Near-final draft

Emotion
Near-final draft for submission to:
Sotirova, V. (ed) (forthcoming 2016) The Routledge Companion to Stylistics,
London: Routledge.

1. Introduction
Until recently, emotion has been largely implicit in stylistic discussions of literary
effect. Since the development of cognitive poetics (or cognitive stylistics), emotion
has received more explicit attention. Although early cognitive poetic work tended to
focus on information processing at the expense of emotion, this strand of stylistics is
strongly influenced by the experientialism of cognitive linguistics, and research in
the last decade or so has seen increasing investigation into the way reading interacts
with a reader s location, mood, disposition, attention, motivation, sensations,
feelings, emotions and self-awareness (e.g. Burke 2011; Freeman 2009; Gavins 2007;
Herman 2009; Kuiken et al. 2004; Lahey 2005; Miall 2006; Oatley 2003; Stockwell
2005, 2009; van Peer 1997; Whiteley 2010, 2011, 2013). One aim of this work is to
develop analytical frameworks capable of accounting for the emotional experiences
which occur as part of literary discourse.

In this chapter, I will present a cognitive poetic analysis of some of the emotional
effects generated by an extract from Kazuo Ishiguro s novel Never Let Me Go (2005).
My analysis will use the cognitive linguistic discourse processing framework Text
World Theory (Werth 1999; Gavins 2007). I will briefly consider some useful models
of emotion from cognitive and social psychology before indicating the suitability of
Text World Theory for the analysis of emotional experience (see also Gavins 2007;
Lahey 2005; Stockwell 2009; Whiteley 2010). As the notion of projection has been
particularly powerful in Text World Theory accounts of emotion to date, I shall go
on to exemplify this in relation to the extract from Never Let Me Go. But the extract
under consideration also creates suspense as readers wonder what will happen next.
I shall argue that these participatory responses Gerrig are an aspect of
emotional response that have yet to be explored in Text World Theory. I will
propose a means of including them in the text-world framework through the
creation of participation-worlds , but also highlight the challenges that they pose to
the framework s parameters. My aim is to illustrate both some of the existing
successes and the future challenges in the stylistic analysis of emotional experience.

2. Emotion

732
Near-final draft

2.1 Defining emotion

As emotion is so fundamental to human experience, research into the phenomenon
spans many disciplines, including anthropology, psychology, computer science,
neuroscience and philosophy. Although the majority of people have an intuitive
sense of what the word emotion means when they use it in everyday discourse,
there is surprisingly little consensus over definitions or models of emotion within
emotion research. It seems likely that when we talk about emotion in everyday
contexts, we are actually talking about an incredibly complex range of phenomena
Griffiths . Discussions of emotion are further vexed by a proliferation of
overlapping terms, such as feeling and affect , which are defined differently by
different researchers, and often refer to phenomena differentiated by duration,
intensity or structure. My aim here is not to define emotion as a phenomenon
outside of discourse, but to consider how some aspects of emotion may be discussed
in cognitive poetic terms and in stylistic textual analysis. However, it is still useful to
bear in mind some of the ways emotion is theorized. Models of emotion from
cognitive and social psychology are especially applicable to cognitive poetic
discussions of emotional experience, and I will briefly review them below.

In cognitive psychology, appraisal theory models of emotion are particularly

influential (e.g. Arnold 1960; Frijda 1986; and see Moors et al. 2013 for a
contemporary review). According to these models, emotion arises when we appraise
or evaluate a particular event or situation as being in some way relevant to our
goals, plans, expectations or desires. This appraisal causes the changes in
physiology, actions and expressions that typically constitute an emotion. Appraisal
theory is particularly useful for explaining how the same stimulus can cause
different emotional responses in different individuals. As Ortony et al (1988: 4) note,
the elation and devastation of the winning and losing fans at a football match is a
good example: both sets of fans are responding to the same situation, but their
appraisal of events differs. In these models, emotion usually functions to activate
bodily resources and direct attention so that we are able to modify or cope with the
situation that triggered the appraisal. But imagined situations such as those found in
literary texts are also able to activate our appraisal processes (see below, and Oatley
1994, 1999a, 1999b, 2002, 2009; Robinson 2005).

Appraisal theories focus on the role of emotions in mediating an individual s goals

and environment, while emotion theories in social psychology emphasize emotion s
role in the interpersonal relationships we have with others. Parkinson et al. (1995,

733
Near-final draft

2005) argue that emotions are inextricably linked to the identities and roles that
people are performing in particular social situations. In these models, emotions
perform a relational, interpersonal function and are seen as communications to others
(Parkinson 1995: 25, 195). The elation and devastation of the football fans mentioned
above would be seen as part of a process of making claims about personal or social
identity in an unfolding social encounter. When we experience emotions alone, such
as during solitary reading, these are regarded as communications to an imagined
audience (Parkinson 1995: 197-8). Other models of emotion incorporate both
cognitive and social aspects, for instance Oatley and Johnson-Laird (1987, 1996) and
Oatley (2009).

In this chapter, I use the term emotion broadly and generally to refer to relatively
brief but intense [felt] experiences which arise in the course of reading Eysenk and
Keane 1995: 435). This includes the experience of recognizable emotions such as
sadness, happiness, fear and so on, and also moments of evaluation such as positive
or negative preferences, likes and dislikes and moments of empathy, sympathy and
identification. Both the cognitive and social views of emotion described above can be
integrated with the Text World Theory framework and applied to literary reading
(e.g. see Whiteley 2010). I will outline some aspects of this synthesis and review
existing text-world approaches to emotion in the following section.

2.2 Text World Theory and Emotion

Text World Theory is particularly suitable for the investigation of emotional
experience because of its holistic approach to discourse (see Werth 1999 and Gavins
2007 for a detailed account of the framework). The theory posits that all linguistic
communication involves two fundamental levels. The first is the level of the
discourse world , containing at least two human discourse participants, their
situational context, and all their stores of knowledge and other mental resources. The
second is the level of mental representation, the text-worlds that those participants
imagine as they create and conceptualize linguistic communication. Linguistic cues
in the discourse known as world-builders and function-advancers specify the
content of text-worlds (see Gavins 2007: 36, 56). These cues also activate a
participant s relevant discourse-world knowledge and inferencing processes. The
principle of text-drivenness specifies that the text determines which areas of
knowledge and experience are relevant to the processing of the discourse (Werth
1999: 149-53; Gavins 2007: 29). When reading a nineteenth-century novel, for
instance, your knowledge of how to reboot a computer is likely to remain redundant

734
Near-final draft

(Gavins 2007: 29). This enables Text World Theory to offer a systematic discussion of
the potentially vast discourse-world context. Participants use their knowledge stores
and inferences to flesh out their mental representations, and like the discourse-
world, text-worlds can be rich, detailed worlds taking place through space and time
and containing life-like entities and objects. But text-worlds exist at a different
ontological level from the discourse world because they are mental representations
created by discourse-world participants (see Gavins 2007: 76).

To date, most text-world approaches to emotion and the experience of reading have
focused on the implications of the ontological boundary between the real discourse-
world and the text-worlds which readers imagine. Sensations of immersion and
emotional involvement are thought to arise from the various ways in which readers
transgress this ontological boundary, psychologically projecting into the text-worlds
they create (Kuiken et al. 2004; Lahey 2005; Gavins 2007; Stockwell 2009; Whiteley
2010). Whiteley (2011) reviews three interrelated forms of narrative projection. In
deictic projection (Bühler 1982; Segal 1995) a reader maps their embodied sense of
space and location from discourse-world to text-world, aligning themselves with the
spatio-temporal perspective in the text. This is thought to create sensations of
immersion in the text-world. In perspective-taking projection, readers use cues in the
text to imaginatively reconstruct the minds of text-world characters; their
worldview, attitudes, emotions, goals and so on (Palmer 2004; Stockwell 2009;
Zunshine 2006). This common form of projection enables discourse-world
participants to treat text-world entities as real , life-like people who have thoughts,
emotions and reactions in the same way as discourse-world participants and
emotionally respond to them as such (Gavins 2007: 42–3). In processes of self-
implication or identification, readers recognize similarities between aspects of their
own personalities and those of text-world characters, so that their sense of self is
involved in mappings between the discourse-world and text-worlds (Kuiken et al
2004; Gavins 2007; Stockwell 2009). Identification is thought to result in close,
emotionally significant connections between readers and text-world entities (see
Gavins 2007; Stockwell 2009; Whiteley 2013, who also discuss the opposite of
identification: disassociation or resistance).

Text-world accounts of the emotional effects of projection align well with the social-
psychological model of emotion described above, as they posit that emotion arises
from the positions or roles that readers adopt in relation to text-world entities.
Literary narrative is thought to involve multi-levelled interaction between authors,

735
Near-final draft

readers, narrators, narratees and characters, making it inherently social (c.f. the
rhetorical approach to narrative, e.g. Phelan 1996, 2005). It is also possible to
incorporate cognitive appraisal theories of emotion in text-world approaches to
literary reading. The discourse-world level of Text World Theory enables the goals
of the reader to be considered during analysis. Although this is a relatively under-
developed area, it is usually assumed that readers want to create a coherent text-
world representation of the discourse (e.g. Gavins 2007: 19, 141-3) or receive some
kind of emotional pay-out from their investment in reading the text Stockwell
2009: 78-102). If the achievement of these goals is problematized by a text, emotional
responses are likely to ensue. Through processes of perspective-taking projection
and identification, readers are also able to infer and adopt the goals of text-world
characters. Oatley argues that during literary reading, we simulate the plans and
goals of characters on our own planning processes (1994, 1999a, 1999b, 2002, 2003;
Oatley and Gholamain , see also Robinson . As a result, character s goals
and plans become relevant to the readers appraisal processes and they experience
emotions as those plans meet vicissitudes.

The text-world approach to emotion is continually being developed through its

application to a range of texts and discourses. In the analysis below, I use Text World
Theory to consider some of the emotional effects of an extract from Never Let Me Go. I
shall argue that projection can only go some way towards explaining my emotional
experience of the extract. The extract also creates suspense, an experience that has
received little attention in Text World Theory to date.

3. Analysis: Never Let Me Go

Never Let Me Go is set in a counterfactual version of the 1990s and is narrated in the
first person by Kathy, a thirty-one-year-old human clone bred for medical science.
Kathy and her childhood friends Ruth and Tommy (also clones) grow up in a school-
like institution called Hailsham , but once they reach adulthood they are destined to
undergo a series of operations to remove their vital organs, which eventually leads
to their death. Some clones, such as Kathy, do not begin donations straight away and
instead are employed as carers for other clones undergoing operations. Within this
dystopian setting, a love story (of sorts) develops between Kathy and Tommy.
Throughout their school years, Kathy and Tommy had a special connection, but
Tommy became Ruth s boyfriend. Ruth is the first to begin donations, and on her
death bed she encourages Kathy and Tommy to finally be together. Amongst the
clones, there is a rumour that couples who are in love can apply for a deferral from

736
Near-final draft

their donations, granting them a few years reprieve before their operations begin.
Filled with hope, Kathy and Tommy seek out the headmistresses from Hailsham,
(Miss Emily and Madame) to apply for a deferral. But Miss Emily and Madame
inform them that deferrals are a myth, and their hopes are unfounded. There is to be
no escape from their fate, and their lives must now run the course that has been set
for them. Indeed, by the end of the novel, Tommy has died and Kathy is due to
begin her donations in eight months time.

The extract for analysis comes from the section of narrative immediately following
Tommy and Kathy s climactic meeting with the Hailsham headmistresses. Kathy is
driving Tommy back to his hospital. This passage stuck out as a moment of
heightened emotion in my reading of the novel because, prior to this extract, the
clone s reaction to the news about deferrals has not been narrated. I realized how
crushingly disappointing the news was for the characters but also wondered what
might happen next. It reads:

Extract 1
1 We hardly discussed our meeting with Miss Emily and Madame on the
journey back. Or if we did, we talked only about the less important things,
like how much we thought they d aged, or the stuff in their house.
I kept us on the most obscure back roads I knew, where only our
5 headlights disturbed the darkness. We d occasionally encounter other
headlights, and then I d get the feeling they belonged to other carers,
driving home alone, or maybe like me, with a donor beside them. I realised,
of course, that other people used these roads, but that night, it seemed to
me these dark byways of the country existed just for the likes of us, while
10 the big glittering motorways with their huge signs and super cafes were for
everyone else. I don t know if Tommy was thinking something similar.
Maybe he was, because at one point, he remarked:
Kath, you really know some weird roads.
He did a little laugh as he said this, but then he seemed to fall deep
15 into thought. Then as we were going down a particularly dark lane in the
back of nowhere, he said suddenly:
I think Miss Lucy was right. Not Miss Emily.
I can t remember if I said anything to that. If I did, it certainly wasn t
anything very profound. But that was the moment I first noticed it,
20 something in his voice, or maybe his manner, that set off distant alarm
bells. I remember taking my eyes off the twisting road to glance at him, but
he was just sitting there quietly, gazing straight ahead into the night.
A few minutes later, he said suddenly Kath, can we stop? I m sorry,

737
Near-final draft

I need to get out a minute.

25 Thinking he was feeling sick again, I pulled up almost immediately,
hard against a hedge. The spot was completely unlit, and even with the car
lights on, I was nervous another vehicle might come round the curve and
run into us. That s why, when Tommy got out and disappeared into the
blackness, I didn t go with him. Also, there was something purposeful
30 about the way he d got out that suggested even if he was feeling ill, he d
prefer to cope with it on his own. Anyway, that s why I was still in the car,
wondering whether to move it a little further up the hill, when I heard the
first scream.
(Ishiguro 2005: 269, my line numbering)

There are two text-worlds underpinning this passage, which reflect the distinction
between narrating-time and narrated-time. At the (unshown) opening of the novel,
Kathy-the-narrator uses the present tense to refer to the moment of her narration and
is established as existing within a text-world set in England in the late 1990s. Extract
1 occurs around twenty pages from the end of the novel, and in it Kathy-the-narrator
is telling us about past events. This cues a world-switch to another text-world
located at a point in time prior to the moment of narration. So in Extract 1, Kathy-
the-narrator creates and updates a text-world containing enactors of herself and
Tommy, located in the countryside in the South of England. Through reference to
object such as headlights , back roads and dark lane[s] , we can infer that the
characters are in a car and that it is evening time. Function-advancing elements in
the narrative focus primarily on the characters speech, thought, actions and
mannerisms. This is apt, because (as noted above) readers are likely to be interested
in finding out the characters reactions to the bad news they have just received.

3.1 Projection and Emotion

According to Text World Theory accounts of projection, readers of this extract are
likely to use textual cues describing behaviour, thoughts and speech to model the
minds of the characters (Gavins 2007; Stockwell 2009; Whiteley 2011; see also Palmer
2004; Zunshine 2006). For instance, the conversation between the characters, in
which they avoid the topic of the deferrals discussing less important things
instead, line 2) and lapse into long pauses (lines 14-6) led me to infer that they find
the implications of news too painful to face. As she describes their route in lines 4-11,
the contrast that Kathy draws between the dark, minor country lanes and huge,
glittering motorways communicates her sense of isolation and loneliness and serves
as a metaphor for her realization of the clone s marginal place in society. To me,

738
Near-final draft

these cues make it clear that Kathy is deeply disappointed and feels helpless in the
face of the wider system governing the lives of clones. If they have adopted the goals
of the characters in the novel (as Oatley suggests they might), readers in the
discourse-world may also share Kathy s disappointment here. This was certainly the
case in my reading of the novel. Readers who recognize similarities between their
own emotions and those of the text-world characters are experiencing a form of
identification. This matching of emotion at the discourse- and text-world level is
thought to be crucial in the experience of empathy (see Keen 2007; Sklar 2013;
Stockwell 2009 for further discussion).

While Kathy s mind is relatively easy to model and even identify with, it is more
difficult to infer Tommy s response to the news. His comment about Miss Lucy and
Miss Emily in line 17 suggests that Hailsham and the headmistresses are on his mind
(Miss Lucy was a Hailsham teacher who believed the clones should be told about
their fate, whilst Miss Emily wanted to keep this from them). But projection into the
mind of Tommy is to some extent wilfully obscured by the use of modality in Extract
1. Never Let Me Go is an example of Simpson s category A narrative with negative
modal shading, meaning that epistemic and perception modality appear regularly
throughout the text. In this extract, however, the use of epistemic modality is
particularly heightened and is centred around Kathy s description of her interaction
with Tommy (e.g. I don t know [line 11], Maybe [line , ], he seemed [line ],
I can t remember... [line ], I...noticed [line ] I remember [line ]). Simpson
associates epistemic modality with a distancing effect and the expression of
uncertainty, bewilderment and alienation , . This uncertainty is
compounded by the use of hypothetical constructions if we did [line 2], if I did
[line 18]) that offer contradictory versions of events. In line 19, the narrative becomes
even more ambiguous, as Kathy states she noticed it but does not explain the
referent of the pronoun it . This uncertainty makes it difficult to infer Tommy s
response to the news, and the failure of perspective-taking projection creates an
absence of information (what has Kathy noticed? what is Tommy thinking?), which
creates suspense.

This suspense is heightened by references to the dark, isolated, creepy environment

in world-building adjectives such as obscure , dark , weird , twisting and unlit
and by Tommy s sudden, unexplained request to stop the car (line 23). Kathy s
attempts to retrospectively explain or justify her actions that s why , lines and
32) and the unexplained scream (line 33) contribute to the sense that something

739
Near-final draft

(probably alarming or negative) is about to happen, and prompts readers to

experience suspense about what that something will be. To date, Text World Theory
has not really considered this form of emotional experience. Below, I offer a
preliminary discussion of suspense from a Text World Theory perspective.

3.2 Experiencing suspense

In various fields of literary study, suspense is recognized as an emotional experience
prompted by particular narrative features (e.g. Sternberg 1978; Gerrig 1993; Ryan
2001; Dannenberg 2008). Gerrig (1993: 79) describes suspense as occurring when a
reader lacks knowledge about some sufficiently important element of a story, and he
notes that it is heightened when that element is particularly challenging and the
author is able to sustain suspense over a period of delay. The experience of suspense
is also associated with a reader s ability to imagine various possible outcomes and to
experience hopes, fears or preferences associated with those imagined possibilities
(Sternberg 1978: 65; Gerrig 1993: 77; Dannenberg 2008: 38). Indeed, Gerrig
hypothesizes that the emotional power of many narratives will arise from readers
abilities to situate the outcomes an author describes with respect to a range of
imaginable alternatives .

In Experiencing Narrative Worlds, Gerrig (1993) discusses suspense as part of a

broader exposition of what he calls participatory responses to narrative.
Participatory responses henceforth p-responses) are a wide class of phenomena
that arise from readers active participation in a particular discourse (see also
Albritton and Gerrig 1991; Gerrig 1996; Gerrig and Rapp 2004; Prentice, Gerrig and
Bailis 1997; Rapp and Gerrig 2002, 2006). Gerrig notes that p-responses are closely
related to inferences but, crucially, do not have the same logically inductive or
deductive gap-filling function (Gerrig 1993: 27, 67). Instead, they represent readers
reactions to the events in a narrative; for example, feeling the impulse to shout
Watch out! when viewing a film in which a character is under threat Gerrig
or hoping Don t die! when a character s life is in danger Rapp and Gerrig 2006:
55). P-responses can include readers hopes and preferences regarding possible
future narrative outcomes or their replotting of events after particular outcomes
have been described (Gerrig 1993: 65-96). Significantly for the present discussion,
Rapp and Gerrig (2006) propose that readers participation in narrative discourse can
often lead to their mental representation of the text being enriched with a variety of
mental contents (2006: 55). They cite empirical evidence to suggest that the contents

740
Near-final draft

of readers p-responses, such as Don t die! or Watch out! , are mentally represented
as part of the ongoing text representation Rapp and Gerrig 006: 55).

Dannenberg (2008) makes similar claims about the mental representations involved
in the experience of suspense. In her cognitive investigation of readers engagement
with plot, she posits that suspense stimulates the reader to imagine multiple
versions of the story s future that flicker on the edge of his[/her] consciousness
(2008: 36, 38-9). Dannenberg calls the creation of such imagined alternate scenarios
liminal plotting , and explains

This term refers to the reader s semiconscious mental images of possible

future events that are logical extrapolations of the action, although they are
not depicted in the text itself. … These images can be called liminal precisely
because they are half-formed responses that are evoked in the recipient s
mind at the same moment as the processes the scene taking place in the
actional present of the narrative. (2008: 38)

Dannenberg argues that liminal plotting divides readers attention between the
events in the narrative world(s) and the liminal images in their imagination and, as
a result, diverts readers attention and becomes a highly immersive activity (2008:
38). The ontological status of these possible future events is unknown, and readers
have to continue reading in order to find out whether any of their imagined
outcomes are realized in the narrative (2008: 39).

In accordance with Gerrig and Dannenberg s descriptions, the suspense generated

by Extract 1 prompted me to consider a range of possibilities regarding what might
happen next. The ones that I can recall include Maybe Tommy will commit suicide?
Maybe Tommy is going to cry? Maybe he is going to attack Kathy? Maybe he is
going to try and escape? The reference to the scream in line 33 prompted a revision
of these possibilities, as I expected that Tommy would be involved in some
unpleasant or horrific situation outside of the car: Maybe Tommy is being attacked?
Maybe Tommy is attacking a human? The questions above represent p-responses,
because although they arise from inferencing processes, they do not function to fill
gaps in the text (Gerrig 1993: 67). It would be entirely possible to construct the text-
worlds of Extract 1 without imagining this array of hypothetical future scenarios or
by imagining different ones. As Gerrig suggests, they are features of my

741
Near-final draft

involvement in the discourse and my desire to know what the reaction of the
characters will be or what is happening in the scene.

Although it is concerned with the mental representations that readers generate

during discourse processing, Text World Theory does not currently have a way of
incorporating the kinds of mental representations that Gerrig, Rapp and Dannenberg
describe. However, Gerrig makes strong suggestions regarding the emotional
significance of p-responses, noting that they often encode highly emotional content
(1993: 67). Providing Text World Theory with a means of incorporating such
responses could be a way to further develop the framework s account of the
emotional experience of discourse. In Text World Theory terms, participatory
responses would best be characterized as fleeting, undeveloped worlds that are
created alongside the text-worlds of the narrative. In a text, the presence of modality
and hypothetical constructions create modal-worlds that hold the modalized
proposition separate from its originating text-world (Gavins 2007: 91-125). In quality,
these fleeting, undeveloped worlds are like modal-worlds because they are the
products of a reader s hypotheses, hopes and desires regarding narrative events. But
they are not jointly constructed by the discourse participants or cued by the text in
the same way as text-worlds and modal-worlds, instead being optional extras in a
readers engagement with discourse (see below for further discussion of this issue). I
propose these hypothetical worlds be called participation-worlds in accordance
with Gerrig s notion of participatory response. Participation-worlds are mental
representations created by the reader (or hearer) during discourse processing which
are not part of the communicated text-worlds, but which both reflect and reinforce
participant s involvement in the discourse.

The multiple hypothetical future scenarios I imagined when reading Extract 1 would
appear in a text-world diagram in the manner shown in Figure A. The text-world on
the left, with a solid border, represents the text-world cued by Extract 1. The
multiple worlds stemming from it are participation-worlds containing
representations of imagined narrative outcomes, generated by my participatory
responses to the extract (for the conventions of text-world diagrams, see Werth 1999:
xvi-ii and Gavins 2007).

742
Near-final draft

Figure 30.1: Participatory responses to lines 1-33 of Extract 1 from Never Let Me Go
(Ishiguro 2005: 269).

Participation-worlds encode highly emotional content Gerrig because

they represent readers hopes, fears, preferences and expectations about future
narrative events. As such, they reflect readers emotional engagement in the
discourse and have implications for their emotional experience of the subsequent
text. Short and van Peer (1989: 66 suggest that reading involves a psychological
mechanism of expectancy in which incoming information is matched against reader
expectations, and they claim that this matching process is important in our textual
evaluations. Clashes or conformities between the content of participation-worlds and
the text-worlds of the discourse could lead to further emotional responses as readers
desires are fulfilled or thwarted.

In my reading of Extract 1, the participation-world that represented Tommy s escape

was the one I most desired to see realized in the text-worlds of the discourse. Other
readers often express a similar desire for this outcome. For example, the quotation
below comes from a recorded reading group discussion of the novel:

at the end when Tommy ran out and started screaming when they d found
out and they d been to see Madame, I thought that he was going to run off
and I was like Go Tommy, go, escape! and that never happened … it just
wasn t going to happen was it

743
Near-final draft

As this reader points out, Tommy does not escape. The source of the screaming in
line 34 is revealed to be Tommy having an emotional outburst rather than anything
more sinister. As a child at Hailsham, Tommy was prone to wild tempers, and this is
why his behaviour set off alarm bells in Kathy s mind. Kathy manages to calm
Tommy down, and then the characters simply return to their roles as carers and
donors (Ishiguro 2005: 260). Several critics and reviewers remark that one of the
unusual things about the clones in Never Let Me Go is that they do not try to escape
their fate or even conceive of escape as a possibility (Kakutani 2005; Montello 2005;
Toker and Chertoff 2008: 166). The desire for Tommy to escape is not a result of
identification with the goals of the characters in the text-world (c.f. Oatley 1994,
1999a, 1999b, 2002), but represents a desire of the reader in the discourse-world. By
the end of Never Let Me Go, the hopes of both the characters in the text-world (who
wanted a deferral) and the reader in the discourse-world are likely to be left
unfulfilled, as the novel avoids a conventional happy ending . This clash between
the contents of readers participation-worlds and the contents of the text-worlds is
likely to be important in terms of the emotional effect of the novel.

The concept of participation-worlds provides Text World Theory with a means of

incorporating liminal plotting Dannenberg and the predictions, hopes and
preferences of discourse-world participants (Gerrig 1993). As such, they have the
potential to enable Text World Theory to model new forms of emotional experience.
But, as noted above, participation-worlds differ from the other types of mental
representations that Text World Theory currently models. I discuss the challenges
that participation-worlds pose to Text World Theory further below.

3.4 Participation-worlds: a caveat

Participation-worlds as I conceive them are not constrained by the text in the same
way as text-worlds and modal-worlds. Participation-worlds are best regarded as
indirectly cued by the text. For instance, the uncertainty, foreboding and suspense
created by the language in Extract 1 makes readers wonder what will happen next
and therefore prompts participatory responses. Readers may imagine a range of
hypothetical outcomes and experience hopes and preferences regarding those
outcomes, but the language of the text does not specify what those hypothetical
outcomes might be. The mental content of readers participation-worlds is highly
dependent upon their cultural and experiential knowledge. For instance, the
participation-worlds displayed in Figure A above rely upon my knowledge about

744
Near-final draft

the way people might react to bad news (e.g. with tears), or my previous experience
of narratives that have happy endings (in which characters escape). As such,
participation worlds can represent mental content that is not cued by specific
linguistic features (e.g. world-builders or function-advancers) in the text.

My suggestion that text-world theory include this additional form of mental

representation is not without precedent. For instance, in an unpublished essay,
Millward posits the existence of suspended text-worlds that contain
information about the questions and unresolved aspects of a particular discourse
and are updated with partial solutions and resolutions as the discourse progresses.
Participation-worlds as I conceive of them differ from those proposed by Millward
(2002) because they represent specific scenarios related to the outcome of the plot,
rather than being a kind of meta-world in which all uncertainties and questions are
stored for future processing. Elsewhere, Gibbons (2010 draws on Dannenberg s
terminology to explain the way discourses can create hypothetical worlds that have
a liminal status . Her analysis describes a short teaser video for the popular U.S.
TV show Lost. The video suggests a possible outcome of the TV show s plot
(involving the romantic union of several characters) but does not confirm whether
the outcome is actually featured in the show. Gibbons argues that the hypothetical
world is thus given a liminal status as readers are likely to bear it in mind as a
potential outcome when they are next watching Lost. Participation-worlds also differ
from the worlds that Gibbons describes because they are not directly created or
referenced by the text itself.

As such, participation-worlds represent a significant departure from the forms of

mental representation described in Text World Theory by Werth (1999) and Gavins
(2007). Werth s principle of text-drivenness is an ingenious method of
streamlining the context of discourse to a level that can be usefully managed by
analysts. If Text World Theory were to begin incorporating mental representations
whose contents were relatively unconstrained by the text in a manner similar to that
demonstrated above, it may be better equipped to explain certain aspects of our
emotional experience of discourse, but it also stretches the boundaries of the
framework to its limits and risks the loss of some of its precision and analytical
appeal. Indeed, Gerrig (1993: 67-8) notes that p-responses are a heterogeneous class
for which there is no detailed taxonomy, so it is difficult to conceive of the limits of
participation-worlds. Whether participation-worlds represent a useful addition to

745
Near-final draft

Text World Theory is a matter for further debate and examination in future
expositions of the framework.

4. Conclusion
In my analysis of an extract from Never Let Me Go, I have argued that processes of
projection and identification are only able to account for some aspects of the text s
emotional effect. This is because mind-modelling is deliberately problematized by
the text s heightened use of epistemic modality (see Whiteley 2014 for discussion of
problematic projection in other areas of the novel). As a result, the text withholds
information and creates suspense, prompting readers to imagine what might happen
next. I have proposed that participation-worlds based on Gerrig s notion of
participatory responses) be added to Text World Theory s existing taxonomy of the
mental representations involved in discourse processing. Participation-worlds reflect
readers hopes, preferences and predictions about possible future narrative
outcomes. I have also suggested that the contents of participation-worlds are
compared with the contents of the ongoing text-world, with potential emotional
effects. The notion of participation-worlds may help Text World Theory to explain
how readers hopes, preferences and expectations enter into their emotional
experience of discourse. But further work is needed in this area in order to determine
the suitability of Text World Theory for the representation of this and other types of
emotional experience.

To conclude, the cognitive stylistic understanding of emotion is an area of stylistics

that is sure to grow and develop in future years. In this chapter, I hope to have
introduced readers to some of the central principles in the Text World Theory
approach to emotion. I also hope to have shown that attempting to integrate
emotional issues into stylistic frameworks is a profitable enterprise, because it
enables the testing and development of those frameworks in accordance with the
progressive spirit of the discipline.

746
Near-final draft

References
Albritton, D. W. and Gerrig, R. J. (1991), Participatory responses in text
understanding , Journal of Memory and Language, 30 (5): 603-26.
Arnold, M. B. (1960), Emotion and Personality: Vol. 1, Psychological Aspects, New York:
Columbia University Press.
Bühler, K. (1982), The deictic field of language and deictic worlds , in R. J. Jarvella
and W. Klein (eds), Speech, Place and Action: Studies in Deixis and Related Topics,
9-30, Chichester: John Wiley and Sons.
Burke, M. (2011), Literary Reading, Cognition and Emotion: An Exploration of the Oceanic
Mind, London: Routledge.
Dannenberg, H. P. (2008), Coincidence and Counterfactuality: Plotting Time and Space in
Narrative Fiction, Nebraska: University of Nebraska Press.
Eysenk, M. W. and Keane, M. T. (1995), Cognitive Psychology, 3rd edn, Hove:
Psychology Press.
Freeman, M. (2009), Minding feeling, form, and meaning in the creation of poetic
iconicity , in G. Brône and J. Vandaele (eds), Cognitive Poetics: Goals, Gains and
Gaps, 169-96, Berlin: Mouton de Gruyter.
Frijda, N. (1986), The Emotions, Cambridge: Cambridge University Press.
Gavins, J. (2007), Text World Theory: An Introduction, Edinburgh: Edinburgh
University Press.
Gerrig, R. J. (1993), Experiencing Narrative Worlds: On the Psychological Activities of
Reading, New Haven: Westview Press.
Gerrig, R. J. (1996), Participatory aspects of narrative understanding , in R. J. Kreuz
and M. S. MacNealy (eds), Empirical Approaches to Literature and Aesthetics, 127-
42, Norwood: Ablex.
Gerrig, R. J. and Rapp, D. N. (2004), Psychological processes underlying literary
impact , Poetics Today, 25 (2): 265-81.
Gibbons, A. (2010), Lost in cognitive narratology , conference paper presented at the
International Association of Literary Semantics conference, University of
Genoa, Italy, 24-28 July.
Griffiths, P. E. (1998), Emotions , in W. Bechtel and G. Graham eds , A Companion to
Cognitive Science, 197-243, Oxford: Blackwell.
Herman, D. (2009), Cognitive approaches to narrative analysis , in G. Brône and J.
Vandaele (eds), Cognitive Poetics: Goals, Gains, and Gaps, Berlin: Mouton de
Gruyter, pp.79-118.
Ishiguro, K. (2005), Never Let Me Go, London: Faber and Faber.

747
Near-final draft

Kakutani, M. (2005), Books of the times sealed in a world that s not as it seems , The
New York Times, April 4,
<http://www.nytimes.com/2005/04/04/books/04kaku.html> (accessed
14/08/2009).
Keen, S. (2007), Empathy and the Novel, Oxford: Oxford University Press.
Kuiken, D., Miall, D. S. and Sikora, S. (2004), Forms of self-implication in literary
reading , Poetics Today, 25 (2): 171-203.
Lahey, E. (2005), Text-World Landscapes and English Canadian National Identity in
the Poetry of Al Purdy, Milton Acorn and Alden Nowlan , unpublished PhD
thesis, University of Nottingham.
Miall, D. S. (2006), Literary Reading: Empirical and Theoretical Studies, New York: Peter
Lang.
Millward, J. (2002), Text world suspension and retrieval in The Blind Assassin ,
unpublished essay, The University of Sheffield.
Montello, M. (2005), Novel perspectives on bioethics , The Chronicle of Higher
Education 51 (6): 6.
Moors, A., Ellsworth, P. C., Scherer, K. R. and Frijda, N. (2013), Appraisal theories of
emotion state of the art and future development , Emotion Review, 5 (3): 119-24.
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. (1999a), Meetings of minds dialogue, sympathy and identification in
reading fiction , Poetics 26: 439-54.
Oatley, K. (1999b), Why fiction may be twice as true as fact: Fiction as cognitive and
emotional simulation , Review of General Psychology, 3 (2): 101-17.
Oatley, K. (2002), Emotions and the story worlds of fiction , in M. C. Green, J. J.
Strange and T. C. Brock Narrative Impact, 36-69, London: Lawrence Erlbaum.
Oatley, K. (2003), Writingandreading the future of cognitive poetics , in J. Gavins
and G. Steen (eds), Cognitive Poetics in Practice, 161-73, London: Routledge.
Oatley, K. (2009), Communications to self and others emotional experience and its
skills , Emotion Review, 1 (3): 206-13.
Oatley, K. and Gholamain, M. (1997), Emotions and identification, connections
between readers and fiction , in M. Hjort and S. Laver eds , Emotion and The
Arts, 263-79, Oxford: Oxford University Press.
Oatley, K. and Johnson-Laird, P. N. (1987), Towards a cognitive theory of emotions ,
Cognition and Emotion, 1: 29-50.
Oatley, K. and Johnson-Laird, P. N. (1996), The communicative theory of emotions
empirical tests, mental models, and implications for social interaction , in L. L.

748
Near-final draft

Martin and A. Tesser (eds), Striving and Feeling: Interactions Among Goals, Affects
and Self-Regulation, 363-93, Mahwah: Lawrence Erlbaum.
Ortony, A., Clore, G. L. and Collins, A. (1988), The Cognitive Structure of Emotions,
Cambridge: Cambridge University Press.
Palmer, A. (2004), Fictional Minds, Lincoln: University of Nebraska Press.
Parkinson, B. (1995), Ideas and Realities of Emotion, London: Routledge.
Parkinson, B., Fischer, A. H. and Manstead, A. S. R. (2005), Emotion in Social Relations:
Cultural, Group and Interpersonal Processes, New York: Psychology Press.
Phelan, J. (1996), Narrative as Rhetoric: Technique, Audience, Ethics, and Ideology,
Columbus: Ohio State University Press.
Phelan, J. (2005), Living to Tell About It: A Rhetoric and Ethics of Character Narration,
Ithaca: Cornell University Press.
Prentice, D. A., Gerrig, R. J. and Bailis, D. S. (1997), What readers bring to the
processing of fictional texts , Psychonomic Bulletin and Review, 4 (3): 416-20.
Rapp, D. N. and Gerrig, R. J. (2002), Readers reality-driven and plot-driven
analyses in narrative comprehension , Memory and Cognition, 30 (5): 779-88.
Rapp, D. N. and Gerrig, R. J. (2006), Predilections for narrative outcomes: The
impact of story contexts and reader preferences , Journal of Memory and
Language, 54: 54-67.
Robinson, J. (2005), Deeper Than Reason: Emotion and its Role in Literature, Music, and
Art, Oxford: Clarendon Press.
Ryan, M-L. (2001), Narrative as Virtual Reality: Immersion and Interactivity in Literature
and Electronic Media, Baltimore: John Hopkins University Press.
Segal, E. (1995), Narrative comprehension and the role of deictic shift theory , in J. F.
Duchan, G. A. Bruder and L. E. Hewitt (eds), Deixis in Narrative, A Cognitive
Science Perspective, 3-17, Hillsdale: Lawrence Erlbaum.
Simpson, P. (1993), Language, Ideology and Point of View, London: Routledge.
Sklar, H. (2013), The Art of Sympathy in Fiction: Forms of Ethical and Emotional
Persuasion, Amsterdam: John Benjamins.
Sternberg, M. (1978), Expositional Modes and Temporal Ordering in Fiction, Baltimore:
John Hopkins University Press.
Stockwell, P. (2005), Texture and identification , European Journal of English Studies, 9
(2): 143-53.
Stockwell, P. (2009), Texture: A Cognitive Aesthetics of Reading, Edinburgh: Edinburgh
University Press.
Toker, L. and Chertoff, D. (2008), Reader response and the recycling of topoi in
Kazuo Ishiguro s Never Let Me Go , Partial Answers, 6 (1): 163-80.

749
Near-final draft

van Peer, W. (1997), Towards a poetics of emotion in M. Hjort and S. Laver (eds)
Emotion and The Arts, 215-24, Oxford: Oxford University Press.
Werth, P. (1999), Text Worlds: Representing Conceptual Space in Discourse, Harlow:
Longman.
Whiteley, S. (2010), Text World Theory and the Emotional Experience of Literary
Discourse . Unpublished PhD thesis, University of Sheffield.
Whiteley, S. (2011), Text world theory, real readers and emotional responses to The
Remains of the Day’, Language and Literature, 20 (1): 23-42.
Whiteley, S. (2014), Ethics , in P. Stockwell and S. Whiteley eds , The Cambridge
Handbook of Stylistics, 393-407, Cambridge: Cambridge University Press.
Zunshine, L. (2006), Why We Read Fiction: Theory of Mind and the Novel, Columbus:
Ohio State University Press.

750
Near-final draft

751