The restricted possible worlds of depression: A stylistic analysis of Janice Galloway’s The Trick is to Keep Breathing using a possible worlds framework

This article uses a theoretical framework of possible worlds to explore the ways in which Janice Galloway’s novel about grief and depression, The Trick is to Keep Breathing, may elicit emotional responses in readers. I give an overview of some of the emotional responses expressed by readers by using online review data, before employing stylistic analysis to demonstrate how emotional effects may be created through the linguistic construction of degrees of possibility. Drawing on Possible Worlds Theory, I demonstrate how readers’ emotional responses may be linked both to the presentation of possibility and to the restriction of possibility. The combination of the empirical methodology utilised here alongside stylistic analysis allows me to harness the capacity of Possible Worlds Theory to cast light on constructions of textual possibility and actuality and to facilitate understanding of some of the mechanisms eliciting readers’ emotions.


Introduction
Janice Galloway's The Trick is to Keep Breathing (1999[1989) is a Scottish novel which was shortlisted for the Whitbread First Novel Award and the Irish Times International Fiction Prize and won the MIND Book of the Year Award (British Council, 2021). The text focuses on the severe depression of its narrator, a drama teacher ironically called 'Joy', who has recently lost both her partner and her mother. A possible worlds framework adapted from Ryan (1991) and Bell (2010) is used here to facilitate stylistic analysis of the ways in which the linguistic construction of depression in this novel may partly work to elicit emotional effects.
Before applying the framework, I outline some of the emotional responses of readers, using data obtained from online reviews. In so doing, I acknowledge the role of the reader in co-constructing meaning in a transaction between reader and text (Rosenblatt, 1978). My stylistic analysis subsequently enables me to harness the flexibility of Possible Worlds Theory as a productive framework used here to illuminate how constructions of possibility may work to affect the emotions experienced by readers. 'Emotions' are conceptualised in this article in accordance with Hogan's (2016: 12) definition of emotions as 'events' resulting from 'the activation of systems by causes' and thus as cognitive responses to stimuli, with the stimulus in this case being the novel itself.

Possible Worlds Theory as a framework
Possible Worlds Theory originated in the philosophical field of modal logic, with the notion of 'possible worlds' used to conceptualise alternative realities existing beyond the 'actual world' of our existence (Kripke, 1959(Kripke, , 1963Lewis, 1973aLewis, , 1973bLewis, , 1986. In philosophy, debate has centred on the extent to which these alternative 'possible worlds' can be viewed as concrete entities just as 'real' as our worldas espoused by Lewis (1973aLewis ( , 1986)or as metaphorical constructions (Rescher 1975;Materna 1998) which enable us to visualise the way the world 'might have been'. These latter conceptualisations of Possible Worlds Theory have lent themselves to its adaptations for literary theory (Eco, 1984;Pavel, 1975); Possible Worlds Theory can enable us to conceptualise texts as containing a set of potentialities existing beyond the text and can thus be used as a vehicle to explore notions of possibility. Possible worlds have been viewed as 'mental constructs' (Bell and Ryan, 2019: 6) with Ryan's influential typology for the analysis of possible worlds (1991) proposing a modal universe that comprises 'three modal systems [the actual universe, the textual universe, and the referential universe], centred around three distinct actual worlds ' (24). In other words, each of these separate systems contains an 'actual world' with 'possible worlds' revolving around a central point of actuality and reality. Hence a fictional text itself will contain a textual actual world as well as textual possible worlds which exist in the minds of characters; for example, within their wishes, dreams and beliefs for the future. The actual universe refers to the same process taking place within the 'real' world, with actual possible worlds within this created in our imagination. The textual referential universe is 'the world for which the text claims facts' (Ryan, 1991: vii); in other words, the world that is established to exist by the text. Ryan's (1991: 26) own suggestion that, in fictional texts, the textual reference world and the textual actual world are 'interchangeable' led Bell (2010) to develop a simplified modal universe for the analysis of hypertext fiction in which the 'textual universe of fiction comprises a textual actual world and alternative possible worlds only' (Bell, 2010: 24). This is the framework adopted by this article, which focuses on the text as containing its own set of textual actual and possible worlds, and acknowledges the existence of the actual world of the reader, but does not use the third concept of the textual reference world.
While Possible Worlds Theory has not previously been used as a specific framework for the exploration of readers' emotions, the experience of the reader is central to Possible Worlds Theory in which possibilities becomes 'recentered' around the textual actual world as constructed in a narrative, thus creating for the reader a new world of 'actuality and possibility' (Ryan, 1991: 2). Textual possible worlds are 'story-like constructs contained in the private worlds of characters' (Ryan, 1991:156) which exist also in the minds of readers. Researchers in Possible Worlds Theory recognise, therefore, the role of readers in co-constructing meaning, such as in Raghunath's (2020) development of the concept of 'Reader-Knowledge Worlds'.
Importantly, the 'text as world' metaphor is not 'indebted to PW theory' (Bell and Ryan, 2019: 8); readers' mental representations of texts have variously been called 'textworlds' (Werth 1999), 'storyworlds' (Herman, 2002;Ryan, 2019), 'mental spaces' (Fauconnier and Turner, 2002: 89), 'narrative worlds' (Gerrig, 1993) and 'fictional worlds' (Doležel, 1998(Doležel, , 2019Fort, 2016). Some criticism of Possible Worlds Theory suggests that its flexibility detracts from its usefulness: Ronen cautions against its use as a 'diffuse metaphor' (1994: 7), while Stockwell (2010: 425) points to a lack of a 'genuine cognitive discourse grammar' in Possible Worlds Theory and suggests that its 'top-down' focus on schematisation and world-building may preclude stylistic analysis. However, I aim to demonstrate here that it is possible to explore the construction of possible worlds in a text linguistically at the micro-level, alongside benefiting from the capacity of Possible Worlds Theory to facilitate exploration of constructions of possibility within narratives.

The study of readers' emotions
The study of readers' emotions is a flourishing area of research both in and beyond stylistics. In Text World Theory, Whiteley (2010) and Canning (2017) have used reading group data to enrich understanding of how fiction elicits readers' emotions, while Gerrig (1993), Oatley (1999Oatley ( , 2004 and Miall (2006) have extensively examined the emotional experience of literature in a broadly empirical sense. Empirical studies of readers' emotions have shown that the experience of emotions when reading is a near-universal experience (Mar et al., 2011: 828). Readers' emotions have been shown to correspond to 'appraisal patterns (objective correlatives)' in texts (Mar et al., 2011: 828); in other words, research has demonstrated that identifiable features of the language of the text can elicit readers' emotions. Oatley and Mar (2008: 173) propose that the 'simulation' of emotion and social experience through reading leads readers to engage in empathetic processes which can engender greater capacity for empathy in their own lives. Stockwell summarises that the feelings experienced when reading are 'fundamentally the same as authentic real-world emotions' (2020: 183); literary emotions can thus be viewed as genuinely felt even though they are engendered through simulation. Character identification has also been explored as one of the key factors eliciting narrative empathy (Keen, 2007: 169;Oatley and Gholamain, 1987). Stockwell suggests that the more richly developed a character, the more readers will be able to 'mind-model' them to 'a rich level of impersonation' (2020: 183), and therefore, these characters are more likely to be the ones whom we feel emotions towards. Stylisticians in Text World Theory suggest that character identification takes place through a process of projection which consequently engenders emotions; Gavins (2005Gavins ( , 2007 suggests that we project enactors of ourselves into a text-world when we encounter it, and that such projection results in 'empathetic identification ' (2007: 64). Adapting this concept, Whiteley posits that readers may project themselves simultaneously into multiple character roles when encountering a text, calling this an act of 'mindreading' (2010: 121) whereby readers imagine the thoughts and feelings of characters and thus experience emotional responses. Therefore, stylistic research into character identification centralises readers' emotional experiences.
Accordingly, Ryan (2019: 74) suggests that the construction of a fictional world which is accessible to the reader, with characters 'perceived as ontologically like us' who undergo recognisable emotions or experiences, contributes to our ability to identify with characters. Ryan (1991: 32-33) also suggests a typology of accessibility relations delineating the distance a textual actual world has from, or how accessible it is, from the actual world, in order to explain how readers' sense of reality is applied in fictional worlds. Character identification, according to Stockwell (2012: 172), is partly created through focalisation patterns throughout a text which lead us to feel emotions 'at points of juncture in the evolution of the plan, as in real life'. The concept of character identification might thus be aligned with a possible worlds approach, because if we become immersed in the world of a character, we may, conceivably, begin to imagine, dream or hope for certain outcomes for that character throughout the process of reading.
Correspondingly, the conceptualisation of reading as immersion may help to shed light on some of the processes eliciting emotions. Our experience of textual actual and possible worlds may generate a sense of immersion due to the 'fictionally complete' nature of a storyworld (Ryan, 2019: 75). Linking to her notion of reading as a process of 'recentering' (Ryan, 1991: 2), Ryan (2015: 73) defines immersion as a process of 'consciousness [which] relocates itself to another world', thus suggesting a process of movement into the world of the narrative. Similarly, metaphors of transportation (Gerrig, 1993;Green and Brock, 2002;Harrison and Nuttall, 2020;Stockwell, 2009) and absorption (Braun and Cupchik, 2001;Kuijpers et al., 2014), used frequently by readers to conceptualise reading, evoke relocation into the world of the text.

Methodology
This article utilises Bell's Possible Worlds framework (2010: 24) adapted from Ryan (1991), which holds that any fictional text contains a textual actual world (henceforth TAW) with other textual possible worlds (henceforth TPWs) existing alongside or revolving around this. TPWs can be understood, therefore, as representing those imaginary or alternative worlds which exist, for example, in a character's wishes, dreams, obligations or desires. The overarching conceptual framework of TAW and TPWs enables me to shape a stylistic analysis of the ways in which possibility, actuality and lack of possibility are depicted and may work to convey the protagonist's emotional states. Within this analysis, my own concept of restricted possible worlds is used to refer to the aspects of the text in which the reader is afforded access to only a limited range of TPWs.
Prior to the stylistic analysis within a possible worlds framework presented here, I outline reader response data from online reviews in order to exemplify some of the emotions reported by readers of this novel. In recent years, several researchers in stylistics have used online reviews as a source of empirical data. Giovanelli (2018) uses excerpts from online reviews to demonstrate readers' views of the narrator in The Girl on the Train, before then using Cognitive Grammar to explore the presentation of the narrator's mind style (Fowler, 1977). A different approach is taken by Harrison and Nuttall (2020) who focus on analysing metaphors for reading in online reviews of the novel Twilight. Their focus is on linguistic analysis of the reviews themselves and the ways readers conceptualise reading, rather than analysis of the novel, and thus provides illuminating insights into the way the text is experienced by readers. Meanwhile Allington (2016) compares Amazon customer reviews with those of literary critics, using online review data to cast light on the ways in which customers' responses differ from those of professional reviewers. Online reviews can be categorised as a 'naturalistic' (Swann and Allington, 2009) form of reader response data rather than 'experimental' data gleaned within an artificial, experimental or laboratory setting; the reader reviews already exist as a form of social reading (Peplow et al., 2017) undertaken after the initial individual reading process.
Six reviews were selected for this analysis from a total of 69 Amazon reviews and seven reviews were selected from a total of 225 GoodReads reviews, with both sites utilised in order to gain access to a range of reviews. Whilst both Amazon and GoodReads invite readers to contribute reviews, there are some key differences in the format and the purpose of reviews, even though both websites are owned by the Amazon company. GoodReads is a community social cataloguing website where readers share and discuss their opinions on books they have read, whilst reviewers on Amazon have often purchased books on the website. Any customer with an Amazon account can leave a review, with reviews of those customers who purchased the book on the website being labelled 'Verified Purchase'. I chose to include both Amazon reviews from verified and nonverified customers for two reasons. Firstly, since there can be no external verification of whether readers have truly read a book on GoodReads, the inclusion of both verified and unverified purchases on Amazon aligned my selection of Amazon reviews with my selection of GoodReads reviews. Secondly, whilst counterfeit Amazon reviews may occasionally be written to promote sales (He et al., 2022), by carefully selecting reviews which discussed feasible emotions or feelings experienced during the process of reading the novel, I was able to ensure as far as possible that reviews were by genuine readers.
A further limitation of using online reviews may be their performative element (Driscoll and Rehberg Sedo, 2019), but arguably any form of verbal or written response to literature intended for any audienceeven if only a single researcheris inherently performative since it involves the expression of one's feelings about reading to others rather than a private emotional experience. Thus, any study of readers' emotionswhether expressed in online reviews or in any other formatis in reality a study of readers' expressions of their emotions, given the impossibility of verifying the veracity of individuals' verbal or written communication.
Reviews were selected for analysis from the dataset which expressed strong emotions towards Galloway's novel. By 'strong emotions', I mean that I focused on reviews that expressed identifiable emotions such as sadness, frustration or sympathy regarding particular aspects of the novel, rather than reviews expressing a greater sense of neutrality or apathy. However, particular emotions were not pre-determined and specific linguistic markers of emotion in reviews were not pre-set, as I wanted to facilitate consideration of a wide range of emotions in my analysis including those which I had not foreseen. For example, R8 described feeling 'trapped in [the main character's] head', an emotion which I later coded as demonstrating the reviewer feeling 'claustrophobic'. I acknowledge that the use of the overarching concept of emotion as a guide for data selection means there is an element of subjectivity in the review selection process, as the reviews identified as demonstrating emotional responses were inevitably partly informed by my own perspective as a reader, as well as by features of the language used by reviewers. The reviews chosen, therefore, should not be viewed as a comprehensive and representative sample of readers' emotional responses to the novel but rather as an exemplification of the way in which the stylistic construction of the novel may affect some readers and also as a tool to guide me as a researcher towards the exploration and analysis of certain emotion-causing elements of the novel referred to in the review data.
Once reviews had been selected and a dataset had been assembled, an inductive approach to coding was utilised. Inductive coding enables researchers to condense raw textual data, thus rendering data a manageable tool in assessing research objectives (Thomas, 2006) and thereby facilitating analysis (Rapley, 2011: 282). In an inductive coding approach, emergent codes are identified during analysis of the dataset rather than selected from an a priori list, and coding should always be 'customized to suit the unique needs and disciplinary concerns of [the] study' (Saldaña, 2016: 64) with researchers recognising that coding is fundamentally a 'decision-making process' (Elliott, 2018(Elliott, : 2850 driven by the context and aims of the research. By identifying readers' emotions in an inductive rather than deductive coding process, I ensured that a broad range of emotions could be examined and identified and did not need to rely on foreseeing a predetermined set of emotions in the data. The first stage of coding involved the initial labelling of data, before these were then grouped into a smaller number of categories identified as prominent. Codes were also used to direct my attention to particular aspects of the novel, although I do not attempt to suggest that particular excerpts have elicited particular emotions in readers but rather use my stylistic analysis to exemplify the way in which the language of the novel is constructed to achieve emotional effects. Since specific phrases or words used by participants can be 'good leads' (Miles et al., 2014: 74) in indicating emerging themes, I did not exclude participants' own terminology from my categorisations, but the emergent codes were descriptive rather than in vivo, as not all readers explicitly used emotion labels to describe their feelings, and sometimes used idioms, metaphors or alternative descriptions of emotions to describe their responses to the novel. By developing my own codes, I was able to utilise coding as a form of 'data condensation' enabling discovery of the 'core content or meaning' of the data (Miles et al., 2014: 73).
While informed consent is not typically sought for the use of pre-published online texts by researchers (Gao and Tao, 2016: 185; British Association for Applied Linguistics Research, 2017), researchers should be mindful of their ethical responsibilities even when data is in the public domain (Boyd and Crawford, 2012: 672). It is problematic to conflate agreement with the terms and conditions of social media platforms with informed consent (Townsend and Wallace, 2016), and therefore, anonymisation of reader data can help to limit the capacity for ethical harm to participants who have not been able to give consent for participation (Townsend and Wallace, ibid.). I have therefore removed reviewers' names and have labelled them as 'R1', 'R2' and so on within this article to limit the risk of identification.
In conclusion, then, my research design can be summarised as a process of datagathering from online reviews to cast light on readers' emotions regarding aspects of the novel, followed by a subsequent process of textual analysis using the possible worlds framework outlined above. I selected excerpts for stylistic analysis in order to explore some of the ways in which these emotional effects of the novel might be elicited; a researcher's own stylistic analysis is in itself inherently empirical (Stockwell 2021: 166;Brône and Vandaele 2009: 7) as a researcher draws on their own personal, felt experiences of texts as a form of data. Therefore, the triangulation of both 'rich first-person phenomenology and rigorous third-person observation' (Brône and Vandaele 2009: 6)in other words, undertaking my own stylistic analysis whilst also acknowledging the responses of other readersmay facilitate a richer understanding of the potential emotional effects of the language of the text. In the analysis below, I outline some key findings from my analysis of reader data, before applying a possible worlds framework to the language of the novel itself.

Reader reviews
As noted in the introduction to this article, The Trick is to Keep Breathing focuses on the depression of its protagonist, Joy. The novel details aspects of her daily life in the aftermath of her lover's death, including her loneliness, alcoholism, multiple affairs, bulimia and treatment for depression. Several reviews analysed for this research discuss the capacity of this novel to elicit sadness. For example, R1 states: 'This isn't a book you'd read if you're feeling down or in winter, it is bleak and even the rare moments of hope are pretty subdued.' The reader's use of the metaphor 'feeling down' suggests a belief in the novel's capacity to cause sadness and therefore potentially exacerbate existing depression, whilst their caution not to read the book 'in winter' might suggest an association of the winter months with feeling depressed, such as in individuals suffering with Seasonal Affective Disorder (NHS 2018a). The reader, then, conceptualises the novel as one that could make readers feel depressed themselves; their warning regarding the content of the novel indicates a perception of its ability to elicit unhappiness for readers, particularly if their own context reflects the experience of the protagonist. Similarly, R11 cautions: 'If you are looking for a fun lighthearted read or just have enough of your own depression to deal with, this may not be the book for you. As for me, I'm off to eat some chocolate in hopes of cheering up a bit'. As well as this reader's comments indicating that the depiction of the protagonist's depression might exacerbate other readers' 'own depression', their comment regarding being in need of 'cheering up' indicates that the novel has elicited sadness. R12's description of the novel as 'depressing and painful' and R2's description of the novel as 'devastating' were also both coded with 'sadness', with the reviewers' words suggesting the power of the novel to incite this emotion.
Two reviewers of The Trick is to Keep Breathing comment negatively on their feelings about being confined to the protagonist's worldview, an emotion which I coded as feeling 'claustrophobic': R8 refers metaphorically to being 'trapped inside Joy's bleak vision, with little sense of what life is really like outside it' and reflects that 'the main problem for me was that we were so trapped in Joy's head that a lot of the world around her remained hazy'. As well as using the metaphor 'trapped' twice to denote the claustrophobic emotions experienced when reading the novel, the reader also specifically uses the term 'claustrophobic' in the review, suggesting the novel is 'horribly claustrophobic' although suggesting that 'this is inevitable, bearing in mind how claustrophobic depression can be'.
Meanwhile, the word 'claustrophobic' is also used by R7 who found the book to be 'claustrophobic and hopelessregardless of the ending'. While this word is used by R7, unlike R8, to describe the novel itself rather than the reader's specific emotions, R7's subsequent discussion of their dislike of the novel suggests a similar sense of feeling claustrophobic: 'This book is horrendous […] 4/5 stars for the consuming and unusual narrative style, and accurate portrayal of mental illness at its worst. 1/5 stars for how terrible this made me feel about life'. The use of 'horrendous', 'unusual' and feeling 'terrible […] about life' work together with 'claustrophobic' to suggest a sense of feeling trapped within the pain of the novel. Evidently, R7 appreciates the technical skill evident in Galloway's writing but finds it difficult to cope emotionally with the strength of the novel's impact on them, to the extent that this limits their reading enjoyment. Similarly, R11 writes that 'this woman's head is NOT a fun place to be. […] I thought I had dealt with depression before, turns out, my head is practically a Disney movie by comparison'. This reader's explanation that they felt inside 'this woman's head' when reading also suggests a sense of feeling restricted to Joy's experience, as well as evoking the conceptual metaphor READING IS TRANSPORTATION (Gerrig 1993;Stockwell 2009).
Finally, several reviews were coded with 'empathy', including the reviews already outlined above, as empathy frequently co-existed with other emotions in readers' reviews. For example, R7's sense of feeling 'trapped' in the narrator's head was also coded with empathy, because their description of feeling inside Joy's head, like R8's, suggests an experiential transportation into Joy's mind and thus the ability to both feel Joy's pain and to experience this as uncomfortable. The word 'empathy' is also used explicitly by R1, who compares Joy's experiences to their own experiences with depression: 'I found the central character a person I have huge sympathy and empathy for, a lot of her brain numbing and frustrating conversations with medical 'experts' ring very true for me, I often think of her as I grate [sic] my teeth speaking with doctors'. As well as the explicit use of 'empathy', the reviewer's description of Joy's conversations about her depression that 'ring true' also evokes an empathetic response. Interestingly, not only does this reader's empathetic response suggest that they relate to Joy's experience, but their reference to thinking of the character during their conversation with doctors indicates that the novel has had an impact on their subsequent experiences and interactions in life.
The corporeal metaphors some readers use to describe their experience of reading the novel also evoke an empathetic response to Joy's pain: R9 comments that 'Joy Stone, the main character, ripped my guts out and took my heart along the way and it was indeed hard to breathe at many points in the book'. The use of physical imagery enables this reviewer to draw parallels between their own experience of reading the book with Joy's depression as depicted within the novel; by stating that it became 'hard to breathe', the reader implies that reading the book enabled them to experience an element of Joy's feelings of depression. Furthermore, the implication of emotions being felt in the 'guts' and 'heart' reflects the way in which Joy's depression is depicted as physically allconsuming, as demonstrated in the novel when she describes her inability to eat and her difficulty in performing everyday tasks. Similarly, R5 states that 'the fierceness of [Joy's] pain feels like a knife in the gut'. This simile evokes the physical experience of emotion normally associated with the experience of personal upset or emotional pain, whilst also implying a favourable opinion of the novel in that the ability to 'feel' the narrator's pain demonstrates its power.
Likewise, R3 explains that Joy's 'mental collapse [was] intimate, and their situation familiar. It also hurts my brain', suggesting a physical identification with Joy's pain by using bodily imagery to describe the psychological pain wrought by empathising with the narrator. Readers' responses here, then, indicate an empathetic response to the ways in which the TAW is constructed as painful, to the degree that readers are able to physically 'feel', to some extent, the narrator's emotions. Even R13, who 'didn't enjoy' the novel, suggests that this is a 'heavy book to get through' in which 'you do feel as though you are wading through deep water right along with the MC' [main character]. The metaphor of 'wading through deep water' might conceivably suggest an empathetic response to the protagonist's depression, in which the reader experiences some of Joy's feelings during the process of reading. On the other hand, others respond positively to the richly detailed depiction of Joy's psychological pain: 'I was intimately involved in ''Joy's'' life and I felt it would have been disrespectful to her to rush through all of her pain as though it really didn't matter' (R10). Some readers relate Joy's depression to their personal experiences: R12 suggests that 'we've all felt the way Joy feels at some time or another' while R6 explains that Joy is 'so relatable… the character is incredibly, worryingly familiar', with both readers therefore suggesting feelings of empathy elicited not only by the novel itself but also by their own personal experiences.
Overall, then, notable emotions expressed in the reviews analysed for this article include sadness, empathy and a sense of claustrophobia. These emotions have been selected from a large dataset and therefore should not be interpreted as exemplifying the most frequent emotions experienced in response to the novel; a code that appears even only once or twice in a dataset may be meaningful (Saldaña, 2016). Rather, the categorisations outlined here represent an exemplification of some of the emotions readers report feeling in response to The Trick is to Keep Breathing. In the analysis below, I analyse several excerpts from the novel using a possible worlds framework in order to explore some of the way in which these emotions may be elicited.

Analysis
The framework of the TAW and TPWs can be used to cast light on the presentation of the protagonist's hopelessness and despair in the following excerpt: I watch myself from the corner of the room sitting in the armchair, at the foot of the stairwell. A small white moon shows over the fencing outside. No matter how dark the room gets I can always see. It looks emptier when I put the lights on so I don't do it if I can help it. Brightness disagrees with me: it hurts my eyes, wastes electricity and encourages moths, all sorts of things. I sit in the dark for a number of reasons. (7) Here, Galloway's use of language anchors the reader in the unhappy TAW of the narrator by presenting her life as dark, small and with limited possibilities for escape to happier TPWs. There is an implicit reference to a TPW existing beyond the TAW in the negation 'No matter how dark the room gets', but this simply works to construct a different TPW in which the room, and therefore the narrator within it, exists in various degrees of darkness. While there also an implicit construction of a TPW of a brighter room in the reference to a world which 'looks emptier' with the lights on, this world is depicted as even less hopeful than the present, dark TAW in its emptiness. This evokes both the TAW and any alternative TPWs as equally miserable: the TAW is dark and thus seems to be a desolate place due to the connotations of darkness with lack of hope, but a brighter room would seem even 'emptier', with both prospects suggesting loneliness and unhappiness. Joy's pain is also emphasised in the way even light itself is constructed as a harmful force for the narrator which 'hurts (her) eyes' and 'disagrees with (her)'. The construction of both the TAW and any alternative TPWs as unhappy conveys the totality of Joy's depression and inability to see beyond her experience of the present. This is reinforced, too, by the fact that as minuscule details of the protagonist's decision-making process regarding sitting in the dark are revealed, the reader becomes aware that this is a character who is spending a great deal of time reflecting and perhaps even obsessing over the smallest elements of their present existence to the extent that even the process of leaving the light off is justified.
Alongside the insight into Joy's state of mind that this level of detail creates, a sense of dissociation is also evoked here with the foregrounded verb phrase 'I watch myself'. The unusual structural placement of 'I watch myself from the corner of the room', which is followed by a line break, highlights the importance of this action of self-observation and indicates a fragmented sense of identity. In positioning the narrator's mind as separate from her body, the locative adverbial phrase constructs the TAW as a disturbing and unreal space, and, in suggesting that her perceptive faculties are operating from 'the corner' of this space, reinforces her desire to retreat from the world. The spatial detail of Joy's location in 'the corner' in the 'armchair' at 'the foot of the stairwell' also emphasises the overpowering nature of her present experience in the TAW and suggests that the mundane details of her environment are her principal focus, thus restricting our viewpoint to the present moment and environment. The repeated use of the simple present tense in 'I sit', 'I put', 'I can' and 'I watch' portrays Joy's existence as centralised around her banal actions in the TAW, thus creating an impression of the TAW as restricted in size and potential for escape. This, then, suggests the unhappiness of Joy's worldview by evoking a constrained life with limited possibilities for happiness and thus conveys the depression that she is experiencing.
In the following excerpt, the narrator begins to plan for the future event of going to work, giving us a glimpse of alternative TPWs beyond this roombut this TPW is developed as a bleak and depressing contemplation: The green numbers on the stereo flash 03.25. But it goes fast. I know perfectly well it doesn't matter what the real time is. This is all beside the point. The fact remains it's so late it's early and I have to move. I have to go upstairs. I have work tomorrow and I have to go upstairs. (7) The repeated use of deontic modality here -'have to'constructs an unrealised TPW of going to work the next day. Rather than constructing multiple, potentially hopeful TPWs in the reader's imaginationas might be developed via other modal formsthere is only one future TPW repeatedly and explicitly constructed here, and it is one which is presented as unfavourable and compulsory. Thus, Galloway makes it clear that not only is Joy's present situation miserable, but so too is her contemplation of the future. By granting the reader explicit access only to the TPW of fulfilling the obligation of work, the author makes it clear that both the TAW and the narrator's conceptualisation of future TPWs are unhappy. This is reiterated by the present tense repeated construction of 'I have to' and the way in which each action in the narrator's planning for work is separately outlined, such as in her statements 'I have to go upstairs' and 'I have to move', which works to suggest that not only does work present a miserable future TPW but the preparation for bedtime in the TAW is also an unhappy, enforced experience. The unhappiness both of the TAW and any alternative TPWs is reinforced by the statement 'it doesn't matter what the real time is' which suggests that even an aspect of life as fundamental as time has become meaningless. This works to develop our sense that Joy's depression colours her interpretation of time itself, meaning all possible worlds available to her are equally dark and restricted. Galloway therefore conveys the pain and unhappiness of Joy's depression by suggesting that she has access only to a narrow spectrum of alternative TPWs. Alongside this, the TAW itself is presented as a dark and unhappy place. This is similarly demonstrated in the following extract, in which the narrator describes the experience of watching television whilst she drinks tea in the morning: There are interviews with junior ministers while I make tea. Always tea in the morning unless I've eaten the night before: then, it's black coffee. Bad mornings, I have only hot water. But I drink something. But I drink something, as much as I can. It helps the headache and the dryness: the weight of fluid is calming. (10)(11) The present tense description of the events of Joy's morning initially seems to refer to only the immediate TAW of that day, as she listens to the radio 'while' she prepares her tea. However, the foregrounding of Joy's daily routine in the verbless phrases 'Always tea' and 'Bad mornings' suggests that this mundane morning procedure is ongoing rather than transient and therefore a monotonous aspect of her life extending to TPWs beyond the TAW. The elision of more typical subject and verb constructions removes agency from Joy as the events of the morning seem to be inflicted upon her, rather than actively chosen, as she follows the restrictive pattern set on previous days. Further to this, the use of the premodifier 'Bad' without any explicit reference to cognition or emotion suggests that these mornings are categorically bad, rather than just momentarily perceived as such, and thus evoke a sense of Joy being trapped within the TAW. The evaluative adjective 'Bad' also works to construct TPWs of 'mornings' beyond the immediate present that are even worse than the TAW, as this is a world in which Joy would only drink 'hot water'. Thus, even though the TAW itself is presented a dark and depressing space, the reader is made aware that life is sometimes even worse than this. Similarly, the irrealis 'unless I've eaten the night before' suggests TPWs that might exist beyond Joy's experience of the current morning, and yet this is an alternative world in which the only difference in Joy's morning routine would be the consumption of 'black coffee' rather than 'tea' and in which she would still not manage to eat anything. Thus, here we see again the linguistic construction of an unhappy TAW alongside a limited spectrum of alternative TPWs which are depicted as even more miserable than the TAW, evoking the restriction of Joy's depression.
Alongside this, Joy explains she drinks as much liquid as possible to help with 'the headaches' and 'the dryness'. By introducing both physical symptoms with the definite article, Galloway evokes both physical states as pre-existing, continuous aspects of Joy's existence rather than transient experiences, thus conveying the pain of the TAW as an ongoing rather than momentary discomfort. Therefore, when we are told that 'the weight of fluid is calming' we can infer that it is not only a single moment in which Joy feels in need of 'calming' but that this is a daily experience. In describing the drink as 'calming' and as helping 'the headaches and dryness', Galloway depicts Joy's psychological and physical malaise as intertwined, thus reiterating the multifaceted discomfort experienced in the TAW and implicitly emphasising the difficulty of accessing alternative TPWs. The metaphor of the calming 'weight' of fluid indicates the extent to which she feels, in contrast to the liquid, adrift and ungrounded, therefore helping to construct a sense of unreality. The notion of the tea as 'calming' also emphasises the extent of Joy's anxiety which means she needs external entities to bring her back to reality, but it is notable that the specific features of her pain are not delineated, and her depression must be inferred from the depiction of her interaction with her surroundings.
The frequent use of material process verbs (Halliday 2013) such as 'have', 'eaten', 'make' and 'drink' to denote practical actions alongside the infrequent use of verbs of cognition also creates a sense of distance from cognitive and emotional processes, potentially suggesting that the narrator may be experiencing the feelings of numbness or emptiness often associated with depression (Mind, 2019). This evokes R12's comment that grief 'renders you a shell of yourself', a statement that suggests an implicit awareness of the way Joy feels emotionally empty and numb in the midst of her depression. The detailing of the banal elements of this morning routine also temporally extends this event, whilst restricting our perspective of the space beyond the house, helping to evoke the monotony of the present and therefore reinforcing the lack of opportunity for psychological escape from the darkness of the TAW, thus potentially contributing to readers' perceptions of the novel as 'claustrophobic' (R7 and R8). This creates a sense of oppression and constriction as our perspective is limited to a constrained view only of Joy's presentperhaps reflecting the way she is, in her depressed state, merely existing in the moment as she struggles to survive each day.
Similarly, when Joy describes the next stage of her morning, the focus of the narrative is on the tedious actions of her daily routine: When I'm ready, I rinse the cup, mop the sink dry and lift my coat. My mouth is still dry as I lock the back door. My mouth is always dry. (11) There is again a lack of mental process verbs here, with material process verbs such as 'rinse ', 'mop', 'lift' and 'lock' depicting the banality of the daily routine and thus locating the narrative firmly in the monotonous processes of the TAW. The tedious tasks of the morning are foregrounded through the syntactical parallelism of 'rinse the cup, mop the sink dry and lift my coat' here, perhaps reflecting the concentration required to undertake ordinary tasks in a clinically depressed state alongside constructing a sense of mundanity, while the repeated use of the present tense active voice creates a sense of constriction to the TAW. The detail of Joy's 'dry' mouth is foregrounded here, both due to repetition and the preceding adverbs 'still' and 'always'. This extends understanding of the 'dryness' (11) established in the previous excerpt, with the emphasis on this element of physical discomfort again conveying her emotional state despite the lack of explicit references to cognition. A dry mouth can be a symptom of anxiety (NHS 2018b) and a side effect of medications for depression (NHS 2018c) but even readers without this knowledge are likely to have schematic awareness that emotions are felt and experienced physically as well as mentally, meaning that the emphasis on an uncomfortable bodily sensation implicitly suggests unhappiness. Furthermore, when considered alongside the emphasis on drinking tea in the previous excerpt as an attempt to help the 'dryness' and also for its calming effect, the fact that Joy's mouth is 'still dry' may indicate that both her physical and mental discomfortor her depressionfeel ultimately unfixable. Again, the reference to physical sensation rather than explicit linguistic references to cognition or emotion evokes numbness and inability to process emotions, emphasising the physicality of depression as well as reinforcing that the TAW is a restricted space in which emotions cannot be easily defined or expressed, and in which happiness is out of reach. The focus on Joy's actions and routine, coupled with the descriptions of bodily sensation, thus implies a schism between her body and mind as she narrates her external experience and physical feelings and avoids explicit reference to thoughts or emotions. The TAW is therefore established as a constricted and painful place, but one in which emotions are not able to be expressed, giving us an insight into the depth of loneliness of the narrator and her imprisonment within her depression.
This focus on physicality as a representation of cognition is also demonstrated in the following excerpt in which the physical consequences of Joy crying are foregrounded in the first-person narrative, rather than her inner thoughts or feelings, and both the TAW and alternative TPWs are presented as unfavourable: Blisters. Little moon craters on the smooth paper. I push the magazine aside and let the tears drip onto the rug until I'm ready to move to the kitchen for some paper towels. My nose fills and drips too, my face will be bloated. (27)(28) The emphasis on the narrator's physical actions here, such as the 'tears' that 'drip', imply that she is acting as witness to her own pain and has become disassociated from cognitive comprehension of her emotions. By describing the immediate results of crying in metaphorical terms that evoke their impact -'blisters' that form 'moon craters' on the paperrather than in terms of the narrator's feelings, Galloway again implies that Joy is observing herself from an external perspective, reinforcing our sense of her withdrawal from society and her inability to express or fully conceptualise her emotions. Furthermore, the focus on the physical details of crying such as Joy's 'face, 'tears' and 'nose' that 'drips' works to separate Joy's body into discrete parts, thus evoking a sense of fragmented identity and dissociation, which is reiterated by the fact that the word 'crying' is not used at all here. The agency afforded, too, to the tears which create 'moon craters' and the nose which 'fills and drips', creates a contrasting impression of Joy's passivity as a narrator who withstands and endures the crying. This is reiterated when she moves the magazine so that the tears fall 'onto the rug' which suggests that while Joy can choose where the tears land, she cannot prevent them from falling.
Again, then, the narrator's body and mind are presented as separate as she observes the results of her crying and her physical symptoms, indicating an inability to psychologically process her feelings. This constructs the TAW as a deeply painful place, with a lack of possibility for mental escape. The only explicit references to alternative TPWs here are upsetting, unfavourable future events which are as unhappy as the present experience of crying; for example, the preposition 'until' merely introduces a world in which Joy will be 'ready to move into the kitchen for some paper towels', therefore constructing a TPW in which she will respond with a practical solution to her crying, but not one in which her emotional pain will dissipate. Similarly, whilst she imagines a TPW in which 'my face will be bloated', again Joy is envisioning a future in which she will have to suffer the consequences of her crying. Not only is the TPW here constructed as even more desolate, perhaps, than the TAW, but the focus on this physical detail also develops Joy's lack of care for herself, as her principal concern is how she will appear to the outside world, rather than how she feels. Therefore, the protagonist's depression is presented as granting access only to a limited, restricted range of alternative TPWs which are equally as unhappy as the TAW.

Conclusion
The psychological pain that is experienced as an inherent aspect of depression is evoked convincingly in this novel through the portrayal of the TAW as a space with limited mental access to alternative TPWs. Where TPWs do exist, they are often constructed as even less favourable or equally as unhappy as the TAW. The fact that readers are granted access, in the main, only to Joy's present experience in the TAWa place of grief, misery and depressionalongside a narrow spectrum of alternative TPWs means that the TAW is evoked as a space of pain, darkness and mental constriction, whilst the alternative TPWs available to the protagonist are portrayed as restricted in their potential to provide happiness. Feasibly, this construction of an unhappy TAW and a restricted range of TPWs may contribute to readers' emotional responses to the text, which include feelings of claustrophobia, sadness and empathy. While extratextual factors as well as emotions elicited by a wide variety of textual factors are also likely to contribute to readers' emotions, within the scope of this article the application of a framework of possible worlds, using the concepts of the TAW and TPW, has helped to cast light on the stylistic construction of the novel and thus to illuminate some of the factors that may contribute to the emotional effects of the novel. Reader responses gleaned from online reviews help to ensure, too, that my assertions regarding the emotional impact of the novel are not simply intuitive and based on my own interpretations, but are also grounded in a broader set of empirical data. Elements of the online reader reviews demonstrate that readers' own personal experiences can intensify their emotional responses to the novel, with some readers explicitly referring to their own knowledge or experience of depression as increasing their emotional responses to the novel. The nuances of readers' experiences were not examined in depth for this article, given the nature of the reader response data gleaned from online reviews and thus the lack of contact with research participants. However, further research in this area using alternative naturalistic social reading methodologies, such as reading groups, might examine the impact of individual contextual and experiential factors in more depth in order to build an enriched comprehension of the role of the reader in co-constructing the TAWs and TPWs of fiction.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.