Erkenn (2014) 79:779–796
DOI 10.1007/s10670-013-9563-z
ORIGINAL ARTICLE
Does the Paradox of Fiction Exist?
Katherine Tullmann • Wesley Buckwalter
Received: 19 March 2013 / Accepted: 11 October 2013 / Published online: 23 October 2013
Ó Springer Science+Business Media Dordrecht 2013
Abstract Many philosophers have attempted to provide a solution to the paradox
of fiction, a triad of sentences that lead to the conclusion that genuine emotional
responses to fiction are irrational. We suggest that disagreement over the best
response to this paradox stems directly from the formulation of the paradox itself.
Our main goal is to show that there is an ambiguity regarding the word ‘exist’
throughout the premises of the paradox. To reveal this ambiguity, we display the
diverse existential commitments of several leading theories of emotion, and argue
that none of the theories we consider are committed to notions of ‘exist’ employed
by the paradox. We conclude that it is unclear whether or not there remains a
paradox of fiction to be solved—rather than to be argued for—once this ambiguity is
addressed.
1 An Emotional Introduction
Suppose you are watching a film and are moved to tears by a protagonist’s horrible
plight. Imagine sitting around a campfire, shaking in fear from a scary ghost story,
or laughing uproariously at the jokes of a Saturday morning cartoon. It seems as
though we frequently have genuine emotional responses to fictions in precisely
these ways. But now examine your beliefs about the very same protagonist whose
situation you previously found so heartrending, or the ghost whose aura you found
K. Tullmann
Department of Philosophy, Graduate Center, City University of New York, New York, NY, USA
e-mail: katherine.tullmann@gmail.com
W. Buckwalter (&)
Department of Philosophy, University of Waterloo, 200 University Avenue West, Waterloo, ON
N2L 3G1, Canada
e-mail: wesleybuckwalter@gmail.com
123
780
K. Tullmann, W. Buckwalter
so frightening. While the goose bumps are real, their elicitors are not. We don’t
believe that the characters of fictions actually exist.
These purported facts about human behavior give rise to what is known as the
paradox of fiction. This paradox, initially formulated by Radford (1975), is
comprised of a triad of inconsistent, yet highly intuitive propositions. First, some
emotional responses to fiction count as genuine emotions. Second, we don’t believe
that the objects of those genuine emotions—the fictional entities—actually exist.
And third, we are only rationally moved by what we believe actually exists. Or in
other words, having a genuine emotional response implies a belief in the existence
of a fictional thing. Radford argues that all three of these premises are true. He
concludes that emotional responses, such as sadness for our protagonist or fear on
account of our ghost, are irrational through and through.
Aestheticians have pursued a number of approaches to rescue people’s ordinary
emotional responses to fiction from this threat of irrationality. Quite often, these
rival approaches reflect substantive philosophical differences regarding the nature
and cognitive science of emotion. And debate among researchers regarding the
nature and structure of emotions is certain to continue. So while defenses and
critiques of the paradox have been numerous, the correct response to Radford’s triad
remains the source of continued and ever broadening controversy. And though in
recent years certain neo-Jamesian theories of emotion have become popular, no one
theory or perspective has emerged the clear victor.
We suggest that the correct response to the paradox of fiction need not reflect any
substantive philosophical differences regarding the nature or cognitive science of
emotion. Nor do we think the correct response need involve defending the
controversial denial of one or more of Radford’s highly intuitive premises. Instead,
the main goal of our paper is to show that current disagreement among philosophers
about the proper response to this paradox stems from the formulation of the paradox
itself. In short, we will argue that despite one’s theoretical commitments, there is
good reason to think there is no such thing as the paradox of fiction.
In the remainder of the paper we defend this bold claim. In Sect. 2 we will
display the purported paradox of fiction, and briefly review a number of influential
responses to it. In Sect. 3, we will identify what we believe to be the central mistake
in the formulation of the paradox, an ambiguity regarding the word ‘exist’ in the
following two premises:
(1)
(2)
One does not believe that fictional entities exist.
A genuine emotion in response to fictional entities implies that one believes
that fictional entities exist.
In Sect. 4, the longest and most important section, we review the commitments of
several leading theories of emotion, and argue that according to these theories, a
uniform notion of ‘exist’ has yet to be deployed in (1) and (2). We conclude in Sect.
5 with a discussion of the implications that this finding has for future debates
concerning the paradox of fiction. We argue that it is unclear whether or not there
remains a paradox of fiction to be solved—rather than to be argued for—once the
ambiguity we point out is addressed. Consequently, we suggest that if we are correct
then attending to Radford’s premises may not be a particularly fruitful way to
123
Does the Paradox of Fiction Exist?
781
answer the central philosophical questions typically of interest to philosophers in
this domain: the ways in which fictions and fictional objects exist, and what
emotions are able to tell us about them.
2 The Paradox and Responses
The paradox of emotional responses to fiction stems from an inconsistent triad
comprised of Radford’s three premises (adapted from Radford 1975):
(PF1) We are genuinely moved by fictions. Emotions in response to fictional
things can be genuine.
(PF2) We do not believe that fictional entities exist. We believe that what is
portrayed in fictions is non-actual.
(PF3) A genuine emotion in response to fictional entities implies that one believes that
fictional entities exist. We are only genuinely moved by what we believe is actual.
For example, when Frodo and Sam finally reach Mt. Doom in Tolkien’s Lord of the
Rings, one may feel relief, exultation, or pride at their accomplishment.
Furthermore, suppose it’s indeed possible that these emotions are genuine tokens
of relief, exultation or pride, and not, quasi or pretend emotions. Even though we are
affected by Frodo and Sam’s story, we know that these characters are not real—they
cannot be found in the actual world. Their journey stems only from the clever
imagination of the author of the fiction. So, if we believe that Frodo and Sam are not
real, why do they affect us at all? The answer, Radford suggests, is that these kinds
of emotional responses to fictional characters are irrational.
Most theorists reject Radford’s conclusion that these kinds of responses are
somehow irrational or inappropriate. Instead, traditional solutions to the paradox of
fiction have proceeded by rejecting one or more of the propositions above. Many
theorists have rejected (PF1) by denying that emotional responses to fictions are
genuine emotions. Perhaps the most well known response along these lines is the
make-believe theory proposed by Walton (1978). In ‘‘Fearing Fictions,’’ Walton
utilizes a complex analogy of a child’s game of make-believe or pretend-play. When
children play ‘house’ for instance, they generally designate certain objects as
something else. And similarly, when engaging in fictions, it might be appropriate to
say that someone ‘‘pities’’ Anna Karenina or ‘‘admires’’ Superman even if what they
say is not literally true ((1978), 9). Rather, it is only a matter of make-believe within
the context of the story.
But theories of make-believe have been heavily criticized because of the
important differences between prototypical instances of pretend play, on the one
hand, and the character of our everyday emotional responses to fiction, on the other.
For instance, Carroll (1990) argues that unlike the make-believe children often
engage in, our emotions are not easily turned on and off at will. When we are
emotionally engaged in a particular work of fiction like a horror film, it seems we
often have very little control over the fear or terror that we feel. And similarly, that
experience has a very different phenomenology than pretend-play (Hartz 1999).
123
782
K. Tullmann, W. Buckwalter
While the latter notion of pretence presupposes some kind of awareness of a game
being played, our emotional responses often do not. So the burden of the pretense
theory is to somehow explain these differences, as well as the possibility of severe
and lasting emotions in response to fictions that feel perfectly real.
Another well-traveled option for dismantling the paradox involves denying the
second proposition (PF2). Often theorists who endorse this approach advocate
illusion, or ‘‘suspension of disbelief’’ arguments (see Coleridge 1985/1817; Hurka
2001). The basic idea is that for the time in which we engage with a particular
fiction we do believe that the fictional entities are real. While reading Lord of the
Rings for instance, illusion theories hold that we temporarily believe that Frodo’s
world exists, and this illusion is responsible for our genuine emotional responses to
fiction. But of course, one straightforward objection to this view is that no matter
how caught up we are in fiction, it seems that we never actually believe that what we
experience is real. If asked about a film later, we never claim that there was some
point at which we stopped believing that the story was fiction, or that the monsters
of a horror film might actually be able to harm us. And at no point were we inclined
to act in the way appropriate for an encounter with a harmful object. We didn’t run
from the theatre or call the police. Nor did we warn our beloved that they were in
danger. As Katherine Thomson-Jones 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).
Thus theorists who support illusion approaches must somehow square suspension of
disbelief arguments with the contrary beliefs and actions people seem to have in
response to fictional characters and situations.1
Based on the challenges faced by these early attempts rejecting (PF1) and (PF2),
most philosophers now attempt to solve the paradox by rejecting the last premise. In
fact, it seems that most of the contention surrounding the paradox arises from
accepting the following trenchant requirement of emotions: we must believe that the
objects of our emotions exist, and that they can affect us in some way in order to be
genuine. And of course, one might think that this is not the case with fictions. Thus
many philosophers have tried to obviate this intuitive requirement of emotions by
arguing—in denial of the third premise—that beliefs of this sort are not necessary
for emotion.
There are two basic approaches for denying that genuine emotional responses
require belief in the existence of fictional things in (PF3). First, non-cognitivists
about emotion—or theorists who deny that cognitive content is a necessary
component of emotions—can simply reject the claim that emotion requires beliefs.
This is the route taken by philosophers such as Robinson (2005) and Prinz (2004a),
1
For more on motivation and the role of behavioral circumscription in theories of belief see Rose et al.
(under review) as well as Tullmann (in prep), Buckwalter et al. (2013) and Buckwalter and Turri (under
review).
123
Does the Paradox of Fiction Exist?
783
who propose non-cognitivist theories of emotions.2 The other approach accepts that
emotions require a cognitive component, but rejects that this component must
involve beliefs. Carroll (1990) favors this solution, arguing that our emotional
responses to fictions presuppose our thought contents about the fiction, rather than
our beliefs about the fiction.
Proper discussion of these responses to (PF3) is beyond the scope of this article.
Nonetheless, one should note that these approaches require one to accept accounts
of emotion that are not universally accepted.3 There is little doubt that a correct
analysis of emotion in modern cognitive science and aesthetics is less forthcoming
than a solution to the paradox itself. Since philosophers are roughly split on the
question supporting cognitive or non-cognitive theories of emotion, methods
involving the denial of (PF3) are even less likely to reach theoretical consensus in
debates about the paradox of fiction.
Lack of consensus above might tempt us to bite the bullet. Radford’s response to
the puzzle was to simply accept the irrationality of our emotional responses towards
fictions. As Radford concludes:
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).
Radford likens this to other everyday situations such as when ‘‘a tennis player who
sees his shot going into the net will often give an involuntary jump to lift it over.’’
Like the tennis player, we often encounter situations in which we are aware that our
beliefs and actions do not match up to reality in the right way. According to
Radford, our emotions concerning fictional characters are one of these instances,
reflecting an important aspect of our cognitive and behavioral makeup.
But we are far from convinced that the appearance and persistence of the paradox
arises as a result of these sorts of behavioral phenomena. Nor do we think that the
solution rests in the responses we have reviewed above. Instead, we think the
paradoxical nature of the premises arises due to an ambiguity in the language
philosophers have used to formulate and respond to this purported paradox. In the
next section, we attempt to reveal this ambiguity.
3 The Ambiguity of ‘‘Existence’’
Our focus is on the use of ‘exist’ throughout the triad. It appears in two premises.
Specifically, (PF2) holds that one does not believe that the fictional objects of our
emotions actually exist. On the other hand, (PF3) stipulates that having a genuine
emotional response implies that one believes the object of that emotion exists. But in
2
Throughout the remainder of the paper we use ‘non-cognitivism’ to label those theories that reject
cognitive content strictly as a necessary condition of emotion. We note however that Prinz and Robinson
do allow for emotions that are evaluative and intentional—and are sometimes thought of as ‘cognitive’ on
this basis.
3
For instance many allege that thought-based theories are overly cognitive theories of emotion, while
others argue that neo-Jamesian accounts are not cognitive enough (see Turvey 1997).
123
784
K. Tullmann, W. Buckwalter
what way do these fictional things exist, and is the same notion of existence being
employed across both of these premises?
From Aristotle to Frege and Kripke, the history of philosophy is shot through
with discussion of existence. And especially since Frege, a considerable amount of
work in the philosophy of language has gone into giving an analysis of existence, its
multiple senses, and the similarities and differences it bears to what is expressed by
the word ‘is’ (see Williams 1981; Miller 1986; Zalta 1988). Using examples like
Sherlock Holmes or Spock, a number of influential philosophers have even argued
that fictional characters do constitute existing entities in the sense that they are
abstract artifacts (Salmon 2002; van Inwagen 1977; Thomasson 1999). But despite
this tradition, such scholarship is typically not reflected in contemporary discussions
in the philosophy of art concerning the paradox of fiction, or in cognitive science
regarding the objects of emotion. Instead, existence (and ‘existence’ for that matter)
is usually left unspecified in these discussions. In fact, it seems little has been said
about the existential nature of the objects of our emotions, both in the literatures on
fiction and emotions.
Perhaps one reason the salient sense of existence being employed in the paradox
of fiction has gone underspecified is because it is meant to capture an ordinary or
non-technical notion. Consider for instance the proposition ‘‘One does not believe
that fictional entities exist’’ in (PF2). As far as we know, very few who have written
on the paradox have explained precisely what they mean when they use the word
‘exist’. But we are left with a few obvious candidates. For instance, on an initial,
commonsense view, it could be that ‘exist’ means something like is a concrete
object. We say that people, trees, paintings, etc. exist because they are currently
composed of matter. Of course, some cases are not so straightforward: consider
electrons, quarks, and musical works, not to mention fictional characters. It is an
obvious point that fictional entities lack a certain kind of physical body. If that’s so,
then it’s probably also the case that this is not the sense in which ‘exist’ is used in
(PF3). But this isn’t a trivial point. Consider how many philosophers handle (PF3).
They reject that audiences suspend their disbelief concerning fictions because if they
did, the audience members would act differently towards fictional objects—they
would run screaming out of the theatre upon discovering that zombies or monsters
occupy their town, for example. A suspension of disbelief in the existence of
zombies implies that we pretend or imagine that zombies exist in the sense
discussed here—concretely. The fact that this view is taken seriously (for example
by such luminaries as Coleridge and Hurka) means that it is worth considering as an
option in the paradox.
Alternatively, ‘exist’ could mean might have existed concretely. This is, in fact, a
variation on the position held most notably by Lewis (1978) and Plantinga (1974)—
fictional entities are possible objects that do not exist in our world, but do exist in
another possible world. In that sense Emma Woodhouse might have existed in the
actual world had the contingent historical events turned out differently. Similarly,
one might think that for all we know Sherlock Holmes could exist, just like black
swans and white crows.
Needless to say, this view has been widely critiqued in the literature on the
ontology of fictional entities (see especially Quine 1953; Thomasson 1999 and
123
Does the Paradox of Fiction Exist?
785
Sainsbury 2010). It is generally accepted by those who think that fictional entities
exist at all (perhaps except for Meinongians) that they are created and causally
dependent. There is no Sherlock Holmes in the actual world, and even if there were
a Sherlock Holmes in a possible world that matched each of the detective’s known
properties, he would not be the same Sherlock Holmes as the one in Sir Arthur
Conan Doyle’s stories. Because of such considerations, we may be tempted to reject
this sense of ‘exist’ out of hand. Nevertheless, we encourage the reader to keep in
mind that this is a plausible notion of how various things exist, and a viable option
in the paradox.
Lastly, we get the sense that when theorists speak of fictional objects in the
paradox, they are suggesting that to be fictional is to be an imaginary object. What
we mean by ‘imaginary’ is simply the inverse of the first two candidates–an object
that is not a concrete particular or even possibly so, in the two senses described
above. This is the most plausible notion of ‘exist’ used in reference to fictional
entities, but also the vaguest. After all, what does it mean for a fictional character to
be imaginary? When we think of an imaginary friend, for instance, we mean
something like ‘‘not having a physical body’’ and ‘‘made up by me’’. An imaginary
friend exists in the sense that it is created by a child in the actual world, is not
corporeal, nor merely a concrete particular in a possible world.
Of course, there are a number of positions philosophers might take on imaginary
objects: whether or not fictional character are real in any sense, whether they are
created or have always existed, and how we use them. While we do not want to
endorse any one particular theory of the ontological status of fictional characters, we
do wish to mention that one major component typically included by theories of
imaginary objects is an account of how they bear meaningful causal relations to
their imaginees—or real human beings. Often, this is said to work through some
kind of simulation theory (Currie 1989) or theory of make-believe (Walton 1993;
see also Kieran 2010) as described above. But one need not posit a theory of makebelieve. As Carroll argues, it seems we can engage in and justify actions, as well as
form propositional attitudes, in regards to non-assertive thoughts about fictional
entities (1990). Alternatively, another possibility is that our responses can result
from unconscious or non-cognitive triggers in our environment, including fictions,
as argued by non-cognitive theories of emotion (see Sect. 4). So a final way to think
about ‘existence’ in the triad is in terms of ‘imaginary’, together with some account
of the causal relationship between imagination on the one hand, and our thoughts
and actions on the other.
Taking stock, we certainly don’t think that any of these suggestions will yield
particularly full-fledged metaphysical theories of existence.4 We’re simply trying to
understand what philosophers of art typically have in mind when they commit to
(PF2). And on that score, we think we’ve identified three relatively good, albeit
basic, candidates:
(3)
(4)
One does not believe that fictional entities are concrete particulars.
One does not believe that fictional entities might exist concretely.
4
To be clear, we remain theory-neutral on the ‘‘reality’’ of fictional entities and reference to them, except
insofar as they relate to objects of emotions.
123
786
K. Tullmann, W. Buckwalter
(5)
One does not believe that fictional entities are non-imaginary (that they exist
concretely, might exist concretely) or are other than imaginary objects.
Our next task is to take these candidates for ‘exist’ and deploy them uniformly for
occurrences of ‘exist’ throughout the paradox. Specifically, we would like to know
how these candidates affect the conditional premise that a genuine emotion in
response to fictional entities implies that one believes that fictional entities exist in
(PF3).
4 The Objects of Theories of Emotion
In the previous section, we’ve proposed three candidates for understanding ‘exist’ in
(PF2). To clarify the sense of ‘exist’ in (PF3) however, we need to know more about
what it means to undergo an emotion. While there is a wide spectrum of emotion
theories advocated in the literature, these accounts are typically classified into three
groups. On one end of this spectrum are pure-cognitive theories. As the name
suggests, these theories maintain that having a belief, judgment, or thought is both
necessary and sufficient for genuine emotion.5 At the other extreme of the emotional
spectrum are the non-cognitive theories. Such theories hold that feelings, bodily
changes, or perceptions of those changes are necessary and sufficient for genuine
emotions. Finally, somewhere in the middle lie hybrid theories. These theories hold
that both feelings of some kind as well as some kind of intellectual, or cognitive
state are necessary for a genuine emotion.6
Our task is to specify the notion of existence at work in premise (PF3). In order to
understand this, we must now determine which sense of ‘exist’ is presupposed by
each type of theory on the emotional spectrum. We will focus on one leading theory
from each general type: feeling (non-cognitivist), judgment (hybrid) and beliefbased (pure-cognitive) theories and examine their commitments about emotional
objects in this debate. In many cases, the same conclusions can be drawn across the
general type of theories as well.
Before proceeding, we should point out that it is actually no easy matter
determining how an object of an emotion must exist for each type of theory. The
reason is that authors rarely spell out the existential commitments of the objects of
their emotive theories. And when they do, it’s not clear in what sense of ‘exist’ they
employ. Nevertheless, what we will attempt to show is that, for each of these
theories, a consistent use of ‘exist’ has yet to be applied across the following two
statements:
(6)
(7)
One does not believe that fictional entities /s.
A genuine emotion in response to fictional entities implies that one believes
that fictional entities /s.
5
Interestingly, it is this kind of theory that the paradox of fiction was traditionally construed to address—
emotions are caused or constituted by beliefs. See also footnote 2.
6
This terminology is based on Prinz’s characterizations of theories of emotions in Gut Reactions
(2004a).
123
Does the Paradox of Fiction Exist?
787
In short, we think it is unlikely that any of the types of theories we consider
below apply the likely candidates we have proposed for ‘exist’ in (3–5) uniformly in
(6–7).
4.1 Feeling Theories
As the name suggests, feeling theories describe a type of non-cognitive position that
emphasizes both the conscious feel of emotion, and often also includes the
conditions of bodies that undergo them. William James, one of the first prominent
proponents of a feeling theory, famously stated that, ‘‘our feeling of the same
changes as they occur IS the emotion’’ (1884). In other words, the felt bodily
changes that we experience in response to various stimuli in our environment
constitute an emotion, rather than any thought or belief about that object. We
experience fear, for example, when we feel the hair stand up on the back of our
neck, our heart pounding in our chests and our hands clutched beside us. We
experience joy by a feeling of lightheartedness and buoyancy of gate. Things like
beliefs, thoughts, or other propositional judgments are not required. Emotions are
primitive in this way—they will occur without reference to conscious, cognitive
mental states (see De Sousa 1987).7 Such theories are associated with the
philosophers Goldie (2004) and Stocker (1983), as well as the neuroscientist
LeDoux (1996). We will focus here on Goldie’s rendition.
Even though emotions may be characterized as ‘‘primitive,’’ most feelingtheorists also maintain that they are not ‘‘brute’’. Emotions don’t simply occur
randomly from any stimuli in one’s environment. Rather, emotions ‘‘tell us
something about the world and how to act in the world’’ (Goldie, 92). Indeed, this is
a basic theme across most theories, cognitive and non-cognitive. Emotions are
intentional or about the world and value-laden. It’s a common misconception that
non-cognitive theories of emotions, including feeling theories, are not intentional. In
contrast, Goldie maintains that the epistemic role of emotions is to reveal things
about the world, provide a subject with relevant information concerning its wellbeing, and gain insight on how to act.
Goldie calls this ‘‘extraspective’’ knowledge-information about the world around
us. Many objects have ‘‘emotion proper-properties’’—the property of being able to
cause a certain bodily state associated with an emotion (94). Some feeling theorists
define these properties in terms of ‘‘core relational themes’’ (Prinz 2004a, b). For
example, a grizzly bear is associated with the core relational theme of fear because
it can cause us harm. Importantly, for all non-cognitive theories of emotions, the
cause of the emotion need not be cognitive. We may only need to perceive an object
in our environment in order for it to trigger an appropriate emotion (a flight
response, for example).
However, emotions can also result in thoughts and introspective knowledge about
our bodily condition. I learn about my own bodily state by the way emotions feel
(93). This can give us information about how we personally respond to various
7
Many non-cognitivist theorists maintain that the associated bodily changes need not be conscious. See
Prinz 2004a.
123
788
K. Tullmann, W. Buckwalter
situations, as well as clue us in on which objects of situations to pursue or avoid.
Importantly for our purposes, neither introspective nor extraspective knowledge
need result in an assertive belief concerning the object, situation or state of affairs
that caused the emotion. We can be seriously wrong about what caused our fear, for
example. We might think, alone in our dark bedroom at night, that a quick-moving
shadow caused by a car passing outside of our window is actually a burglar
attempting to break in. Just the perception of this shadow is enough to cause an
emotion. But we need not have any thoughts or beliefs about what this shadow
actually is or what harm it can cause us in order for us to we feel fear.
This is the key to understanding how we can have genuine emotions about
fictions. We perceive objects on a screen, or in a play, or even in our imagination (as
when reading a novel) as possessing an emotion-proper property. The result is a felt
bodily change—an emotion. The feeling is real even if the object that caused it is
not, and so the emotion is genuine. This gives us an indication of how feeling
theories typically think about existence of emotional objects. Because emotions are
primitive in the way described, the object of an emotion need not be concrete
particulars, possibly concrete particulars or non-imaginary. In other words, a nonconcrete, possibly concrete, or imaginary object may cause our emotion by
possessing the right kind of emotion-proper property.8
What does this mean for the paradox? We gave three candidates in (3–5) about
what existence might mean in premise (PF3). And if those candidates are applied
consistently throughout the paradox, it would imply that an object of an emotion
must be either concrete or non-imaginary. But according to feeling theories, we can
have genuine emotions about all of these things. This is established quite
independently of the paradox of fiction. The paradox need not arise for these noncognitivist theories because none of the common-sense uses of ‘exist’ in (PF2)
reflect the feeling theorist’s use of ‘exist’ in (PF3).
4.2 Judgment Theories
The judgment theory of emotions is probably most associated with Solomon (1993),
(2004). Solomon, like the feeling theorists, argues that emotions are evaluations
about the world. Unlike the feeling theorists however, he posits that emotions
involve more than bodily reactions to stimuli, but also a cognitive judgment (or
judgments) of them. The judgment serves as a kind of appraisal, evaluating objects
of value to us in our environment.
While his early work reflected a pure-cognitive stance, Solomon’s later work
repeals that notion in favor of a hybrid view ((2004), 85). He still maintains that
judgments are necessary for emotions to occur, as well as to cause an emotion.
However Solomon also holds that these judgments can be both intellectual and bodily.
And perhaps most importantly, these judgments need not be conscious, reflective or
propositional. We might note that if a judgment is either bodily or intellectual (in some
8
Of course, the question of perceiving fictions becomes a bit more difficult when talking about multiple
genres, and for instance, if experiencing literature counts as direct perception (Goldie 100). We think that
a case can be made that imagining an emotion-proper property of an object in such cases is reasonable,
though will postpone this argument for another occasion.
123
Does the Paradox of Fiction Exist?
789
way), as well as not necessarily conscious, propositional or reflective, then it’s rather
difficult to see what exactly a judgment is at all. Solomon more or less sidesteps this
issue, instead focusing on an emotional judgment’s relation to a subject’s world. Like
Goldie, Solomon argues that emotions are an ‘‘appraisal of the world and our place in
it’’ (81). Emotions are about the world, an evaluative process that includes dispositions
to respond in certain ways to various stimuli, perceptions of bodily states and, often,
thoughts or beliefs about our environment.
Solomon’s characterization of emotions as ‘‘appraisal[s] of the world and our
place in it’’ seems to suggest that the relation between self and world required
cannot be met by fictions. But again, it’s unclear what exactly Solomon means by
‘‘world and our place in it.’’ Can we only have emotions regarding the actual,
present world? This doesn’t seem completely correct, since he also claims that
emotions can be appropriate appraisals of propositions, ordinary objects or
perceptions, and imaginary or remembered things (82). So ‘‘the world’’ must be
understood in a broad sense, including imaginary and plausibly, fictional entities.
This understanding of ‘‘world’’ gives us a relatively straightforward way to analyze
Solomon’s existential commitments in terms of fictions. A judgment theorist may
deny that emotional judgments must be about concrete or possibly concrete things
[statements (3) and (4)]. Solomon explicitly says that we can have emotions about
imaginary things (5).
How would an appraisal of a fictional entity work, according to this view?
Consider our reaction to Anna Karenina’s plight. How can we say that we genuinely
experience pity for Anna when she is not allowed to visit her son, for example?
Perhaps our judgment would be something like the following: we grow attached to
this character as we read about Anna’s trials, judge her to be a sympathetic character
and formulate different opinions about the other characters in the novel based on
this regard. As we ‘‘witness’’ Anna’s misfortunes, we formulate appraisals of her
situation much as we would with any friend with whom we sympathize in the actual
world who undergoes the same sorts of experiences. In a way, this appraisal
concerns us because it is a reflection of our own sentiments and tacit beliefs about
how women should be treated, what constitutes unjust behavior, etc. Since Solomon
maintains that emotions are also bodily, we may experience a variety of bodily
changes associated with pity, sorrow or anger about Anna and her peers which, in
part, constitute the emotion.
Again, Solomon never plainly states what it would mean for us to have an
emotional appraisal about fictions, so it’s difficult to say exactly how the judgment
process would work. But we think that something like the process outlined here is
consistent with Solomon’s view. So as we saw in the case of feeling theories of
emotion, the paradox of fiction need not arise for judgment theories similarly
construed to Solomon’s. We can have genuine emotions about fictions, where the
emotion is some sort of appraisal about the fiction that concerns us. We do not
believe that fictions are real, in either sense of concrete. We may have genuine
emotions about things that we don’t believe to be real because we may have genuine
emotions about imaginary things. The commonsense uses of ‘exist’ in (PF3) are just
not the same notion of ‘exist’ that the judgment theorists requires of emotional
objects in order for one to have a genuine emotion.
123
790
K. Tullmann, W. Buckwalter
4.3 Pure-cognitive Belief Theories
Kendall Walton, a pure-cognitivist, is one of the most out-spoken deniers that we
have genuine emotional responses towards fiction. Pure-cognitivist theories are
often expressed in terms of belief. Typically, belief-based theories of emotion hold
that our emotional responses are caused by the beliefs we have concerning objects in
our environment that bear on our well-being (see Nussbaum 2004). On this type of
view for example, one fears a tornado because one believes that the tornado may do
harm. For the belief-theorist this cognitive content is a necessary component of
emotions. Subsequently, belief-based theories like Walton’s robust view on fictions
are perhaps the most difficult type of theory for us to explain to reach the conclusion
that the paradox of fiction is really no paradox at all.
Before we begin however, we would like to point out that while Walton commits
to a hard line cognitivist theory, he does not himself propose any specific theory of
emotions over and above his work relating emotion to action. But since Walton
provides no specific theory of emotions, one natural question to ask here is whether
or not there is something about belief-based theories in general that forces one to
accept Walton’s claims about the existence of fictional objects tout court. Here, we
will argue that the answer is ‘no’ insofar as prominent belief theories today—such
as the one proposed by Martha Nussbaum—do not seem to be subject to the
existential commitments we have been considering.
In his famous ‘‘Charles and the Green Slime’’ example (1978), Walton argues
that Charles does not feel genuine fear about the oozing monster, even though he
displayed all the typically signs of fear—clutching his armrest or covering his eyes,
jumping out of his seat, etc. But if it looks like Charles fears the slime by displaying
all of the relevant behaviors associated with fear, then why shouldn’t we say that his
fear is real? Walton’s answer, of course, is that Charles does not have the right sort
of belief about the green slime in order to constitute a genuine emotion. Namely,
Charles does not believe that the slime can actually cause him harm. If he did, he
would surely act on this belief. In other words, emotions are motivating. And this is
the significant difference between emotions about the fictional world and those
about the real world. Real-world emotions motivate us to act in response to them—
in response to whatever object or state of affairs actually caused the emotion to
begin with. Charles was never motivated to act on his ‘‘quasi-fear.’’ He never runs
out of the theater, putting as much ground between him and the slime as possible.
He doesn’t call the fire brigade or army reserve to put the slime down once and for
all. These things simply do not happen in regards to fictions. Charles lacks the
motivation to act, and so is missing an important element of a genuine emotion.
Nussbaum’s ‘‘neo-Stoic’’ theory of emotions works in a similar way. Nussbaum
highlights four different features of an emotion. First, emotions must have an object.
Second, that object must be interpreted in a certain way. Third, emotions are based
on beliefs about the world. Fourth, these beliefs are value-laden (190). The picture
of emotions we get from Nussbaum places special emphasis on the idea that
emotions are evaluative and ‘‘eudaimonistic’’—they have to do with human
flourishing (189). An emotion occurs, or will occur, when we evaluate some object
123
Does the Paradox of Fiction Exist?
791
or state of affairs as having a ‘‘non-trivial’’ (188) bearing on ourselves or on those
we care about, and believe that this is so.9
But alas, the ambiguity in ‘exist’ can be found here as well. Walton and
Nussbaum do not clarify how they use this term. The general assumption in premise
(PF2) seems to be in line with (3) or (4)—that the object of an emotion must be a
concrete particular or possibly so.10 However, Nussbaum explicitly claims that the
object of a genuine emotion need not exist in the world, presently or ever. While this
idea is not developed in detail, we take Nussbaum’s position to be that these kinds
of entities nevertheless do instill in us the right sort of belief or elicit the right sort of
evaluation to constitute an emotion. Conversely, it seems like Walton would
adamantly deny this, claiming that entities that are not concrete, possibly concrete
or non-imaginary can never cause the right sort of evaluative belief in order to give
rise to genuine emotion.
So the crucial question for the emotion theorist who adopts the pure cognitivist
theory about belief is to decide which kinds of entities can give rise to the right sort
of belief for emotion. And despite Walton’s hardline stance, we think that, in
principle, it is possible that imaginary objects can cause genuine emotions under
this type of pure-cognitive theory. Importantly, as we saw in the example of Charles
and the Green Slime, the right kind of belief (according to Walton) has to with
motivation. The key is to determine what kind of existence an object must have in
order to achieve a belief that inspires the right kind of value or motivation to act.
This is illustrated in the following example:
Becky has bad luck in her romantic relationships. Recently she read Jane
Austen’s Pride and Prejudice for the first time. Becky is stuck by Austen’s
description of the depth, intelligence, sincerity, and, let’s face it, wealth and
physique of this fictional character. She imagines a person just like Mr. Darcy
that exists in her world, and has certain emotional responses towards him:
longing, hope, etc. Becky even goes so far as to change her appearance,
reading list, and locale in order to appeal to men like the ones in her fantasy
(perhaps he is out there now).
It’s not hard to imagine real situations like this. The question is, would Becky’s
imagined, Mr. Darcy-like entity fall under the first, second, or third notion of
existence that we’ve discussed? Clearly not the first; much to Becky’s dismay, the
imagined Mr. Darcy is not concrete. Whether the imagined Mr. Darcy would be
possibly concrete or strictly imaginary may depend on one’s theory of fictional
entities. Walton would claim that the imagined Mr. Darcy here is not imaginary in
our third sense and so does not present a counterexample to his theory. Rather,
Becky has genuine emotions about a possible actual person, which, he would argue,
is perfectly fine for his theory (see Walton 1993).
9
Note that neither Walton nor Nussbaum discuss the possibility of unconscious emotions. Perhaps we
can have an unconscious belief about a fictional character that triggers an emotion. This, however, seems
to go against the spirit of the pure-cognitive project.
10
On this view one might also wonder about the status of abstract systems, mathematical and logical
entities that sometimes evoke emotional responses—we find them beautiful.
123
792
K. Tullmann, W. Buckwalter
This reply might be suitable for possibly concrete objects—an imagined tornado,
an imagined unicorn, etc. But does it work for fictional entities? Consider the
following adaptation to the example above:
The unlucky-in-love Becky restricts her imagination to the character Mr.
Darcy, just as he is described in Jane Austen’s novel. It is this Mr. Darcy that
serves as the object of her emotions (longing, wistfulness—but also, this time,
a strange jealousy for Elizabeth Bennett and regret that such a man doesn’t
exist). These emotions may even cause Becky to engage in certain peculiar
actions: she places her copy of Pride and Prejudice on her nightstand in an
oddly affectionate manner and makes caustic remarks about Elizabeth
Bennett’s contrariness to her friends.
What might we say about this example? If we agree with Lewis and Plantinga’s
view of the nature of fictional characters, then we might be inclined to say that the
two Becky cases are similar in at least one significant respect: they are both entities
that exist in another possible world, but not the actual one. We hold, in contrast, that
there is a marked dissimilarity in the two cases. Recall that there are many
arguments against that claim that fictional entities are possible objects. We can
describe two of the main critiques without delving into too far into this literature.
First, there are too many possible objects that are relevantly similar to a fictional
character to identify them. As Thomasson notes, the descriptions of fictional characters
in works of art are incomplete in detail—they do not typically describe the character’s
weight, blood type, or ordinary everyday activities (1999). On the other hand, possible
worlds are thought of as complete (Sainsbury 2010). If we try to identify a fictional
character with a possible entity, we run into trouble: the details left out by the story can
be filled in an infinite number of ways by the possible entities, and we would have no
way of identifying some with Sherlock Holmes, for example, but not others. Secondly,
most people have the intuition that fictional characters are created by an author; they
come into being via certain intentional actions by an actual person. The possibilist view
has no way of accounting for this. Thomasson notes: ‘‘Even if we could find a single
candidate possible detective to identify as Sherlock Holmes, this would be a possible
man with the property of being born in the nineteenth century, not of being created by
Arthur Conan Doyle’’ (18). For these reasons, as well as others, both Plantinga and
Kripke eventually rejected this view (Thomasson 1999; Kripke 1980). If the objections
to the possible world view hold water, then it may make more sense to say that in the
second case, the object of Becky’s emotions is an imaginary object rather than a possible
one. Furthermore, this imaginary object has motivated her behavior in the way Walton
claims that genuine emotions do: the book by her bedside, the remarks to her friends.
There may be cases, then, in which our reactions to fictional stories or characters
do motivate us to act. And while we do grant that it’s unclear how to best interpret
the precise existential commitments of motivation in terms of imaginary things, our
point is that it’s perfectly possible for belief-based theorists to grant that they can
generate genuine emotions.11
11
We also retain the option of simply denying that belief theorists must be committed to Walton’s notion
of motivation.
123
Does the Paradox of Fiction Exist?
793
If that’s the case, then we have enough to extend our overall critique to these
sorts of pure cognitive theories as well. In other words, it is once again false that the
candidates for ‘exist’ provided in (3–5) are the same sense of exist required by
typical belief theories in order to have genuine emotion. Of course, Waltonians are
free to propose their own theory of emotion under which one is forced to accept the
existential commitments we have been considering. And such a theory should be
evaluated on its own independent merits. But without such a theory, and by
attending to this fact, we see that even for pure cognitivist accounts, there’s nothing
inherently paradoxical about the paradox of fiction.
5 Requiem for a Paradox
Our discussion of the three types of theories of emotion—feeling, judgment, and
belief-based theories—has revealed an inconsistent use of ‘exist’ in the second and
third premises of the paradox of fiction. According to our argument, the only way in
which the paradox arises is if ‘exist’ denotes the same type of existence across the
second and third premises. But as we have argued in Sect. 4, it is doubtful there
actually is a theory of emotions advocated in the literature today committed to the
same type of existence across these two premises. Furthermore, we think it’s equally
implausible that any theory of emotion would want to grant that ‘exist’ must be
understood consistently across these premises. Of course, it is possible that such an
account may be developed in the future. But this just goes to show that if the
paradox of fiction does exist, it remains to be discovered or argued for—not solved.
We conclude that the formulation of the paradox of fiction, as traditionally
construed by Radford and subsequent philosophers, is largely responsible for its
own persistence. That said, there is no doubt that some theorists will continue to
argue for or against the basic ideas represented in Radford’s three sentences—it just
need not be motivated by an attempt to avoid inconsistency, or to rescue people’s
behavior from the threat of irrationality. The novel contribution of this paper has
been to reveal a crucial ambiguity in the formulation of the paradox of fiction, and
subsequently, to provide a theory-neutral and minimally controversial solution by
denying that it exists as a problem to be solved in the first place.
Our solution to the paradox of fiction clears the way for a cluster of interesting
psychological questions regarding the nature of fictional entities, the kinds of
emotional objects they represent, and the affective responses they generate as we
experience them. For instance, there is a natural analogy between the affect
generated when engaging with works of art and the phenomenology or perceptual
seemings generated when engaging with known illusions. To illustrate this
comparison, recall the feelings that many report when they experience powerful
optical illusions like the Müller-Lyer illusion. Although we know that the MüllerLyer lines are of equal lengths, we still cannot help but feel that one line is longer
than the other. Similarly, our emotional responses to characters and situations do not
suddenly cease after we discover that they are purely fictional. Our experiences of
fear when we see the killer enter the heroine’s bedroom remain after we remind
ourselves that what we are watching is just a film. Knowledge that these characters
123
794
K. Tullmann, W. Buckwalter
are fictional or that Müller-Lyer is an optical illusion often makes little difference to
our immediate reactions in these matters.
Returning to Radford’s question of rationality, note that the affect accompanying
fictional responses is no more or less ‘‘irrational’’ than the seemings typically
accompanying the standard Müller-Lyer illusion. In the case of the Müller-Lyer
illusion our perceptions appear to be largely cognitively impenetrable, and
independent of deliberate control. It is simply a matter of psychology and empirical
cognitive science to explain why the feelings associated with the Müller-Lyer lines
persist or dissipate after the illusion is discovered and the belief that both lines are of
equal lengths is formed. Similarly, it is an interesting psychological question why
certain kinds of affect persist after one learns of and forms the belief that objects of
an emotional response are fictional. We would only emphasize that the answer to
this interesting question does not bear on the central issue taken up in this essay
regarding what the paradox of fiction tells us about the rationality of emotion.12
But having set the question of irrationality aside, the analogy between fictions
and engagements with known illusions can still be instructive. With respect to fear
for our heroine for instance, some have noted that our bodies are primed to respond
with reactions to perceived stimuli whether they are real or not (Harris 2000) and
that these feelings are instigated before and alongside the cognitive components of
an emotion, such as the belief that something in our environment can cause us harm
(see Schroeder and Matheson 2006). Future research on the cognitive science of
emotion might fruitfully study these kinds of bodily responses for fictional and nonfictional objects and compare them to our experiences with known illusions as one
of way learning more about the affective nature of these responses.
Alternatively further work on the affective responses generated by certain kinds
of artistic works over others might begin to strain the basic analogy between fictions
and illusions. For instance, parables and fables often evoke very complex and
sophisticated affective responses connected to motivation, action and behaviour.
These affective responses seem to extend far beyond the seemings generated by
Müller-Lyer or the tinge of fear we feel during a horror film. No longer clouded by
Radford’s threat of irrationality discharged above, comparisons between the
affective responses generated by various kinds of fictions over others may again tell
us more about fictional objects of emotions, their relationship to other kinds of
emotions and theories of emotion. We flag this as another profitable area for future
research in the cognitive science of emotion motivated by our affective responses to
fiction.
Given these interesting psychological research questions what is next for the
philosophical literature on emotions and fiction? We close with a policy
recommendation. In short—forgo solutions to the paradox of fiction. If what we
have said is correct, its persistence has traded on an ambiguity. But uncovering this
ambiguity has been instructive. By questioning the notion of existence throughout
Radford’s premises, it has led us to reconsider the different ontological commitments held by several different leading theories of emotion. We suggest that
researchers interested in the paradox should simply study these commitments
12
We thank an anonymous reviewer for discussion on this point.
123
Does the Paradox of Fiction Exist?
795
directly, or the manner of existence that objects of emotions need have according to
each theory on the whole. We suspect that this new strategy will give us a better
picture of what theories of emotions can actually tell us about fictions. And, perhaps
with this new focus, new progress is possible in the quest to determine whether
fictions really posit a unique case in the literature or whether they warrant the same
consideration as other objects of emotions.13
References
Buckwalter, W., Rose, D., & Turri, J. (2013). Belief through thick and thin. Noûs. doi:10.1111/nous.
12048.
Buckwalter, & Turri, J. (Under review). In the thick of moral motivation.
Carroll, N. (1990). The philosophy of horror; or, paradoxes of the heart. New York: Routledge.
Coleridge, S. T. (1985/1817). Biographia Literaria. In H. J. Jackson (Ed.), Samuel Taylor Coleridge.
Oxford.
Currie, G. (1989). The nature of fiction. Cambridge: Cambridge University Press.
De Sousa, R. D. (1987). The rationality of emotion. Cambridge: MIT Press.
Goldie, P. (2004). Emotion, feeling, and knowledge of the world. In Robert. C. Solomon (Ed.), Thinking
about feeling (pp. 91–106). Oxford: OUP.
Harris, P. (2000). The work of the imagination. Malden, MA: Blackwell Publishing.
Hartz, G. (1999). How we can be moved by Anna Karenina, green slime, and a red pony. Philosophy, 74,
557–578.
Hurka, T. (2001). Virtue, vice, and value. Oxford: Oxford University Press.
James, W. (1884). What is an emotion? In C. Calhoun & R. C. Solomon (Eds.), What is an emotion?.
New York: Oxford University Press.
Kieran, M. (2010). Emotions, art, and immorality. In P. Goldie (Ed.), The Oxford handbook of philosophy
of emotion (pp. 681–703). New York: Oxford University Press.
LeDoux, J. (1996). The Emotional Brain. New York: Simon & Schuster.
Kripke, S. (1980). Naming and necessity. Cambridge, MA: Harvard University Press.
Lewis, D. (1978). Truth in fiction. American Philosophical Quarterly, 15, 37–46.
Miller, B. (1986). Exists’ and ‘existence. Review of Metaphysics, 40, 237–270.
Nussbaum, M. (2004). Emotions as judgments of value and importance. In R. C. Solomon (Ed.), Thinking
about feeling (pp. 183–199). Oxford: OUP.
Plantinga, A. (1974). The nature of necessity. Oxford: Oxford University Press.
Prinz, J. (2004a). Gut reactions: A perceptual theory of emotion. Oxford: Oxford University Press.
Prinz, J. (2004b). Embodied appraisals. In R. C. Solomon (Ed.), Thinking about feeling (pp. 44–58).
Oxford: OUP.
Quine, W. V. (1953). On what there is, in from a logical point of view (pp. 1–19). Cambridge, MA:
Harvard University Press.
Radford, C. (1975). How can we be moved by the fate of Anna Karenina? Proceedings of the Aristotelian
Society, Supplemental, 49, 67–80.
Robinson, J. (2005). Deeper than reason: Emotion and its role in literature, music, and art. Oxford:
Oxford University Press.
Rose, D., Buckwalter, W., & Turri, J. (Under review). When words speak louder than actions: Delusion,
belief and the power of assertion.
Sainsbury, R. M. (2010). Fictional and fictionalism. New York: Routledge.
Salmon, N. (2002). Mythical objects. In J. K. Campbell, M. O’Rourke, & D. Shier (Eds.), Meaning and
truth: Investigations in philosophical semantics (pp. 105–123). New York: Seven Bridges Press.
Schroeder, T., & Matheson, C. (2006). Imagination and Emotion. In S. Nichols (Ed.), The architecture of
the imagination (pp. 19–40). Oxford: Oxford University Press.
Solomon, R. (1993). The passions. Indianapolis: Hackett Publishing Company Inc.
13
For helpful feedback and discussion we thank Berit Brogaard, Noël Carroll, Jesse Prinz, Andy Specht
and John Turri.
123
796
K. Tullmann, W. Buckwalter
Solomon, R. (2004). Emotions, thoughts and feelings: Emotions as engagements with the world. In R.
C. Solomon (Ed.), Thinking about feeling (pp. 76–90). Oxford: OUP.
Stocker, M. (1983). Psychic feelings: Their importance and irreducibility. Australian Journal of
Philosophy, 61, 5–26.
Thomasson, A. (1999). Fiction and metaphysics. Cambridge: Cambridge University Press.
Thomson-Jones, K. (2008). Aesthetics and film. London: Continuum Press.
Tullmann, K. (In preparation). The moral experience of fiction. The City University of New York
Graduate Center.
Turvey, M. (1997). Seeing theory: On perception and emotional response in current film theory. In R.
Allen & M. Smith (Eds.), Film theory and philosophy (pp. 431–457). Oxford: Oxford University
Press.
Van Inwagen, P. (1977). Creatures of fiction. American Philosophical Quarterly, 14, 299–308.
Walton, K. (1978). Fearing fictions. Journal of Philosophy, 75, 5–27.
Walton, K. (1993). Mimesis as make-believe. Cambridge, MA: Harvard University Press.
Williams, C. J. F. (1981). What is existence?. Oxford: Oxford University Press.
Zalta, E. (1988). Intensional logic and the metaphysics of intentionality. Cambridge, MA: MIT Press.
123