A Feeling for Fiction: Becoming What We Behold

David S. Miall and Don Kuiken

© 2001 David S. Miall & Don Kuiken

preprint version: March 14 2001

Paper prepared for The Work of Fiction: Cognitive Perspectives
Conference, Bar-Ilan University, Ramat-Gan, Israel, 4-7 June 2001 http://cogweb.english.ucsb.edu/Culture/WoF/

A revised version appeared in Poetics, 30, 2002, 221-241.

Abstract | Overview | The contributions of feeling to literary reading | Three properties of modifying feelings | The generative power of feelings | The cathartic relationship | References


Feelings during literary reading can be characterized at four levels. First, feelings such as suspense and amusement are reactions to an already interpreted narrative (Hansson, 1990). While providing an incentive to sustain reading, these feelings play no significant role in the distinctively literary aspects of text interpretation. Second, feelings that derive from perceived affinity with an author, narrator, or narrative figure are the outcome of an interpretive process by which a fictional representation is developed. Although important in the reader’s development of a situation model (Kneepens & Zwaan, 1994), these feelings, too, do not derive from distinctively literary aspects of text interpretation. Third, feelings of appreciation (aesthetic pleasure or interest) are an initial moment in readers’ response to the formal components of literary texts (narrative, stylistic, or generic). Although serving to capture and hold readers’ attention (Miall & Kuiken, 1994), these aesthetic reactions only anticipate the level of feeling that will be the main focus of our discussion. This fourth level of analysis involves the modifying powers of feeling that appear to be triggered primarily by the formal and narrative components of literary texts. We will argue that these components interact during reading to produce composite and interactive metaphors of personal identification that modify self-understanding, and that the familiar concept of catharsis (the conflict of tragic feelings identified by Aristotle) identifies one particular form of this more general pattern of response in which feelings evoked during reading interact to modify the reader. During this process, in Coleridge’s words, we "become that which we understandly [sic] behold and hear" (1804).


1. Overview

Poetry [is] a rationalized dream dealing . . . to manifold Forms our own Feelings, that never perhaps were attached by us consciously to our own personal Selves. . . . O there are Truths below the Surface in the subject of Sympathy, & how we become that which we understandly behold & hear, having, how much God perhaps only knows, created part even of the Form.-- (Coleridge, Notebooks, II, 2086)

Feeling has been a focus for empirical studies of literary reading for over ten years. In that time, several different aspects of feeling have been studied, and important proposals have been made, notably by Kneepens and Zwaan (1994) and by Cupchik (1994), both presenting arguments for a typology of feeling responses, and by Oatley (1999; in press), who has presented reading as a form of simulation in which emotion is central. Our own research, beginning with Miall (1988, 1989), has focused on the dynamic aspects of feeling: we have tried to understand the role that feeling performs during reading. We have asked to what extent feeling may be said to guide the reader’s interpretive activity, doing so at a level more fundamental than the cognitive processes involved in reading about which much more is known. One of our proposals has been that, during literary reading, processes of feeling may operate that are distinctive to the literary domain, although this is clearly a contentious proposal and one that is far from being well established empirically. In the present paper we try to integrate within one framework several aspects of the recent empirical work on feeling as experience and as a dynamic process. We describe how during reading, in Coleridge’s words, "we become that which we understandly behold & hear."

Before turning to the literary issues, we should point out that the general literature on feeling in psychology and philosophy is, of course, also far from decisive. To this day several quite fundamental issues remain in dispute. To mention only three: first, the argument about the extent to which feelings are culturally determined continues to be debated with social constructivist positions on the one hand (Harré, Averill) and pan cultural positions on the other (Izard, Epstein). Second, the debate begun by Zajonc (1980) about the "primacy" of feeling over cognition continues unresolved; while opposition to Zajonc’s position has been presented by Oatley and Johnson-Laird (1987), evidence from neuroscience has offered Zajonc some support (LeDoux, 1986). Third, while some scholars (Plutchik, Russell) continue to propose models of emotion in which valence and intensity are the primary variables, other commentators (Griffiths, 1997) have argued that emotions are modular and operate in distinctive ways, thus ruling out the common sense notion that feelings and emotions form a unified system. It is not to be expected, then, that psychological research can offer any straightforward guidance regarding the role of feeling in literary response. Indeed, the position may rather be the reverse: given the detail and complexity that literary response affords to the study of feeling, the conclusions that we will eventually reach about feeling may become important in pointing psychological investigators of feeling in new and more productive directions. Even the distinction between emotions and feelings is rather unclear, but in what follows we will not take time to lay out our own position on this problem but speak of each as the occasion warrants.

The plan of the present paper is as follows. First we will briefly review some of the salient accounts of feeling from the empirical research on literary reading to illustrate the types of feeling that have been proposed and their implications. Second, we will outline the main proposals that have been made for understanding the processes that feeling initiates during literary reading – and here we will primarily describe our own work. Finally, we will offer a more comprehensive view of the modifying processes of feeling that, we believe, better accounts for the power of literary texts and the distinctive mode of engagement that such texts are able to evoke in their readers.

2. The contributions of feeling to literary reading

Feeling in literary response can roughly be sorted into four domains: (1) evaluative feelings toward the text as a whole, such as the overall enjoyment, pleasure, or satisfaction of reading a short story; (2) narrative feelings toward specific aspects of the fictional event sequence, such as empathy with a character or resonance with the mood of a setting; (3) aesthetic feelings in response to the formal (generic, narrative, or stylistic) components of a text, such as being struck by an apt metaphor; and (4) self-modifying feelings that restructure the reader’s understanding of the textual narrative and, simultaneously, the reader’s sense of self. While there is no sharp demarcation between these four domains in readers’ experience – a given moment may contain elements of more than one feeling process – we propose that each feeling domain depends upon characteristically different structures and processes. We will argue that the fourth domain, that of self-modifying feelings, is most likely the one in which we can locate what is distinctive to literary response. In the remainder of this section we will discuss the first three domains.

Readers often explicitly report evaluative feelings of enjoyment, pleasure, or satisfaction, and the pursuit of such feelings may even be the reader’s primary goal. Evaluative feelings are not unique to reading, however, and may regularly be experienced while watching a movie or a television drama. Hansson’s (1990) study of readers of popular fiction is perhaps one of the clearest accounts of these feelings. While Hansson overthrows the common prejudice that such readers are passive, that they read for compensation, or that they are manipulated by their reading, he finds that "relaxation, rest, entertainment, and diversion" are the primary reasons given for reading (p. 285). The main feelings expressed, as indicated by the ratings Hansson collected, are that the books are entertaining, exciting, engaging, or amusing. No questioning of conventional views of life is reflected in these requirements: readers commonly say "they feel as though they had taken part in something they are already familiar with from the reading of other books" (p. 287); they don’t feel after reading that they "have taken part in something new and different" or that they would like "to ponder over the book" (p. 288). These comments affirm that evaluative feelings are not involved in text interpretation; their primary role is to sustain reading.

Because evaluative feelings concern the text as a whole, they often entail feelings about feelings. That is, enjoyment, pleasure, or satisfaction is a common response to narrative feelings, aesthetic feelings, or even self-modifying feelings. For example, readers can take pleasure in the emergence of suspense within the progression of events in a mystery story. In fact, Brewer and his colleagues have found (e.g., Brewer & Lichtenstein, 1982) that readers judge narratives as "well-formed" only when they evoke suspense and satisfy those feelings. Similarly, readers can take pleasure in the aesthetic feelings that arise from their engagement with formal (generic, narrative, or stylistic) features of literary texts; this pattern defines a particular kind of aestheticism. Finally, readers can take satisfaction in the self-modifying feelings that are at work in tragic drama; indeed, such satisfaction can surpass in importance the disturbing feelings that are also initiated by that genre.

Evaluative feelings can be distinguished from narrative feelings, which have been usefully discussed by Kneepens and Zwaan (1994) as "fiction emotions." Narrative feelings are prompted by events and characters in the imagined world of the text. Following Kneepens and Zwaan, it seems useful to subdivide narrative feelings into those involving reactions to other characters (e.g., sympathy with a character) and those that are shared with other characters (e.g., empathy with a character). In either form, narrative feelings regularly mirror the text rather closely. When readers feel sympathy for a story character, for example, their feelings often reflect the attitudes expressed by a sympathetic narrator. Or, when empathizing in a lively manner with a character’s fear, feeling functions to mimetically represent that aspect of the offered narrative. Oatley (1994) similarly suggests that feeling plays a mimetic role: reading, in his term, is a "simulation": "the central process is that the reader runs the actions of the character on his own planning processes, taking on the character’s goals, and experiencing emotions as these plans meet vicissitudes. . . ." (p. 66).

To mimetically engage a scene in a text, such as imagining oneself in the position of a character, draws upon the same social skills that enable us to understand others and maintain an appropriate stance towards them. For that reason, it seems likely that narrative feelings will show emotions running on conventional, pre-scripted lines. However, as Oatley (in press) points out, through fiction "our emotions may be transformed by having them deepened or understood better, and they may be extended towards people of kinds for whom we might previously have felt nothing." So, unlike Kneepens and Zwaan, in our discussion of modifying feelings, we will go beyond narrative feeling to examine the processes through which transformations of understanding might occur.

Kneepens and Zwaan (1994) suggest that the more indeterminacy a reader encounters in a text – through what they call "abstract and conceptually vague descriptions" – the more the reader will draw on personal feelings to provide orientation. Without disputing their claim about indeterminacy, we also suggest that aesthetic feelings (what Kneepen and Zwaan call "artefact emotions") play a critical role in this process. Aesthetic feelings reflect heightened interest, what readers have in mind when they report that passages within a text are so striking that they capture and hold their attention. We have found that aesthetic feelings in this sense are prompted by the formal (generic, narrative, and stylistic) features of a text and that, in response to such features, readers slow their reading and report greater uncertainty (Miall & Kuiken, 1994). Since it is these moments that especially challenge the reader’s existing framework for understanding, aesthetic feeling may motivate attempts to revise and reconstruct this interpretive framework. Thus, aesthetic feelings may indicate more than appreciation of the formal aspects of a text; they may initiate changes in the reader’s grasp of textual meaning.

Evidence compatible with this proposal is provided in several studies by Cupchik and his colleagues. For example, Cupchik, Oatley, and Vorderer (1998) looked at the role of emotion in response to short stories by James Joyce. After reading each of four story segments (totalling approximately 700 words) that were focused either on description or on the emotions of the main characters, readers answered questions about any emotions they experienced, and whether these were fresh or remembered emotions. Early within the sequence of segments, remembered emotions were more frequent, whereas fresh emotions became more frequent later during reading. If a text such as Joyce’s prompts fresh emotions, however, we might begin to doubt whether it is sufficient to characterize literary reading simply as a mimetic experience. Oatley (in press) himself suggests this, remarking that during reading a fiction may enable a reader to reflect on an emotion: "then the reader may reach an insight, and build a new piece of his or her model of the self and its relations. In other words, some cognitive transformation may result." Although aesthetic feelings, as we noted, may initiate such change, we now want to consider more broadly how feelings may play a determining role in modifying readers’ self-understanding, doing so in ways that are distinctive to the literary domain.

In the following discussion of the modifying power of feelings, we will refer to specific comments about the feelings evoked by a literary text, drawing from two empirical studies with readers. Both studies invited readers to comment on their experience of reading "The Trout" by Sean O’Faoláin. In this story, Julia, a 12-year old girl at her family’s summer home, revisits the Dark Walk in her garden with her younger brother, where she soon finds a live trout trapped in a well, or small pool; after hearing improbable accounts of how it came to be there, and considering its predicament, she eventually goes out alone at night with a container, locates the trout in the dark, and releases it into a nearby river; she tells no-one what she has done. We divided this story into 84 segments (roughly one sentence each). In one study (for which participant numbers below are indicated by P) we asked readers to think aloud in response to each segment of the story as it appeared on a computer screen. In a second study (indicated by R), readers made marginal marks on a paper copy of the story whenever they read a passage that they found striking or evocative; after reading the whole story they returned to each marked passage and commented at length on their experience of it. Readers’ commentaries were tape recorded and later transcribed for analysis.

3. Three properties of modifying feelings

The distinction between remembered emotions and fresh emotions seems to mark two contrasting modes of feeling in literary response. In remembering an emotion during reading we find a similarity between the world of the text and some similarly scripted situation in memory. We recognize a setting as something familiar; a character’s behaviour reminds us of something we have done; or a narrative event is reminiscent of a similar scene from another text. In one typical example, one participant [R209] commented as follows on the opening description of Julia and her brother running through the "Dark Walk": "I particularly related that to myself and to my family, just the images that came back to me were very much of my own family and my brothers." This autobiographical comparison enabled the reader to partially reinstate feelings from an earlier time in life.

In experiencing a fresh emotion, in contrast, we realize something in a literary text that we have not previously experienced – or at least that we have not experienced in the form provided by the text. Another participant [R208] seems to have responded to the description of the "Dark Walk" with fresh emotion, his tone and choice of words suggesting a measure of surprise at his own response: "This brings into my mind the image of a train tunnel, which is hard to shake for the rest of the entire story. I feel it’s a symbol of [a] train and departure, and growth and the development of the character herself." The participant repeatedly returned to the implications of this "symbol," offering an increasingly novel and intricate characterization of Julia’s maturation. What processes enable the development of such a shift in characterization, or, in Oatley’s term, such a "cognitive transformation"? In a previous paper (Miall, 1989) we outlined how three properties of feeling, which have no corresponding mechanism in standard accounts of cognitive processing, enable feeling to be a vehicle for cognitive change. In what follows, we will extend that analysis.

First, in general, the experience of feelings in one situation leads to the re-experiencing of those feelings in situations that are perceived as similar. This tendency toward the reinstatement of previously experienced feelings within familiar settings has given rise to the notion of affective scripts (Tomkins, 1979). However, sometimes the perceived similarities that reinstate certain feelings cut across conventionally scripted boundaries, and such boundary crossings are often unique to the individual. Thus, the reason I might feel the same way about an argument with antagonistic colleague as I do about taking a driving test is that in both cases I sense that I am being evaluated unjustly, on inappropriate grounds; the result is that the feelings and reactions that were present in the first situation, the argument, are now reinstated in a seemingly different one, the driving test (although here they might seem rather "out of place," even to me). As Bower and Cohen (1982, p. 329) rather aptly point out, such boundary crossing can become the source of affective similes and metaphors: for me, the driving test "is" (or "is like") an argument with a distrusted colleague.

During literary reading, we suggest, aesthetic feeling (feeling struck, captured, held) in response to foregrounding provides the reading experience with a diffusely heightened feeling tone, an affective context within which narrative feelings are more likely to cross conventionally scripted boundaries (cf. Kuiken, 1986, for an analogous argument concerning dream formation). In this way, feelings reinstated while reading about a conventionally scripted scene may provide a framework for understanding a subsequent but not "obviously" related scene. The subsequent scene may then become imbued with feelings that the reader had not previously experienced in such situations. For example, for one of our readers [R235], the description of "the fish panting in his tiny prison" at first evoked a closely related image of an elephant confined in the city zoo. This initial image remained within a conventionally scripted scenario for the suffering of animals that are removed from their ecological niche to artificial confinements. However, this reader went on to remark that the trout’s predicament was "like being in a wheelchair or something, you know. You . . . can’t move your limbs or do anything for yourself." This extension of the entrapped animal script to the realm of human disabilities reflects the ease with which boundaries can be crossed during literary reading. Moreover, this particular boundary crossing allowed feelings associated with the entrapped animals script (e.g., claustrophobic constriction) to be reinstated during reflection on human disabilities—and perhaps beyond: this reader remained alert throughout to the affective implications of the trout’s personified predicament.

Second, feeling often exercises anticipatory effects; it alerts us to the fuller significance of an event as that event begins to unfold and prepares us to interpret incoming evidence in a specific way. Anticipation of this kind seems to be one of the fundamental properties of feeling. In an emotion such as fear, the anticipatory component is quite clear: fear mobilizes us to foresee and avoid imminent threat. But a less primal feeling such as nostalgia is also anticipatory: despite its apparently backward-looking stance, the experience of nostalgia unfolds within an experiential scenario in which we sense and then regret the irretrievable disappearance of something we still value. In general, feeling, with its script-like qualities, leads us to act as if certain contingent events were likely to unfold in the immediate or near future.

During reading, this property of feeling enables readers to monitor their ongoing response and shape its significance, as new events fall within the scope of the anticipation or fail to do so. However, during literary reading, if, as we suggest, aesthetic feeling initiates a diffusely heightened feeling tone, the anticipatory function of feeling will alert readers to the novel feeling connotations that emerge when narrative feelings cross conventionally scripted boundaries. For example, the reversible simile relating the person "in a wheelchair" back to "the fish panting in his tiny prison" oriented reader R235 to the trout’s dependency, its inability to do "anything" independently of Julia’s intervention. In this case, that is exactly what story comprehension required, i.e., Julia is the one who eventually responds to the trout’s dependency. Thus, in this case, feeling enables anticipation of an aspect of the trout’s predicament (i.e., dependency) that challenges and reshapes the conventional entrapped animals script.

Third, because it is so central to our sense of ourselves, feeling is generally self-implicating. Thus, during literary reading, when foregrounding accentuates aesthetic feeling and narrative feelings cross conventionally scripted boundaries, the readers’ sense of self will sometimes be challenged. One reader [R220] concluded that Julia exemplified aspects of herself when she was young: "I ended up, after the finishing of the reading, with admiration for the character, because I guess I felt a real kinship with her. She was a character, not unlike myself as a child. I would have liked to save that trout." This reader was gently surprised by the emergence of her admiration for Julia; it recalled a sometimes submerged "heroic" aspect of her younger herself. Through such challenges and departures, literary reading has the capacity to alter the narratives we weave about who we are or wish to become. This self-transformative aspect of reading is not always explicitly manifest; among some of the most sensitive readers, it emerges during reflection on the completed reading experience (Kuiken & Miall, 2001).

In sum, although feeling during reading may be called into play mimetically, as remembered feeling, remembered feelings may evolve in new directions as reading proceeds. Remembered feeling, in other words, does not remain merely replicative; what began as remembered feeling may eventually become fresh feeling. Either the original feeling is modified, or limitations of the original feeling may be shown in such a way that a fresh feeling is created in its place. In several previous studies, we have provided evidence of the modifying power of feeling, in particular showing how aesthetic feelings, i.e., moments of defamiliarization in response to foregrounding, instigate an affectively guided search for alternative interpretations that, in turn, shape subsequent understanding (e.g., Miall, 1989; Miall & Kuiken, 1999; Miall & Kuiken, in press). We have suggested that this sequence, with its affective elaborations, anticipations, and confirmations, is experienced in a pulsing temporal pattern that, to use a musical analogy, has the structure of a fugue (Kuiken & Miall, 2000). We will now specify beyond this model an outline of what we believe to be the dynamics of the modifying process embodied by feeling.

4. The generative power of feelings

Aesthetic feeling, we have suggested, creates a diffusely heightened feeling tone. Within this context each of the three properties of feeling we have outlined endows feeling during literary reading with a certain generalizing power: the cross-domain property enables generalization to other, apparently unrelated instances; the anticipatory property gives feeling a governing role in revealing unexpected aspects of an unfolding scenario; the self-referential indicates the larger implications for the self, whether these are negative or positive. In each case, feeling enables novel aspects of the unfolding narrative world to emerge. In playing this generative role, feeling may thus create unexpected challenges to the reader’s sense of self. This suggests that feeling acts in part by placing a specific instance within a class. This may be a familiar, already established class, in the case of a remembered feeling, or a created, ad hoc class, in the case of a fresh feeling. This complex of processes can be usefully compared with those by which metaphors exert their generalizing and generative effects. We can conceive of literary reading as a source of feeling-guided metaphors, providing in particular metaphors of personal identification (Cohen, 2000). We will now try to clarify that suggestion.

The description of the "Dark Walk" in the first paragraph of "The Trout" reads:

It is a laurel walk, very old, almost gone wild, a lofty midnight tunnel of smooth, sinewy branches. Underfoot the tough brown leaves are never dry enough to crackle: there is always a suggestion of damp and cool trickle.

The foregrounded features here, especially the words "smooth sinewy," and the verbal echo set up by "crackle" and "damp and cool trickle," arouses aesthetic feeling in most of the readers we have studied, increasing the likelihood that narrative feelings cross conventionally scripted boundaries. One reader [P40], for example, commented that it "sounds like a dream, or a nightmare of some sort," and because Julia is said to visit it every time she comes to the garden, "it’s a lot like a dream where leaves are never drying up to crackle, where . . . the same thing happens over and over again." Through the feeling it evokes, the Dark Walk becomes "like" a nightmare, although qualifying this observation with the word "like" suggests that an ad hoc category has probably been created, which involves some (but not all) of the attributes of the nightmare. If this is correct, then the generative power of the category allows a range of additional possible meanings to be activated.

For this reader, one possibility is that reading the passage about the Dark Walk reinstates the apprehensions about inexplicably repetitive events characteristic of a nightmare. Thus, within the reading moment, the emerging impression of the Dark Walk might seem to involve a transfer of those aspects of nightmares to that story setting. However, this makes the reader’s reflection seem like simple generalization, and the generative capacity of this emerging impression is not addressed. In fact, there seems to be a much more complicated interaction underway between the meaning of the Dark Walk and the dreamer’s understanding of nightmares. First, compared to fear, for example, the repetitive nature of events is not a salient characteristic of our conventional understanding of nightmares. Instead, the repetition emphasized by this reader seems to come from the text, as an echo of the "never dry enough" and the "always a suggestion of damp and cool trickle." Second, the fear that is a salient characteristic of nightmares is not straightforwardly a characteristic of Julia’s run through the Dark Walk; the description of her run in the story suggests a mitigating fascination with fear such that the Dark Walk is the "first place" that Julia sought out during her summer visits. More carefully considered, this reader’s experience of the Dark Walk seems like a more complicated confluence of the Dark Walk and nightmares.

This complexity may be compared to that ascribed to metaphor by Glucksberg and Keysar (1990). Objecting to the notions of transfer or generalization that drive comparison or similarity theories, they propose that a metaphor is better understood as a case of class inclusion, as is suggested by familiar nominal metaphors such as "My job is a jail." They point out that a noun can often be used to represent—through exemplification—a class of objects or situations that is temporarily understood to include both the metaphoric vehicle and topic (jail may, according to context, exemplify the class of confining, oppressive places; or, the class of multi-occupant buildings, etc.). A metaphor promotes the prototypic meaning of such an ad hoc category.

They also point out that the ad hoc class has two types of structure. First, it is hierarchical, in the sense that a superordinate term such as food includes the subordinate category vegetable, which in turn includes subordinate categories such as broccoli. To say "My job is a jail" places the job at a subordinate position in the hierarchy. Second, it offers a horizontal structure: for example, where prototypes are well established, broccoli is more prototypic as a vegetable than tomato. The metaphor thus endows "My job" with some of the attributes of the ad hoc class of situations exemplified by "jail" but also including "my job." Similarly, then, the reader’s identification of the Dark Walk as a nightmare nominates the Dark Walk as an instance of the ad hoc class exemplified by nightmare but also including the Dark Walk. It also configures the properties of this class horizontally (sinewy branches, cool trickle) as an ordered series of signs, determined by an interaction between the prototypic attributes of nightmares (e.g., their frightening and repetitive nature) and the constraints imposed by those properties along which Dark Walks may vary (e.g., their enclosing and repetitive nature).

The metaphoric creation of a class in which feelings (e.g., frightening, enclosing) or their cognates are exemplified by the metaphoric vehicle is not, of course, unique to the literary domain. What may be important to literary reading is when those ad hoc categories, allowing the interaction of vehicle and topic, reflect the crossing of relatively "distant" domains under the influence of aesthetic feelings. The distance of the Dark Walk-nightmare relation suggests an incompatibility with conventionally scripted similarities; distance ensures the novel construal of narrative feelings and the related challenge to the reader’s sense of self. In Coleridge’s words, literary reading deals "to manifold Forms our own Feelings, that never perhaps were attached by us consciously to our own personal Selves" – clearly an argument for fresh feelings and their constructive role for the self.

5. The cathartic relationship

Our second main proposal concerns conflicts of feeling. Since a literary text will usually evoke more than one feeling, or evoke opposed feelings, it will also be possible for one feeling to modify another. In our think-aloud transcripts, this may be signalled by a reader’s sense of surprise. For example, when Julia is in the act of rescuing the trout at night, the trout seems violent: "When the body lashed they were both mad with fright." One reader [R228] commented on this phrase, "it was just a shock to find that, . . . as a reaction of the fish, that he’d be lashing so violently. . . . I guess it set up a, sort of a sense of apprehension in me." This was a surprise to the reader, as she said, "compared to . . . the rest of the story." Her opening response to the Dark Walk passage, for example, was "a nice image here, although it may be a bit on the common side"; she was reminded of the trees in Stanley Park at Vancouver; and she had earlier found some of Julia’s behavior irritating. But a few phrases after the surprise at the violence, the reader experiences another surprise. When Julia releases the trout into the river, "She hoped he was not dizzy." "I loved this one," the reader commented. "I immediately thought, will the fish get dizzy, which is perhaps a slightly strange thought . . . [T]hat this child would be thinking that she hopes the fish wasn't dizzy, strikes me as such a nice innocent thought, and child-like." Her tone in this description, suggests a corresponding sense of release in herself as reader, and pleasure in the childish thought that has followed so soon after the intimations of violence.

We will suggest that this sequence of responses constitutes the modifying of a first feeling by a second, and that it can be considered a form of catharsis, although it lacks the more dramatic dimensions of traditional examples. But in this respect the standard accounts of catharsis seem to us to have consistently overlooked one important feature of Aristotle’s somewhat enigmatic account of the tragic emotions. Apart from James Joyce, who pointed this out in Portrait of the Artist, commentators have paid little attention to the opposed nature of pity and fear. Joyce suggested that both terror and pity arrest the viewer, confronting us at the same with the individual being pitied and the "secret cause" of the terror; thus he construed the tragic emotion as static (Joyce, p. 205). In practice, however, as shown in the dramas of Sophocles that were Aristotle’s main examples, fear in the end appears to be modified by pity.

According to Belfiore (1992), Aristotle appears to have meant by catharsis the action of the emotions (of pity and fear) in the service of restoring order, that is, to bring about the "proper state" of the audience. Catharsis is said to act allotropically, separating the worse from the better; thus, it modifies inappropriate emotions. As Belfiore summarizes it, "This shock of tragic emotion opposes and counterbalances our shameless desires and beliefs" (p. 345). Both Belfiore (1992) and Nussbaum (1986), two important recent commentators, have seen hubris, or shamelessness, as the emotion that the Greek dramatists aimed to cure; thus pity and fear are administered as opposites, not because they are of especial value in themselves, but because they convert the aggressive and potentially dangerous effects of hubris to a more respectful, balanced stance towards the world. In Belfiore’s account, in tragic catharsis "the initial shock of extreme pity and fear passes off, along with the shameless emotions they have ‘mastered’" (p. 345).

In terms of class inclusion, however, what appears to be taking place in a drama such as Oedipus is, that with the disaster of the play, hubris is relocated within the context of fear. The pride and prowess of Oedipus have, in the words of the Attendant, brought about "Calamity, death, ruin, tears, and shame, / All ills that there are names for" (Sophocles, 1947, p. 61). This radical manoeuvre, which literary texts are able to carry out for their audience, redefines the emotion of hubris in order to impose upon it the prototypical attributes of fear, thus transforming its familiar meaning (presumably a chastening experience for the ancient Greeks who watched this happen on stage). But fear in turn is then modified by its relocation within the context of pity in the final scene of the play. Most poignantly, this is witnessed in Oedipus’s appeal to Creon over his daughters: "Creon – If I could touch them once, and weep" (p. 66).

A similar process can be observed in the tragic process enacted in King Lear, where (roughly speaking) hubris predominates in the figure of Lear in Acts I and II, fear in the mad scenes of Acts III and most of IV, and pity finally taking over in Act IV, scene vii, when Lear awakens from his madness and seeks forgiveness from the estranged Cordelia. If feeling considered as class inclusion instantiates a process similar to metaphor, then the apparent reversal of meaning we have described here can also be illustrated by metaphor: for example, as Glucksberg and Keysar (1990) point out, the metaphor "the surgeon was a butcher" reverses the customary attributes of surgeon by including the surgeon in the class of butchers. Thus, to contextualize fear within the class represented by the feeling of pity, radically qualifies our understanding of fear in plays such as Oedipus and King Lear, appearing to humanize it and to ameliorate the immoderate aspects of its power.

This radical qualification of one emotion by another in our rereading of catharsis, suggests that Aristotle’s catharsis is likely to be a special case of a more general process in literary reading. But in looking for examples of this process, we should not limit ourselves only to those that are formally structured into the text itself, such as Oedipus or King Lear, since the indeterminate nature of response to literature will often ensure that readers’ feelings vary, so that different, specific confrontations of one emotion with another will occur. In the case of the "Dark Walk," for example, what was felt as nightmare for one reader, as we have seen, was experienced by another as a pleasurable reminder of the woods of Stanley Park; so too, the subsequent feeling that confronts or modifies the first feeling may also be distinctive to that reader. But the process will depend in part on how compelling are the emotional qualities of the text in question, and what personal concerns, memories, and experiences a reader brings to the text.

Most literary reading is less arresting than witnessing Oedipus or King Lear, yet narratives can unobtrusively implicate our own feelings as readers, leading us to moments of recognition or acknowledgement that signal change. We now examine more closely the cathartic process, as we have redefined it, in the moments of transition towards the end of the story, confining our examples to just one of the protocols cited above.

For this reader, as we noted, the opening "Dalk Walk" passage "sounds like a dream, or a nightmare," attended with some foreboding – a felt but unspecified threat. For this reader Julia’s position as a maturing child is a recurring theme. When Julia is shown considering the trout in the middle of the story, the reader notes her preoccupation and suggests "She’s subconsciously thinking about herself, and she wants her freedom." At the same time, he finds the description of the trout "very dreamlike": "the light [is] shining off the silver belly, on the smooth stones and water, and the rest is all dark." However, when Julia at last goes to rescue the trout at night, the "nice little dreamworld" of daytime has become "more of a nightmare," and the entrance to the Dark Walk, described as a "maw," now "sounds very foreboding." At this point the maturity issue becomes central, as the reader comments, "So she’s out there in her pyjamas. So maybe she’s outgrown her childhood." When Julia releases the trout, the reader notes both Julia’s hope that the trout would not be "dizzy" and several other images that seem childlike, yet Julia’s childhood nature has now been recontexualized by her ability to set aside her possessiveness about the trout: "She realized that he would die if she left him there, kept him trapped, so she’s let him [go], sacrificed her enjoyment for his life. She’s become more, uh, aware of other things around her." In his response to the next sentence, where Julia is shown noticing her surroundings, the reader comments "suddenly she, she sees everything. It’s as if she sees the truth. The river, the moon, trees, mountains, what life is all about, I think."

Earlier in the story, despite his awareness of the maturity theme, this reader had at times been impatient with Julia, seeing her, as he put it, "clinging onto her childhood in some ways." Julia’s growth and the life or death of the trout was, in his words, perhaps only a part of the "cycle of life." The outcome at this climactic point of the trout’s release, however, appears to exceed his expectations. If Julia looks around and "sees the truth," it is for this reader as though Julia’s action brings her into alignment with a larger reality, beyond the egotism of the child made evident earlier. What she sees at that moment stands for the class of insights into "what life is all about." The reader’s comments imply that, like her regard for the trout itself, Julia’s understanding has shifted from an instrumental, self-serving attitude towards the world around her to a more inclusive recognition that things have an end in themselves.

Thus in this phase, towards the end of the story, the reader acknowledges a shift in his feelings about the character, from a certain detached impatience to admiration. In a number of protocols readers’ feelings about Julia are typically more complex at this point than during the opening segments of the story, and responsive to aspects that are incongruous with their earlier assessments of her character. A cathartic shift seems to occur at this point for a number of readers, as earlier feelings are recontextualized by other more inclusive feelings.

Finally, in construing Julia’s maturation as a transition into responsibility, not only for her actions, but for her way of construing the world, reader P40 also seems to realize that something is at stake in his own position. In his final comment after reading the story he remarks of Julia:

she’s gained, she’s made the first step towards maturity, although we can’t, uh, you don’t become mature overnight, that you, uh, it takes time, and you’re not aware that you’re becoming mature until many years down the road when you look back and, you can understand what was happening to you.

If the previous phase represented a shift in feeling, here the reader’s reflections on the implications of this cathartic shift have come into focus. The thematic significance of the story has become superimposed upon the reader’s own sense of self. As the repeated use of the generalized pronoun "you" indicates, the reader is enacting the implications of the story for his own understanding, merging the identities of Julia and himself in a single, although complex, perspective on the question of what the story might mean: "she’s begun to understand what it means to be an adult," but "you don’t become mature overnight." Thus a reconfiguration of feelings and beliefs has occurred for the reader by the end of the story. Although, given the nature and relative shortness of the story, this is neither a particularly dramatic nor a striking process, it is one in which readers seem to derive a degree of satisfaction, characteristic of aesthetic experience. The "you" form of response suggests, in other words, how we make the stories we read our own, "becoming what we . . . behold and hear."

The mode of analysis we have attempted here enables us to characterize some of the more salient moments in the evolution of readers’ feelings, as these exhibit specific developments or transitions towards a (provisional) final response. Of particular interest, the role of the self in this process, given the potentially self-referential nature of feeling, remains to be explored. It is apparent that a number of the specific moments we have cited have involved feelings relating to the motives and behaviour of the central character. Self-implicating feelings, occur less often, although the last reader, as we noted, appears to be experiencing such feeling in his closing comments. However, this and other protocols contain references to readers’ feelings frequently enough that we can assume readers regularly recognize and incorporate their own feelings into their reading (in the first study of readers of "The Trout" that we have drawn on here, we find that 3.1% of comments at the segment level refer to readers’ feelings: Miall & Kuiken, 1999, p. 132); they also refer about as often to personal memories evoked by some aspect of the setting or a character (3.6% of comments). Thus, in the background of the response as a whole, it seems likely that readers are often making implicit comparisons between the world of the fiction and what they know, believe, or feel about their own lives, since these comparisons surface from time to time in the form of either explicit feelings or memories. Readers who, in metaphoric fashion, situated their conception of Julia running through the Dark Walk in relation to death-rebirth or a nightmare may, in this manner, be locating the story within a personally significant complex of meaning. If so, the response already signifies the beginning of a modifying process dependent upon the cross-domain role of feeling. In some cases, as the last protocol we have cited shows, the engagement of the self of the reader is manifested explicitly in the occurrence of the "you" form of reflection in which the self and the character seem to have merged. The reader has, we might speculate, confronted some long-standing prior feeling about the self, and recontextualized it in the light of the fresh feelings aroused during his reading of the narrative. This process, if it involves a transformation of existing feelings about the self, thus represents a type of catharsis that, as we suggested, is shaped by the distinctive feelings that a given reader brings to the narrative. This may be one of the most interesting and characteristic effects of literary reading.



Belfiore, E. S. (1992). Tragic pleasures: Aristotle on plot and emotion. Princeton, NJ: Princeton UP.

Bower, G. H. & Cohen, P. R. (1982). Emotional influences in memory and thinking: Data and theory. In M. S. Clark & S. T. Fiske (Eds), Affect and cognition: The 17th Annual Carnegie Symposium on Cognition (pp. 291-331). Hillsdale, N.J.: Lawrence Erlbaum Associates Inc.

Brewer, W. F., & Lichtenstein, E. H. (1982). Stories are to entertain: A structural-affect theory of stories. Journal of Pragmatics, 6, 473-486.

Cohen, T. (1999). Identifying with metaphor: Metaphors of personal identification. Journal of Aesthetics and Art Criticism, 57, 399-409.

Coleridge, S. T. (1957-1990). The Notebooks. Kathleen Coburn (Ed.). 4 vols. London: Routledge.

Cupchik, G. C. (1994). Emotion in aesthetics: Reactive and reflective models. Poetics, 23, 177-188.

Cupchik, G. C., Oatley, K., & Vorderer, P. (1998). Emotional effects of reading excerpts from short stories by James Joyce. Poetics, 25, 363-377.

Glucksberg, S., & Keysar, B. (1990). Understanding metaphorical comparisons: Beyond similarity. Psychological Review, 97, 3-18.

Griffiths, P. E. (1997). What emotions really are. Chicago: Chicago UP.

Hansson, G. (1990). Not a day without a book. Spiel, 9, 277-293.

Joyce, J. (1976). A portrait of the artist as a young man. Harmondsworth, Middlesex: Penguin Books. (Original work published 1916)

Kneepens, E. W. E. M., & Zwaan, R. A. (1994). Emotions and literary text comprehension. Poetics, 23, 125-138.

Kuiken, D. (1986). Dreams and self-knowledge. In J. Gackenbach (Ed.), Sleep and dreams: A sourcebook (pp. 222-247). New York: Garland Press.

Kuiken, D., & Miall, D. S. (2000). Reading expressively through "tears of light." Paper presented at the 7th Conference of the International Association for the Empirical Study of Literature, University of Toronto, July 31-August 6.

LeDoux, J. E. (1986). The neurobiology of emotion. In J. E. LeDoux & W. Hirst (Eds.), Mind and brain: Dialogues in cognitive neuroscience (pp. 301-354). Cambridge: Cambridge University Press.

Miall, D. S. (1988). Affect and narrative: A model of response to stories. Poetics, 17, 259-272.

Miall, D. S. (1989). Beyond the schema given: Affective comprehension of literary narratives. Cognition and Emotion, 3, 55-78.

Miall, D. S., & Kuiken, D. (1994). Foregrounding, defamiliarization, and affect: Response to literary stories. Poetics, 22, 389-407.

Miall, D. S., & Kuiken, D. (1999). What is literariness? Three components of literary reading. Discourse Processes, 28, 121-138.

Miall, D. S., & Kuiken, D. (in press). Shifting perspectives: Readers’ feelings and literary response. In W. Van Peer & S. Chatman, Eds., New Perspectives on Narrative Perspective. Albany, NY: SUNY Press.

Nussbaum, M. C. (1986). The fragility of goodness: Luck and ethics in Greek tragedy and philosophy. New York: Cambridge University Press.

Oatley, K. (1994). A taxonomy of the emotions of literary response and a theory of identification in fictional narrative. Poetics, 23, 53-74.

Oatley, K. (1999). Meetings of minds: Dialogue, sympathy, and identification, in reading fiction. Poetics, 26, 439-454.

Oatley, K. (in press). Emotions and the story worlds of fiction. In T. Brock, J. Strange, & M. Green (Eds.), Narrative impact. Mahwah, NJ: Erlbaum.

Oatley, K. & Johnson-Laird, P. N. (1987). Towards a cognitive theory of emotions. Cognition and Emotion, 1, 29-50.

O’Faoláin, S. (1980-82). The trout. In The collected stories of Seán O’Faoláin, Vol. I (pp. 383-386). London: Constable.

Sophocles (1947). The Theban plays, trans. E. F. Watling. Harmondsworth, Middlesex: Penguin Books.

Tomkins, S. S. (1979). Script theory: Differential magnification effects. In H. E. Howland and R. A. Dienstbier (Eds.), Nebraska symposium on motivation 1978: Human emotion (Vol. 26) (pp. 201-236). Lincoln, Nebraska: University of Nebraska Press.

Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151-175.

Return to Reading

Document prepared March 22nd 2001 / updated October 12th 2007