‘It's in the air here’: Atmosphere(s) of incarceration

Contrary to descriptions of a desensitising situation – with restrictions on movement, monotonous regimes and sparse surroundings – much research highlights imprisonment as sensorially and emotionally powerful. Following work within the ‘turn to affect’ that focuses on non-verbal, non-conscious and, often, non-human embodied experiences, scholars have attended to how such elements cohere into ‘atmospheres’. Whilst the language of atmosphere is synonymous with the prison – a space that is widely anecdotally considered to conjure a particular ‘feeling’ – discussion of the mechanisms for and experiences of atmospheric production and consumption in this space has, thus far, evaded scholarly attention. Atmosphere is a word often used in prison literature, but it is rarely analytically unpacked. Accordingly, drawing on qualitative research data from individuals designing, and working and living in prisons, we focus on how various components – including aesthetics, olfaction, temperature, and the performances that arise from them – comprise sensory atmospheric affects in prison. In doing so, we find atmosphere(s) emerge – not simply from the materiality of the prison itself, but from cultural constructions of carceral and non-carceral landscapes and in conjunction with personal practice and preference. Accordingly, the prison is tied to particular constructions about what a prison should feel like and how people should (re)act to/in such spaces. In some cases, prison designers attempt to engineer particular atmospheres that cohere with wider political motivations around penal philosophies. However, despite the common reflection that prisons generate some kind of atmosphere, respondents are unable to offer a concrete description of what this may be, and much of our data highlights a definite precarity and changeability to atmospheric affect, which is likely to raise ambiguity around attempts to design carceral atmospheres.


Introduction
What does it mean when an atmosphere is 'wrong'? How is this sensed? What bodily affects arise in the person who feels something isn't 'right'? How is this emergent 'bubbling' a combination of a placeits design features of lightening, colours, materials and the movements possible thereinand lived experiences of that place through the bodythrough sound, smell, taste, feeling? There is a growing literature in criminology, carceral geography, and architecture studies, among other related fields, that seeks to explore prison architecture, its aesthetic components and/or its relationship with the lived experience of carceral space (see, e.g. Beijersbergen et al., 2016;Fransson et al., 2018;Karthaus et al., 2019;Moran et al., 2016). Indeed, previous prison research has explored how carceral spaces are filled with a cacophony of sounds, a potent concoction of smells, and brimming with competing emotions (Crewe et al., 2014;Herrity et al., 2021). Such work is underscored by a recognition that the component characteristics of a place or buildingof lighting, colours, temperature, materials, movement and use of spaceintersect with such sounds, smells and emotions. Yet, following from broader literature within the 'turn to affect' (see Thrift, 2004;Pile, 2010) that focuses on non-verbal, non-conscious and, often, non-human embodied experiences, scholars have attended to how 'these elementsand relations between themmight cohere or assemble into something far less obvious, but far more pervasive: atmospheres' (Turner and Peters, 2015: 313).
As the epigraph intimates, an atmosphere is a physical molecular form that surrounds us, permeating the body through the air we breathe (McCormack, 2008). It is also something experientialsomething we feel (whether, e.g. an atmosphere is tense, relaxed, fun or frightening). Atmospheres hang in the air, metaphorically and actually. And, as Turner and Peters (2015) have argued, atmospheres can stabilise around places, developing from characteristics associated with that place. In this sense, we may associate certain spaces with certain atmospheres: a bar with an atmosphere of bustle, or a library with an atmosphere of calm. To this end, there has been discussion on whether atmospheres can be engineereddesigned to evoke specific feelings through particularly sensory designs generated in space and place (Adey, 2008;Waterton and Dittmer, 2014). Indeed, lighting, sounds, and sensory engagements with them, together with personal relationships cohere (and come apart) to 'do greatly productive work in terms of engineering atmospheres' (Waterton and Dittmer, 2014: 122).
The prison is a space where experience and engineering of atmospheric qualities demands interrogation. To date, whilst there is work that brings the term 'atmosphere' into prison scholarship, it remains under-explored as a means of making sense of engagements between people and place; between prisoners and the condition of incarceration. This article offers this interrogation. It first outlines, builds from, and develops this work to examine how visible, tangible and experiential components of prison architecturefelt through the sensescoalesce to produce atmospheric affects. By considering the production and consumption of prison space in examples from individuals designing, working and living in these environments, we 1 explore how aspects of architecture and design have a pervasive impact on sensory experience, which emerges from both tangible and more intangible interactions with these spaces: in other words, physical and emotional engagements with the prison. We do so to both develop the importance of thinking of prison space in atmospheric terms, and to take seriously what such atmospheres may mean for the lived experience of differing groups in prison space, and for those designing and planning spaces of incarceration.
The article begins with a review of the literature on atmosphere(s) as they have been explored in the context of incarceration, demonstrating how such work could be further interrogated to better understand the intersections between prison experience, design and (micro)architectural forms. It further introduces the cross-disciplinary scholarship on the senses, emotion and affect in relation to atmosphere, particularly drawing from the rich work in human geography on these themes, where intersections of affect, spatial engineering and atmosphere have been keenly examined (see, e.g. Adey, 2008;Edensor, 2012). Following a summary of our methodological approach, we present empirical data from a large UK Economic and Social Research Council (ESRC) project that serves to highlight how working with the atmosphere analytically can help us better engage with understanding prison environments, experiences in them and design of them. It furthermore helps us to attend to the tensions between design and lived experience. Architectural desires aim to 'fix' or 'do good' through an engineering of space that hopes for specific intended uses and 'feelings' in place. Yet, whilst places can be designed, atmospheres often remain elusivethey can spill over, upset, or disrupt planned spaces because of their undesignable, unplannable nature, which is, by its nature, unpredictable, changing, volatile and difficult to pin down (see Adey, 2008;Turner and Peters, 2015a). In the first substantive section that follows, we theorise carceral atmospherics and outline the deployment of this term in relation to the architecture of incarceration.

Theorising carceral atmospherics, senses, affects and engineering
Atmosphere, as noted previously, is a word often found in prison literature. A number of publications bring the terms 'prison' and 'atmosphere' together in the interrogation of empirical material (see, e.g. Hemsworth, 2016, on the 'sonic atmosphere' of prison; Liebling, 1999, on 'deteriorating' atmospheres during fieldwork in a dispersal prison; Pool and Regoli, 1981, on atmospheres of 'tension and fear' felt by prison staff; Shefer and Liebling, 2008, on the business-like atmosphere of privatised prisons; and Van der Helm et al., 2011, on the 'group climate' in a Dutch psychiatric prison, among others). Taking an even more micro-focused approach, Martin (2021) discusses the 'politics of prison air'. Yet, whilst acknowledging that air is 'an intense indicator of a particularly charged penal atmosphere' (ibid, 2021: 6), a discussion of atmosphere specifically is not pursued. Accordingly, atmosphere is oftentimes used as a mere descriptor of a given state (a deteriorating atmosphere; the weight of atmosphere). Indeed, Turner and Peters have examined 'carceral atmospheres' in relation to the production of tourist spectacle at former prison sites such as museums (Peters and Turner, 2017;Peters, 2015a, 2015b); and the prison has been considered to evoke certain atmospheric affects, including feelings of claustrophobia and oppression, in the work of Young (2021, see also Young, 2019). This literature highlights that 'atmosphere' certainly merits further attention but it is rarely employed in an analytic sense. Accordingly, this article pushes for a deeper interrogation of atmosphere as a means of further grappling with 'what is in the air' in the prison experience.
An atmosphere is not simply the air itself, but something held in the air that, when brought together with people and place, sparks 'affects'. As Ben Anderson's important work notes, affect refers to 'a transpersonal capacity which a body has to be affected (through an affection) and to affect (as the result of modifications)' (Anderson, 2006: 735). Affects reside not in bodies, but in the space between, as emergent sensorial responses that arise when bodies come into touch with an atmosphere. Kathleen Stewart's work on atmospheres notes how they are lived through the experiencing 'subject' and through the 'body' (2011). As Anderson explains: 'atmospheres are shared ground from which subjective states and their attendant feelings and emotions emerge ' (2009: 78). However, as Turner and Peters, drawing on Massumi, remind, humans can be both affected by and affectual in atmospheres (2015b: 324). As Edensor and Sumartojo write, within atmospheres, 'personal memory and subjective experience come together with designed and regulated environments through sensory perception ' (2015: 261). Lees (2001) further comments that our atmospheric connection with the material environment is not simply developed at the moment and/ or derived from that specific materiality itself. We must acknowledge how our other place-based attachments and assumptions intersect to produce lived and practical experiences of the built environment. Bille and Simonsen (2019) take this one step further to consider atmosphere itself as the relative and relational process that allows our understanding of the world to develop. Accordingly, whilst atmospheres are physical entities, they are also metaphorical and are (co) created in circumstances that both accidentally and deliberately evoke or are evoked by a raft of other societal situations, memories, and constructions. In this way, Turner and Peters consider that we can be similarly alerted to the 'elusive, intangible, felt, aspects of carceral space' and pay attention to the way these aspects 'seep from, and are designed, engineered and co-constituted around material and visual components …' of spaces (2015b: 315).
An all-too-common assumption is that prison space presents itself as a flat, texture-less landscape, often purposefully devoid of sensory engagements and their affects, which is altogether reasonable when we consider the stark imagery of, in particular, maximum security custodial spaces. However, recent literature from criminology has fruitfully called into question the sensory and embodied experiences of prison space and the affective bodily responses they elicit. For example, Herrity's research posits the prison as a richly-aural landscape, where music (2018) and sound (2020) are very much intertwined with technologies of self-governance and spatial management. Natali et al. (2020) deployed a visual and sensory methodology, which relied on walking to engage 'often-ignored' perceptions and narratives about the prison landscape. Indeed, the burgeoning interest in research at the nexus of senses, atmospheres and aesthetics has prompted a whole edited collection devoted to this theme. Herrity et al.'s (2021) Sensory Penalities, as its subtitle suggests, 'explores the senses' in punitive landscapes, most notably prison spaces, to respond to what was hitherto a prioritising of only visual encounters. Taking influence from such sensorially-oriented work, we use this article to consider specifically how carceral spaces evoke affective atmospheres for those who engage with such spaces and the role of design in this process.
Yet, even if we subscribe to the more staid, sparse or sense-less versions of prison space, perhaps those which Weegels (2021) outlines as sensorially 'muted', 'even buildings and details that hardly possess any aesthetic values manage to create a sensorially rich … atmosphere', according to Pallasmaa (2014: 241). In the context of carceral space, it is the arguably taken-for-granted and banal nature of the prison aesthetic that is so disconcertingly powerful. Indeed, as Jewkes et al. (2017: 293) summarise, 'prisons are frequently characterized as blank canvasses onto which we project our deepest fears and fascinations'. As Jewkes and Young write, in prison, the air itself 'can be part of the punitive nature of the institution' (2021: 187). Their work on sensory engagements in Kyoto Prison, Japan considers the atmospheric qualities of prisonthe something 'in the air' that often has powerful impact upon those contained within these spaces. Crewe's (2011: 522) deployment of 'tightness' as a metaphor to describe the power that prisons have to 'wrap … up, smother … and incite' the individuals within them shares parallels with the notion of an all-encompassing, body-engulfing, invisible-but-tangible aether that shapes experiences of incarceration. Certainly, by acknowledging that the prison itself evokes atmosphere(s), we consider how it may have immediate and lingering impacts for prisoners and staff both within and once they leave a physical place.
As previously noted, atmosphere in the context of prison is rarely deployed as a way of seriously engaging with the design of such spaces that hold people within them (prisons, but also detention centres, camps etc.). Yet, much academic literature highlights the 'complicit' role that designers have in the creation of the atmosphere of a particular space (McKim, 2012). For Böhme, it is architects who create 'the conditions in which the atmosphere appears ' (2008: 4). These conditions are called generators, which as Edensor and Sumartojo explain, allow designers to create '"tuned" spaces with tones, hues and shapes, and it is not a stretch to identify the multiple ways in which spaces of all kinds are managed so as to produce a particular feel, mood or ambience ' (2015: 253). This process is particularly evident in what we might term the 'experience economy' (Pine and Gilmore, 1999), where sites of consumption (such as shopping centres), leisure (e.g. cinemas, theme parks), tourism (e.g. hotels) and the workplace are aestheticised in a certain way through lighting, design, performance, music and so on, to generate expectations about a place (Thrift, 2004). Such 'cultivation' of affect often includes attention to 'the size or positioning of windows, the proportions of the room, the number and forms of doors and other circulation apertures, the materials used, and the acoustics', as well as decorations, objects and 'those less knowing but nonetheless purposeful agents such as mould, damp, dust, rust, and fading' that all form part of an architectural atmosphere (Jacobs et al., 2012: 136). In the context of prisons, it does not take a huge leap of imagination to comprehend the ways in which designers 'avoid sensory and semiotic ambiguity' (Edensor and Sumartojo, 2015: 254) in other words, engineering spaces to meet the expectations and demands of their purpose.
Additionally, as Borch reminds us, 'very often, architectural atmospheres are carefully and consciously designed in order to achieve specific political objectives ' (2014: 82). This is particularly relevant in the realm of incarceration, where the process of prison procurement, design and construction is underscored by a much wider politics in relation to the agenda of the penal system in which the institution will operate (Moran and Jewkes, 2015;Moran et al., 2016). Scholars in geography have argued that 'mood design' can have political, social and economic outcomes (Adey, 2008;Kraftl and Adey, 2008;Bissel, 2010), which can range from 'turning a buck' to coercion or persuasion (Jacobs and Merriman, 2011: 218). This is critical in the prison environment where, among other things, a prison is a result of design choices that concern security and safety, discipline and control. For example, in Nordic prisons, an exceptional penal philosophy goes hand in hand with a less punitive, and more rehabilitative, architectural philosophy (Pratt and Eriksson, 2014). Here, a different visual aesthetic appears to conjure, or demand, a deliberate change in the atmospheric characteristics of that space. This raises the question of whether a particular type of atmosphere can be purposefully engineered and experienced. In other words, does a less punitive design equal a less punitive atmosphere? Would people in such spaces sensorially and affectively experience them differently to spaces that might be more retributive? This also raises an ethical dilemma of whether 'atmosphere' may be a means of contributing to the improvement and, therefore also, the persistence of the prison institutiona conundrum at the heart of contemporary abolitionist scholarship (Loyd et al., 2012;Gilmore, 2007).
Although this article cannot answer this question, it does aim to consider the tensions implicit in the workings of atmosphere as something 'felt'arising in a given situation, space and/or timeand its design potential. Adey's work on atmosphere likewise grapples with the contradictions of atmosphere in his study of the megacity (2013). He shows how atmosphere is a material, elemental state, which has an atomic, particular and molecular form. It forms part of the air and, as such, surrounds us and permeates our bodies as we breathe it in and out. More than this, whilst air is invisible to us, it is powerful, it has agency. Winds uproot trees, propel ships and impel turbines; and the pressure exerted by the weight of the air in the atmosphere of the earth shapes weather and dictates the boiling point of water. Despite, then, its seeming intangibility, atmosphere has matter and force. It also, though, has force for humansit is 'affective' -'a quality of environmental immersion that registers in and through sensing bodies while also remaining diffuse, in the air, ethereal' (McCormack, 2008: 413).
In sum, in the prison context, whilst atmosphere appears quite tangible, the language of affective atmosphere is 'partial, inchoate and fuzzy' (Fraser, 2021: 227). Our aim here is to extend Jewkes and Young's (2021) conceptualisation of prison atmospheres, as well as the other aforementioned work that brings the term 'atmosphere' into currency for making sense of environments of incarceration. The article also brings this discussion into conversation with the rich literatures on the spatial dimensions of affect and the sensesand their relation to atmospherefrom the cross-disciplinary literature in geography and design (see Adey, 2008Adey, , 2013Anderson, 2009;Edensor, 2012;Kraftl and Adey, 2008). In what follows, we present empirical material generated through an ESRC-funded comparative project based in the UK and Nordic region investigating how penal aims and philosophies (i.e. what prison is 'for') are expressed in the design of new prisons (completed in or after 2010); and how those prisons are experienced by prisoners and staff. 2 Our analysis draws upon selected data collected across the project, which included: 6 weeks of ethnographic observations by the lead author; 29 focus groups (of between two and six participants, with 75 participants in total), and interviews with 36 members of staff and 42 prisoners (including both genders, sentenced and remand prisoners and a variety of age groups who had also participated in the focus groups) at a prison in the UK, which housed adult male offenders and adult and young offender females. Under the terms of research access, no further identifying information can be given. The observations included cells, residential units, special care units, health centres, visiting suites, education spaces, outside spaces and workshops. Nineteen architects, landscape architects, contractors and justice sector professionals involved in prison design and construction in the justice context within and outside of the UK were also interviewed.
The mixed-methods approach enabled collective experiences to be heard in the focus groups, alongside the more personal and individual accounts generated through interviews, and both to be supplemented by ethnographic observations of interior and exterior spaces. Audio recordings were transcribed and analysed with NVivo, using an inductive coding method. Using this approach, 'atmosphere' emerged as a key node in an analysis of the wide-ranging discussions about the design of all elements of a new prison. In the following section, these findings are described.

Atmosphere(s) from, in and around prison architecture
Prison architecture has long been associated with attempts to deliberately engineer or evoke affectual responses. In their earliest guises, prisons exhibited a threatening exterior, where unassailable walls, towers and parapets were often decorated by gargoyles or figures pictured behind bars (see Figure 1(A)). This meant that although the bricks screened what was taking place within, the public could still be reminded of the sombre nature of the building and the detrimental aspects of committing crimesa key component of what was presented as a crime-deterrence strategy (Pratt, 2002: 37). As such, barred-windows, imposing architecture and thick exterior walls have become emblems of the traditional prison environment, evoking impressions of discipline and punishment whilst also bringing to bear discourses of life inside the prison. Media representations of prisons typically compound an impression of bleakness and severity. For example, the much-acclaimed BBC drama Time, broadcast in June 2021 was widely praised for its authentic portrayal of incarcerationyet the real, decommissioned prison in which it was set, HMP Shrewsbury, was painted grey so that interior shots would appear even more gloomy and oppressive than in reality (Carr, 2021). These mediated constructions shape public assumptions about poor conditions (coldness, dampness, darkness etc.) and restricted lifestyles (harsh diet, minimal personal belongings, and so on).
Historically, prison design has been inflected with quasi-utopian ideals of reform and rehabilitation, famously in Bentham's Panopticon that inspired Foucault's (1977) Discipline and Punish.
Here, a drum-shaped prison with a single central watchtower had cells lining the outer wall to enable constant surveillance. This design was premised on a belief that architecture could affect human behaviour and that a building could enable moral reform. Similarly, Eastern State Penitentiary (Philadelphia, USA, built in 1829) was the physical manifestation of an 'evangelical' rational and humanistic reformatory mindset, promoting physical separation to enable reflection, solitude and prayer (Haviland, 1999). Criticised for its damaging psychological impacts, this 'separate' system was later abandoned. Yet this follows Borch's (2014) point that atmospheres can be political; that is, there is a sensory politics to space where atmospheres harness a multitude of senses to govern or induce particular behaviours (see also Jewkes and Young, 2021).
As our architect-respondents in Scandinavia demonstrated, the way that a prison looksits visual, sensory cuesoften delivers a direct message in relation to how 'prison-like' a space should be: Well first of all for example a simple thing, we don't have iron bars. There's one place in the scheme, a special security prison in the prison, they have iron bars but they are actually not needed. It's more for visual, for telling the prisoner that it's going to be difficult to get out … It's symbolic yeah. It's ordinary glazed windows as everywhere else. (Interview, architect, Scandinavia) This point regarding visual design was echoed by our UK-based participants, who commented on the changes to the overall aesthetic appearance of the exterior of the most recently built prisons in the UK. Commenting specifically on a photograph of the front entrance to HMP Low Moss (a 784-bed, male prison near Glasgow, Scotland opened in 2012) (see Figure 1(B)), one prisoner's response is representative of the general appreciation for this design: It doesn't look like a jail. It looks like the front of a shopping centre, the way that is with the glass. So, it's very inviting for visitors, and it's not an ugly building … It maybe gives the Prison Service … maybe it's in their favour because they're saying 'we don't have our prisoners in a horrible place anymore'. Similarly, participants were often able to appraise whether the facility in which they were housed met the visual expectations of a prison. As Pallasmaa explains, the judgement of an environment is often 'immediately and synthetically grasped' (2014: 23), or 'felt' through visual affects. First impressions are therefore a central component of architectural atmospheres. In the following example, the interviewee suggests that the prison doesn't look 'too bad' but, in this case, 'too bad' is equated with not being punitive-looking enough: Being honest with you, from the outside of the main gate it doesn't look too bad but from [a distance] it looks like a Hilton hotel. It doesn't even look like a prison.
For this respondent, notions of what a prison should look like stem from wider cultural constructions placed in juxtaposition. Here, although the prisoner was unable to say for sure what visual image a prison should conjure, they were certain that it should not be aligned with something, such as a hotel, that connotes leisure, pleasure and freedom.
For Pallasmaa, 'we grasp the atmosphere before we identify its details or understand it intellectually ' (2014: 232). It is clear that in addition to their visual affects, '[m]aterials, colour, rhythm and illuminations' are considered to be 'strongly atmospheric' due to their capacity to generate embodied or haptic 'envelopments' (Pallasmaa, 2014: 241). Such tools have long been considered vital when appraising the atmosphere in the arts, which has a much longer history than in discussions in architecture. We might consider the 'atmospheric painting' demonstrated by the likes of JMW Turner and Claude Monet through their use of light and tone. In this way, atmosphere becomes 'both the subject matter and the expressive means of these paintings' (Pallasmaa, 2014: 234). Similar mechanisms may have been employed in an attempt to deliberately alter the atmosphere of the prison environment. In prison, colour palettes have historically comprised what Wener has termed 'traditional institutional shades ' (2012: 228). Indeed, as our participants frequently commented, 'prison grey' had become the prevailing colour of the visual carceral landscape.
However, as Wener further intimates, '[a]lthough it is clear than one cannot reasonably use color to fine-tune mood or behaviour … colors can and should be part of an overall design plan that can counteract the monotony and boredom of a placea potentially positive and stimulating effect for all users, inmates, as well as staff ' (2012: 228). In view of this, colours are considered to be a vital component of creating a mood or generating atmospheric e/affect. They are therefore a means through which to change the design choices made about the prison environment and the sensory affects and atmospheres emergent therein. Colour is also something that can be altered without having (significant in the grand scheme of things) cost, security or safety implications. As such, designers and building consultants often include colour choices as part of their appraisals of their innovation in the prison environment. One Scandinavian architect explained how designs for newbuild prisons are moving away from 'traditional institutional shades' such as 'prison grey': We used the colouring as the concepts; they made an entire booklet about that … so in the long corridors instead of having an institutional grey, whitish hospital-like feel to it, we worked with colour. But also, from a mood point of view, blue was used in places where probably there was more sunlight coming in. Opposite then; so shady areas we would use warmer colours. So, we tried to even out the composition so to speak so it had that feel of joy as you're walking around in the facility.
When considering such circumstances, we recognise that colour (and also lightsee Edensor, 2012) is not simply a visual characteristic, but one bound up in other haptic, sensorial contexts. It is entangled in the feeling or 'mood' of that space. As one UK-based architect concluded, colour was part of a design brief that is outlined by the Ministry of Justice for a new prison that required interpretations and consideration in relation to the 'feel' of prison space: It's quite interesting because it doesn't just give you a tick list or a round chart, or a Dulux guide if you like those. It does actually talk more about the environment it engenders. It talks more about daylight; it talks about appropriate colours, colour philosophy, engendering the right feelings in prisoners in incarceration. So, it does describe more than just 'this is the colour we want'. (emphasis added) Further, colours not only appear to have the potential to change the designed atmosphere of the environment, but to act as navigational devices and cultural reference points. In this example, the 'feel' of the health centre is purposefully engineered to harness expectations about healthcare environments outside of prison: [P]sychologists … have been employed to give us colour schemes and to say 'this is a mood changer'. And there are good examples of that … The house-blocks and the accommodation and all the activity areas have slightly different colouring to signpost so you'd know where you're going. It also changes the look and feel of the place. Waiting areas in healthcare and well-being areas are mimicked to be a bit like a doctor's surgery waiting room. Probably better than most actually … (Interview, building consultant, UK) (2015) we may argue that without the 'visual communication' methods of a traditional aesthetic, contemporary prisons leave those entering such spaces disorientated by the disruption of their expectations of these spaces. By changing the visual aspects/affects of prison space, we can arguably engineer a considerably different 'feel' or atmosphere of that space too. Yet, to return to Thrift's appraisal of the criteria that 'cultivate' affect, visual aspects are not the only vehicle for atmospheric interactions with space. Another such sensorial contributor to the carceral atmosphere is smell.

Following Edensor and Sumartojo
Prison is often considered to have a particular odour that pervades assumptions of the prison environment so directly that it is a central feature of narrative accounts of prison space such as Rudesind's (2006) Bang-up for Men: The Smell of Prison. Ventilation systems have been used in new-built prisons to deliberately 'engineer' the very air of the atmosphere in relation to this smell. This recalls Healy's (2014) discussion of the design of shopping malls, where ambient 'platform' air conditioning creates an affective atmosphere in an attempt to induce individuals to shop.
In new-build prisons, it was commonly recognised by our research participants that 'the smell of prison' was different in recently-built facilities. This, in turn, created a different sensorial atmosphere. Technological advances contribute to altering the ambience of the residential environment. Building Information Modelling (BIM) technologies are used at the outset to perform thermal energy simulation or modelling to calculate the performance of a building in relation to its ability to retain sufficient/radiate excess heat. In principle, such modelling can facilitate air circulation, which has an impact on the retention of odour in an environment with multiple bodies. And, the inclusion of in-cell sanitation in the form of toilets and showering facilities affords good hygiene practices. Additionally, materials used in the cell environments are considered to be robust, with durable surfaces that increase the likelihood of good maintenance. Individuals housed in new-build prisons are more likely to 'have the facilities on hand to keep their cells clean' (Interview, staff member, UK). As such, new builds are considered less malodorous than older prisons: How does the jail smell? It's certainly a different smell to this jail than there has been from other jails. It's not as stagnant … It's not so bad here. I don't have the same deep intake of breath when I walk out the front door of this jail, probably because it's more open and it's less stagnant. So, there is no overwhelming smell that I can get from this, other than your standard cleaning materials and … maybe not quite as austere as a hospital but certainly a clean smell to it … So, the smell in this jail is far better than before but old jails always had a stagnant smell to them, and the air on the outside of a jail used to be very different from the inside. I would suggest that that has changed in this jail. (Interview, staff member) In short, although perhaps cleaner, fresher, and healthier, several participants commented that their generic anchor pointthe 'stagnant' smell of prisonswas indeed missing in newly-built facilities; underscoring olfactory elements as both a material and metaphorical component of atmospheric expectations of prison.
To draw this section to a close, we argue that, when considering aspects such as visual imagery, colour and smell, it is evident that such components may be vital in the experience of, and more specifically, the engineering of carceral atmospheres. Although each criterion may be singularly e/affective in its contribution, what is evoked is an overall sense of a building that engenders specific atmospheric responses. Here, one staff member encapsulates these sentiments: I don't know how you've found it, but I'm used to prisons. Most of my adult life. But there's a feeling in this prison, there's an atmosphere in this prison that it's good. It's like a house. You know you get good houses and bad houses. You walk in here and you go 'this is alright. This is okay'. And I feel that. And a lot of us have commented on that. And I think that is partly the building. A significant chunk of it is the building.
Our participants, echoing much of the literature on prisons that mentions atmosphere, do so with confident acknowledgement of the very existence of a, the or some kind of prison atmosphere: This quote, like the excerpt from fieldnotes with which we began this article, is indicative of the many interviewees who recounted feeling something palpable emerging from the prison environmentthe 'bubbling' that prisoners and officers frequently mention. In almost all of these cases, atmosphere was affirmed concretelysomething did existbut almost always alongside the statement that it couldn't be described exactly and participants were unable to explain why or, in particular, how that atmosphere emerges. Accordingly, in these accounts, atmosphere is something of a known unknownpresent, acknowledged but largely unexplained.
Despite these appraisals of the carceral atmosphere as something that is 'stabilised' or 'repeated' through architectural expectations and associations, we recognise that an atmosphere may 'also have emergent properties that can change based on its configuration' as described previously (Turner and Peters, 2015: 325). Returning to the quote above, the interviewee posits the atmosphere as 'a time bomb'no doubt ready to change state, quite violently at any given time. In that sense, this prisonerlike many others sharing similar sentimentintroduces the notion that the prison atmosphere is malleable, and has the potential to transform, depending on the circumstances at play and the individuals involved. As one respondent explained: The atmosphere changes daily. Plus, if they put somebody in, I don't know, certain characters or whatever, then the atmosphere would change then. And different shifts can change it, with the staff. Different staff shifts can change it. (Interview, male prisoner) We can note that there is an anticipated or expected 'atmosphere' created in a prison environmenthow we might expect a prison to feelbut that this may be 'co-produced' or open to change depending on the people or circumstances. Multiple atmospheres may exist simultaneously, emergent from and within the same prison space. It is this latter potential, which demands consideration of whether it is possible to design or engineer a particular atmosphere in the prison setting. Such aspirations may be the raison d'etre of architectural philosophy but beg the question of whether the intentions are successfully transmitted to the end user, particularly in an environment that is often contingent upon individual-and societal expectations of prison space. In the following section, we explore what might be termed the atmospheric deviations of the prison environment, drawing attention in particular to the resulting implications that this may have in the carceral environment.

Atmospheric deviations
Although architects may envisage the creation of a particular feeling or mood for the design of a building, its manifestation may change in the 'architectural production' of that space (see Moran et al., 2016) and in the lived experiences of individuals who encounter it. We may draw upon our own experiences of how a space that is associated with happiness can be transformed into one of sadness depending on particular events; places that hold joyful memories might become tinged with melancholy after the unexpected death of someone who spent time there, for example. As such, buildings may well become the 'container' of atmosphere, but they are 'not "air-tight": the atmosphere will change' (Turner and Peters, 2015: 326). As one staff member described in a manner that is, like this article's epigraph, as vague as it is evocative: It changes. It really does change. I'm able to gauge when there's a little bit of an atmosphere or a little bit of a mood. You just get a feeling. I can't really describe it. You know when you get that feeling like 'oh, something's about to happen' or something's not right. It could be anywhere in your life, you just get that feeling. I've had that feeling many, many times in here. Something's not right.
In the prison, a variety of different factors (from the excitement around the supplementary food and small provisions [canteen] that prisoners can purchase from their weekly allowance to a particular event such as a film showing) can dramatically alter the atmosphere: Obviously, there's more of an atmosphere on canteen day for some reason. I don't know if that's because people are hyped up because they've just had their tobacco, everything's coming in the door. But it's very loud on a Monday, canteen. It's very loud on the hall. And certainly, on a Sunday night it's completely dead because nobody's got nothing: nobody's got no smokes, people are begging for stuff. It's always quite loud on a Monday. The rest of the week it … doesn't seem like there's much of an atmosphere in here. You always know when something's happening. Like that sort of thing, the atmosphere changes. Like there's a heightened … I'm quite sensitive with stuff like that. You can feel there's a different … it's hard to explain. It's a sort of sixth sense. You can tell when something's up, or when there's something just not quite right. I don't know if it's because we're so used to the way things are, that it's different, all of a sudden you notice it a lot more. (Interview, male prisoner) Therefore, in addition to being expounded as both something that one 'cannot put your finger on' and something intrinsically changeable under various circumstances, the atmosphere is also impacted by (or 'co-produced' to follow Edensor and Sumartojo, 2015: 252) by different people too. Never more so than in a prison environment can we consider Turner and Peters' suggestion that atmosphere will 'seep out as people leave and take something of it with them, and it will change again as other [people] enter in their place ' (2015: 326). It was frequently reported by staff-and prisoner-participants that the equilibrated environment of a particular residential unit or wing can be disrupted by the entry into or removal of certain individuals from that environment. For example, a unit atmosphere might transform from calm to fractious with the introduction of a protected prisoner into a setting which also houses mainstream prisoners; or encompass feelings of low mood created by the loss of prison friendships (see Caine, 2006) when long-term prisoners are released.
People are not the only thing that might be considered to change atmosphere in unexpected ways. In the prison setting, the weather was also noted as having a considerable impact on emergent atmospheres. Particularly hot weather was reported to provoke testy, irritable feelings and engender a more hostile environment: When it was really hot outside there were heaps of fights all of a sudden, like people's tempers. It's like it was hot outside, they were hot in here, kind of thing. It's weird. It seemed to have that effect. There were [sic] a heap of fights. (Interview, male prisoner) These situations were particularly compounded when there was a malfunction in temperature control mechanisms such as heating or ventilation systems. In this way, factors contributing to the more embodied, haptical appreciation of atmospheric conditions, such as those felt by the skin (temperature) and encountered by other organs such as the nose (as in smell, as previously mentioned) were considered to have a much more intense impact upon the individual. Therefore, it is crucial to acknowledge that atmospheres may be experienced in different ways.
As such, the extent to which designed atmospheres can 'close down or open up meaning and practices to improvisation, contestation, interactivity and experimentation' can 'vary widely' (Edensor and Sumartojo, 2015: 253). In prison, we might consider whether the atmospheric messages purported by architects in their engineering of these environments are conveyed to, and therefore experienced by, the individuals who spend time in them. Here, we return to colour, which we introduced earlier as something architects considered to be a vital component in engineering atmosphere. However, colour is also a clear example of contestation in designed atmospheres. Our research indicated a wide disparity in perceptions of colour in the prison landscape, particularly demonstrating indifference to these purported innovations. It was common for interviewees, when asked to describe the colour palette of the prison space, to reply that they 'didn't know', 'would have to think' or, like the following participant, 'didn't care': 'Probably green or something … I don't really care. I'm in jail, so I've just got to get the head down and get on with my sentence' (Interview, male prisoner). However, despite this apparent indifference, it is clear that colour does have (arguably more subconscious) affects with the potential to generate very negative atmospheric attachments. In spite of a cheerful colour palette, the following prisoner explains: I feel like the mood on the wing is just dark and gloomy, even though there is [sic] bright colours everywhere, but everyone is just in a zombie mode … I feel like I am in a hospital. I just feel like 'down' in here for some reason … I feel like there may be colours but everywhere should just be white walls and everyone should be just strapped to a straightjacket, that's how.
Here, the bright colours do little to erode a 'dark and gloomy atmosphere'. Instead, the prisoner considers white to be a more fitting descriptor of their world: the colourful landscape has conversely rendered their intimate spaces and emotions colourless.
Similarly, in relation to the removal of odour from prison, some respondents reported feelings of disorientationwhere the smell was part of a memoryscape (Moulton, 2015). Although, for many people, memories of prison evoke negative connotations, repeat offenders often demonstrate positive associations or even feelings of comfort with the regime of imprisonment. Here, the male prisoner interviewee comments on how the smell of prison, combined with its acoustic properties was essential in generating the particular atmosphere of prison: It gave good character. It was just like a prison, the smell of it, it took me back to the old buildings where they'd be like the clinic, a hospital and you'd have a clinic … It spoke to you. It was just the smell of it, the echo with doors shutting and slamming. It just spoke to me, you know what I mean?
As Edensor and Sumartojo explain, 'the inapposite of ineffective staging of atmosphere' can result in it being 'inimical to immersion and co-participation by those for whom it is designed' (2015: 257). Yet, more than just the dissolution of familiar atmospheres of incarceration, attempts to engineer such environments may be altogether contradictory to the intended outcome and purpose of that space. In other words, designers intending to engineer an overarching environment of respect, trust and humanity may not succeed in meeting all of those aims in relation to the expected associations of how carceral environments (or atmospheres) should function. Our participants were quick to comment upon situations where they felt that architects had produced spaces which were too far-removed from what they thought the 'prison atmosphere' should be like. In the following examples, the creation of 'homely' or 'comfortable' spaces was considered to have implications on the likelihood of increased recidivism: Because it's not designed in a way to make you think that you're actually in prison. To me … if I was homeless and I was outside and I saw something like that outside I would want to live in somewhere like that because it's like a proper bedsit. (Focus group, female prisoners) We may conclude that this is because atmosphere is contingent upon historical and cultural contexts. Pre-conceived ideas about what we/prisoners think a prison should feel like might differ according to where a person is from. Part of the affective qualities of atmosphere is that it can form different capacities and experiences in relation to different individuals and actors (Conradson and Latham, 2007: 238). In this respect 'atmospheres are multiply composed out of phenomenological and sensual elements, and the social and cultural contexts in which they are consumed, interpreted and engage with emotionally as well as affectively' (Edensor and Sumartojo, 2015: 252).
These ideas were also clearly discernible in the disparity between responses from individuals in the UK and Scandinaviaboth staff and prisoners as part of the wider project dataabout what prisons should look and feel like.

Conclusions
This article has been concerned with extending, expanding and taking seriously atmosphere as an analytic tool for making sense of prison experience and design. Atmosphere is a word often used in prison literature, but it is rarely unpacked or explored in its complexityas something elusive and felt (sensorial and affectual) and in tension with architectural desires and plans for prison estates and the 'goal' of rehabilitation. Drawing on qualitative research data from individuals designing, working and living in prisons, we have focused on how various componentsincluding aesthetics, olfaction, temperature, and the performances that arise from themcomprise sensory atmospheric affects in prison. In doing so, we find atmosphere(s) emergenot simply from the materiality of the prison itself, but from cultural constructions of carceral and non-carceral landscapes and in conjunction with personal practice and preference.
In exploring engagements with prison, we acknowledge the contingency of carceral atmosphere(s)where atmosphere is variously conceived and differently consumed. In doing so, we unravel the complexity of such understandings in the particular context of incarceration, leading us to conclusions that raise several challenges for prison design. In brief, we find that prisons do indeed evoke atmosphere(s); an ambiance or aura that pervades not simply from the material landscape of the prison itself, but from a range of other sources, including: experiences of media portrayals of such places; interactions with other physical landscapes of incarceration (prison walls, fences and gates etc.); and from personal assumptions about such places generated in relation to individual feelings about isolation, punishment or even confined spaces. Accordingly, the prison is tied to particular constructions about what a prison should feel like and how people should (re)act to/in such spaces. In some cases, prison designers attempt to use such understanding to engineer particular atmospheres that cohere with wider political motivations around penal philosophies. However, despite the common reflection that prisons generate some kind of atmosphere, respondents are unable to offer concrete descriptions of what this may exactly be, and much of our data highlights a precarity and changeability to atmospheric affect, which is likely to raise ambiguity around attempts to design carceral atmosphere.
To echo Pallasmaa, then, understanding atmospheres can 'teach us about the secret power of architecture and how it can influence entire societies ' (2014: 244). It can also help us stretch the ontologies that foreground much prison scholarship (i.e. notably Foucauldian scholarship for example) by providing opportunities to understand the very place of prisons, differently. What is a prison? How is it known, understood, lived and engaged? All of these questions are brought into new focus through a lens of atmosphere. This leads us to pose two final questions, which we offer in lieu of definitive conclusions.
First, despite the common reflection that prisons generate some kind of atmosphere, respondents are unable, often, to pin such atmospheres down. These reported feelings were not always consciously linked to specific design choices. This poses an interesting question of whether prison design, however innovative or far from traditional it is, can change what it means to be in prison. In other words, it raises the crucial tension between the desires of designers to create spaces of specific purpose, and the work of elusive affective atmospheres that counter such attempts. Indeed, much of our data highlights a precarity and changeability in atmospheric affect, which has the potential for multiple atmospheres at any one time in the prison environment.
Such appreciation of the precarity of the prison atmosphere (despite its lack of concrete descriptors) is still useful nonetheless. Participants often reported that they were able to 'feel' the potential changeability of their immediate environments, and reported a clear sense of knowing that the atmosphere of the prison had changed, regardless of whether they were able to fundamentally say how and why. Such efforts to conceive these atmospheric 'tipping points' (Anderson and Ash, 2015: 47) are likely to be critical for prison staff in their efforts to know and understand 'the changes in the health and atmosphere of the prison', which facilitates the 'operational grip' that staff can exert in prison space (Gooch, 2020: 18). Nevertheless, such variability is likely to raise ambiguity around attempts to design a carceral atmosphere, which leads us to our second question.
If atmospheres are characteristic manifestations of the co-presence of subject and object (2014)the 'passionate liaison' between bodies and environment (Bachelard, 1964) what does this mean for the relationship between prison and prisoner? Prison employees and prison architects will likely experience carceral atmospheres differently to those incarcerated. Without the inclusion of the end-user as a key stakeholder in the design process, we question whether architects are able to deploy atmospheric intentions for their final creation. Particular devices are designed into carceral space, but whether they are consumed as intended highlights an even more complex amalgam of atmosphere(s) than those that architects ordinarily grapple with. What certainly hits a note in this discussion is the notion that atmosphere is a 'heuristic device' (Martin et al., 2020: 64) that can arguably be deployed as a powerful tool to assist designers in communicating the intentions for a particular space. In the prison environment, questions of power and powerlessness are omnipresent. Accordingly, the question remains as to whether atmosphere should be harnessed in carceral space, and whether its engineering represents just one more tool for punitive power. situated position (see Rose, 1997) as a group of female academics, across career stages and disciplines thus offering a unique lens of approaches and theories in our co-working. Although we together have extensive experience working on/in prisons, our findings and readings of atmospheres are no doubt shaped by the fact we have not had the experience of either working directly as prison officers, or of being incarcerated. That said, 'our' interpretations are led inductively from the voices of those with whom we spoke so as to put the participants at the centre of the analysis. 2. ESRC grant ES/K011081/1: "Fear-suffused environments" or potential to rehabilitate? Prison architecture, design and technology and the lived experience of carceral spaces (2014-17).