1. The psychology of emotions about fiction
Typically, if harm comes to someone I care about, I would feel a great
deal of sorrow and concern for that person. I may also feel anger at
whomever or whatever harmed them and would be motivated to act on
their behalf. However, we generally won’t have such reactions
during our emotional engagement with works of fiction. To borrow an
example from Radford (1975: 70), I do not weep over Mercutio’s
body after he is killed while watching a theatrical production of
Romeo & Juliet
. I do not seek revenge or challenge Tybalt
to a duel. I do, however, feel very strongly about Mercutio’s
death. It’s the
behavioral
responses to our emotions
about fictional characters and reality that vary wildly. In this
section, we will explore some of the ways to explain this asymmetry
and, in general, the nature of our emotional responses to fiction.
Ultimately, one’s response to the “asymmetry
problem” depends on further commitments concerning the nature of
emotions and other mental states. The views expounded
here—theories of make-believe, simulation theory, and theories
of empathy—all explain emotional responses to fiction in
relation to real-life emotional responses. The views differ with
respect to the psychological framework they emphasize, be it
imagination, offline processing, or a range of empathetic
practices.
1.1 The distinct attitude view (DAV)
Some scholars argue that a solution to the asymmetry problem requires
a
distinctive mental attitude
that we employ during our
engagements with fiction. The nature of the distinct attitude varies
by the theorist. Quite often, a mental “box” or mechanism
is posited that we utilize exclusively when considering non-actual
objects, including fictional ones (see, for example, Nichols &
Stich 2003), as well as hypothetical and counterfactual thought,
mental activities involving deliberation and decision-making, mental
imagery, and mindreading (attributing mental states to other
people).
Alternatively, we may use largely the same mental processes as in
real-life situations, but these processes are run
“offline”, disconnected from their typical functional and
inferential output. The result is a different kind of mental state
than we would have if we considered a real-life or actual object. We
adopt
pretend
(Searle 1975; see also Kripke 2011 & 2013),
simulated
(e.g., Currie & Ravenscroft 2002; Walton 1997),
or
imaginary
beliefs, desires, emotions, and thoughts toward
non-actual objects (e.g., Schroeder & Matheson 2006; Weinberg
& Meskin 2006). These states are isomorphic to genuine mental
attitudes and may sometimes be phenomenologically indistinguishable
from them. While pretense, simulation, and imagination differ in
important ways, each view trades on the idea that our mental attitudes
concerning fiction differ in content and functional role from those
concerning actual events or objects—even if, phenomenologically,
the experience of each is similar (see Friend 2022 for a comment on
and critique of this asymmetry).
All proponents of the distinct attitude view argue that mental states
such as emotions can be identified in terms of their characteristic
functional roles. Consider beliefs. It is generally understood that
our beliefs about real-life and non-actual objects utilize many of the
same causal and inferential pathways. However, they have significantly
different inputs and result in very different outputs. Our real-life,
everyday beliefs are about things that we can or might in principle
see, touch, and hear—things that exist concretely, or
spatiotemporally. These objects act as the input for our everyday,
stereotypical beliefs (see Fitzpatrick 2016; García-Carpintero
2019; Lewis 1978; Plantinga 1974; Quine 1948 [1953]; Sainsbury 2010;
Salmon 1998; Schiffer 1996; B. Smith 1980; Thomasson 1999 & 2003;
Walton 1990). In contrast, beliefs about fiction (and other non-actual
objects) do not take actual, concrete objects as their object. Rather,
they are about fictional things, however that might be cashed out
ontologically. The output of our mental processing about actual and
fictional objects is also different; they result in different kinds of
behaviors. So, while our beliefs about real life and our beliefs about
works of fiction are
similar
in many ways, proponents of the
DAV hold that they are different enough to constitute a distinct kind
of mental state. The result, on this view, is that we have
belief-
like
(or imaginary, or simulated) attitudes towards
the content of fiction, but not beliefs simpliciter.
Consider Currie and Ravenscroft’s (2002) argument for imaginary
mental states. They maintain that mental states such as emotions and
beliefs serve as part of an inferential network, motivating not only
action, but also other mental states. The authors posit that beliefs
about fiction are run offline, disconnected from their normal
behavioral and cognitive networks. The same goes for other activities
with non-actual content, such as hypothetical thought and pretend
play. These states are not stereotypical beliefs, but rather
imaginative
ones because stereotypical beliefs are understood
in terms of the behavior that they produce.
The psychologist Paul Harris (2000) provides further evidence for this
view, particularly in terms of our emotional responses to imaginings.
Very young children tend to be overcome by the emotions caused by
their imaginative activities—e.g., fearing the monster under the
bed or witches that they saw in a movie—even if they know that
the fiction is not real. Older children and adults are generally able
to regulate and override these emotions, but still may often become
absorbed in their imaginative activity. We get “lost” in a
film or novel and experience strong emotional reactions, emotions that
color our real-life activities. We generally do not act on these
emotions, though, which suggests that they are distinct from the
everyday emotions we experience in response to real-life objects and
events.
Further reading: Camp 2009; Damasio 1994; Fontaine & Rahman 2014;
Gendler & Kovakovich 2006; Gilmore 2020; Goldman 2006a; Kosslyn
1997; Kosslyn et al. 1993; LeDoux 1996; Meinong 1904 [1960]; Moran
1994; O’Craven & Kanwisher 2000; Stecker 2011; Toon 2010a,
2010b, 2012; Tullmann 2022; van Leeuwen 2013.
1.2 Theories of make-believe
One way of explaining the asymmetry problem from a distinct attitude
perspective is a fictional theory of make-believe, popularized by
Kendall Walton (1978 & 1990). According to a theory of
make-believe, our engagements with fiction draw on our capacities for
imagination and pretense. While reading a novel, for example, we play
a game of make-believe, creating a “fictional world” in
which the propositions presented in the novel are fictionally true. On
this view, each reader is the participant and creator of her own
fiction-based game. Fictional worlds are similar to the imaginary
worlds that children create during their pretend play (Walton 1990;
Currie 1990, Harris 2000). Works of fiction prescribe imaginings:
words on a page, for instance, invite imaginative engagement with a
work.
Pretense
, here, involves acts of behaving and thinking
as if some proposition or state of affairs is true while knowing that
it is not. In their games of make-believe, a child may pretend that a
couch is a house, underneath a dining room table is a dungeon, and a
baseball bat is a magnificent sword—all while knowing that none
of these are the case. Rather, the imaginative game of make-believe
makes it fictionally true that the dining room contains a dungeon and
that the baseball bat is a sword.
Theories of make-believe hold that something similar occurs when
adults engage with fiction, but this time the props are largely
imaginative. While reading the
Lord of the Rings
series, we
pretend that there is a world like our own in which wizards, elves,
and dwarves exist alongside good-natured hobbits and an evil dark
lord. The novel itself acts as a prop, each line feeding into our
fictional world and adding layer upon layer of detail to our game of
make-believe. The game of make-believe extends to our psychological
states: we pretend to believe that Frodo defeated the Dark Lord by
tossing the Ring of Power into the fires of Mount Doom, for instance.
Importantly, we have fictional
emotions
about Frodo and his
crew and fictional
desires
for the young hobbit to vanquish
Sauron. Stacie Friend (2022) points out that Walton does not deny that
we have some genuine emotional responses to fiction. Rather, he seems
to deny that we have genuine fear, pity, joy, etc. Instead, we
experience a different kind of emotion that is bound by the
imaginative game; these emotions are different in kind from everyday
emotions because they are functionally and cognitively
“quarantined” (to borrow Friend’s term) from
cognitive and functional states about real life things.
A theory of make-believe attempts to explain two things: the ontology
of fiction and the nature of our psychological states about them.
Let’s begin with the ontological question. Do fictions and
fictional characters exist? What are fictional entities such that we
can think about them, speak about them, etc.? Theories of make-believe
have a ready response to these questions: fictional entities do not,
strictly speaking, exist outside of one’s game of make-believe.
Instead, when we discuss fictional entities, we make pretend
illocutions concerning pretend objects. Like an actor in a performance
of
Hamlet
, we do not make genuine assertions when we speak
about the goings-on of fiction. We merely pretend to do so as a part
of the game. Fictional entities are not objects that can be found in
space or time, even if the images or words used as props can be (a
painted figure, a film image of a person, or words that describe a
villain).
Walton contrasts his view with a realist Meinongian theory, according
to which fictional entities exist eternally as abstracta, similar to
Platonic forms (1990; see also Sainsbury 2010 and Wolterstorff 1980).
On this view, fiction entities are not created, but rather drawn upon
and put together in creative ways by authors, filmmakers, dramatists,
etc. Meinongian theories have come under strong attack, and rightly
so—it is counterintuitive that fictional entities are Platonic
ideals that are not created by the author, for instance (see Thomasson
1999). However, contrasting a pretense theory with the Meinongian
position ignores the more fine-grained issues concerning the existence
of fictional entities. The middle ground includes other broadly
realist theories, such as the possible objects view (Lewis 1978, which
holds that fictional entities are denizens of far-off possible worlds)
or the abstract artifact view favored by Kripke (2011 & 2013),
Salmon (1998), Schiffer (1996), and Thomasson (1999), which holds that
fictional entities are non-corporeal but created entities whose
existence and persistence depends on the practices of people in the
real world.
However, neither realist nor pretense-based ontology entails that our
mental states about fiction are distinct in type. It could be that we
merely pretend that fictional objects exist when we think about and
discuss them. Our mental states about those objects are the standard
mental states,
about
something fictional. That is one
possible way in which to understand the ontological and psychological
questions of fiction: an anti-realist ontology of fictional entities
coupled with a genuine attitude view of mental states. Unfortunately,
this view raises a host of questions concerning the possibility of
referring to nonexistent objects.
For this reason, many anti-realists about fictional entities favor the
DAV according to which our mental states towards fictional entities
are not stereotypical mental states. One popular way in which to
ground a psychology of make-believe is to adopt some form of
simulation theory
(ST), which we will explore further in the
next subsection. ST is typically used in cognitive science to explain
how we attribute mental states to others, especially to predict their
behaviors and understand their emotional state (see Goldman 2006a;
Gordon 1986; Currie & Ravenscroft 2002; Nichols, Stich, et al.
1996; Prinz 2002). Currie and Ravenscroft (2002) argue that
imagination essentially involves the capacity to put ourselves in the
place of another or our self in another place or time. Importantly,
both Walton and Currie/Ravenscroft hold that the mental attitudes we
adopt in our imaginings are substitutes for genuine ones; we have
imaginative
or
fictional
beliefs about fiction
instead of beliefs simpliciter. On their view, imagining simulates the
role of other states, such as the role beliefs play in inferential
processes (2002: 49).
The upshot of a theory of make-believe—combining anti-realism
about fictional entities and ST to explain our psychological states
towards them—is that it allows proponents of the distinct
attitude view a theoretically cohesive and elegant means by which to
solve the puzzles of fiction described below. For example, we can
solve the paradox of fiction by arguing that we do not possess
ordinary types of beliefs about fictional entities, rejecting the
premise of the paradox that states we have genuine beliefs about
fictions. We feel “sympathy for the devil” because that
emotion is distinct from our normal cognitive processing, which allows
us to experience unique emotional and moral responses towards entities
that do not match how we would respond to their real-life
counterparts.
Further reading: Kroon 1994; Levinson 1993; Mothersill 2002; Searle
1975; Summa 2019; Toon 2010a & b, 2012.
1.3 Simulation theory
Simulation theory gained prominence in the late 1980s with the work of
philosophers like Alvin Goldman (1989, 1993, 2006a,b), Robert Gordon
(1986 & 1996), and Jane Heal (1996) and has been adopted in a
variety of fields, including aesthetics, to explain our mental
processing about other persons’ mental states. ST holds that we
utilize our own perceptual, emotional, and cognitive mechanisms while
mindreading
(attempting to understand the mental states of
others). We project or imagine ourselves in the situation of the
observed person. As with the cases of mental states about fiction
mentioned above, these mental mechanisms are run offline, disconnected
from their typical functional output (see Currie & Ravenscroft
2002). This results in simulated mental states that are distinct from
stereotypical mental states. In fictional cases, audiences utilize
simulated input (non-genuine emotions, for example) and garner
simulated output (a pretend emotion about what a character thinks or
feels) when simulating the mental states of a fictional entity.
Consider Alvin Goldman’s version of ST (1993 & 2006a).
Goldman argues that we attribute mental states to another after we
recognize our
own
mental states under actual or imagined
conditions. I
transform
myself, imaginatively, into her based
on my understanding of how I would feel or think in the same
situation. I imaginatively take on what I imagine to be her relevant
beliefs, desires, emotions, and perspective to determine further
mental states and behavioral predictions. This is called
enactive
imagining
—or, e-imagining, for short—because I
utilize my own mental processes for simulative imagining. Once I
e-imagine myself as the target, I can introspect what I feel during
this particular situation. I then judge that the target feels the same
way.
Gregory Currie’s version of ST provides another elegant
explanation for psychological states about fiction. Currie has backed
away from a strong simulative approach in some of his recent work, but
the general assumption of offline ST remains the same (see Currie
& Ravenscroft 2002). Like Walton, Currie argues that our basic
psychological interactions with fiction involve games of make-believe,
games that are based on imaginatively simulating fictional actions and
the minds of fictional characters. On this view, audiences
imaginatively take on the mental states of others as closely as
possible and run their own “mental economy” to see how
they would feel in a comparable situation. This view also adopts a
version of the DAV; our mental attitudes about imaginings are not
stereotypical mental attitudes, but rather pretend/imaginary ones.
This is because these states lack their typical functional role.
According to Currie’s ST, audiences are intended to adopt
imaginative attitudes toward fictional characters and situations
(Currie 1990). Currie distinguishes between
primary imagining
and
secondary imagining
. Primary imaginings involve what is
true in a story, “those things which it makes fictional”
(1995: 255). We adopt the fictional beliefs necessary to maintain a
coherent fictional world while disregarding those which contradict it.
I disregard my real belief that eagles are not large enough to carry
humans while reading
The Lord of the Rings
, for instance. I
also adopt the imaginative belief that there exists large,
perambulating tree folk. When I watch the film adaptation of
The
Fellowship of the Ring
and come to learn of the danger and power
of the One Ring, I do not acquire a new belief that “there is
One Ring desired by Sauron”, but rather a “belief-like
imagining” of this proposition.
Secondary imagining
is a form of simulation that occurs when
we engage in an empathetic reenactment of a character’s
situation (Currie 1995: 256). First, I put myself into a fictional
character’s position: I imagine what it would be like to be
Frodo learning about the ring from the wizard Gandalf. Then I reflect
on what I currently feel as the result of this imagining: surprised
and scared. Once I identify my thoughts, desires, and feelings, I then
imagine that that is how Frodo feels in this situation as well. In
this sense, secondary imagining helps us to identify and empathize
with fictional characters. I then remove myself from the simulation,
so to speak, and attribute these same feelings to Frodo. The
assumption is that all or most of this mental processing occurs both
offline and unconsciously, so we do not act on our emotion-like
imaginings and are not necessarily aware when they occur.
Further reading: Blanchet 2020; Cochrane 2010; Gallese 2019; Gordon
1992; LeBar 2001; Nichols & Stitch 2003; Short 2015; Spaulding
2016.
1.4 Empathy
Finally, some philosophers contend that the primary way by which we
emotionally engage with fictional entities is through empathy (see
Agosta 2010; Bailey 2021; Coplan & Goldie 2011; Gallagher 2012;
Maibom 2014; Prinz 2011; Stueber 2011; Vetlesen 1994; Walton 2015).
Arguments about fiction and empathy are very similar to those
concerning simulation and make-believe; in each case, audiences are
thought to imaginatively put themselves in the situation of the
character, consider how they would feel in a similar context, and
infer that must be how the character feels as well However, empathy
can take on a variety of forms, some which are quick, seemingly
automatic and outside of conscious control and appear rather different
from the simulation theories described above. Other forms of empathy
are slower and more deliberate, like those that may be part of games
of make-believe.
Martin Hoffman (2008) identifies processes like mimicry, mirroring,
and direct association as three forms of fast, unconsciously activated
empathy. Mimicry involves an “innate, involuntary, isomorphic
response to another’s expression of emotion” (2008: 441).
There are two steps involved in mimicking the emotion of a target.
First, there is an automatic change in the subject’s facial
expression, voice, and posture at the same time as a corresponding
change in the target’s facial expression, vocal intonation,
posture, etc. These changes then trigger similar feelings in the
target as those present in the target.
Mirror neurons
may be
the neural basis of mimicry. These neurons are triggered when one
person observes the actions or emotional expressions of another. This
results in the same kind of neural pattern in the subject as if she
were performing the observed action or having the same emotion herself
(2008: 441; see also Freedberg & Gallese 2007; Clay & Iacoboni
2011; Decety & Meltzoff 2011; Goldman 2006a).
There is also some evidence that subjects can mirror the motor
intentions of a target; witnessing a movement in another (or even in a
statue!) results in the subject’s motor cortex being activated
as if
she
were the one moving (Freedberg & Gallese
2007).
Finally,
direct association
occurs when we perceive a target
who undergoes an event that is similar to one that we have experienced
in the past (Hoffman 2008: 441). For example, a friend of mine has
recently lost her pet dog and displays sorrow-related signals (crying,
having a “long face”, slouched posture, etc.). My memory
of a similar situation in which I lost my pet parrot makes me also
express these signals and even consciously experience sadness. I make
this association unconsciously, without thinking about it or planning
to do so.
While some of our social cognitive abilities seem to occur
automatically, there are other cases in which understanding a
target’s mental state requires slower, more thoughtful, and
deliberate processes. Sometimes we may need to deliberately take on a
target’s perspective to know how they feel or what they will do
next.
Perspective-taking
seems especially important in
ambiguous scenes in which we do not know enough about a person or her
context and so we have difficulty judging how she feels or what she
believes or desires. Sometimes perspective-taking is simply equated
with empathy. When subject
X
empathizes with a target,
Y
, they imaginatively take on
Y
’s mental
states as closely as possible.
X
shares in
Y
’s
mental state and, further, that
X
’s responses are
caused by
and involve the same
type
of state as
Y
’s. In taking another’s perspective, I
“put myself in their shoes”, and imagine what they must
think or feel in a particular situation.
Each of these forms of empathy is relevant to how audiences
emotionally engage with fictional characters. To continue the
Lord
of the Rings
example, audiences may have to take Frodo’s
perspective to fully appreciate how he feels the moment he discovers
how dangerous Bilbo’s old ring really is. Mirroring
Frodo’s facial expression (while watching the movie) may
engender similar feelings of fear and surprise in me. These are both
instances of empathy, each of which, arguably, helps audiences better
appreciate the work of fiction and characters within.
Further reading: Carruthers 1996; Goldman 1993& 2006a; Freedberg
& Gallese 2007; Clay & Iacoboni 2011; Decety & Meltzoff
2011; Hoffman 2008.
1.5 Alternate views
The above views—make-believe, simulation, and
empathy—explain interactions with fiction in a similar manner:
we engage with fiction by utilizing similar pathways to real-life
interactions but with non-real objects. Our mental states about
fiction may not be genuine sorrow, anger, etc. The differences between
the views are in the details of how the psychological states
concerning fiction are cashed out—whether the best explanation
for engagement with fiction stems from simulation theory, a theory of
imagination, or a theory of empathy.
Some philosophers have rejected simulation theories and theories of
make-believe as the primary way to understand the psychological
mechanisms involved in our emotional responses to fiction (Carroll
2008; Matravers 2014; Tullmann 2022; Wilson 2011). Philosophers such
as Noël Carroll (2008) contend that simulation, make-believe, and
empathy are generally not required for understanding fictional
entities. Authors, filmmakers, and other fictional content creators
often make the inner lives of characters obvious to their audiences.
For instance, in the Frodo and the Ring example, it is clear that
Frodo is scared by his words, actions, and facial expression—no
make-believe or simulation is required!
The main theoretical support for the DAV stems from a folk
psychological and functionalist understanding of the nature of mental
attitudes. Beliefs, desires, judgments, etc. have characteristic
behavioral and inferential roles, understood in terms of inputs from
stimuli and their cognitive and behavioral outputs. As we have
discussed, real-life emotions and their imaginative counterparts might
utilize much of the same causal and inferential pathways to bring
about certain responses, but they employ significantly different
inputs and result in very different outputs (see, e.g., Currie 1990;
Currie & Ravenscroft 2002; Weinberg & Meskin 2006; Schroeder
& Matheson 2006). Furthermore, research in psychology and
cognitive neuroscience seems to support the idea that engaging with
non-actual objects utilizes
many,
but not
all
, of
the same neural pathways as our mental activities concerning actual
things (Kosslyn, Thompson, & Alpert 1997; Kosslyn, Thompson, &
Ganis 2006). These studies help to explain why our reactions to
imaginative activity are often so robust but may also suggest that we
utilize a distinct attitude in our engagements with fiction (Damasio
1994; Schroeder & Matheson 2006).
As we’ve also seen, theorists generally try to explain the
distinctness of our imaginative attitudes in terms of functional role
(behavioral outputs) or inferential role (mental outputs); our mental
states towards fiction do not lead to the kinds of thoughts and
behaviors that they would for real-life objects. This problematically
assumes a straightforward functionalist view of mental states that,
while attractive in terms of folk psychology, may not accurately
capture the nature of how mental states motivate action or inferential
processes. However, if mental states are not individuated in terms of
functional role, then the case could be made we may in fact have
standard mental states towards fiction, despite the fact that we do
not act towards fictional objects as we do towards real-life ones (see
Buckwalter & Tullmann 2017). One can argue from a principle of
parsimony that there is no need to posit a distinct mental attitude if
typical ones have the same explanatory power.
Finally, the DAV (especially one utilizing a theory of make-believe or
simulation theory) does not seem to be able to account for our actual
phenomenological—that is, conscious, possibly
introspectable—experiences with fiction. Our emotions, beliefs,
desires, and other mental states towards fiction feel natural and
relatively automatic, not like we are playing a game of make-believe
or simulating a possible course of action. The assumption, here, is
that there is
something it is like
to engage in a game of
make-believe; we knowingly and willingly begin games of make-believe
and do not explicitly do so with works of fiction. The proponent of
the DAV may dismiss the phenomenological worry by arguing that the
game of make-believe or simulation takes place unconsciously and,
after some practice, quite rapidly. Some of the imaginings involved in
a game of make-believe are deliberate and consist of conscious,
occurrent mental states. But others are spontaneous, unconscious, and
automatic. We do not tell ourselves to begin imagining what is going
to happen to our favorite television character. We simply do it,
sometimes without realizing it. Walton says that when this happens our
imaginings “have a life of our own” and we feel less like
an author than a spectator to the imagining (Walton 1990: 14).
Stacie Friend (2022) has recently put forth another approach to the
psychology of fiction that provides nuance to the debate concerning
genuine and fictional emotions. For Friend, emotions about fiction and
those about real life cannot be so easily separated in terms of
functional role, motivation, and phenomenological experience. Some
emotions about fiction are motivational and some emotions about real
life are not (admiration, for instance). Walton and Friend deny that
there is a phenomenological difference in kind between the two types
of emotion: we cannot say, for instance, that our sorrow over the
death of Anna Karenina is less intense than sorrow about the death of
a real-life friend—or, at least, the intensity does not differ
so much to warrant a distinct type of mental attitude. In general,
Friend states:
Emotions are multidimensional, and each dimension—physiological,
phenomenological, motivational or evaluative—is complex,
admitting of a variety of degrees and distinctions. There is no
dimension along which a dichotomy between “fictional” and
“ordinary” emotions can be sustained. (2022: 262).
The upshot of this view is that discussions concerning emotions and
fiction can move past debates concerning the nature of fictional
emotions to questions about the appropriateness, rationality, and
social/moral implications of those emotions, as we will see in
§3
below.
Further reading: Carruthers 1996 & 2011; Friend 2020; Gopnik 1993;
Gopnik & Schulz 2004; Fodor 1983; Robinson 2005.
2. The Paradox of Fiction
The nature of the mental architecture for imaginary and fictional
contexts grounds some of the pressing puzzles concerning our emotional
responses to fiction—most prominently, the so-called paradox of
fiction. Cognitive belief-based theories of emotions were in full sway
when the paradox was first introduced (Radford 1975; Walton 1978;
Currie 1990). According to these views, an emotion about an object
X
requires that we have some relevant belief
Y
concerning
X
’s relation to our well-being. For example,
experiencing fear requires that I believe that there is an object in
my environment that could harm me or someone I care about. We lack the
emotion if the relevant belief is absent (Solomon 1976 [1993]).
The wording of the paradox reveals an adherence to a belief-based
theory of emotions. One version of the paradox states:
- We have genuine emotions about fiction all the time.
- We do not believe that fictional characters exist.
- We can only have genuine emotions about things we believe to
exist.
Different authors word the three propositions slightly
differently—and, indeed, this formulation is not found in
Radford’s original 1975 piece—but the tension between
belief and fiction is the same across versions. The paradox is
intended to capture a very natural thought concerning our emotions: if
we know that we are engaged with a work of fiction, we should not have
the emotionally relevant belief. No emotion
should
arise.
Nevertheless, we have emotional experiences about fiction all the
time, whether these are genuine emotions or not.
Responses to the paradox typically proceed by either accepting the
irrationality of emotions about fiction or dismantling one of the
propositions of the paradox. Some reject the paradox out of hand. This
section explores each of these responses in turn.
Further reading: Buckwalter & Tullmann 2017; Langland-Hassan 2020;
Cova & Teroni 2016.
2.1 Emotional responses are irrational
One interpretation of the paradox states that there is something
fundamentally
irrational
about our responses to fictional
entities. Colin Radford (1975) accepts (variations on) each of the
paradox’s propositions, arguing that our emotions towards
fiction force the reader into adopting two contradictory beliefs: we
both believe and do not believe that the fictional object of our
emotion exists. Radford states:
I am left with the conclusion that our being moved in certain ways by
works of art, though very “natural” to us and in that way
only too intelligible, involves us in inconsistency and so incoherence
(1975: 78).
It isn’t immediately clear what Radford meant by this
inconsistency and incoherence. One way of understanding this statement
is to suggest that Radford contends that we hold two contradictory
beliefs (about the existence of a fictional entity) at the same time.
Fabrice Teroni (2019) suggests that Radford simply means that our
emotions about fiction do not make sense, because they ought to
dissipate once we acknowledge that the object is
fictional—similar to Radford’s case in which it would be
irrational to continue to feel sorrow about the death of one’s
sister once one comes to realize that the belief in one’s
sister’s death is false. Still, it does seem that, to Radford,
it is just a brute fact of human psychology that we have emotion-like
responses to fiction. We are sad when our favorite character dies and
are righteously angry when they are harmed. But these are not genuine
emotions, because most, if not all, genuine emotions require a belief
in the existence of their object. Radford doesn’t treat the
irrationality of emotional responses to fiction as a problem, however.
In fact, he argues that we have these sorts of incoherent responses
all the time: when we cheer for our favorite sports team, fully
knowing that nothing we do on our couch at home influences the game,
or when we fear death even while acknowledging that it is nothing more
than a dreamless sleep (to use Radford’s example, 1975: 79).
While not explicitly addressing the paradox of fiction, other
philosophers have taken up the question of the rationality of emotions
in ways that could be relevant to a response to the paradox of fiction
(Gendler & Kovakovich 2006; de Sousa 2002 & 2004; D’Arms
& Jacobson 2000a & 2000b). There are several ways in which
emotions can be understood as being rational. First, emotions can be
correct
. Friend (2022) states that “It is correct to
respond emotionally only if the object of emotion exists” (2022:
264). This would seem to imply that emotions about fictional entities
are incorrect since the object of the emotion does not exist. In
contrast, Teroni (2019) states “Correct emotions for fictional
entities are emotions that correspond to truths about these fictional
entities supplied by the relevant fictions” (2019: 125). To use
Teroni’s example, one’s fear of a dog is correct if the
dog
is
dangerous. For the fictional case, emotions are
correct if the work of fiction gives reasons to believe that the
fictional entity or event warrants that emotion. Fear about the fate
of the hobbits at the end of the
Lord of the Rings
trilogy is
correct if the text provides evidence to suggest that the hobbits are
in danger.
Even if an emotion with a fictional object is incorrect (on
Friend’s view), that doesn’t entail that the emotion is
irrational. Teroni’s conception of emotional correctness has a
stronger normative component than Friend’s—that is,
conditions for which emotions might be apt. The former view might be
likened to an understanding of emotional
fit
: how an
emotional response may fit its object (see D’Arms & Jacobson
2000a). An emotion fits its object if we have some good reason to feel
it; the emotion accurately represents its object. We can compare the
fit between an emotion and its object to that of a true belief and a
state of the world. Both spiders and battlefields may be fitting
objects of fear; this evaluation is apt in some way, as being proper
formal objects. Our colleague’s promotion may be a fitting
object of jealousy. An off-color joke may be a fitting object of
amusement. On this view, emotions fit their object in case we have
some reason to have them for that object (compare this to D’Arms
and Jacobson’s slightly different account of fit, which they
characterize in terms of a response-dependent feature of the object
that does not require reasons or norms, 2000a).
Finally, we can also speak of an emotion’s
propriety
.
“Propriety” carries significant normative implications; it
suggests that there are appropriate contexts in which one can or
should have certain emotions. This way of understanding the aptness of
emotions moves beyond reasons for belief and into the appropriateness
of those emotions. The latter has significant social and moral
implications, some of which are captured in the puzzles mentioned
below in
§3
.
Importantly, correctness, fit, and propriety may not always match in
any particular object. An emotion may fit its object, but not
necessarily be the proper response to take. For example, even if a
battlefield is a fitting object of fear, it may not be
proper
for a soldier to feel if he or she has an important task to fulfill.
If our colleague is also our friend it may be improper for us to be
jealous of her promotion—we
should
be happy for
her—even if it is fitting for us to be, since, perhaps, we were
also due for a promotion and did not get one. When we conflate the fit
and propriety of emotions, we commit
the moralistic fallacy
:
taking the morally normative implications to be built into our
emotional responses towards things in our environment (D’Arms
& Jacobson 2000a & b).
Recent literature in cognitive science has begun to debunk the
traditional bifurcation between rationality and emotions, showing that
emotions are often necessary (or at least useful) for planning, making
important decisions, and making moral judgments (Ben-Ze’ev 2000;
Damasio 1994; Gordon 1986; Nichols 2004; Solomon 1976 [1993], etc.).
However, it is debatable whether these benefits extend to our emotions
about fiction. The central question remains whether it is rational
(fit or proper) to have emotional responses to things that have no
bearing on our actual lives. Moreover, Currie (1995) points out that
the concept of epistemic emotional rationality is puzzling because
fiction presents audiences with a great deal of false information.
This could undermine one’s ability to function in the real world
if one takes it literally (see Best 2020 & 2021). If this view is
right, then Radford’s paradox highlights an important aspect of
human psychology.
Further reading: Adair 2019; Greenspan 1988; Roberts 1992; Song
2020.
2.2 Against the first proposition
One popular way of responding to the paradox is to argue that our
emotions about fictions are not genuine—we do not have genuine
emotions about fictional entities all the time. While initially
counter-intuitive, this view makes sense when thought of in terms of a
theory of make-believe or simulation as described in
§1
.
Perhaps the most influential view on these lines comes from Kendall
Walton (1978 & 1990), who, as we have seen, has argued that our
engagements with fiction are akin to childhood games of make-believe.
Some of our emotional responses to fiction are genuine, but some are
imaginary or make-believe. As described in
§2
,
Walton holds that emotional responses we have towards fictional
objects are very similar to emotions we have about real-life objects
but are not typical emotions. This explains both why we do not act on
those emotional responses and why we may seek out fictions that elicit
negative emotions such as anger or sadness.
The central issue with this view is that it certainly
feels
as though our emotional responses to works of fiction are real. When I
watch a horror film such as
Get Out
or
Hereditary
, I
may cover my eyes with my face, my heart rate accelerates, I break out
into a cold sweat, and involuntarily scream. These are all embodied
indications of fear that I’m not faking. I really do
feel
terrified. According to the current response to the
paradox, however, the fear responses and phenomenal states that I
experience in such cases aren’t full-fledged emotions. They are
“quasi” emotions. Importantly, on this view, feelings and
bodily responses are not emotions themselves. While Walton is not
explicitly cognitivist about emotions, he does make certain statements
that seem to suggest cognitivism. For instance, in “Fearing
Fictions” (1978), Walton states:
It seems a principle of common sense … that fear must be
accompanied by, or must involve, a belief that one is in danger;
(1978: 6, 7)
and also that “Charles does not believe that he is in danger; so
he is not afraid” (1978: 7; see Friend 2022 for more on
Walton’s view and reception). This view suggests that emotions
proper
are
cognitive. They involve a belief or other
cognitive mental state, such as a judgment or thought (more on this in
the next subsection). Specifically, this view accepts the idea that a
genuine emotion—say, sorrow over a friend’s
death—requires a belief that our friend actually died. Imagine
that a family member informs you of the death of a beloved family
friend. You feel sad because you believe that your friend no longer
lives. Suppose now that your family member was merely playing a cruel
trick on you—the family friend isn’t dead! Your belief
about the friend’s death would be overruled. However, the
feelings of sorrow may linger, the way that anxiety or fear lingers
after waking from a dream that we know isn’t real. Would it
still make sense to say that you are sad about your friend’s
death? No; most people would likely say that emotion goes
away—probably replaced with anger about your family
member’s cruel trick.
According to belief-based theories of emotion, something similar is
happening with our emotional responses to works of fiction. One may
have all the embodied reactions related to fear while watching a
fictional film, but without the belief in the actuality of the fearful
object, none of those reactions amount to genuine fear.
Further reading: Dos Santos 2017; Vendrell Ferran 2022; Humbert- Droz
et al. 2020; Williams 2019.
2.3 Against the second proposition
Few contemporary philosophers opt to eliminate the second proposition
of the paradox of fiction. Doing so implies that a reader or viewer of
fiction would genuinely believe that the fiction is real while she
reads or watches it. Theorists who opt to reject the second
proposition of the paradox must somehow square the “suspension
of disbelief” with the contradictory beliefs and actions we seem
to have in response to fiction. The notion of suspending disbelief was
first introduced by the British Romantic poet Samuel Coleridge, who
argued that we suspend our disbelief in the nonexistence of fictional
objects during our engagement with fictional stories (Coleridge 1817;
see also Hurka 2001). This supposedly explains our emotional responses
to fictional entities; we emotionally respond to them because we
believe that they concretely exist in the time that we engage with the
artwork.
While it’s certainly true that we sometimes become very
absorbed
in fiction, it is a genuine question whether we
forget or are tricked into believing that fictional characters and
events exist. Some contemporary philosophers have followed this line
of thought and argued for some kind of illusion theory about fiction
(see Quilty Dunn 2015; Kivy 2011). We know that the objects in
fictional films are nonexistent, just like we know that a
magician’s “magic” isn’t real. The question is
whether we can be tricked, perhaps momentarily, into believing
otherwise. Noël Carroll (Carroll 2008) characterizes this
challenge as
the illusion thesis
: we fall prey to some kind
of illusion during our engagements with fiction. So far, we have been
discussing one type of illusion, concerning belief. Carroll
characterizes
two
types of illusions:
- The Cognitive Illusion Thesis: one might come to believe that
fictional events or characters actually occur or exist.
- The Perceptual Illusion Thesis: we are committed to the existence
of the represented fictional objects on a perceptual level.
Different artistic media may commit us to one or both of these
illusions. For example, reading a novel might subject one to a
cognitive illusion, but not a perceptual illusion. Perceptual
illusions generally apply to visual fiction, although we can imagine
someone listening to an audiobook in her car falling prey to the
perceptual illusion that the fictional person being narrated is an
actual person. Moreover, while neither of these theses addresses
emotional responses
per se
, we can think of a similar type of
argument in which emotions about fiction are illusory, like the
phantom limb pain phenomenon or the rubber hand illusion (Richardson
2009; de Vignemont 2007).
Many philosophers reject both versions of the illusion thesis out of
hand since they seem to entail that we would act towards a character
in just the same way that we would act towards a real person. As
Katherine Thomson-Jones (2008) points out:
I am able to appreciate the vivid depiction of an army of zombies
surging forward with arms outstretched, the use of special effects or
highly emotive music, the importance of the scene for the narrative,
and so on. Surely, if I had suspended my belief that the zombies are
fictional, I would be too frightened to appreciate film in this way.
(2008: 107)
Moreover, most of our behaviors towards fiction (or lack thereof) are
inconsistent with the idea that we even temporarily suspend our
disbelief about the reality of fiction—this is the asymmetry
problem, once again. We do not act as if we believe that the fiction
is real. The same idea works for other mental attitudes, such as
desires and emotions. Moreover, our conscious experience of watching a
film is also antithetical to the cognitive illusion thesis. If asked,
we would deny that fictional entities are real. We would also deny
that we have been tricked into believing otherwise.
Carroll also rejects the perceptual illusion thesis, arguing that it
our visual experiences do not meet the first criterion. Our perception
of movie screens and actual objects are not identical, or even
sufficiently similar to the perceptual experience of real-life
objects, to suggest a perceptual illusion. There are surface
interferences—scratches and dirt on a film strip, hair on the
projector slide, the size and shape of the screen, framing devices,
etc.—which make the viewer aware of the screen and remind her
that the objects in the movie are not really in front of her. Carroll
also points out that we typically perceive edge phenomena; we can see
around the edge of an object as we move but we don’t experience
such phenomena in our visual perception of film. We cannot look around
a character to see what is going on behind her. We can provide a
similar explanation for other fictional media. We do not visually
perceive plays in the same way that we do real-life people and events,
because of the stage and other spatial and physical discrepancies
between them. Pictures are always framed and are not subject to edge
phenomena, just like films (see also Derrida 1978 [1987] and Foucault
1966 [1970]). Even listening to an audiobook will probably not sound
identical to listening to real people give an account of their
lives.
In contrast, Jake Quilty-Dunn (2015) provides one way to argue for a
version of the perceptual illusion thesis. In this view, film viewers
deploy many of the same perceptual capacities that we do in real life.
We may form isomorphically similar
perceptual
beliefs
(beliefs formed on the basis of perception rather than other cognitive
states) about a face we encounter while watching a film that we would
the same face in real-life. The visual processing that leads to such
beliefs is, in effect, under the illusion that there is an actual face
being perceived. This leads to contradictory beliefs: the cognitive
belief that the person we perceive does not exist and the perceptual
belief that, implicitly, suggests that they do. In turn, our emotional
responses to that face are genuine emotions, but the belief in the
existence of the person is perceptual, not cognitive (see Siegel
2010). In this case, the perceptual illusion theory may stand to
explain certain aspects of the paradox of fiction (at least for visual
artworks).
Further reading: Fish 2009 & 2010; Lopes 2015; McMahon 1996;
Stokes 2014.
2.4 Against the third proposition
Many philosophers opt to eliminate the third proposition of the
paradox. There are several ways to do this. First, one can deny that
beliefs are a necessary component of emotions, but still maintain a
cognitivist position that emotions are comprised of thoughts (Carroll
1990 & Lamarque 1981) or judgments (Solomon 1976 [1993]). For
example, when we engage with fiction, we generally have various
thoughts about the characters. While watching
The Conjuring
,
I may contemplate the nature of the demon that possesses one woman.
This thought fills me with terror. Importantly, thoughts do not have
the same assertoric requirement that beliefs do; we do not need to
believe that the object of our thought exists in order to contemplate
and respond emotionally to it.
Alternatively, one can deny that
any
higher-order cognition
is required for emotions. This is the route taken by non-cognitive
perception and feeling-based theories of emotions. According to these
views, emotion does not require that we have a thought, judgment, or
belief about an object in our environment. Conscious feelings, bodily
changes, or perceptions of those changes, constitute an emotion
(Goldie 2000; James 1890; LeDoux 1996; Prinz 2004a & b; Robinson
2005). Here, the ontological status of the emotion’s object is
irrelevant to whether the emotion itself is a stereotypical state; if
the feeling or perception of bodily changes is genuine, then the
emotion is as well.
Non-cognitive theories of emotion track with the phenomenological
experience of emotional responses to fiction described in
§2.1
.
in response to the first proposition of the paradox.
A folk psychology of emotions contends that emotions involve a
conscious feeling.
Prima facie
, we identify our emotions by
how they feel; sorrow, anger, joy, jealousy, pride, etc. all feel a
certain way to us. One could ask whether someone really does feel
sorrow over the loss of a family member if they never consciously felt
that sorrow. For our purposes, feelings are the qualitative bodily
responses that are typically consciously experienced (but see Prinz
2004a; Berridge & Winkielman 2003; Rosenthal 2008). Chocolate has
a particular conscious taste and red has a specific qualitative look;
similarly, emotions have a conscious qualitative character. As William
James (1890) noted, feelings put the “emotionality” in the
emotion, making it salient and important to our lives.
Philosophers, psychologists, and neuroscientists alike all generally
accept that emotions involve
some kind
of judgment. Emotions
are evaluative. When we have an emotion, it is because something in
our environment—or something that we think, remember, or
imagine—bears significance on our lives or the life of someone
we care about. This may be a very quick, automatic evaluation, like
when we suddenly fear a loud noise behind us or are afraid that we
will slip on an unseen staircase. Or the evaluation could be quite
complex, like when we experience jealousy towards someone in our
workplace.
In cases of fiction, the non-cognitivist about emotion would suggest
that our feelings and other bodily reactions to a terrifying monster
are sufficient evidence for genuine fear. These emotions are
evaluative in that they indicate and are responses to things we care
about. I feel elated when Frodo finally throws the ring into Mount
Doom because I care about the narrative and character. Still, one
could argue that we do not have the right kind of evaluative
relationship with fictional objects to justify that we have genuine
emotions about them. Fictional characters may not be the kind of thing
that we can genuinely care about, empathize with, feel sympathy for,
etc. One might respond that we
seem
to care about and
identify with fictional objects all the time. We feel very strongly
for our favorite television, film, and literary heroes. We want them
to succeed, and we feel frustrated, sad, or angry when they do not.
This perspective ties back to the epistemic significance of emotional
responses to fiction: whether it is appropriate or rational to care
about fictional entities to begin with.
Further reading: Helm 2010; Huebner et al. 2009; LeDoux 1996; Loaiza
forthcoming; Zajonc 1984.
3. Further puzzles
While the paradox of fiction still holds sway in contemporary research
and thought, other interesting philosophical puzzles concerning
fiction and emotions have gained prominence over the past several
decades. This section introduces a few of them: the puzzle of
imaginative resistance, the “sympathy for the devil”
phenomenon, and the paradox of painful art. While not strictly puzzles
concerning emotions, the puzzle of imaginative resistance and sympathy
for the devil phenomenon is nevertheless related to emotional
responses to fiction, insofar as emotions such as disgust, anger, or
pride may coincide with, cause, or even constitute moral judgments
(Prinz 2007; Nichols 2004; Slote 2009; Schroeder 2011; Gill 2007).
3.1 The puzzle of imaginative resistance
Near the conclusion of “Of the Standard of Taste”, David
Hume makes several remarks on the moral status of fiction that are
particularly relevant when thinking about emotional and moral
responses to fiction:
[Where] the ideas of morality and decency alter from one age to
another, and where vicious manners are described, without being marked
with the proper characters of blame and disapprobation; this must be
allowed to disfigure the poem, and to be a real deformity. I cannot,
nor is it proper that I should, enter into such sentiments; and
however I may excuse the poet, on account of the manners of his age, I
never can relish the composition. And whatever indulgence we may give
to the writer on account of his prejudices, we cannot prevail on
ourself to enter into his sentiments, or bear an affection to
characters, which we plainly discover to be blameable.
The case is not the same with moral principles, as with speculative
opinions of any kind. These are in continual flux and revolution. The
son embraces a different system from the father. Nay, there scarcely
is any man, who can boast of great constancy and uniformity in this
particular. Whatever speculative errors may be found in the polite
writings of any age or country, they detract but little from the value
of those compositions. There needs but a certain turn of thought or
imagination to make us enter into all the opinions, which then
prevailed, and relish the sentiments or conclusions derived from them.
But a very violent effort is requisite to change our judgment of
manners, and excite sentiments of approbation or blame, love or
hatred, different from those to which the mind from long custom has
been familiarized (Hume 1757a, paragraph 32–33 [1994:
90–91]).
Hume’s statement here has been taken to capture an interesting
puzzle about moral responses to fiction, what Tamar Gendler (2000)
termed
the puzzle of imaginative resistance
. Although we may
be willing to accept factual or metaphysical discrepancies in fiction,
we may be loath to accept deviant moral values and practices that are
treated positively by the work.
Several philosophers have pointed out that the puzzle of imaginative
resistance if it is to be considered a puzzle at all (see Walton
2006), should be thought of as several discrete puzzles (Walton 2006;
Weatherson 2004). The first is the
aesthetic puzzle
: if an
artwork in some way embodies moral defects, do those defects detract
from the aesthetic value of the work? Walton believes that this puzzle
may be only indirectly related to moral resistance (Walton 2006).
Second, the
fictionality puzzle
states that there are certain
propositions can or should be made fictional. Walton writes:
We easily accept that princes become frogs, or that people travel in
time, in the world of a story, even, sometimes, that blatant
contradictions are fictions. But we balk…at interpretations of
stories of other fictions according to which it is fictional that
(absent extraordinary circumstances) female infanticide is right and
proper…or that a dumb knock-knock joke is actually hilarious.
(2006: 140).
The fictionality puzzle concerns any sort of value judgment, not just
moral ones. People may often deny that a value that they reject in the
real world is correct in the fictional world (or vice versa). We may
refuse to accept that the dumb knock-knock joke could possibly be
funny, even in a fictional world. We may also be unable to accept that
female infanticide is morally permissible in another world because we
don’t believe that it is in ours.
Finally, the
imaginative puzzle
does not concern what is or
is not fictional, but rather what we can or can’t imagine to
begin with. We might be able to imagine a situation in which female
infanticide is morally acceptable, even if we do not accept that it
this could ever be fictionally true. Alternatively, we might not even
be able to imagine that female infanticide is morally acceptable. We
are unable to conceive of a world in which it is morally acceptable to
kill one’s female child because she is female. In other words,
the fictionality puzzle concerns what we can
accept as true
in the world of the work. The imaginative puzzle concerns the limits
of our imagination.
Philosophers have responded to the puzzle(s) of imaginative resistance
in several ways. Gendler (2000) argues that there are two basic ways
to explain the imaginative puzzles: we can be “cantians”,
“wontians”, or some hybrid of the two.
Cantians
about resistance argue that we are often
unable
to imagine
certain kinds of impossibilities or evaluative deviances.
Wontians
hold that resistance arises because we are unwilling
to imagine a situation in which a certain impossibility or deviance is
acceptable. Gendler argues that imaginative barriers arise when the
principles and background knowledge the reader has accepted in the
story leave no way for the impossible or deviant proposition or
situation to be true (Gendler 2006). This makes Gendler a cantian
about the imaginative puzzle. We are unable to imagine some deviant
moral scenarios. However, Gendler is a
wontian
concerning the
fictionality puzzle. Even if we could imagine some deviant evaluative
response in a fiction, we often
will not allow ourselves
to
believe that the value judgment is true in the fiction. That is, we
may
be able to
imagine that some moral or aesthetic value
could conceivably differ from that we hold in the real world, but to
actually imagine that, for instance, female infanticide is morally
acceptable even in a fictional world would make some readers balk.
Walton takes the opposite approach: whereas Gendler is a cantian about
the imaginative puzzle and a wontian about the fictionality one, he
argues for the reverse (Walton 2006). I may not imagine a solid gold
mountain or a round square, because I have an inability to imagine
such a thing. The difficulty in imagining in these cases has to do
with one’s imagistic and conceptual limitations. I do not
imagine female infanticide is right, because I am
unwilling
to do so. So, Walton is a wontian when it comes to the imaginative
puzzle (see also Moran 1994). Graham Priest (1997), another wontian,
argues that we can understand stories that contain inconsistencies
like both occupied and unoccupied boxes; if we do not imagine them, it
is because we are unwilling to.
Further reading: Black & Barnes 2017; Flory 2013; Levy 2005; Liao
2016; Miyazono & Liao 2016; Nanay 2009; Tooming 2018; Tuna 2020;
Stock 2005; Yablo 2002
3.2 Sympathy for the devil
Why do consumers of fiction find themselves drawn to morally deviant
characters, whose real-life counterparts we would find abhorrent?
Following Noël Carroll (2004 & 2008), we call this the
sympathy for the devil phenomenon
(hereafter, SDP). The SDP
covers any of our pro-attitudes towards immoral or unlikeable
fictional entities including, but not limited to, sympathy. Other
pro-attitudes include emotions such as admiration, compassion,
empathy, pity, pride, and joy. We may admire morally deviant
characters for their wily ways or feel compassion for them once we
learn of their difficult upbringing. Sometimes audiences come to feel
pro-attitudes toward an anti-hero, a hero who is characterized by
moral and personal flaws, but nevertheless is shown throughout a text
or film to have sympathetic characteristics (Carroll 2008). Anne Eaton
also describes the “Rough Hero”: a flawed protagonist
whose flaws “are always moral, conspicuous and grievous”
and whose flaws, in contrast with the anti-hero, are “an
integral part of his person” (Eaton 2010: 516). It is often an
aesthetic achievement, according to Eaton, for an artwork to get an
audience to feel pro-attitudes toward Rough Heroes, whose character
and actions ought to be reviled.
Philosophers have responded to the SDP in a variety of ways. One way
to explain the SDP is to adopt a simulation/distinct attitude
approach, along the lines of Gregory Currie (1997). One this view, we
imagine or simulate moral propositions and judgments that we normally
would not in our actual lives. This would allow us to feel (imaginary
or make-believe) sympathy for a morally bad character, such as
Milton’s Satan,
Mad Men
’s Don Draper, or
Breaking Bad
’s Walter White.
Matthew Kieran (2006 & 2010) points out that we can sympathize
with an immoral character because we suppose that the character
inhabits a fictional world that is quite different from our own. This
fictional world encompasses a different land with different rules,
including moral rules. Call this the
distancing approach
.
Kieran contends that imaginative distancing amounts to a psychological
distance between an audience and a devilish character (Kieran 2010).
We feel free to allow ourselves to feel pity, compassion, and sympathy
for someone like Hannibal Lecter or Milton’s Satan because of
this psychological distance.
Carroll’s own view is that authors of fiction intend for their
audiences to sympathize with a fictional character, and so create
their work in such a way to achieve this end. Carroll calls this
process “emotional prefocusing” (Carroll 2008; see also
Smuts 2014). An author may intend for her audience to feel
pro-attitudes toward a particular character. This character is often
morally corrupt or deviant in some way but, for whatever reason, the
author desires the reader to sympathize with her. So, the author
“prefocuses” the work to highlight some of the
character’s more morally positive attributes or goes to lengths
to suggest ways that their behavior might be justified or excused. Our
emotional responses to characters are often a matter of how a
narrative is constructed and how the details of a story influence our
feelings about particular characters. Carroll suggests that the reason
we feel sympathy for Tony Soprano of
The Sopranos
, Tyrion
Lannister of
A Song of Ice and Fire
, or Ethan Edwards in
The Searchers
is because, despite their flaws, they are
morally better off than the other characters in the fiction. Tony is
surrounded by an astonishing array of violent, manipulative,
power-hungry mobsters. Tyrion is a clever, witty, well-meaning louse
with rotten family members. Ethan Edwards is gruff and brutal, but
also loyal and in possession of a certain code of honor. So, when we
search the fictional world for an emotional allegiance, these are the
characters we choose.
Another view holds that
fascination
is the key to
understanding our pro-attitudes towards immoral, fictional characters
(see also M. Smith 1999). Katie Tullmann (2016) argues that immoral
characters in works of fiction are often attractive, interesting
curiosities. Immoral characters are compelling, often more compelling
than the more morally good characters. Take the Count of Monte Cristo:
a revenge-driven, cruel, mysterious, and charming noble. Many readers
probably think that the Count’s tactics of carefully planned
revenge against those who had him imprisoned are not morally
justified. Still, audiences are fascinated by immoral actions
(consider also the recent popularity of the true crime genre). Readers
are willing to imaginatively explore immoral actions in the safe
environment of fiction (see Mar 2018 for a similar view). Perhaps we
think that by taking an interest in this character we can expand our
folk psychology to include the vengeful and obsessive mindset the
Count represents. So, on this view, fascination is the pre-condition
for our sympathy towards immoral characters. Audiences need to be
fascinated by immoral characters before we feel sympathy for them.
Fascination is achieved by how the character is portrayed in the
narrative as possessing exotic and curious traits. Once this is
achieved, other aspects of the narrative will cause us to feel
sympathy for them.
Further reading: Clavel-Vázquez 2023; Dain 2021; Friend 2022;
Harold 2008; Setiya 2010.
3.3 Paradox of painful art
The ability of fiction to engage with an audience’s emotions is
irrefutable. We seek out certain works of fiction for exactly this
reason: to make us joyful, to make us laugh, to make us feel connected
to others around us. Less intuitive, but no less deniable, is the idea
that audiences seek out works of fiction that make them feel negative
emotions such as sorrow or fear. Consider Hume, again, this time on
tragedy:
It seems an unaccountable pleasure, which the spectators of a
well-written tragedy receive from sorrow, terror, anxiety, and other
passions, that are in themselves disagreeable and uneasy. The more
they are touched and affected, the more are they delighted with the
spectacle; and as soon as the uneasy passions cease to operate, the
piece is at an end (1757b, paragraph 1 [1987: 216])
The “paradox” of painful art is just this: we seek out
works of art with the intention of feeling negative emotions such as
anger, sorrow, or fear, when we wouldn’t seek out comparable
real-life situations. Indeed, people try to avoid these negative
emotions as much as possible. Hume himself argued that the pleasures
of the imagination and emotional expression outweigh the negative
emotional responses such as sorrow—the former
“predominates” over the latter. Susan Feagin (1983) argues
that there are two types of (emotional) responses to fiction: direct
responses and meta-responses. Feagin defines the meta-response as
“how one feels about and what one thinks about one’s
responding (directly) in the way one does to the qualities and content
of the work” (1983: 97). Direct responses to tragedies are often
negative: sorrow over the death of a beloved character; anxiety about
the future of another character. The
pleasure
of tragedy,
Feagin argues, stems from our meta-response to our direct, negative
responses: we receive satisfaction from the fact that we are the kind
of people that can feel anger, sorrow, fear, etc., and can respond
negatively to moral ills in works of fiction.
Carroll (1990) posits another response to what he calls the paradox of
horror: if horror fictions are characterized by fearful monsters and
disgusting circumstances, why do many of us seek them out? Carroll
states: “the imagery of horror fiction seems to be necessarily
repulsive and, yet, the genre has no lack of consumers” (1990:
160). Like other forms of tragedy, works of horror fiction are both
repulsive and attractive. One potential explanation for this puzzle is
that audiences can experience all the terror of horrifying situations
but with none of the personal risk (1990: 167). The emotion is
“detached” from reality. Carroll’s own account is
that works of horror are constructed in a way to engage other
emotions, in addition to fear and disgust—namely, fascination
with the imagery and plot structures that surround the fearful
elements.
Each of the additional puzzles concerning emotions about fiction
highlights the curious ways in which an audience’s emotional
responses to fiction belie one’s endorsed moral beliefs about
the actual world in which we live. The puzzles continue to hold sway
in our cultural understanding of the importance of fictional artworks.
Relationships with fictional entities are often deeply important to
us—consider the widespread emotional responses to scenes like
the “Red Wedding” in the
Game of Thrones
television show, or the death of a certain prominent character in the
final season of the show
Succession
. Such responses implicate
and reveal our deepest values and beliefs. It is no wonder, then, that
philosophers continually try to explain and, indeed, justify these
emotions’ appropriateness and authenticity.
Further reading: Bantinaki 2012; Contesi 2016; Gaut 1993; Friend 2007;
Schwarz 2022; Smuts 2009.