Text World Theory and real world readers: From literature to life in a Belfast prison.

Cognitive stylistics offers a range of frameworks for understanding (amongst other things) what producers of literary texts 'do' with language and how they 'do' it. Less prevalent, however, is an understanding of the ways in which these same frameworks offer insights into what readers 'do' (and how they 'do' it). Text World Theory (Werth, 1999; Gavins, 2007; Whiteley, 2011) has proved useful for understanding how and why readers construct mental representations engendered by the act of reading. However, research on readers' responses to literature has largely focused on an 'idealised' reader or an 'experimental' subject-reader often derived from within the academy and conducted using contrived or amended literary fiction. Moreover, the format of traditional book groups (participants read texts privately and discuss them at a later date) as well as online community forums such as Goodreads, means that such studies derive data from post-hoc, rather than real-time textual encounters and discussions. The current study is the first of its kind in analysing real-time reading contexts with real readers during a researcher-led literary project ('read.live.learn') in Northern Ireland's only female prison. In doing so, the study is unique in addressing experimental and post hoc bias. Using Text World Theory, the paper considers the personal and social impact of reader engagement in the talk of the participants. As such, it has three interrelated aims: to argue for the social and personal benefits of reading stylistically rich literature in real-time reading groups; to demonstrate the efficacy of stylistics for understanding how those benefits come about, and to demonstrate the inter-disciplinary value of stylistics, particularly its potential for traversing traditional research parameters.

1. The reading context

Reading is an inherently social activity during which ‘readers are engaged prospectively in social encounters from the moment they pick up the new book and open the cover’ (Myers, 2009: 339). The current study derives from the author’s social literary initiative, ‘read.live.learn’,[1] which was established as a way of introducing meaningful and stimulating activity into Northern Ireland’s only female prison. Prior to the project’s inception, the prison environment was found to be apathetic and disheartening, constituted by a ‘debilitating and demoralising regime’ (Scraton and Moore, 2009; see also a prison inspection report HM Inspectorate of Prisons (2013). It was anticipated that the project would improve wellbeing by encouraging social interaction, developing empathy and confidence through reading and responding to literary fiction in real-time read-aloud groups (see for example, Billington, 2011). The weekly groups were held in a small unfurnished room in the education block of the prison and attended voluntarily by between three and twelve prisoners for around two hours. Some of the women knew each other through communal activities in the prison, and some shared the same landing (but cells were not shared). The structure of each session was as follows: as group facilitator I read prose (for example, short stories or chapters from books) and poetry aloud pausing often to allow responses from within the group, and to encourage discussion about real-life issues triggered by the characters and situations represented in the texts. Although not its primary aim, the project also provided a unique opportunity to explore first hand the complex relationships between texts, real readers and real-time reading contexts. This study, therefore, focuses on such interactions and is grounded in stylistics through an application of Text World Theory as a way of understanding and expressing these relationships. Using data collected from detailed observational notes, interviews and written responses to evaluative questionnaires, the paper offers an appraisal of Text World Theory and suggests developments to the framework to account more fully for what I will propose is both a top-down and a bottom-up interpretative process.

2. The framework

Rooted in cognitive linguistics, Text World Theory offers a holistic framework through which to analyse how discourse is conceptually constructed and negotiated by discourse participants. The model is fundamentally concerned with coherence, in that it can be used as a tool for understanding how readers make sense of texts. In so doing, Text World Theory employs the notion of mental spaces (Fauconnier, 1994; Fauconnier and Turner, 2002) or ‘worlds’ that are ‘text-driven’ (Werth, 1999: 103); that is, they are derived from the propositions in a text established through deictic parameters and referential elements. The resulting text-worlds develop through ‘function-advancing’ elements, new information which is added or incremented as the text progresses, which according to Werth (1999: 346) ‘give the text its point’. Participants are salient entities in text-world formation who in a reading situation occupy either the ‘here and now’ of the reading context (readers) or the ‘then’ of the writing context (authors). Both these contexts sit outside the world of the text (although they inform it) and constitute a ‘(split) discourse-world’ (that is, a space that represents the context of a communicative interaction – in the case of reader / author this world is separated by time and space, hence ‘split’). Discourse-world participants construct text-worlds, relying in part on their immediate discourse situation and on what they know or can infer from their own knowledge frames (Werth, 1999: 83). Participants within the text itself are defined as ‘characters’ (and further subdivided into ‘enactors’ and ‘bystanders’) by Werth (1999: 82; see also Emmott, 1992 and Gavins, 2007) whose own mental representations within the text-world often generate other ontologically distinct worlds through the use of, for example, modality – such as when readers are given access to characters’ mental representations – negation, hypotheticality, reported speech or focalization. These other worlds are termed ‘sub-worlds’ (Werth 1999: 55) or ‘world-switches’ (Gavins, 2007). Such world-switches occur when the narrative adopts a different temporal or spatial perspective and are usually triggered deictically (for example, through adverbs such as ‘meanwhile’, ‘yesterday’, ‘then’, ‘now’, ‘here’, ‘there’, and so on).

My rationale for using Text World Theory lies in its ability to account for both the language event and the context or situation in which it occurs, which makes it a very serviceable framework for analyzing readers’ responses to literature as Whiteley (2011) and Peplow et al. (2016) convincingly demonstrate. My analysis departs from current research that posits Text World Theory as a top-down processing model that operates unidirectionally so that discourse-world knowledge is imported to the text-world, rather than text-worlds contributing to discourse-world understanding. Lahey (2015) acknowledges as much when she writes that ‘upward influence from the text-world to the discourse-world level is not predicted by contemporary Text World Theory, in which the text-world is always considered to be a product of the discourse-world, rather than an element in its formation’. I argue that through the concept of ‘projection’, resulting text-worlds can be incremented back into the discourse-world, thus functioning bi-directionally, potentially reifying discourse-world knowledge. Stockwell alludes to such circularity when he talks of a ‘feedback loop’ from the text-world to the discourse-world in the production of ‘empathy’ that ‘indicates a shift back […] towards a realignment of [a] readerly stance’ (2009: 95). However, whilst his suggestion of bidirectionality points to the capacity for text-worlds to inform and reshape discourse-world narratives, his study does not consider how such bidirectionality works in practice. In what follows, I address this gap, and in so doing discuss how the ontological divide between the text-world and discourse-world is blurred and negotiated during reading. The following sections examine the shared reading experiences of a group of female prisoners reading Kate Chopin’s short story, The Story of an Hour (1894). I begin with a short summary of the text, followed by a stylistic analysis of the story that employs the Text World model, before discussing the reception of the text in a real-time reading group in Hydebank prison.

3. The Story of an Hour

In The Story of an Hour, the protagonist, Mrs Mallard, is told that her husband, Brently, has died in an accident.[2] The remainder of the story describes Mrs Mallard’s contemplation of the potential (and thus, hypothetical) impact of Brently Mallard’s death after she retires to her bedroom upon hearing the news. During this period, the heterodiegetic narrator reports some conflicting messages; there is weeping with ‘wild abandonment’ (paragraph 3) which a reader might safely assume implies grief – indeed such a conclusion can be confirmed a few lines later when ‘the storm of grief had spent itself’ [my emphasis] – and then there is ‘exalted perception’ (paragraph 12) followed by an epiphanic realisation (relatively later) that she, Mrs Mallard, is now ‘free’. Shortly afterwards, Mrs Mallard leaves her room and descends the stairs in her home to be greeted by none other than her husband who, to everyone’s surprise (including the reader’s), has just walked in through the front door. Consequently, Mrs Mallard is overcome with shock and in an ironic twist, drops dead of a heart attack.

Stylistically, the text is written in third person narration through an external focaliser (but see below) and opens with a non-finite embedded clause: ‘Knowing that Mrs Mallard was afflicted with a heart trouble, great care was taken to break to her as gently as possible the news of her husband’s death.’ In text-world terms, this constitutes an epistemic modal-world (Gavins, 2007: 110) in which the truth value of a proposition (its epistemic commitment) is encoded in lexical modal verbs (triggered by ‘knowing’). Information about Mr Mallard’s death is enactor-accessible (and therefore unverifiable) but it is incremented nonetheless and is, of course, crucial to setting up the twist. In this emerging text-world the world-builders are the noun phrases, Mrs Mallard, and less so, Brently Mallard; the latter’s referent is what I will call a ghost enactor as he is not a sentient being. Therefore, incrementing this information means accounting for a logical presupposition (that Brently was previously alive) and so requires a second text-world that precedes the initial text-world. This can be seen diagrammatically in Figure 1.

Figure 1.

‘Knowing that Mrs Mallard was afflicted with a heart trouble, great care was taken to break to her as gently as possible the news of her husband's death.’ (In the diagram, ‘c’ refers to characters, ‘t’ to time and ‘l’ to location.).

The presupposition is firstly accommodated then backgrounded. Other information likely to be incremented by discourse participants includes (a) there is ‘news’, (b) that Mrs Mallard has a health issue, (c) that there are text-world enactors who exist who know of Mrs Mallard’s health issue, (d) that the health issue is likely to be adversely impacted by bad news, (e) that there are people who are sensitive to her health issue, and so on.

After the news is broken there is a world-switch in narrative time back to the enactor Richards’s earlier discovery of the accident, ‘it was he [Richards] who had been at the newspaper office when intelligence of the accident had been received’. This world-switch constitutes an enactor accessible sub-world in which Richards ‘receives’ the intelligence via a telegram (Figure 2) which is implied rather than stated (it was ‘confirmed’ through a second telegram that presupposes the first) and are in grey type in Figure 2. The world building (WB) elements in grey type are inferences.

Figure 2.

‘It was he who had been in the newspaper office when intelligence of the railroad disaster was received, with Brently Mallard's name leading the list of "killed." He had only taken the time to assure himself of its truth by a second telegram, and had hastened to forestall any less careful, less tender friend in bearing the sad message.’

Later there is a shift in narrative perspective when Mrs Mallard becomes the focaliser established through the lines ‘she could see in the open square before her house’ in which the prepositional locative adverb ‘before’ along with the non-finite verb form ‘see’ act as world-builders conveying Mrs Mallard’s visual perspective. Her point of view is developed through the deictic parameters of the non-finite verb forms; ‘a peddler was crying his wares’, ‘[…] someone was singing’, and ‘countless sparrows were twittering’ [my emphasis], which position Mrs Mallard as the deictic centre around whom the text-world is being built (Figure 3).

Figure 3.

‘She could see in the open square before her house […]’.

The verb tense forms bring the discourse-world participant closer to the text-world enactor’s present time, the ‘now’ of Mrs Mallard’s subjective experience, which up to this point is mediated through an external focaliser; ‘She was young […]’, before the mode of narration changes with ‘but now […] there was a dull stare in her eyes, whose gaze was fixed away off yonder on one of those patches of blue sky. […]’ [my emphasis]. The proximal deictic adverb ‘now’ along with ‘those’ further aligns the reader with Mrs Mallard’s point of view, narrowing the conceptual distance between enactor and narrator. Finally, a diegetic shift into free indirect discourse generates an enactor-accessible epistemic modal-world:

There was something coming to her and she was waiting for it, fearfully. What was it? She did not know; it was too subtle and elusive to name. But she felt it, creeping out of the sky, reaching toward her through the sounds, the scents, the color that filled the air (paragraphs 9 and 10).

The epistemic modal-world (Figure 4) represents Mrs Mallard’s confusion (signaled by the lexically under-specific ‘something’, the free indirect ‘what was it?’, and the indefinite pronoun ‘it’), and contrasts with (and is emphasised by) the epistemic certainty of the opening line. Thus, the confusion is shared by the enactor and the narrator, and is the point at which readers of this text in my read-aloud groups most often signal projection into the text-world (I will deal with this below).

Figure 4.

Epistemic modal-world: ‘There was something coming to her […]’.

The narrative then briefly returns to external focalization to capture Mrs Mallard’s epiphany which is mediated through direct speech: ‘When she abandoned herself a little whispered word escaped her slightly parted lips. She said it over and over under her breath: "free, free, free!"’ Following this, Mrs Mallard imagines what her life might be like in the future and reflects on how it has been up to the narrative present, resulting in a series of temporal and modal-world-switches (paragraphs 13, 14 and 15).

As the outline shows, text-world networks can account, amongst other things, for various enactor-perspectives, deictic switches, and even narrated but not realized events. However, the analysis above predominantly models an individual’s interpretation of the language of the text during solitary reading. During the shared reading situation of the prison reading group, text-worlds are jointly negotiated by multiple discourse-world participants, and those co-constructed text-worlds feed back into the discourse situation. This latter type of reading has not yet received attention in applications of Text-World Theory. In the following sections I will discuss the interpretative processes that derive from this shared reading situation.

4. Collaborative text-worlds and bi-directionality

The discourse-world of the prison reading group under study involved five participants (a facilitator and four prisoners) sharing the same spatio-temporal environment, a room in the education block of a Belfast prison. As was the case for every reading session, I read the text aloud stopping often at points to allow for reflection and responses. These pauses are naturally occurring in that the prisoners often interject when they want to comment or question the goings-on in the text. As facilitator, I also initiated pauses at key events in the plot as a way of encouraging interaction, or as a response to non-verbal cues from within the group that signalled a desire to probe the text further. The goings-on in Chopin’s The Story of an Hour generated a range of responses that arose during some of these brief pauses, from suspicion, ‘do you think Richards had something to do with it?’ (prisoner A), to surprise at ‘Mrs Mallard’s reaction’ upon hearing the news of her husband’s death. This ‘reaction’ is not Mrs Mallard’s per se, but relates to the ‘happy’ (as one prisoner put it) narrative descriptions of the setting (paragraphs 5 and 6 outlined below) which the prisoners believed were incongruous when placed alongside Mrs Mallard’s solitary grieving:

‘She could see in the open square before her house the tops of trees that were all aquiver with the new spring life. The delicious breath of rain was in the air. In the street below a peddler was crying his wares. The notes of a distant song which someone was singing reached her faintly, and countless sparrows were twittering in the eaves.

There were patches of blue sky […]’ [my emphasis]

Although still in third person narration, presumably (for we did not discuss it) the text in italics could have been perceived by the prisoners to be representative of Mrs Mallard’s thoughts, helped perhaps by the internal focalization and the adjective ‘delicious’ that indicates the move towards free indirect discourse. Whether or not this was the case, it was at this point that one prisoner remarked that ‘you wouldn’t think of rain as “delicious”, or see “new spring life” if your husband just died, would you?’ Her comment drew nods of agreement (although, one woman did quip ‘well you would if you were glad to be rid of him’). The following comments are part of a discussion that followed (the anonymised prisoner responses are taken from my notes):

A1: ‘I can see why she would remember the birds though [when you receive bad news], you just remember the small stupid stuff –’

B1 (interrupts A): ‘Maybe it’s a protection thing’

C1: ‘I remember the sun being out at my da’s funeral – I thought it was lovely and I was pissed off at myself for thinking that’

A2: ‘Yeah, but it’s those wee things you remember – like what was on the TV or songs or stuff, it seems wrong because somebody’s just died’

C2: ‘She’s [Mrs Mallard] probably just in shock’

These comments suggest that the narrative description of the setting (‘new spring life’, ‘delicious’ rain, and so on) mediated through Mrs Mallard’s point of view jars with the theme of the story (grief and death). Moreover, they indicate that through analogical reasoning (in the responses of C1 and A2), and modalised projections (‘she’s probably just in shock’, ‘maybe it’s a protection thing’) prisoners identified with (and evaluated) the text-world enactor’s response to her husband’s loss from differing perspectives in a bid to help the text-world into coherence. In text-world terms, the women constructed what I will call ‘plausible’ text-worlds in which inferences were made as to the (yet unknown) motivation for Mrs Mallard’s incongruous sensory awareness (Figure 5). It is important to note here that the enactor, Mrs Mallard, is not the narrator, and so it is interesting that the group attributed the narrative description as representing the enactor’s direct perception. This is perhaps a result of the narrative mode and the fact that Mrs Mallard is the focaliser in this section.

Figure 5.

Plausible text-worlds.

Of course, we have not been given much detail about Mrs Mallard’s marriage up to this point and so in the absence of what Gerrig and Rapp (2004) call ‘trait-consistency’ predictions, readers must do more cognitive work to interpret Mrs Mallard’s reaction as a plausible response in the circumstances. In order to do this the group exemplified what scholars have called psychological projection (Mar et al., 2008) that is, ‘projecting ourselves into the minds of actual others – inferring their desires, beliefs, and emotions’ (130), and which according to Gavins is guided in Text World Theory by deixis. This requires the ‘reading participant to project their notion of a zero reference point onto someone or something else in the text-world’ (Gavins, 2007: 40-1). Gavins characterises this experience as ‘vicarious’ given the ontological division of text- and discourse-world in a reading situation. However, the prisoners’ comments in the reading group suggest that projection was much more immersive than deictic projection, demonstrating a deeper engagement with the text. Whiteley (2011) terms this level of engagement ‘perspective-taking projection’ which she argues is ‘a kind of imaginative reconstruction of other psychological aspects of that entity’s perspective’, and ‘which involves the mapping of particular human characteristics onto enactors in order to flesh out their representation within the text-world’ (2011: 27).[3] In the reading group, the women used their own discourse-world knowledge to reconcile the anomaly they perceived in the text-world (e.g. Mrs Mallard is focusing on ‘new spring life’ because ‘she’s probably in shock’); if there were gaps in their discourse-world knowledge they worked collaboratively to find a plausible resolution, projecting from different perspectives to co-construct a text-world network that offered a much wider frame of interpretation.

One of the prisoners, who I will call Mary, was initially a reluctant participant in the group (Canning, 2012). A quiet, brooding woman, Mary had been a victim of domestic violence for years preceding her incarceration. She was convicted of the manslaughter of her partner and was described in court by probation officers, psychologists and in the judge’s summing up as emotionless and without any remorse. When I met her she had been in prison for two years and had never acknowledged her offence. At the reading of the Chopin story it was Mary’s seventh visit with the group and as we reached paragraphs 9, 10 and 11, she became very agitated. She tried to speak many times, clearly struggling to formulate and articulate her thoughts, uttering only disjointed sentences such as ‘guilt’, ‘love him’, ‘I can’t grieve’ and ‘free’. For a few minutes Mary continued in this way until she finally pointed at the text in her hand, saying ‘That’s me – that’s how I feel’. As I read the lines ‘she saw beyond that bitter moment a long procession of years to come that would belong to her absolutely’, and ‘free, body and soul free’, Mary was eagerly nodding and visibly shaking. When she was able to speak, Mary kept referring to the word ‘something’ in the text which she felt she could now identify as ‘guilt’ and ‘relief’, saying she had no right to grieve for her partner because ‘I killed him’.[4] Throughout the remainder of the session and at various points, she reported that while she felt ‘free’ (from her situation in a violent relationship), she also felt loss (‘I loved him’). Mary’s sudden and physical response suggests that her emotions were what Scheff (1981) refers to as ‘underdistanced’ (where ‘distance’ is defined as ‘the ratio of observation of, to participation in, one’s emotions’) signaling a ‘return to repressed emotions’ (Scheff 1981: 46). Mary’s response was triggered by the text-world and given full presence by the interaction between the text-world events and her own autobiographical experience, that is, her inability to recognize or deal with what she felt since her husband’s death, and her own guilt at causing it. In particular, she had not expressed any such recognition to the prison staff (including the psychologist), nor to the court throughout her trial. From her comments noted above, it seemed that Mary processed the remainder of the story through this newly accessed interpretative frame, expressing recognition of her own experiences of the conflicting emotions of guilt, freedom and loss which up to that point she had rejected or resisted since her trial and incarceration.

This exemplifies how the text-worlds of a literary work can productively interact with the discourse-worlds of the participants during reading. Throughout the group discussion, Mary presents scenarios from her past as relevant to the text being read (e.g. ‘I loved him’) and recognizes aspects of her present feelings in the literary work (‘that’s how I feel’). Mary’s responses blur the distinction between text- and discourse-world by bridging the reading situation (the immediate discourse-world), Mary’s distant past (the event of her crime) and the fictional text. In terms of the group dynamic, Mary’s contributions explicitly add information to the discourse-world(s) being constructed by each group member (including me) about how she feels at the moment of reading, as well as her memories and past experiences. Because it was shared during the reading, this information from Mary’s point of view constituted part of the common ground that eventually informed the group’s interpretation of the text, and thus the construction of the text-world network.

The literary text (or more specifically, the text-worlds being created during the shared reading) provided what McNay (2010: 105) has called a ‘distancing buffer’ (which reduces the feeling of ‘being exposed’ when sharing personal connections to the text), that allowed Mary to recognize and reflect on her own catastrophic experiences in the discourse-world through the safe medium of the text-world (safe, because at a remove from her actual world reality), enabling her to acknowledge what she had previously suppressed. In other words, she used the text-world network bi-directionally to help make sense of her discourse-world; ‘Them’s [sic] my words’, she said. This was reflected in the prison psychologist’s personal correspondence to me in which she reported that in the days after the reading session Mary ‘had finally opened up to begin working out her feelings over what put her in [prison]’. In so doing, Mary had begun to ‘assimilate’ (Mar et al., 2011: 825) what appeared to be a marked return of repressed emotions.[5]

For the other discourse-world participants, Mrs Mallard’s subjective experience was not as easily interpreted prior to Mary’s exposition: the ‘blank’ (Iser, 1978) rendered by the lexically underspecific ‘something’ required information from the discourse-world in the form of inference in order to fill the gap (the ‘something’ was finally processed as an admixture of ‘guilt’, ‘loss’, and ‘release’). The autobiographical information that Mary incremented into the discourse-world provided a way of making sense by helping the other group participants to understand both the ‘something’ and the seemingly incongruous clauses (‘new spring life’, ‘birds twittering’, ‘delicious breath of rain’, ‘free, free, free’). In short, the group co-constructed an enriched text-world network, which offered a more empathetic reference frame through which to read both the text-world enactor’s response to a traumatic event, and to understand Mary’s life experiences.

To try to represent this collaborative process diagrammatically, consider the four shapes in Figure 6 below as four discourse-world participants where the grey dots represent their different discourse-world knowledge. This is fed into the text-world(s) (which anchor the discussion), and the line connecting the participants represents the interaction involved in negotiating the text-world (through establishing ‘relevant’ common ground).

Figure 6.

The collaborative negotiation of common ground.

As discourse-world knowledge is shared, meaning is generated; as the text-worlds are co-constructed, anomalies are ironed out, gaps filled. What results is an enriched text-world (one that has been constructed through multiple perspectives) that feeds back to the discourse-world(s) offering a broader perspective (represented here by the black dots), a wider reference frame through which to read not only the literary text, but both their own and each other’s actual world experiences (Figure 7).

Figure 7.

Reified discourse-world knowledge.

As a result of this process, the group demonstrably adopted a wider perspective, reflected in one participant’s comment, ‘It’s not a bit of wonder your woman thinks them [sic] trees are full of new life’. The women weren’t just reading Mrs Mallard, they were reading Mary.

5. Text-worlds and expressive enactment

Whilst perspective-taking was imaginative for the others (in that they represented it that way in their discursive contributions), for Mary it was visceral; her response to the text went much deeper, exemplifying what Kuiken et al. (2004) call ‘expressive enactment’, namely ‘a form of reading that penetrates and alters a reader’s understanding of everyday life’ (171–172). Expressive enactment has its roots in psychology and concerns the therapeutic recognition and expression of primary adaptive emotions. According to Greenberg (2008: 90), ‘emotion awareness is not thinking about feeling; it involves feeling the feeling in awareness’, for which articulation in language seems to be an important component. Psychotherapists use expressive emotion when they encourage patients to enact a feeling that is not currently being experienced in order to facilitate access to a ‘suppressed, disallowed experience’ (97). They do this, for example, by ‘coaching’ the patient to express an emotion ‘until the emotion actually begins to be experienced’ (97).

The aim of such emotional (re)enactment is to help ‘people make sense of their experience and promote its assimilation into their ongoing self-narratives’ (Greenberg, 2008: 93; see also Leahy’s (2002) work on ‘emotional schemas’). A very similar process occurs in the discourse of the reading group, but one of the key differences in the latter is that the process is not contrived (by which I mean that it is not a therapist-patient goal-orientated exchange) and there is no expectation to contribute, no spotlight, something the women resented in their infrequent appointments with the prison psychologist; for instance, one prisoner told me: ‘I love [the reading group] because I WANT to talk here, I can understand myself better here. In psychology, you have about half an hour to get everything out, and maybe I’m just not in the mood to talk about all this when the appointment comes up’. In fact, my reading group data, both from the prison reading groups and from four mental healthcare unit patient groups, consistently records that participants felt more able to reach a new, more productive understanding of their own and others’ actual world experiences by reading fiction; as one reader from a forensic mental health unit put it, ‘stories […] kindle the imagination; the picture becomes sharper’. For Mary, participating in the reading session prompted her to construct text-worlds that were not only informed by but which reified her own actual world experiences and the memories of those experiences. Crucially, her acknowledgement of the emotions she exhibited was ‘buffered’ by the ontological distance of a fictional text-world enactor. Ultimately, negotiating the story’s text-world(s) acted as a catalyst triggering Mary’s ‘felt’ experience and facilitating its assimilation. This proved to be cathartic for Mary (as her psychologist pointed out) and demonstrates Greenberg’s contention that ‘once such emotions are in words, they allow people to reflect on what they are feeling, create new meanings, and evaluate their own emotional experience’ (2008: 96). Expounding a dialectical-constructivist view of human functioning, Greenberg cites the importance of narrative (specifically personal narrative) on our ability to ‘make sense’ of emotional experiences. She writes:

[S]ymbol and bodily felt referents are viewed as interacting to carry meaning forward, and newly symbolized experience is organized in different ways to construct new views. Attending to and discovering preconceptual elements of emotional experience influence the process of meaning construction. New experiential elements from many sources from within, and sometimes from without, can be integrated into this process. People are then viewed as constantly striving toward making sense of their preconceptual emotional experience by symbolizing it, explaining it, and putting it into narrative form (Greenberg, 2008: 94, my emphasis).

There are striking parallels here with the earlier study by Kuiken et al. (2004) specifically with their concept of self-implication, which they report characterised some readers’ immersive responses to literary fiction, and which, through expressive enactment, often resulted in ‘self-modifying feelings’ (insofar as the discourse-participant is changed in some way because the impact of expressive enactment ‘carries forward […] and externalizes a freshly conceived sense of self’ (185)). Indeed, Mary and others have expressed this very quality: their comments on evaluation questionnaires suggest that their experience in the group has changed how they will think about things in future actions and how they think about their own experiences retrospectively: ‘It [the reading group] helps me to understand people in the stories and in the reading group because I understand what they’re experiencing’; ‘I enjoy listening to other people and can understand other people and most of all I WANT TO ENJOY AND UNDERSTAND VIEWS now’; ‘Other people’s views can help [me] to understand some things better’ [capitals in original]. To offer a quantitative overview, responses (n=38) collected from my prison reading group on a Likert five-point wellbeing scale (0-5, where 0 = disagree and 5 = strongly agree) relating to six key wellbeing criteria showed the highest mean score (4.4) was for ‘understanding others views’.

By jointly negotiating and co-creating text-worlds, readers of literary fiction can go some way to understanding not just the literary work, but their own and others’ views and experiences (the ‘within’ and ‘without’ of Greenberg’s exposition). In fact, it is often the case in our groups that ‘symbolizing traumatic emotion memories in words’ allows discourse participants to organise their experience ‘into a coherent story’ (Greenberg, 2008: 93). Moreover, where participants are unable to articulate their traumatic emotion memories, the literary text itself often constitutes (rather than substitutes for) personal expression (recall Mary’s assertion ‘them’s my words’, and as another group member in a mental health unit reading group puts it, ‘it helps me because somebody else has written something that understands how you feel’).

However, I argue that direct experience is not a necessary precondition for expressive enactment; in a different reading group outside the criminal justice context (with healthcare practitioners) the same story prompted responses which suggested that participants fill blanks by incrementing analogous experiences to help them to fully construct coherent text-worlds, resulting in equally ‘felt emotions’: for instance, two group members talked tearfully about the guilt they felt when their respective elderly parents for whom they were carers were placed permanently into residential care homes. Through analogous reasoning, they identified with Mrs Mallard’s conflict – freedom and loss, and what they also called ‘guilt’ – through the immediacy of the text-world experience in paragraph 9. Incidentally, it is very interesting that the word ‘guilt’ never appears in the Chopin story, yet these participants seem to have inferred the concept, perhaps from the lexically underspecific term ‘something’ (‘too subtle and elusive to name’) and the representation of its movement towards the protagonist (‘creeping’, ‘reaching’, ‘approaching’).

6. Conclusion

By looking more closely at how discourse-worlds inform text-world networks and the potential for this interaction to function bidirectionally (from the text-world out), I hope to have developed Text World Theory, particularly its utility in shared reading contexts (which have never been studied through a Text World framework before), and demonstrated its value beyond traditional academic boundaries. In situations where people struggle with some element of their emotional histories (and I am guessing that accounts for much of the population), literary fiction can help unlock inaccessible or challenging personal narratives. For example, Mar et al. (2011) suggest that ‘literary narratives could provide a method for circumventing a person’s natural defences […] and provide a useful tool for studying those with affective disorders’ (830). Certainly, this has been the case for much of my own work in read-aloud groups; the text often serves as a catalyst enabling participants to connect the literary material to their own personal experiences and vice versa. In most of these discussions, the participant evaluates or critiques their actual world behaviour, often perceiving it through a new perspective. In my personal experiences of reading in prisons, hospitals, mental health institutions and schools, it has consistently been the case that perspective-taking is cited as one of the key benefits of participation (reflected on participants’ evaluation questionnaires). Text World Theory can be instrumental in understanding how those differing perspectives help formulate and reify text-world and discourse-world narratives (personal and literary) and how they interact. More generally, it conveys the value of reader response research and offers insights into the way that we, as stylisticians, as readers and as humans, come to make sense of our own multi-layered stories.