About CA
Contact CA

Search Journal

Editorial Board

Permission to Reprint


Site Map


HOT Emotions: Dissolving the Paradox of Fiction

  Katherine Tullman

This essay critiques two of the main theories in the philosophy of emotions, the pure-cognitive theory and the neo-Jamesian theory, through the paradox of fiction.  After explaining the different kinds of emotions we experience when engaging with fictions, I argue that a middle-ground, hybrid theory more adequately accounts for current scientific research and the paradox of fiction than either of the previous two.  I propose a “HOT” theory of emotions (higher-order thought) specifically to explain complex emotions about fictions.

Key Words
emotion, higher-order thoughts, James-Lange theory of emotions, judgment theory of emotions, paradox of fiction, thought-theory of emotions

1.  “Nobody can eat fifty eggs”

It’s another rainy day in prison, and the boys are getting restless.  One inmate casually claims that he can eat fifty eggs.  The bored prisoners place bets; the man’s eager “trainer” grabs the cash.  You watch on, at first in anticipation—can he do it?  Surely nobody can eat fifty eggs—then disgust as his belly swells up “like a watermelon” as the man forces down egg after egg, and finally, with a subtle pride, admiration and pity for this brave man as he lies forgotten by his peers, arms spread like Christ on the cross.

Except, of course, that this “hero” isn’t real.  Many readers will know him as the maverick protagonist of Cool Hand Luke (1967) played by Paul Newman.  But weren’t our emotions real?  Your reaction was sincere; the gagging sensation in your throat physically present.  You empathized with Luke’s sacrifice and nonchalance throughout his ordeal and resent the other prisoner’s greed.  These emotions vary from the simple and reactionary (disgust) to the rationally complex (pride and empathy).

The question of whether it is possible to experience emotional reactions towards fictional entities is often characterized in terms of the paradox of fiction.  Traditionally, three propositions comprise this discordant triad:

            1. We are genuinely moved by fictions;

            2. We know that that which is portrayed in
                fictions is not actual; and

            3. We are only genuinely moved by what we
                believe is actual.[1]

This paradox reveals an inconsistency in our emotional responses to fictional entities and situations.  At the heart of this inconsistency lies an assumption about the nature of emotions:  what causes emotions and what they are comprised of.  Taken together, these three premises are a hurdle for any leading theory of emotions:  pure-cognitive theories, non-cognitive theories, or hybrid theories.  Simply put, cognitive theorists hold that emotions are constituted by some form of cognition:  beliefs, judgments, thoughts, and so on, though the particular form varies by theory. Non-cognitivists, in contrast, hold that such mental states do not constitute an emotion; rather, either a perception or bodily response to stimuli is the emotion.  Hybrid theories are a mixture of the two:  both some form of cognition and bodily responses are necessary components of emotions.

The Cool Hand Luke example above reveals another intuition underlying the paradox.  When engaging in a fiction, we are moved not only by sensory stimuli in the art media, like reactions to sudden noises, sharp movements, and bright colors, but also by the work’s narrative.  As stated, these emotions range from the seemingly cognitively complex to the automatic and subconsciously realized.

Any theory of aesthetic emotions must explain the paradox of fiction and this intuition about the breadth of our emotional responses in order to serve as an adequate account of our affective responses towards fictional entities. Furthermore, a theory of emotions in general must do the same; I take the paradox as a paradigm of sorts concerning one significant aspect of our emotional repertoire.  I argue that the pure-cognitive theory, held most notably by Martha Nussbaum and Kendall Walton, fails on this account.  At the same time, strictly non-cognitivist theories of emotions also cannot adequately explain our experiences with fictions, especially in terms of more complex emotions.

My goal is to supplement a non-cognitivist theory with another component that solves the paradox and explains the above intuitions while still maintaining a strong foundation in a general theory of emotions:  a middle-ground, hybrid theory that encompasses the full range of our emotional responses toward fictions.  To this end, I will argue that the typically troublesome case of complex emotions to fictions are best understood in terms of a higher-order thought-theory of emotions in which one must have an intentional evaluation or belief concerning a thought that is itself the formal object of the emotion.[2]  As we’ll see, though, I will explain the role that thoughts play in the emotional process in a slightly different way from other hybrid theories. 

2.  Fiction’s emotions proper

Engaging with fictions can elicit different types of emotions from an audience.  Many of these emotions include involuntary or subconscious affect responses, such as increased heart rate or involuntary screams.  We can also have more complex emotions, such as jealousy or empathy, which seem to require conscious thought.  Somewhere in between lie the more basic emotions, such as joy and anger. We may often be aware of these feelings as well as what causes them, but not always.

Which of these types of emotions does the paradox address?  Since the first type, the automatic emotional responses, happen involuntarily, it might seem that they should not be included as genuine emotions.  Noël Carroll describes the involuntary or sub-cognitive affect responses and moods as “cognitively impenetrable,” taking place on a level not accessible to or affected by one’s reason or other cognitive resources.  For example, “despite the fact that you know that you are not in danger [while watching a film], your body will respond otherwise and prepare you affectively for flight or some other self-protective behavior.  Adrenaline will rush into your veins as we are primed for action.  Furthermore, this high can be enjoyable—can be savored—if there really is no real danger in the vicinity.”[3]  In other words, we will have these responses whether we know that we are in real danger or not.  Such responses can be brought about by sudden noises or movements in a film, musical composition, or play.  The startle response is an example of a sub-cognitive reaction common to films—especially horror films—because they induce the right kind of affective response in the audience, the fearful “flight or fight” reaction which Carroll describes.

The philosopher Carl Plantinga provides an interesting example of these automatic affective responses.  He argues that films often evoke an emotional response by using certain techniques that emphasize emotions displayed in the human face.[4]  Plantinga bases his argument on the influential work of the psychologist Paul Ekman, who proposed that there are six universal emotions (anger, fear, enjoyment, sadness, disgust, surprise), each of which can be recognized by specific, cross-cultural facial expressions.[5]  In cases of affective mimicry, one automatically mimics another’s facial expression, posture, motions, and so on. This, in turn, results in an automatic affective response in the viewer.  Plantinga argues that a similar situation occurs when engaging with films.[6]  A film viewer responds to the emotions on an actor’s face that, if performed well, will in turn evoke an emotion in the viewer, either the same emotion, in the case of affective mimicry, or a similar sympathetic emotion.

No philosopher denies that these sorts of automatic emotional responses towards fictions take place.  The question is whether they constitute real emotions.  Here, I argue, the distinction between cognitive and non-cognitive theories of emotions begins to become apparent.  The cognitivist view is that since these automatic emotional responses are “cognitively impenetrable,” not involving a conscious judgment or belief, they are not real emotions.  If they aren’t real emotions, then they are not the focus of the paradox of fiction.  Non-cognitivist theories will deny this, as we’ll see, but for now we can grant Carroll’s distinction.  Opposed to the automatic responses, Carroll calls emotions such as anger, joy, sympathy and pride “emotions proper” or “garden-variety” emotions because they are the ones typically brought to mind in reference to the term “emotion.”[7]  Although emotions such as anger and joy may have characteristic bodily responses that accompany them, the emotion itself is constituted by some form of cognition, separating them from the automatic affective responses.

Cognitive theories of emotions can also more fully explain the distinction between basic emotions, such as anger and fear, and more complex emotions, such as pride and envy. The complexity of the emotion reflects the complexity of the thought or judgment.  These may arise on account of our contemplation of a story’s narrative or a character’s situation, history, character flaws, and so on.  Consider the following example.  One evening, I decide to watch a scary film.  Unfortunately, the speakers on my television are not working, so I watch the whole film without sound.  I don’t understand what’s going on in the story, but I still see a monster sneak up behind the film’s hero and I feel fear in response.  Would we still respond emotionally to the character in a film if the sound was turned off and we only saw the actress’s face but did not know what caused her to show the tell-tale signs of fear?  We may for some emotions—fear, perhaps, or disgust.  However, in order to feel the complex emotions pity, sympathy, pride, and so on, one would also have to consider the narrative as a whole, and that information would be missing in this case.  Imagine how we would react to Luke’s plight while he eats those eggs if we didn’t know the context; we would certainly be very confused if we didn’t know what was happening and probably not feel the range of emotions that we do when we are familiar with his plight.  This would suggest that some sort of cognitive appraisal of the fiction is required for certain emotions.

The paradox of fiction concerns all of these types of emotions, from the basic to the complex.  Nevertheless, it’s important to make the distinction between different kinds of emotions if we want to be clear about the nature of the problem the paradox poses for our understanding of fictions, because different theories of emotions have different explanations for the different types of emotions distinguished here.

3.  Cognitive vs. non-cognitive theories of emotion

Ultimately, my own view is to strike a balance between pure-cognitive and purely non-cognitive theories of emotions, forming a hybrid theory that explains the paradox in terms of all three types of emotions discussed above.  To begin, I will consider two theories of everyday emotions with which I disagree on how they address the various types of emotions and how the theory can be applied to fictions.

            3.1.  Pure-cognitive theories

Many philosophers have stated, more or less, that they adhere to some form of cognitivist theory of emotions.  For example, Kendall Walton argues “that fear necessarily involves a belief or judgment that the feared object poses a threat is a natural supposition which many standard theories of emotions endorse.”[8]  It is the relevant belief or judgment that constitutes the emotion, not the affective responses, though those may accompany the belief.  Walton’s position is best described as a pure-cognitive theory.  Only a certain type of belief or judgment  is necessary for an emotion to obtain; what type exactly, we’ll explore shortly.

Cognitivists argue that the cognitive element of emotions is at least a necessary component of an emotion.  What separates pure-cognitivists like Walton from hybrid theorists like Robert Solomon and Noël Carroll, for example, is that they argue that a belief, thought, or judgment is also sufficient for an emotion;  it all that is required in order for an emotion to obtain.  Martha Nussbaum, for example rejects the claim that all emotions have a felt bodily response.  She argues that:

There usually will be bodily sensations involved in grieving, but if we discovered that my blood pressure was quite low…or that my pulse rate never went above sixty, there would not, I think, be the slightest reason to conclude that I am not grieving…We do not withdraw emotion-ascriptions otherwise grounded if we discover that the subject is not in a certain brain-state.[9]   

The important component of one’s grief is not a bodily change or perception of that change, for such a response need not be present for that emotion to occur.  In other words, bodily changes are not a constitutive element of an emotion’s identifying characteristics.  Instead, it is the judgment, belief, or conscious recognition of something in our environment affecting our well-being that constitutes the emotion.[10]  We can imagine a situation similar to Nussbaum’s in which this is possible.  I recently broke up with my boyfriend.  I have confused thoughts, am distracted, and involuntarily think about my ex but experience no bodily sensations relating to my heartbreak.  Instead, my emotion results from a judgment directed towards the confused thoughts that I experience.  As Nussbaum argues, I still have the emotion even without bodily sensations.  This seems especially fitting for long-term emotions, such as love, pride or grief, which do not always seem to include a felt bodily response to a situation.

Pure-cognitivists such as Nussbaum and Walton posit several important claims about the nature of emotions.  The first involves intentionality:  emotions are directed towards or are about something in one’s environment.  Nussbaum argues that emotions do not merely point out a formal object in one’s environment but rather involve an internal and partial way of seeing.[11]  I perceive an object in my environment as causing my grief because of my unique perspective and way of interpreting the world.  Solomon summarizes this point nicely:

All emotions presuppose or have as their preconditions…certain sorts of cognitions—an awareness of danger in fear, recognition of an offense in anger, appreciation of someone or something as loveable in love.  Even the most hard-headed neurological or behavioral theory must take account of the fact that no matter what the neurology or the behavior, if a person is demonstrably ignorant of a certain state of affairs or facts, he or she cannot have certain emotions.[12]

Cognitivists also hold that emotions have to do with the subject’s well-being.  As Nussbaum says, emotions are “eudaimonistic,” having to do with the agent’s own flourishing or the flourishing of those we care about.[13]  Finally, as we’ve seen, emotions involve a belief or judgment about their objects that causes bodily changes associated with the emotion but which are not themselves a necessary component of it.  These beliefs can be complex or simple.  Being afraid, for example, involves the belief that an object, such as a snarling dog, is dangerous and can do me harm. Guilt or jealousy may involve very complex series of beliefs, such as I might (falsely) belief that my partner is unfaithful and take trivial events or objects in my environment as supporting that belief.

Pure-cognitivist theories of emotions seem to be able to handle many of our emotional responses, including those concerning fictions.  Walton, and others, dissolve this paradox by rejecting the first proposition.  We do not have genuine emotions concerning fictions.[14]  We do have the automatic affective responses towards fictions described above, which might be why we feel like we have genuine emotions about fictions.  But those feelings are misleading. We do not have the right kind of belief about fictions to constitute a real emotion; we don’t actually believe that a film monster can do us or others harm, for example. We are merely caught up in what Walton calls a “game of make-believe.”[15]

There are problems with this theory, however.  One argument against pure-cognitivists has to do with the intentionality component.  Nussbaum stresses that the beliefs involved in emotion are “not trivial, but serious,”[16] meaning that to find a situation frightening, the object must really be potentially dangerous to me or someone I care about.  If the object of my fear turns out not to be dangerous, then my fear will subside.[17]  Consider our emotional response to the protagonist in Cool Hand Luke.  While watching the film, one might feel fear for Luke’s sake while he eats the eggs, wondering whether or not he can possibly (safely) finish.  One might also have a set of beliefs about Luke’s stubborn personality, past actions, and current situation that all amount to a feeling of anxiety on his behalf. The recognition, though, that Luke is a fictional character should eliminate our belief that he is in any real danger and thus eliminate the emotion, which doesn’t seem to capture our actual experiences with fictions.

Walton would argue, perhaps, that the reason the emotion does not subside is that we are so caught up in our game of make-believe with the fiction that we experience a wide-range of make-believe emotions, even though we know that the characters and situations aren’t real.  There’s a more pressing concern with pure-cognitivist theories, though, stemming from recent work in the neuroscience and biology of emotions themselves.  While it might seem strange to appeal to scientists when discussing fictions, emotions themselves are cognitive or biological states.  A theory of emotions about fictions should at least be in accord with general scientific theories of emotions!

            3.2.  Two pathways of emotional processing

The neuroscientist Joseph Ledoux’s recent work on the brain systems involved in emotional processing has shown that there’s an affective or non-cognitive appraisal in emotions that takes place below the level of awareness and produces physiological changes in the subject.  In fact, many of our emotions happen too quickly for cognitive evaluation, skipping the cortex altogether, which is the part of the brain that processes this information for higher-level thought.[18] In his work on fear responses in rats, LeDoux has shown that information about an emotional stimulus in one’s environment reaches the amygdala, the part of the limbic system associated with emotions, through a direct pathway from the thalamus.  This information also goes through another, slower pathway from the thalamus to the cortex, and then to the amygdala.  Once the information reaches the amygdala from either path, emotional responses occur.[19] This is significant because the neocortex is associated with higher processing systems, such as those involved with thought, reasoning, and consciousness.  Why might we have two emotional pathways instead of one?  The direct thalamic pathway is faster than the cortical pathway, which allows for much quicker responses to potential threats in our environment.  This is a major advantage in terms of evolutionary fitness.  However, the response from the thalamic pathway may be inaccurate, as when we mistake a shadow for a predator.  This is where the cortical pathway steps in.  LeDoux argues that the role of the cortex is to prevent or stop an inappropriate emotional response rather than to bring about a response:  a shift from reacting to a stimuli to producing a conscious action.[20]

If LeDoux’s findings are correct, the pure-cognitivist’s claim that beliefs constitute an emotion is undermined; emotions can occur without one first having a belief about one’s environment.  Of course, a pure-cognitivist could argue that the quick, subcortical processing and responses don’t result in genuine emotions; real emotions still may require the slower beliefs.  Yet LeDoux argues that all of our emotions are based on this sort of subcortical processing, which necessarily results in felt bodily responses.  LeDoux argues that “feelings do involve conscious content, but we don’t necessarily have conscious access to the processes that produce the content.  And even when we do have introspective access, the conscious content is not likely to be what triggered the emotional responses in the first place.”[21]  A pure-cognitivist would be hard pressed to explain away these bodily responses when arguing that only beliefs are necessary for emotions to occur, and to explain why the response can occur without being caused by a belief.

LeDoux’s findings may all be well and good for simple or automatic emotions, such as fear or disgust.  But what about the more complex emotions:  pride, jealousy, and so on?  As it turns out, we can appeal to LeDoux’s model of emotional processing here as well.  LeDoux identified two mechanisms at work in the emotional process:  one automatic and fast, the other slower and more discriminating.  His model suggests that the latter process is at work for complex emotions, such as the pride we feel for Luke’s feat.  Some form of cognition may be required for some emotions.  Perhaps the complex emotions involved in the paradox are not the initial sub-cognitive events but rather their cognitive reappraisal, which takes place shortly following the initial reaction.[22]  Ekman suggests that certain “emotional plots,” such as grief, jealousy, and infatuation, do specify a “particular context within which specific emotions will be felt by specific persons, casting the actors and what has or is about to transpire.”[23]  In other words, in order for these emotional plots to occur, we must recognize ourselves to be in a certain context.  If we feel pride for Luke it’s because we recognize that he has done something difficult that is worthy of praise.  This is an important point, and one to which I will return momentarily.

This shifts the burden of proof from the hybrid theorists to the non-cognitivists.  They must be able to explain how we can undergo complex emotions without requiring thoughts, judgments, or beliefs.  Most non-cognitivists feel that they are up to the task.  Here, I will examine one prominent view: the philosopher Jesse Prinz’s neo-Jamesian theory of emotions.

            3.3.  Prinz’s perception-based theory

LeDoux argues that thoughts are not necessary for all emotions to occur, a point with which Prinz and other non-cognitivists would readily agree.  These theories hold that all emotions have a bodily component, or at least a disposition for one.  These bodily responses, or our perception of them, are the emotion.  Historically, such theories follow a Jamesian theory of emotions, or feeling theory.  James’ argues that an emotion occurs when we have a perception of a bodily change, and that “our feeling of the same changes as they occur IS the emotion.”[24] Prinz, similarly, identifies emotions as “perceptions (conscious or unconscious) of patterned changes in the body.”[25]  In Prinz’s updated Jamesian theory, perceptions of bodily changes rather than a judgment or belief are the necessary and sufficient conditions for emotions.  Even emotions that seem to be cognitively complex, such as guilt or loneliness, have some sort of bodily response.  A guilt-ridden person might have a “down-trodden body” or “guilt pangs and agonies,” along with a “blush of guilt,” downcast eyes, and a lowered chin.[26]  Long-standing emotions, such as love, predispose us to “enter into patterned bodily responses.” [27]  Prinz goes so far as to state, “I would defy the critic of James and Lange to identify a single emotion that lacks a bodily mark, at least dispositionally.”[28]

Central to Prinz’s claims about the structure of emotions is the concept of “core relational themes,” borrowed from the psychologist Richard Lazarus (1991).  He utilizes this concept to explain emotion’s intentionality. The core relational themes involve an organism/environment relation that bears on the subject’s well-being.  For example, the bodily responses related to fear result from one’s perception of a threat in one’s environment.  We feel sorrow when we recognize that we have lost something.  Prinz argues that we should only consider bodily feelings as emotions if they involve this kind of core relational theme between an object and its environment.  Other bodily feelings, such as hunger and fatigue, do not have the relation between the organism and its environment, even if they do bear on its well-being.[29]  Importantly, core relational themes can be triggered both consciously and unconsciously, as when we have automatic disgust or fear reactions that occur before we have a thought about what causes those reactions.

Prinz’s view may seem plausible in terms of seemingly basic emotions, such as fear and sadness, which have typical physiological bodily responses along with the subconscious affective responses described in Section Two.  But what about complex emotions?  While Prinz’s explanation of complex emotions is itself quite complicated, his basic point is simple. [30]  He argues that we must distinguish between the causes of emotions and what constitutes them.  Both basic and complex emotions are bodily perceptions but the causes of emotions, such as pride and empathy, are typically more complex than those involved in emotions like fear or sadness.  In other words, the core relational themes required may need more information or context than those needed for a basic fear response.  This would explain why we think that thoughts, beliefs, or judgments are part of the emotion, when they are merely the cause of the emotion.

This can apply to emotions about fictions as well.  Non-cognitivists, such as Prinz, deny the third proposition of the paradox.  Beliefs are not necessary for emotions.  The perceived bodily changes that we undergo when exposed to emotionally-charged stimuli, not our beliefs about them, are the necessary component.  When watching a film, we may cover our eyes, clutch our armrests in fear, gnash our teeth or frown in anger, or smile or cry from joy.  These bodily changes are obviously real and perceiving them alone constitutes our emotions.

Does this adequately answer the paradox?  Prinz’s theory accepts that even automatic responses about fictions that do not involve a conscious belief, thought, or judgment about a situation are emotions.  Basic emotions, such as fear and sorrow, have characteristic core relational themes that trigger them both in reality and fiction without necessitating a conscious judgment.  It’s more difficult to see how core relational themes apply to complex emotions about fictions. How can we feel guilty about what happens in a fiction or envious of a fictional character, not just concerned on behalf of a character with whom we sympathize?  It seems like appealing to core relational themes just shifts the problem. For emotions such as pride or empathy, we may need to make a judgment or have a thought or belief about a situation to trigger these emotions.  For example, the core relational theme of guilt involves recognizing that one has “transgressed a moral imperative,” and envy involves “wanting what someone else has.”[31]  These judgments, thoughts, or beliefs need not be conscious but they are still required, and so are necessary for the emotion to occur. These thoughts, judgments, and beliefs about our environment are a part of the processing involved in bringing about certain emotions, in line with LeDoux’s point about the two emotional pathways described above.

Again, some form of cognition seems necessary in order to explain our complex emotions concerning fictions.  Perhaps, then, a hybrid account of emotions would best solve the paradox, a theory that does not necessitate a belief, judgment, or thought for all emotions but only for some. Let’s consider one approach.

4. Carroll’s thought-theory of emotions

Noël Carroll denies the third proposition of the paradox by ridding his theory of beliefs, which, he argues, are the problematic element of the proposition.  Similarly to Walton, Carroll argues that it is not the fictional world itself towards which we have an emotional response.  Instead, it is our thought content about the fiction.  Thoughts, according to this theory, are the major cognitive constituent of emotions. Carroll argues that this is beneficial because we can, and do, have genuine emotions about the contents of our thoughts all of the time, so we need not propose a theory of make-believe or some other pretense theory to account for fictions.[32]

The thought-theory hinges on two main points.  The first concerns the possibility of having emotions about things we imagine.  Carroll illustrates this point with the following example:

Standing on a precipice, though in no way precariously, one might fleetingly entertain the thought of falling over the edge.  Commonly, this can be accompanied by a sudden chill or tremor which is brought about, I submit, not by our belief that we are about to fall over the edge of the precipice, but by our thought of falling, which, of course, we regard as a particularly uninviting prospect.  It need not be a prospect we believe is probable, there is no one around to push us, and we have no intention of jumping.  But we can scare ourselves by imagining a sequence of events that we know to be highly unlikely.[33]

It is not the event of our falling that scares us in this situation but rather the content of our thought:  a mental image or attitude of the scenario, perhaps.  Carroll argues that this sort of emotional response happens quite frequently in our everyday lives and is also what happens when we engage in fictions.

Secondly, Carroll shows that thoughts do not necessitate beliefs:  “The thought of a fearsome and disgusting creature like Dracula is something that can be entertained without believing that Dracula exists…Thus, if we grant that thought contents can frighten, then we shall have no problem saying that standard readers and viewers of fictions about the Count do not believe the Count exists.”[34]  This point is especially helpful in solving the paradox.  It is possible to have genuine emotions about fictions without believing in the existence of the emotion’s formal object.  The emotions are still intentional—not about the fiction itself but about our thought content.  The first two propositions of the paradox hold, but the tricky third proposition is eliminated.

Perhaps, though, Carroll’s thought-theory is too intellectual. Many of our emotions about fictions are too instantaneous and involuntary to involve thought.  Because the thought-theory does not involve changes in bodily states, one could argue that it cannot serve as a general theory of emotions. Solomon argues that the problem with the thought-theory, like the pure-cognitive theories, is that it excludes bodily changes from emotions, which we very clearly have.[35] Carroll, however, does not exclude bodily changes from his account of emotions.  In his view, both thoughts and bodily changes can constitute emotions.  Theories such as this are what Prinz calls hybrid appraisal theories.  The question is, how do the bodily changes fit into the picture?  Not all of our emotions are based on thoughts, perhaps not even most, as the two pathways model shows.  How can a thought-theory explain them, besides denying that they are emotions to begin with?

Despite these concerns, I think that Carroll has highlighted an important point in positing a solution to the paradox in terms of thoughts instead of beliefs.  His theory accounts for our complex theories of emotions.  Our pride for Luke, for example, is based on a thought about his trials.  It has also eliminated the third proposition of the paradox, maintaining that we have many genuine emotions about fictions.

Ultimately, though, I agree with Solomon that a thought-theory cannot serve as a complete theory of emotions about fictions.  In the next section, I will modify Carroll’s view to one that is less problematic in light of what LeDoux and others have taught us about how the emotional process works.

5. HOT emotions

There are two aspects of Carroll’s thought-theory that I believe need further elucidation. The first has already been noted. As Carroll understands them, emotions necessarily involve a cognitive appraisal and bodily change, and the appraisal causes the bodily change.  But research by LeDoux and others has shown that bodily changes are not caused by cognitive appraisals; rather, the perceived bodily change takes place before the appraisal.  This is not a serious objection to Carroll; he simply has the order of the emotional process backwards.  Carroll still seems right in saying that appraisals have a place in our emotional responses to fictions, especially the complex emotions, but his view must be expanded to account for what we now know about how emotions work physiologically.  Second, we need to get clear on the extent of our awareness of our thoughts concerning fictions.  Are we aware of our sorrow over Anna Karenina’s death or our fear of Dracula?  In other words, are these thoughts conscious?

We find ourselves entering the nebulous realm of emotional consciousness.  Is this journey really necessary?  Consider again the problematic emotions we’ve been discussing thus far:  complex emotions, such as pride and jealousy, which, while arguably constituted in part by perceived bodily changes, also seem to require a cognitive appraisal. It’s this appraisal which I will now attempt to elucidate.  In order to explain how thoughts may be the cause of an emotion, I utilize a HOT (higher-order thought) theory.  “HOTs” are common parlance in consciousness studies, most actively employed by David Rosenthal in his HOT theory of consciousness.[36]  Basically, a higher-order thought is a thought about a (first-order) mental state:  a sensation, belief, another thought, an emotional state, and so on. Rosenthal describes this as follows:

Sensing is not…the only way we are conscious of things.  We are also conscious of something when we have a thought about that thing as being present.  I need not see somebody in the audience to be conscious of that person; it’s enough just to have a thought that the person is here.  There is, moreover, no other way we know about of being conscious of things.  So, if we are not conscious of our conscious states by sensing them, the only alternative is that we have thoughts about them—what I have called elsewhere higher-order thoughts.[37]

Here Rosenthal raises two important points for our purposes.  The first has to do with the object of our mental states and is especially important for understanding the paradox of fiction.  Rosenthal makes it clear that we can have HOTs about things that aren’t physically present and that we don’t believe are physically present.  We can have HOTs about our family member across the country, about people long-gone, about hypothetical or counterfactual cases, and even fictional characters.[38]

The second, more important point has to do with higher-order awareness of a mental state.  Arguably, a mental state of which we are not or can never be aware is not a conscious state.  First-order thoughts are conscious in the sense that we are awake while they happen, for example, we are awake while watching a film, and we will always be in some mental state throughout it, but we are not always aware of that state.  Basic emotional responses, such as raised blood pressure, muscular tension, and so on, can take place without one’s being aware of them; recall that Carroll describes these reactions as cognitively impenetrable.  In order for the first-order state, for example, the thought about the film, to be conscious, we must be aware of it.  Note that we need not be conscious of our awareness of a state; the HOT itself need not be conscious. This would involve, Rosenthal argues, a second-order thought:  a thought about the higher-order thought. Rosenthal describes these cases as introspection:  “in which we are deliberately and attentively conscious of our mental states.”[39]

Rosenthal’s basic line is that we must have a higher-order awareness (HOA) of a state in order to be conscious of it. HOA is the key to explaining our complex emotions about fictions.  The HOT itself is not the emotion; remember that emotions need not be cognitive or conscious.  I take the same line of reasoning:  the HOT can cause an emotion by acting as the intentional object of the emotion.[40]  Recall that, according to Carroll’s position, an emotion involves an evaluative attitude towards an internal mental state, rather than towards an external object or state of affairs. Borrowing Carroll’s example, I fear the character Dracula in cases where I have an evaluative attitude about the Count.  The thought that “Count Dracula is dangerous” is the formal object of my emotion, not the state of affairs concerning the Count.  We maintain Carroll’s basic point against the pure-cognitive theory; emotions do not necessary involve beliefs.  I also, like Carroll, reject the third proposition of the paradox; according to a HOT theory we need not have a genuine belief about the existence of a situation in order to have an emotion about it.

We can also go further and show how complex fictional emotions fit into a theory of emotions in general.  According to non-cognitivist theories, we need not be conscious of the cause of an emotion, and thoughts themselves involve a cortical reasoning process that is too slow to account for more basic emotions.  This doesn’t count against a HOT theory, however.  Not all emotions necessitate thoughts, and not all emotions must be conscious.  Subcortical responses may have unconscious causes and never result in thoughts, but they are still emotions.  Thus, this remains a hybrid theory of emotions rather than a pure-cognitivist one.  A HOT’s role becomes apparent for the complex emotions.  Like Ekman, I maintain that many complex emotions require a similarly complex emotional-plot in order for the emotion to occur.  I take the HOTs as the additional step needed to explain the complex emotions about fictions that comprise the paradox of fiction and life in general.  The HOTs represent certain environmental situations and stimuli, something like Prinz’s theory of core relational themes:  a thought about a moral transgression for guilt, “fearing the worst but yearning for better” for hope or “representing a third party for loss of threat to another’s affection” for jealousy.[41]  In these cases, a HOT is required for the emotion to occur.  We need to be aware of those factors in the environment that affect our own well-being and that of people we care about.  I need to have the thought that my partner is being a touch too friendly with another woman in order to feel jealousy; in order to feel pride towards Luke, I need to first have the conscious thought about Luke’s plight.

In sum, a HOT theory of emotions accounts for a wide range of our emotional responses to fictions, from the subcortical, automatic, and unconscious emotions to the complex, rational, and conscious ones, without denying that they are genuine emotions.  Like Carroll’s thought theory, a HOT view denies the third proposition of the paradox of fiction.  We have myriad emotions about fictions in our daily lives, all of which are genuine.

6.  Final thoughts

I have argued in this paper that pure-cognitive judgment theories of emotion cannot solve the paradox of fiction, despite the assumption by several prominent aestheticians, nor do these theories suffice as explanations of the nature of emotions in general.  Because of its adherence to recent neurological findings, Prinz’s neo-Jamesian theory fares better in both of these respects, but it needs further elucidation to account for the problematic complex emotions concerning fictions.

I propose my own account of narrative emotions as a “best of,” building from non-cognitivist theories and Carroll’s thought theory, and modifying both in order to develop a hybrid account of emotions.  A HOT theory can adequately solve the perennial paradox of fiction and account for our intuitions about our emotional responses to fiction:  that we respond to the complex narrative of a fiction and that these emotions are appropriate.

As I’ve stressed throughout this paper, emotions about fictions are a significant part of our daily emotional lives in their own right.  We watch films, view paintings or sculptures, and listen to music, and we are moved by them. In fact, having an emotional experience is one of the primary reasons why we engage in artworks.  As such, it is important for any theory of emotions to be able to account for them adequately.[42]


Katherine Tullman

Katherine Tullmann received her MA in philosophy from University of Missouri-St. Louis and is currently working towards her PhD at the City University of New York Graduate Center.  She works in philosophy of art, mind and cognitive science.

Published on 6 June 2012.


[1]Noël Carroll, The Philosophy of Horror (New York: Routledge, Chapman and Hall, Inc., 1990), p. 87.

[2] Jenefer Robinson, “Emotion:  Biological Fact or Social Construction?” in Thinking About Feeling, ed. Robert C. Solomon (Oxford:  Oxford University Press, 2004) pp. 28-43; ref. on p. 28.

[3] Noël Carroll, The Philosophy of Motion Pictures (Malden, MA:  Blackwell Publishing Ltd, 2008), p. 149.

[4] Carl Plantinga, “Scene of Empathy and the Human Face on Film,” in Passionate views:  film, cognition, and emotion, ed. Carl Plantinga and Greg M. Smith (Baltimore:  Johns Hopkins University Press, 1999) pp. 239-256; ref. on p. 240.

[5] Ibid., p. 242.

[6] Ibid., pp. 242, 243.

[7] Noël Carroll, “Film, Emotion, and Genre,” in Philosophy of Film and Motion Pictures, eds. Noël Carroll and Jinhee Choi (Malden, MA:  Blackwell  Publishing Ltd., 2006) pp. 217-233; ref. on pp. 217-218.

[8] Kendall Walton, Mimesis as Make-believe (Cambridge, MA:  Harvard University Press, 1990), p. 197.

[9] Martha Nussbaum, “Emotions as Judgments of Value and Importance” in Thinking About Feeling, ed. Robert C. Solomon (Oxford:  Oxford University Press, 2004), pp. 183-199; ref. on p. 195.

[10] Ibid., p. 194.

[11] Ibid., p. 188.

[12] Quoted in Paul Griffiths, “Is Emotion a Natural Kind?” in Thinking About Feeling, ed. Robert C. Solomon (Oxford: Oxford University Press, 2004), p. 245.

[13] Nussbaum, p. 189.

[14] Walton (1990), p. 196.

[15] Walton, Kendall, “Fearing Fictions,” in Philosophy of Film and Motion Pictures, ed. Noël Carroll and Jinhee Choi (Malden, MA:  Blackwell Publishing), pp. 234-246.

[16] Nussbaum, p. 188.

[17] Ibid., p. 188.

[18] Robinson, p. 36.

[19] Joseph LeDuox, The Emotional Brain (New York:  Simon & Schuster, 1996), pp. 163-65.

[20] Ibid., p. 163

[21] Ibid., p. 299.

[22] Ibid., p. 36.

[23] Ekman attributes this point to Philip Johnson-Laird and Keith Oatley.

[24] Ibid., p. 44.

[25] Ibid., p. 45.

[26] Jesse Prinz, “Embodied Emotions” in Thinking About Feeling, ed. Robert C. Solomon (Oxford:  Oxford University Press, 2004), pp. 44-58; ref. on p. 49, 50.

[27] Ibid., p. 50.

[28] Ibid., p. 50.

[29] Ibid., p. 53.

[30] See Prinz, Gut Reactions, chapter 4.

[31] Ibid., p. 16.

[32] Carroll (1990), p. 88.

[33] Ibid., p. 80.

[34] Ibid., p. 81.

[35] Robert Solomon, “Emotions, Thoughts and Feelings,” in Thinking About Feeling, ed. Robert C. Solomon (Oxford: Oxford University Press, 2004), pp. 76-90, ref. on p. 80.

[36] See David Rosenthal, Consciousness and Mind (New York: OUP, 2005) for a complete description of this theory.

[37] David Rosenthal, “The Timing of Conscious States,” Consciousness and Cognition 11, (2002) 215–220, 658.

[38] Note that this account of the object of our thoughts is also available to first-order theories of consciousness; I take this to be a basic assumption across theories of consciousness.

[39] Ibid., p. 658.

[40] Carroll specifically discusses emotions about artworks, but also seems to suggest that a thought theory applies to emotions in general.  See Carroll (1990), chapter 2.

[41] Prinz, p. 16.

[42] Thank you to Brit Brogaard and the anonymous reviewer for Contemporary Aesthetics for their helpful comments on earlier drafts of this paper.