portrayed (at
least within a certain degree of tolerance). The further a fiction moves
away from these conditions, the less we will be inclined to think of it as
“realistic” in this sense. On such a definition, then, fictions as various
as The Archers, The Hurt Locker (Kathryn Bigelow, 2009), Mad Men (Mat-
thew Weiner, 2007–), The Good Wife (Robert and Michelle King, 2009–),
Lionel Shriver’s We Need to Talk About About Kevin (2003), Jonathan Fran-
zen’s The Corrections (2001), Will Eisner’s A Contract with God, and Other
Tenement Stories (1978), and Daniel Clowes’s Ghost World (1993–7) are
all realist fictions. Thus in Mad Men, to take one example, Don Draper
(Jon Hamm) and company are fictional inventions, but the Great De-
pression, World War II, the Korean War, the beat movement, Kennedy’s
victory over Nixon, the gradual tightening of regulation over cigarette
advertising, and the invention of the Kodak slide carousel are all real
events and trends in the history of the society in which the story takes
place. The fictional universe of Mad Men also carries over the physical
laws of our universe—people get drunk when they consume alcohol;
they get injured and are likely to die if they get caught in a bomb blast.
It is in relation to such fictions that our readiness to speak of fictional
characters just as if they were real is most evident.
on the twofoldness of character 279

And yet, even for realist fiction of this type, fiction that seems to
invite a mimetic attitude towards its characters, that attitude is—so to
speak—only half of the story. Or, more precisely, the attitude itself is only
half-described if we accept the sketch I have offered. In the extensive
fan discussions of The Archers following the death of Nigel Pargetter, at
least as much emotion—usually in the form of upset, annoyance, and
anger—was expressed towards the show and its creators as towards
the characters. Throughout these debates, listeners of the show reveal
themselves to be sophisticated dramaturges, highly aware of the show’s
artifice. Their willingness and ability to respond to The Archers’ charac-
ters as if they were real is matched by a thorough understanding of the
characters’ irreality. Thus we need to probe our stance towards fictional
characters more deeply; to aid that task, I draw upon Richard Wollheim’s
notion of twofoldness.6

II

Wollheim coined the term “twofoldness” in his account of depiction,

which turns principally on the idea of “seeing-in.” According to Woll-
heim, when we look at a picture, we see what the picture represents in
the surface marked by the artist, and such seeing-in contrasts strongly
with our ordinary experience of “seeing face-to-face.” Painting, and
depiction in general, builds upon our natural capacity to see objects
in uneven, differentiated surfaces, as when we fancy that we see a face
in the knot of a tree, a mushroom in a cloud, or a head in the frost
markings on a window.7 Twofoldness describes a crucial quality of our
experience of such seeing-in—our apprehension, at once, of both the
depicted object and the marked surface, or what Wollheim refers to as
the recognitional and configurational aspects of seeing-in (73). Wollheim
stresses that these are “two aspects of a single experience that I have . . .
two aspects [that are] distinguishable but also inseparable. They are two
aspects of a single experience, they are not two experiences” (46). As
an example of the “twofold” stance in action—somewhat self-conscious
action, given its status as a piece of art criticism—consider the follow-
ing analysis of Francisco Goya’s The Third of May 1808, Execution of the
Revolutionaries (1814):

Goya’s concern is to confront us with and condemn the unadulterated harshness

of war. The structural composition guides our visual attention from the faceless
row of soldiers on the right, along their converging muskets, towards a brightly
lit central figure. His posture is reminiscent of the crucifixion, and the lighting
emphasizes his vitality and individuality against the mass of ranked soldiers.
280 new literary history

Even Goya’s use of paint shapes our visual and cognitive responses. If we look
closely we see the roughly treated surface, the lightning stripe of the soldier’s
sword sheath in the foreground, the scraped, coagulated, dark splotches of
crimson blood and the smeared facial features of the foremost prostrate figure,
his individuality blotted out.8

In this passage, Matthew Kieran focuses precisely on the channeling

of our attention to the “what” of the painting by its “how”—on the sweet
spot of twofoldness, the point of perfect integration of attention to the
recognitional and configurational aspects of the painting. Although
Wollheim’s theory is a theory of visual depiction, and—as will be obvious
from my opening reference to radio drama—my interest here ranges
beyond visually rendered characters, I believe his theory finds a strong
echo in our experience of characters in general. Moreover, beyond the
central parallel that I will point to, various details and ramifications of
Wollheim’s theory of seeing-in also have relevance for a properly nuanced
understanding of character, so I will spend a little more time with it here.
To begin with the core parallel: Wollheim’s “recognitional aspect” of
seeing-in equates with the aspect of our response to character captured
by my opening description of The Archers. So readily do we recognize
in fictions those “virtual persons” we call characters that we can speak
of them and respond to them in many ways just as if they were actual
persons—as if we were able to interact with them “face-to-face.” That is
one side of the twofoldness of characters; Wollheim’s account of seeing-in
points us towards the other side, towards the “configurational aspect” of
our apprehension of characters. What does this consist in? We exhibit
awareness of the configurational aspect of character whenever we note
or notice something bearing upon the designed status of a character,
when we see a character as an element in a representation, just as this
aspect is manifest for Wollheim when see how the frost on the pane
forms the outline of a head, or the markings on a cave wall limn the
outline of a bison. One of the most routine and explicit ways in which
we exhibit awareness of the configurational aspect of film characters is
via their embodiment by performers, and stars in particular. We move
fluidly between reference to the character, and to the actor or actress
embodying the character; sometimes the difference matters, but it is the
ease of movement between the two that is important here. Watching
Alfred Hitchcock’s North by Northwest (1959), I come to recognize Roger
Thornhill as a middle-aged advertising executive in a fix, but I also see
how Thornhill is “configured” through Cary Grant, and I apprehend both
of these aspects of the film at once. This is the twofoldness of character.
Wollheim goes on to argue that “[t]he twofoldness of seeing-in does not
. . . preclude one aspect of the complex experience being emphasized at
on the twofoldness of character 281

the expense of the other.” Thus, on the one hand, I can focus my atten-
tion on the marked surface, scrutinizing the canvas, the brushstrokes, or
(in the case of film) the grain of the image, to the point where I “lose
all but a shadowy awareness” of whatever is depicted by the image.9 On
the other hand, I can focus on what I see in the image, marveling at the
pristine, clean-cut appearance of Roger Thornhill, retaining only the
“vaguest sense” of the surface of the image that allows me to see Roger
Thornhill in it. And at the extremes, according to Wollheim, twofold-
ness can “evaporate” altogether, so that we find ourselves either seeing
the marked surface “face-to-face,” that is, simply as a textured surface in
which we see nothing depicted, or (at the other extreme) drifting into
a mode of visualizing “in the mind’s eye” the contents of the depiction
(47). In this latter state, while the picture has clearly been vital in getting
this process of private visualizing off the ground, once in it we cease to
engage with the picture as a picture. Wollheim also notes that, as a mat-
ter of fact, neither of these extreme states are likely to be stable when
we are engaged with a depiction: the marked surface draws us back to
the twofold nature of seeing-in. So while Wollheim’s theory accommo-
dates the intuition that it is possible to lose awareness of the artifactual
status of a depiction, or to lose awareness of the depictive function of a
marked surface, he underlines both the natural and the normative force
of twofoldness. Looking at a marked surface that has been successfully
designed to trigger seeing-in, we find it difficult to resist that invitation;
and it is a norm for the proper appreciation of such depictions that our
perceptual engagement with them should involve an interplay between
their recognitional and configurational aspects—between attention to
the design and to the depicted content (see also Wollheim 166, and my
discussion in section V below). Thus, in the passage of critical analysis
on Goya’s The Third of May quoted above, Kieran praises the way in
which the configurational aspects of the painting expressively shape
our experience and understanding of its recognitional elements, that
is, what the painting depicts.
In speaking of the “interplay” of the two aspects of twofoldness, it is
important to contrast the position I argue for here with the rival theory
of depiction propounded by E. H. Gombrich. Gombrich—with whom
Wollheim sustained a dialogue over many years—also recognized that our
perception of both the marked surface and the depicted object figure
in our experience of engaging with depictions. But Gombrich’s theory
was based on a notion of illusion: when we perceive what a depiction
represents, we are subject to an illusion that we are seeing that object
face-to-face.10 Gombrich likens our experience of depictions to our expe-
rience of Robert Jastrow’s duck-rabbit illusion, made famous by Ludwig
282 new literary history

Wittgenstein: looking at this ambiguous figure, we can see the duck or

the rabbit, but we cannot see them simultaneously. So, for Gombrich,
we can see what a picture depicts, and we can see the marked surface,
but we cannot see both at once. We are able to flip-flop between these
states, so rapidly that our phenomenal experience may give us the mis-
leading impression that we can see both simultaneously. Wollheim notes,
however, that the duck-rabbit figure is a poor analogy for our experi-
ence of pictures, for the two aspects of the duck-rabbit figure are both
recognitional—they are both experiences of seeing something (a duck,
a rabbit) in the simple but cunning array of lines provided by Jastrow
(360, note 6). Wollheim’s argument that when see the duck we are also
aware of the configuration of lines allowing us to see the duck—and
similarly for the rabbit—is left standing. In any event, the important
point to bear in mind here is that in using such terms as “interplay,” I
do not mean to fragment or cleave recognition and configuration from
one another, dispersing them into temporally distinct (if ever so rapidly
alternating) experiences. Recall how emphatic Wollheim is on this point:
the two aspects of twofoldness “are two aspects of a single experience,
they are not two experiences” (46).
The model here is one based on recognition of the complex and mul-
tilayered nature of our conscious experience in general. The quality of
our attention to objects comes in different grades, ranging from low-level,
peripheral awareness to focal attentiveness. And at any given moment, I
may be attending—to different degrees—to a variety of distinct things: as
I dry and put away the dishes on “autopilot”—that is, with only minimal
conscious attention to this “overrehearsed” task—my ear is caught by a
compelling song that pops up on the radio. Alongside listening to this
music and drying the dishes, I am also peripherally monitoring sound
from an adjoining room, where the kids are watching a movie and a
squabble may erupt at any moment. A minute earlier, before the song
started, my ear was focused on the sound of the movie in the other
room, and the radio functioned as little more than “furniture music,”
but now the hierarchy of my attention has shifted to the put the song
coming from the radio in the foreground of my attention. The notion
of twofoldness—regarded as a property of both seeing-in, and of our
apprehension of characters across all media—sits naturally within such
a model of conscious awareness.11 The quality of my attention to the
marked surface and the depicted content may ebb and flow relative
to one another, but under normal conditions, I retain an ineradicable
awareness, if only peripheral, of the artifactual status of a depiction. And
so, ceteris paribus, it goes for our attention to the dual facets of fictional
characters.
on the twofoldness of character 283

It might be objected that Wollheim’s concept of twofoldness picks out,

quite specifically, a property of a type of perception—when we look at a
depiction, we have visual awareness of both the subject depicted and
the marked surface. Speaking of twofoldness in relation to characters
in general—characters represented verbally and performatively as well
as pictorially—certainly does increase the extension of the concept, but
justifiably so. When we engage with a literary character, our awareness
of that character brings together an understanding of their place in the
fictional world with an appreciation of their place in the design of the
work. Lisbeth Salander is an expert computer hacker, and one of the
protagonists of Stieg Larsson’s The Girl with the Dragon Tattoo (2005). The
title of the novel (a part of the novel’s design) makes reference to the
tattoo adorning Lisbeth’s back (a feature in the novel’s fictional world);
we attend to both The Girl with the Dragon Tattoo and the girl with the
dragon tattoo. We hold these two sorts of fact about Lisbeth in mind
as we read the novel, and we must do so in order to understand and
respond to it appropriately. Although Wollheim arrived at the concept
of twofoldness through an examination of visual perception, then, the
term can legitimately be extended to refer to a more pervasive feature
of cognition. The twofoldness of depiction should be thought of as a
species of the twofoldness characteristic of our engagement with repre-
sentation in general.

III

Talk of film characters makes it imperative to clarify exactly how Woll-

heim’s theory of depiction is relevant for a general theory of character.
It is relevant in two ways: in the last section, my goal was to draw out
the parallels between the recognitional and the configurational aspects
of seeing-in—the two faces of “twofoldness”—and the two equivalent
aspects of our apprehension of character, irrespective of the medium
through which a particular character is represented. But there is a fur-
ther, distinct way in which Wollheim’s account of seeing-in is relevant
for characters in film in particular—perhaps not surprisingly, given that
Wollheim is concerned above all with visual representation.
In my introductory remarks I referred to the “representational trans-
parency” of realist fictions with respect to characters, noting the fluency
with which we recognize and grasp them using the same folk psychology
and language that we deploy in understanding those we interact with
in the actual world. In the domain of film theory, however, the word
“transparency” carries a special additional charge, where it refers to the
284 new literary history

idea that photographic and filmic representation allows and encourages

the viewer to see through the film frame, precisely as if it were a trans-
parent window onto a fully specified and “ontologically self-subsistent”
world of action.12 In its strongest form, the claim is that photographic
and filmic depiction is metaphysically distinct from traditional forms
of depiction like painting, and that as a consequence, these modern
depictive technologies give rise to a phenomenology distinct from the
“seeing-in” Wollheim attributes to painting. George Wilson, who has
explored this idea in relation to film at greatest length and in the most
nuanced fashion, teases out a variety of factors that, he argues, contribute
to the impression of transparent access to the fictional world of a film.
According to Wilson, one of the factors that subtends transparency is the
relative absence of visual texture or “facture” in the moving photographic
image: “The projected screen image contains a vast wealth of diversified
visual information within a two-dimensional surface that is apparently
‘unworked’ and relatively untextured. Viewing a comparably realistic
painting of the same scene, a viewer’s attention tends to oscillate between
an inspection of the items pictured and a scrutiny of the handling of the
paint upon the canvas. The painting’s facture is the ever-present mark
of the agency that produced it. In the standard photographic image,
the visible facture is either eliminated or reduced to a minimum. The
items photographed appear transparently.”13 Note that Wilson’s under-
standing of our experience of a painting points towards Gombrich’s
rather than Wollheim’s account, insofar as he writes of an “oscillation”
between attention to the painted surface of the canvas and attention
to “the items pictured.” With that bifurcated model of our attention
to painted depictions in place, Wilson then argues that the relatively
“unworked” and “untextured” surface of a film allows and encourages
us to hold our attention on what is depicted to a greater extent, and
in a more sustained fashion, than is normally possible with painting.14
How much water does Wilson’s argument hold? I think it can be
challenged on two fronts. First, while there surely is a contrast between
the surface texture of a painting and a film, it would be easy to over-
generalize here, and make too much of whatever contrast remains once
various qualifications have been acknowledged. The degree of facture
in a painting varies with the school and individual style of the artist;
photorealist painters mimic the very untextured and transparent sur-
face that Wilson identifies as a property of photographic film. But the
same is true for film: different movements and individual filmmakers
are characterized by different degrees and types of overt visual texture.
Filmmakers obviously cannot achieve texture in the same way as painters,
and in particular they cannot (using standard filmmaking equipment)
on the twofoldness of character 285

achieve literal three-dimensional effects, obtained by painters who daub

paint onto the canvas in thick layers or globs. But they can and do work
with visual texture using the means available to them, including varying
the grain or the contrast of the image.
Wilson is right to underscore the importance of the “wealth of diversi-
fied visual information” in a film image. Such information arises from
the “automatic” dimension of film and photography. That is, although
filmmakers and photographers must, minimally, set up a camera before
a space, and to that extent everything in the frame is there by dint of
this action, in another sense much of what we see in the framed image
is there just by virtue of the fact that the camera was running and point-
ing in the right direction to capture it. Say that I take a snapshot of a
friend, prompted by an extraordinary new haircut. I want to capture that
haircut just as it is, right now. But in taking this shot and inspecting it
later, I see that the photograph also renders the crow’s feet around the
subject’s eyes, his five o’clock shadow, and the distinctive texture of the
mohair cardigan he was wearing that day. (Incidentally, he doesn’t ex-
ist—but see how readily I write, and you think, of this fictional “friend”
in terms apt for a real person.) I didn’t intend to capture any of these
features—it was the haircut that mattered to me. But, this being a pho-
tograph rather than a painting, I got these additional features for free.
Given the boundaries of the framing, the quality of the light, the aper-
ture setting and other such technical matters, whatever was within the
camera’s field of view was going to be photographed. It is in this sense
that much of the information in a photograph bypasses the intentional
agency of the photographer in a way that no information in a painting
can bypass the agency of the painter. This is the feature of photography
that led André Bazin to argue that photographic family albums embody
“the disturbing presence of lives halted at a set moment in their dura-
tion, freed from their destiny; not, however, by the prestige of art but
by the power of an impassive mechanical process.”15 Granted that most
filmmakers and art photographers—as distinct from casual video diarists
and snapshooters—will be more attentive to the composition as a whole,
and to the visual content of its constituent areas, still, the filmmaker
does not paint each grain or pixel as the painter paints each part of the
canvas.16 And yet—to reiterate the crucial point here—it does not follow
that this information will be “untextured.” An entire film may be shot
using a high-grain film stock, and to that extent such imagery may resist
transparency. The “automaticity” of photography and film—its purely
causal underlay—provides no guarantee of transparency.
Even on the terms of Wilson’s argument, then, we might dispute at
least the degree of contrast between painted and filmed depictions, and
286 new literary history

thus the impact this has on the presumed transparency of the film image.
To this we may add the second point: even if we grant that there is some
general and important contrast between painted and filmed depictions
in respect of their degree of transparency, we should not conceive of
the viewer’s attention or experience in terms of an “oscillation” between
the fictional world depicted, and the “configurational” features of the
image through which the depiction is achieved. The model of multiple
and variable levels of attention developed above, in which our “twofold”
awareness of the image as both designed surface and depiction is the
norm, furnishes us with a more plausible and supple account of both
painted and filmed depiction.

IV

Earlier I stated that Wollheim’s theory of depiction has important

implications not only for the representation of characters in visual media
like film, but for our understanding of character across all media. One
way to see the relevance of Wollheim’s theory to the general theory of
character is to relate the concept of twofoldness to a distinction made
by a number of philosophers between two dimensions of fictional char-
acters. Francis Dauer, for example, argues for a distinction between
the “formal” and “referential” aspects of character: thus, with Charlie
Chaplin’s Modern Times (1936) in mind as an example, “Charlie wants
the flower girl to see again” speaks to the referential dimension of a
character; “Charlie is the protagonist of Modern Times ” speaks to the
formal dimension of that character.17 The formal dimension of character
thus gives rise to the plethora of metafictional statements uttered in rela-
tion to fictional characters, statements like “North by Northwest features
a character called Roger Thornhill,” or “Roger Thornhill is played by
Cary Grant,” or indeed the second of Dauer’s remarks about Modern
Times. Part of the point of the distinction between the referential and
the formal dimensions is to insist upon the limitations of a purely or
naïvely mimetic account of character: no matter how significant the
carryover of real-world assumptions to our engagement with fictional
characters, there is a crucial aspect of character beyond such mimesis.
I hope it is fairly obvious how the referential-formal distinction maps
onto Wollheim’s recognitional-configurational distinction.18 If not identi-
cal, the distinctions mirror one another precisely, each pointing to the
contrast between the content and the design of a representation. What
Wollheim’s theory brings to the table is the insistence upon twofoldness,
that is, the insistence that, at the level of psychological response, we
on the twofoldness of character 287

hold these two dimensions of character in mind simultaneously as we

engage with a character in a fiction. The model of attention I elaborated
above—a model implicit in or at least compatible with Wollheim’s theory
of depiction—allows for the fact that the relative intensity of our focus
on one or the other dimension may vary, and to that extent the two
aspects may “hang together” in a variety of ways (determined in part by
the form of the particular fiction itself, and in part by the interests and
stance of the person engaging with that fiction). But, as with twofoldness
in the context of depiction, under normal psychological conditions and
with respect to our standard appreciative norms, we apprehend both
aspects of character simultaneously; neither aspect is eliminable from
our experience.
So far, my examples of the ways in which our awareness of the con-
figurational or formal aspect of a character manifests itself have mostly
pointed to elements of visual design in a film. Such awareness is very
pronounced in cases such as Jean-Luc Godard’s Pierrot le fou (1965) or
Wong Kar-Wai’s In the Mood for Love (2000)—films which, in their dif-
ferent ways and to different ends, foreground their composition, color,
set design, casting, performance, and costuming. But our awareness of
the configurational aspects of character is not absent from plainer fare.
Even Wilson, in making the case for the transparency of film—a feature
of filmic representation that, on his argument, contrasts strongly with
the “opaque” nature of painting—acknowledges the fragility of that
transparency. He notes that

this attempted portrayal of an autonomous objectivity [of the fictional world]

can be damaged or eclipsed by an audience’s conscious or subliminal aware-
ness that the shots that bring the fictional world into existence are consistently
being addressed to them as “ideal” or, at least, especially perspicuous visual com-
munication. Characters X and Y are situated in such-and-such a relation to each
other and to the camera in shot Z precisely because the resulting configuration
is meant to make clear to the viewer the nature of their psychological interac-
tion. Their being positioned as they are is dimly felt by the audience to be an
incident that has been designed for their inspection. (59)

Put positively, that “dim” feeling is nothing other than evidence of

twofold awareness even in the classical Hollywood style of filmmaking,
the tradition that arguably works hardest to exploit and maintain the
transparency built into the medium (on Wilson’s view). Wilson also
notes the constant adjustments and acknowledgements of technical fac-
tors that we must make in order to properly understand the action of a
film. Generally speaking, an out-of-focus object is not a blurred object
in the fictional world of a film, but a product of the limited depth of
288 new literary history

focus of the camera lens. Objects in a black and white film are not to
be understood as lacking colour in the fictional world; rather, they ap-
pear to us in that form as a consequence of the film stock used by the
filmmakers (58). In all these ways, our appreciation of films that strive
for transparency show that awareness of configuration does and must
accompany proper attention to and understanding of the elements
comprising the fictional world.
In order to show how the theory here extends to include characters
in literary and other media, however, we need to detail other ways in
which our awareness of the formal aspect of character may be manifest.
Narrative structure and generic form are perhaps the most significant
of these. We have already encountered some basic examples of formal
claims about characters bearing upon narrative structure: the statements
that North by Northwest features a character called Roger Thornhill, and
that Charlie is the protagonist of Modern Times. In a similar spirit, we
might say that the subplot concerning Don Draper’s half-brother Adam
(Jay Paulson), in Mad Men, reaches its climax when Don discovers that
Adam has committed suicide. We might claim that the narrative arc of
the first season of the show as a whole completes itself with Draper’s
presentation of his proposed ad campaign for the Kodak slide carousel
(based on its role as an engine of nostalgia). Or we might speak of
the final scene in The Silence of the Lambs (Jonathan Demme, 1991), in
which Hannibal Lecter (Anthony Hopkins) phones Clarice Starling
(Jodie Foster) from a Caribbean island, as a blackly humorous coda to
the main action of the film. Whenever we think in such terms about
characters—pondering their degree of significance within the narrative,
or the way in which they figure in the “parts” constituting the narrative
whole—we are thinking formally about the fiction. And if we go on to
think about the types of character and narrative that the fiction embod-
ies—recognizing the figure of the little tramp from other Chaplin movies,
and wondering if this fiction will unfold in the same direction as those
other works by Chaplin; recognizing a character as a hero, an antihero,
an antagonist, a bounder, or a femme fatale—we are also responding to
the formal dimensions of character, now in a way that brings in generic
considerations alongside those specific to the particular fiction in hand.

Greg Currie marks out a similar contrast to that between the recogni-
tional/referential and the configurational/formal aspects of character,
speaking of the “external” and “internal” facets of narrative and character.
In a related discussion, he notes that
on the twofoldness of character 289

when we learn something about a fictional character because the work itself tells
or suggests it, this is not at all like learning something about someone in the
real world by casual observation. For what the fiction tells us comes labeled as
having a special evidential relevance; we are told this for a reason . . . Thus the
world of a fiction is not one that we engage with by applying ordinary standards
of evidence and probability; what is evidence is inseperable from what is a sign
of authorial intention. And so we may rightly draw conclusions about motive
or behavior that would be unwarranted on the same purely evidential basis in
real world situations.19

When we engage with a narrative representation—especially a fictional

narrative—we are disposed to finding meaningful connections to a degree
that we are not in ordinary life, just because we know that a narrative
work is an artifact, a product of purposeful design. We seek to maximize
the meaningful inferences and connections that we can make, in a way
that would be pathological if applied routinely to our interpretation
of the actions of actual persons. In ordinary, real-world interactions,
we know that what we perceive in the behavior of others will always be
partial and often accidental, and thus may not be reliably representative
of their beliefs, emotions, or personality traits. We do, of course, make
inferences based on the limited information available to us; and our
understanding of narratives is parasitic on that habit. But knowing that
there is no grand author or designer responsible for shaping our access
to real individuals, at some level we understand that the inferences we
make in real life are provisional, and subject to error, in a qualitatively
different way to the provisionality of our inferences in the context of
understanding works of narrative. Thus Currie’s remarks bring out the
way in which our apprehension of character is always conditioned by an
understanding that characters are artifacts, even when they strike us as
realistic. Recognition of the purposeful nature of a narrative work is built
too deeply into the transaction between the author and the perceiver
of such a work to be erased by even the most potent realism (whether
conceived as a matter of style or content or both).
Were this recognition not built in so deeply—were our apprehension
of characters not marked by the twofold recognition of both their for-
mal and referential dimensions—we would be apt to fall into the trap
of routinely posing what Kendall Walton terms “silly questions.” As one
example, Walton notes that “it is fictional in William Luce’s play The Belle
of Amherst [1976] that Emily Dickinson is an extraordinarily shy person
who keeps to herself. Yet she is onstage throughout the play, speaking
constantly. . . . How can it be fictional that Dickinson says all that she
does, all of what Julie Harris actually says while impersonating her, yet
fictional that she is not gregarious?”20 These questions arise because of
290 new literary history

a failure to recognize—or actively to erase—the distinction between the

formal and referential dimensions of character. Referentially speaking,
the Emily Dickinson of the play is shy; but we come to understand that
through the formal design of the play (essentially an extended mono-
logue). In seeking a referential answer to a formal question, we turn
a potentially sensible query—why did Luce choose this form?—into a
silly question, a question which is “not only irrelevant to appreciation
and criticism but also distracting and destructive” (176). Happily, how-
ever, the crucial point here is that this is typically a vice not of ordinary
readers and viewers, but of a certain wrongheaded, pseudosophisticated
professional criticism and theory. In our ordinary traffic with fictional
and nonfictional narratives, confusing the referential and the formal is
the exception rather than the rule. The twofoldness of our perception
of characters and their narratives ensures this.
Walton also draws on the visual arts to conjure up other examples of
misguided, silly questions. “Why do all thirteen of the diners in Leon-
ardo’s Last Supper [1498] line up in a row on the same side of the table?
. . . Must we suspect that they are fearful of facing one another—of kicks
under the table or bad breath?” (175) The initial question requires a
formal answer; but once channeled into the second question (“Must
we suspect . . .”), it has become (in Walton’s sense) absurd—“pointless,
inappropriate, out of order” (176). Buried (Rodrigo Cortés, 2010) af-
fords us a more contemporary example. This film tells the story of an
American truck driver in Iraq who is abducted, buried alive in a cof-
fin, and held there for ransom. The visual action of the film is entirely
restricted to the space of the coffin, and for the most part, the camera
occupies a physically plausible position within the confines of the cof-
fin. At one moment, however, the composition of the action gives us
the impression that the camera is suspended several feet above the
entombed driver. What’s happened to the lid of the coffin and all the
earth that, just a few seconds earlier, we understood was pressing down
upon it? Bad question. The composition works expressively, rather than
denotatively, picturing the driver as isolated at the bottom of a deep well
of darkness. The shot is expressive of his state of mind; it appears at a
moment of particularly intense despair. There is no suggestion that we
are to conclude that the driver’s physical situation has changed. And,
unless we are an exceptionally naïve viewer or perhaps an exceptionally
clever theorist, we know this perfectly well.
These visual examples allow us to return to Wollheim and pick up a
further significant development of his account of the twofoldness of our
experience of depictions—and specifically naturalistic or realistic depic-
tions. Wollheim argues that the error of most theories of naturalistic
on the twofoldness of character 291

depiction is that they assume that the naturalistic effect arises from some
form of suppression of the configurational aspect of a depiction, so as to
allow for an especially fluid or fast uptake of the recognitional content
of the depiction. But, Wollheim insists, it is the reciprocal interplay be-
tween awareness of the marked surface, and of what we see in a picture,
that creates the naturalistic effect. And naturalism is something that is
continually reinvented in the context of new materials and new types of
subject matter; that is why there is no single style of depiction which can
lay claim to being the naturalistic style (73–74). Further on in Wollheim’s
discussion, the normative aspect of twofoldness—that is, the sense in
which it is not only something that usually characterizes our experience of
pictures, but something that should be part of that experience—surfaces
again. Any picture that seeks to maximize our absorption in the action
depicted must, reciprocally, find a way of returning the spectator “from
imagination to perception: twofoldness must be reactivated” (166). So
again we see that twofoldness is a quality of our experience that typically
is and should be prompted not only by the varieties of nonnaturalistic
depiction—caricatural, expressionistic, reflexive, and so forth—but by
prototypically realist and naturalist pictures as well. Judging a depiction
to be realistic is not at all like mistaking it for a real scene, because our
awareness of the differentiated surface runs alongside—is a part of the
same experience as—our awareness of what is depicted.
The same holds, once more, for nondepictive forms of representa-
tion. An integral part of appreciating the realism of a novel or a stage
performance is an awareness of the configurational features of the
object of our attention, of its status as an artifact, a product of design.
So we return finally to the central claim of this essay: that the property
of twofoldness characterizes not only our experience of depictions, but
our experience of representations—and the characters that loom large
within them—in general. Understanding and appreciating fictional
characters in any medium involves “a complicated blend of imagina-
tive engagement and metafictional reflection.”21 We may talk freely of
fictional characters as if they literally inhabited our own world, but in so
doing we do not lose sight of their invented status. Far from suggesting
naïveté, such talk betrays the ease, sophistication, and naturalness with
which we handle fictional characters.

University of Kent

Notes

1 The comment is contained within the eulogy to Pargetter posted at: http://onelifecer-
emonies.co.uk/funeral-blog/index.php/2011/01/nigel-pargetter-the-archers-a-eulogy/
292 new literary history

2 The pilot series for the show began in May 1950; the confirmed series in January
1951, with five fifteen minute episodes broadcast each week since then.
3 Whether such affective responses are genuine emotions has been actively debated
over the past thirty years or more. In a well-known paper, Kendall Walton argued that such
responses are “quasi-emotions”; others have defended the idea that our affective responses
to fictional characters are authentic emotions. See Kendall L. Walton, “Fearing Fictions,”
Journal of Philosophy 75, no. 1 (1978): 5–27, and Mimesis as Make-Believe: On the Foundations
of the Representational Arts (Cambridge, MA: Harvard Univ. Press, 1990), 241–55.
4 See note 1 for the link to the eulogy for Pargetter. The actor who had played Pargetter
for 30 years, Graham Seed, also spoke of his grieving over the death of the character. Given
the extended nature of his imaginative inhabitation of the character, one might think of
Seed’s situation and response as qualitatively different from that of committed listeners
of The Archers. I would be more inclined to think of his response as on a continuum with
similar responses from such listeners, though I will not pursue that claim here.
5 A four-part series broadcast on BBC Two, February–March 2011; Faulks makes this
claim in the introduction to the first episode, The Hero. Recently a debate has emerged on
the implications of research in social psychology, most famously associated with Stanley
Milgram, concerning the power of situations—rather than character traits—in the deter-
mination of action of actual persons. On one interpretation, experimental work in this
tradition shows that human individuals do not possess the kind of stable clusters of traits
embodied in the idea of “character.” If real individuals lack character in this sense—a
substantial degree of trait consistency over time—then it follows, on this view, that fictional
characters possessing such consistency cannot give us insight into actual human psychol-
ogy in the way that Faulks, for example, assumes. I think this worry is overstated, but in
any event, note that Faulks’s claim bears on the movement from fictional characters to
actual persons—on the role that fictional characters play in our own real lives, on what
we may learn from such characters. By contrast, the focus of my argument here is on the
movement from actual persons to fictional characters—on, that is, our readiness to think
and speak about fictional characters using the same cognitive and linguistic tools that we
deploy in relation to real persons.
6 Richard Wollheim, Painting as an Art (London: Thames and Hudson, 1987) (hereaf-
ter cited in text); the notion of “twofoldness” is introduced on page 21. Although I draw
upon Wollheim in this essay, the seeds of the idea that our apprehension of characters
is marked by “twofoldness” are in my Engaging Characters: Fiction, Emotion, and the Cinema
(Oxford: Clarendon, 1995), 42–44, 138, and “Engaging Characters : Further Reflections,”
in Characters in Fictional Worlds: Understanding Imaginary Beings in Literature, Film and Other
Media, ed. Jens Eder, Fotis Jannidis, and Ralf Schneider (Berlin: DeGruyter, 2010), 237.
7 The first example is my own; the second two examples are inspired by two photographs
used by Wollheim to illustrate seeing-in: Cumulus Cloud, Mushrooming Head (1930s), by E.
E. Barnard, and Empty Head, Frost on Window, Rochester, New York (1962) by Minor White.
Painting as an Art, 46–47.
8 Matthew Kieran, Revealing Art (London: Routledge, 2005), 59.
9 Wollheim, Painting, 47. Dominic Lopes makes a further distinction, related to this
point in Wollheim’s discussion, between “design-seeing” (seeing the marks on the surface
as parts of a representational or abstract design) and “surface-seeing” (seeing the marks
on the surface simply as marks). See his Sight and Sensibility: Evaluating Pictures (Oxford:
Oxford Univ. Press, 2005), chap. 1.
10 Gombrich’s arguments were originally laid out in his classic Art and Illusion: A Study
in the Psychology of Pictorial Representation, rev. ed. (Princeton, NJ: Princeton Univ. Press,
1969), first published in 1960. Wollheim’s review of that work initiated a dialogue be-
tween them, the comments in Painting as an Art representing a much later moment in
on the twofoldness of character 293

that conversation. It is also worth mentioning at this juncture that another major player
in these debates over the nature of depiction, the perceptual psychologist J. J. Gibson,
argued that our apprehension of still and moving pictures involves a dual, simultaneous
perception of the two-dimensional surfaces of images, and the three-dimensional spaces
represented by them. Gibson’s position is thus (in this respect, at least) remarkably similar
to Wollheim’s. See Gibson, The Ecological Approach to Visual Perception (Boston: Houghton
Mifflin, 1979), 276.
11 For more on consciousness and our experience of film, see my “Film and Conscious-
ness,” in The Routledge Companion to Philosophy and Film, ed. Paisley Livingston and Carl
Plantinga (London: Routledge, 2009), 39–51.
12 On the phenomenon of “ontological self-subsistence,” and its relationship with filmic
transparency, see George Wilson, “The Transfiguration of Classical Hollywood Norms: On
Von Sternberg’s Last Films with Dietrich,” in The Creation of Art: New Essays in Philosophi-
cal Aesthetics, ed. Berys Gaut and Paisley Livingston (Cambridge: Cambridge Univ. Press,
2003), 270–71.
13 George M. Wilson, “Coherence and Transparency in Classical Narrative Film,” in
Narration in Light: Studies in Cinematic Point of View (Baltimore: Johns Hopkins Univ. Press,
1986), 56–57 (hereafter cited in text). In a later essay, Wilson restates this argument:
Because motion picture images are formed on a fixed screen by means of the projection of
light, they, unlike paintings, do not exhibit the sort of worked surface produced, for example,
by strokes of paint on canvas. As a consequence, film images do not have the same potential for
eliciting the experience of seeing the drawn or painted scene as arising out of the fine-grained
configurations of material on the displayed surface—an aspect of our total experience of paint-
ing whose aesthetic importance Richard Wollheim has done so much to elucidate. The absence
in film images of this property, foundational for representational painting, both enhances our
impression of the “immediacy” of their depictive power and denies to them the special artistic
possibilities of well-crafted facture.

Wilson, “Le Grand Imagier Steps Out: The Primitive Basis of Film Narration,” Philosophical
Topics 25, no. 1 (1997): 313.
14 Note also that Wilson’s account diverges here from Walton’s account of photographic
transparency, to which Wilson alludes. Walton argues that in looking at a photograph,
we see both the photograph and the object that it depicts. As Wilson notes (212, note
13), however, Walton’s argument is directed at documentary photographs and films in
the first instance, while Wilson’s is concerned with the transparency characteristic of cin-
ematographic fiction films. See Kendall L. Walton, “Transparent Pictures,” Critical Inquiry
11, no. 2 (1984): 246–77.
15 André Bazin, “The Ontology of the Photographic Image,” trans. Hugh Gray, in What
is Cinema? vol. 1 (Berkeley and Los Angeles: Univ. of California Press, 1967), 1:14.
16 Of course, in many animated films and some films using CGI, the filmmaker does cre-
ate every part of the image. But these special cases leave standing the more general points
regarding the differences between painting on the one hand, and film and photography
on the other, that I discuss here.
17 Francis Dauer, “The Nature of Fictional Characters and the Referential Fallacy,” Journal
of Aesthetics and Art Criticism 53, no. 1 (1995): 36.
18 It is also, of course, reminiscent of the classical structuralist distinction between the récit
and discours of a text, as elaborated in the work of Gérard Genette and Seymour Chatman,
for example; and of the distinction in theories of emotional response to fictions between
“fiction emotions” (responses to events in a fiction) and “artifact emotions” (responses to
the work representing the fiction). On the latter distinction, see Ed S. Tan, Emotion and
the Structure of Narrative Film: Film as an Emotion Machine, trans. Barbara Fasting (Mahwah,
NJ: Erlbaum, 1996).
294 new literary history

19 Gregory Currie, “The Capacities that Enable Us to Produce and Consume Art,” in
Matthew Kieran and Dominic McIver Lopes, Imagination, Philosophy and the Arts (London:
Routledge, 2003), 296–97; the related discussion of the “external” and “internal” facets of
narrative is to be found in “Narrative Representation of Causes,” The Journal of Aesthetics
and Art Criticism 64, no. 3 (2006): 312.
20 Kendall L. Walton, Mimesis as Make-Believe: On the Foundations of the Representational
Arts (Cambridge, MA: Harvard Univ. Press, 1990), 175–76 (hereafter cited in text).
21 Paisley Livingston and Andrea Sauchelli, “Philosophical Perspectives on Fictional
Characters,” this volume, 355.
 CONTRIBUTORS

Sara Ahmed is Professor of Race and Cultural Studies at Goldsmiths College,

University of London. Her books include: Differences that Matter: Feminist Theory
and Postmodernism (1998); Strange Encounters: Embodied Others in Post-Coloniality
(2000); The Cultural Politics of Emotion (2004); Queer Phenomenology: Orientations,
Objects, Others (2006); The Promise of Happiness (2010); and On Being Included:
Racism and Diversity in Institutional Life (forthcoming). She has begun a new
research project on the will and willfulness.

Amanda Anderson is Caroline Donovan Professor of English Literature at Johns

Hopkins University and Director of the School of Criticism and Theory at Cor-
nell University. She is the author of Tainted Souls and Painted Faces: The Rhetoric
of Fallenness in Victorian Culture (1993), The Powers of Distance: Cosmopolitanism
and the Cultivation of Detachment (2001), and The Way We Argue Now: A Study in
the Cultures of Theory (2006).

Catherine Gallagher is the Ida May and William J. Eggers Professor of English
at the University of California–Berkeley. She has published several books on
the British novel and is currently writing on narratives of counterfactual history.

Suzanne Keen is Thomas Broadus Professor of English at Washington and Lee

University, where she teaches the novel and narrative. Her most recent book
is Empathy and the Novel (2007). She is guest editor of a special double issue of
Poetics Today (2011) on “Narrative and the Emotions.”

Paisley Livingston is Chair Professor and Head of Philosophy at Lingnan Univer-

sity in Hong Kong. His most recent book was Cinema, Philosophy, Bergman (2009).

Julian Murphet is Professor in Modern Film and English and Director of the
Centre for Modernism Studies in Australia at the University of New South Wales.
He is the author of Literature and Race in Los Angeles (2001) and Multimedia
Modernism (2009) and is writing a book on William Faulkner’s relations with
the modern media ecology.

Andrea Sauchelli is a research assistant and part-time lecturer in philosophy

at Lingnan University.

Murray Smith is Professor of Film Studies at the University of Kent in Can-

terbury. His publications include: Engaging Characters: Fiction, Emotion, and the
Cinema (1995); Film Theory and Philosophy, coedited with Richard Allen (1998);
Contemporary Hollywood Cinema, coedited with Steve Neale (1998); Trainspotting
(2002); and Thinking through Cinema, coedited with Tom Wartenberg (2006).
New Literary History, 2011, 42: 361
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.