Inside the Belly of the Penal Beast: Understanding the Experience of Imprisonment

One of the striking characteristics of much ‘big picture’ penal scholarship is that it stops at the gates of the prison, or breaches its surface somewhat barely or briefly. This article proposes that such work could be advanced and made more compelling if its insights were married with – and modified through – those provided by empirical and ethnographic analyses of the practice and experience of penal power. It then sets out a framework which would enable this form of engagement and analysis, first providing an account of the development of the main components of the framework, before elaborating in more detail how its constituent parts – depth, weight, tightness, and breadth – might be conceptualised. It concludes by offering some reflections on the practical implications of this agenda for prison researchers.


Introduction
According to a range of highly influential accounts, the aims and functions of punishment have been radically transformed. Feeley and Simon's (1992) ground-breaking 'new penology' thesis posits that concerns to rehabilitate prisoners have been supplanted by discourses that prioritise retributive austerity, risk management, and low-cost containment (see also Simon 1998;Wacquant 2001). Other prominent theorists have described a 'crisis in penal modernism' (Garland 1990: 4), in which penal sensibilities have hardened, the retributive and expressive dimensions of punishment have been revived (Pratt 2000) and the rehabilitative ideals that were emblematic of an era of penal welfarism have been reconfigured or virtually abandoned.
However, such 'grand narrative' accounts have significant blind spots and limitations. As Nicola Lacey (2008: 26) notes, much of the literature on 'late-modern' or 'neo-liberal' penality is too schematic, and risks 'elevating an explanatory framework largely informed by the specificities of the US situation to the status of a general theory'. In his 2012 Sutherland address to the American Society of Criminology annual meeting, David Garland (2013) himself makes a similar (self-)critique: by remaining largely at the macro-level, such broad theorisations fail to explain IJCJ&SD 51 Online version via www.crimejusticejournal.com © 2015 4(1) how structural trends play out locally; that is, why similar kinds of social and economic determinants are 'translated' into very different penal outcomes on the ground (Garland 2013: 483). In part, this tendency reflects assumptions of growing convergence in penal practices and the global spread of neo-liberal ideals (Garland 2001;Wacquant 2009). It also exposes a wider pattern in contemporary criminology whereby penal theorists and prison ethnographers work in domains that are largely disconnected from each other (Carrabine 2000(Carrabine , 2004, so that theorists tend to operate with an impoverished sense of empirical complexity, while on-theground prison research is too rarely informed by or oriented towards debates about wider trends in penal values and practices. This 'analytical division of labour' (Carrabine 2004: 3) is seen just as clearly in relation to the body of work which has explored the relationship between political-economy and penal severity (inter alia, Cavadino and Dignan 2006;Lacey 2008), seeking to move beyond more abstract and sweeping analyses of contemporary penality. Such research has exposed what it claims are 'almost watertight dividing lines' between different types of national political-economies and their respective penal practices (Cavadino and Dignan 2006: 446), explaining variations with reference to broader social arrangements, national identities, and the strength of state bureaucracies. John Pratt's publications on 'Nordic exceptionalism' (for example , Pratt 2008a, Pratt , 2008b are exemplars of this kind of account, arguing that the Nordic countries' distinctive histories and socio-cultural arrangements shape penal systems that are unusually mild and humane. Writing more recently with Anna Eriksson (Pratt and Eriksson 2013), Pratt has gone much further than most in trying to detail the roots of different penal sensibilities and in seeking to evidence claims of penal tolerance through the inclusion of some measures that are 'experiential and subjective' (Ugelvik 2013: 580). In Contrasts in Punishment, for example, Pratt and Eriksson (2013) identify a number of broad indicators of penal difference, which include 'qualitative' factors, such as the closeness and respectfulness of staff interactions with prisoners, and the nature of reception processes, as well as more conventional measures such as food quality and hours out of cell.
Yet one of the striking characteristics of much of this 'big picture' penal scholarship is that it stops at the gates of the prison, or breaches its surface somewhat barely or briefly, judging levels of harshness or humanity largely by metrics such as imprisonment rates and prison conditions, or official sources and first appearances. Some reviewers of Pratt and Eriksson's research have been critical of a method which relies more on literary references, official reports and accompanied prison tours than on an immersive engagement with prison life and prisoners themselves (Minogue 2009;Ugelvik 2013). Such criticisms, while legitimate, risk asking too much. It is churlish to expect anyone to combine detailed historical-theoretical work with the serious empirical graft of prison ethnography. Moreover, key writers in this field have acknowledged that imprisonment rates are 'crude' (Cavadino and Dignan 2006: 452), 'have limitations' (Lacey 2008: 43) and are 'only one strand of penal exceptionalism' (Pratt 2008b: 14). Some have attempted to look beyond the most basic indices of penal severity. Lacey (2008: 31) notes that conditions in many US prisons are 'staggeringly harsh: overcrowding is widespread, rape and other forms of violence are endemic and constructive prison regimes are rare'. Pratt (2008a: 123) interrupts his narrative to note that 'one in eight inmates at Helsinki prison still request to be placed in isolation at some stage of their sentence'. There is some recognition, then, that the prisoner experience has social dimensions beyond issues of crowding, human rights and material provision.
Nonetheless, for scholars whose research lives are spent inside prisons, peering out, rather than outside them, squinting in, the deficiencies of these measures are -for a range of reasons -all the more arresting. First, imprisonment rates, while undoubtedly telling, are poor indicators of the prisoner experience: it seems perfectly possible that a state might imprison relatively few people, but in a highly punitive manner, or relatively many, in conditions that are decent. Valverde's (2010: 118)  them there are' is perhaps over-stating it, but it emphasises the fact that punitiveness is about more than quantity.
Second, what actually occurs in prisons cannot be derived from official aims and policy rhetoric. As Garland and Young (1983: 18) argue, 'The actual practices of the prison cannot be deduced from the Prison Rules [...] The objectives, purposes and effects of specific practices can only be identified on the basis of concrete empirical analyses'. In the UK, for example, it is only because of the primacy given to official proclamations over insider testimonies that it was possible to characterise the post-war period in terms of 'penal-welfarism'. For while policy documents and pockets of penal liberalism gave an impression of progressive penality, first-hand accounts of prisoners and staff belied such characterisations (see, for example, Caird 1974). Rehabilitative ideals were never embedded in staff practices, leading to gross disparities between official discourse and the reality of prisoners' daily lives ; see also Nilsson 2012).
Third, it is a mistake to confuse the material conditions of imprisonment with prisoner quality of life (Liebling assisted by Arnold 2004), or, as the classic works of prison sociology emphasised, to neglect the relevance of the prisoner social world to the prisoner experience (Sykes 1958). Conditions and facilities matter a great deal, in that prisons that are grossly overcrowded, physically dilapidated and highly austere are unlikely to be experienced by prisoners in positive ways. 1 But prison disorders often occur in prisons that are newly opened, with high quality accommodation and facilities; and some prisons that are physically rather decrepit demonstrate comparatively high levels of prisoner wellbeing due to positive relationships between prisoners and staff. As Liebling's work has demonstrated (Liebling assisted by Arnold 2004), this is because how a regime is delivered matters as much as its substance (see below). The relationship between frontline staff and prisoners does as much to determine whether a sentence is survivable or de-humanising as the frequency and quality of visits, or the availability of educational opportunities. Indeed, as Liebling and Arnold (2004: 333) note, 'Even "numbers of hours unlocked" [...] varies in practice, according to staff behaviour on the day'. Furthermore, whether 'time out of cell' is experienced as a positive thing or not will depend on a prisoner's position within the prisoner community: whether he or she is being victimised by his or her peers, for example.
Fourth, most standardised and quantifiable measures get us only so far in understanding the prisoner experience, which is far more nuanced than most metrics can convey. Some, such as the length of prison officer training and the availability of education courses and structured programmes, are 'inputs' or 'outputs'. Whether they lead to a better quality of experience for prisoners remains an empirical question: prisoners do not always welcome offending behaviour courses, for example, if their content is considered infantilising, or if attendance is in effect mandatory (see Crewe 2009;Shammas 2014). Others, like the length of prison visits or the cost and availability of prison phone calls, are good starting points for cross-jurisdictional comparison, but may occlude as much as they expose, unless they are supplemented with greater detail. The impact of overcrowding will depend on how long prisoners have to spend in their cells and, while some prisoners might find cell-sharing preferable to solitary confinement (particularly if hours of 'bang-up' are long), others will value isolation over the perils and irritations of company (see King and Morgan 1980). To give another example, Shammas (2014: 116) notes that, in closed prisons in Norway, 'inmates may be prohibited from using their native language', a practice which might well make factors such as the opportunity to access phones and the cost of outward calls somewhat irrelevant. Moreover, the frustrations of imprisonment are subjectively experienced in ways that are dependent on context (what is lost as a result of being imprisoned) and relative expectations: in open conditions, prisoners may experience the near-reach of freedom as bittersweet, while those in the most progressive parts of any system 'may feel worse-off the greater the access to goods and services' (Shammas 2014: 116), as their sense of entitlement changes. It is difficult, therefore, to be satisfied with accounts of the prison experience that dip only a toe into the swirling waters of the carceral sea. To feel such misgivings is not to diminish the value of recent work on either 'neo-liberal' or comparative penality, nor to dispute that the amount of punishment meted out by the state is a key indicator of its punitiveness. Rather, it is to propose that such work could be advanced and made more compelling if its insights were married withand modified through -those provided by empirical and ethnographic analyses of the practice and experience of penal power. For current purposes, such studies are vital for two further reasons. The first is reflected in the scepticism that has been expressed about 'Nordic exceptionalism', primarily by Nordic scholars themselves (see Ugelvik and Dullum 2012). Most notably, Barker (2012) highlights a range of practices, including the use of pre-trial detention (often involving solitary confinement), 2 and the compulsory treatment of drug addicts and alcoholics, which represent highly intrusive and oppressive forms of penal power. Notably too, such interventions are disproportionally used against certain categories of people -particularly those perceived as 'outsiders' -in the name of the 'public good'. The inclusionary concept of the Nordic welfare state as a 'people's home', with a place for everyone, might in fact rely on (and mask) practices and processes which are in many ways, and for certain people, decidedly exclusionary (see Ugelvik 2012). Nordic penality, then, appears more 'Janus-faced' (Barker 2012) and paradoxical than orthodox accounts have claimed. Any penal system should be interrogated for such contradictions, especially where their impact might be felt differentially by different sub-groups and populations.
The second reason relates to some key changes in penal practice, which can be summarised as the 'softening' of penal power. In many jurisdictions, psychological power has superseded coercion as the primary basis for control and compliance. Accordingly, many of the conventional pains of imprisonment -those shaped by austere and authoritarian regimes -have become less onerous and less conspicuous. But as penal power has been reconfigured, it has generated a distinctive set of demands and burdens, which are less visible, harder to quantify and more difficult to 'read' than the more familiar traits and traumas of confinement. Many of the forces that operate upon prisoners have an ambiguous quality, inciting and 'incentivising' them to behave in particular ways without the need for direct or immediate compulsion. Staff-prisoner relationships have become more relaxed, but in many ways more complicated (see Crewe 2011): the greater use of staff discretion adds a layer of complexity to daily interactions, as does the fact that prisoners rely on staff for positive reports. Prison life is described as 'easier but hard' or 'softer but shitter' (see Crewe 2009: 111-112), highlighting some of the experiential contradictions that result from attempts to de-rigidify the system, and make it work more accountably and transparently.
These changes and their consequences -described in greater detail below -require us, as researchers, to adjust some of the lenses through which we have traditionally thought about penal power and its outcomes. To quote Shammas (2014: 115; emphasis added): While viewing prison conditions through the optic of 'humane' or 'inhumane' conditions arguably produces a flattening, one-dimensional gaze -the task of prison scholars might better be understood as studying how punitive power varies in kind rather than degree, how the nature of pain-imposition varies qualitatively, producing incommensurable pains that are to some extent not easily given to cross-national comparisons.
The difficulties of comparison should not be under-stated. But they make it all the more important to develop a conceptual apparatus which allows us to depict and compare prison systems and the prisoner experience in all of their complexities, going beyond standard measures of the quantity and severity of incarceration, and recognising that the experience of any penal state -the balance of frustrations -will vary by sentence stage, establishment, and variables such as sex, ethnicity, citizenship and age.
IJCJ&SD 54 Online version via www.crimejusticejournal.com © 2015 4(1) To conduct this kind of comparative penology, and to engage seriously with fundamental questions about what it feels like to be imprisoned, requires considerable conceptual and empirical effort. The remainder of this article first seeks to set out a framework which would enable this form of engagement and analysis. It begins with a brief historical account of the development of the main components of the framework, before elaborating in more detail on how its component parts -depth, weight, tightness, and breadth -might be conceptualised. It concludes by offering some reflections on the practical implications of this agenda for prison researchers.

Depth
In his 1988 analysis of British and Dutch penality, Contrasts in Tolerance, David Downes observed that comparative work on the experience of imprisonment had tended to concentrate on the length of prison sentences. As Downes noted, sentence length leant itself to quantitative analysis, and the shorter average sentences given out in the Netherlands compared to England and Wales helped account for its proportionately lower prison population. Yet as Downes (1988: 165-6) highlighted, for the individual prisoner, the experience of imprisonment was about what it felt like as well as its 'amount'. Some elements of what Downes referred to as the relative 'depth' of imprisonment reflected systemic issues, including the quality of food and clothing, levels of prisoner pay, facilities for washing and using the toilet, and rules governing letters, phone calls and visits. Others were what might now be called 'relational' issues: how staff used their power; the degree to which they treated prisoners with humanity; and whether they enacted the institution's disciplinary system in a manner that was strict and arbitrary or flexible and fair. In English prisons, Downes (1988: 179) summarised, imprisonment was experienced as 'an ordeal, an assault on the self to be survived, time out of life', whereas in the Netherlands, the rupture of confinement was 'not so marked, the passage of time less prolonged, the sense of social distance from society less acute, and the problems of psychological survival less chronic' (p. 179).
This summary hinted at two relatively discrete dimensions of the prisoner experience that Downes (1992) himself differentiated: first, the degree to which it is psychologically onerous and oppressive; second, a definition of depth that was centred on the relationship between the institution and the outside world: By depth of imprisonment is meant the openness of the prison life to the outside world, both in terms of the actual opportunities for contact with family and friends by visits, home leave, letters and the telephone, and also by the permeation of the institution by outside world agencies, whether recreational (visiting pop groups, etc.), informational (access to the media, newspapers, etc.) or social (visits by students, politicians, academics, etc.). (Downes 1992: 15-16) As King and McDermott (1995) subsequently highlighted, this definition of depth was much closer to the way that prisoners used the term. When talking of 'deep end' imprisonment, they generally meant being in high-security prisons or highly-controlled conditions, 'deeply' embedded in the system, and many years from release. The term 'depth' also conveyed a sense of being buried way beneath the surface of freedom -almost subterranean. Conceived in this way, the simplest definition of depth might be the distance or polarity between the prison and the outside world, with distance having an almost literal as well as a metaphorical meaning. Such a definition (which can be applied both to prison systems and individual circumstances) encompasses sentence length (that is, temporal distance from freedom), as well as levels of security and control, which contribute to the prisoner's sense of the remoteness of their existence, and their degree of physical freedom and bodily autonomy. Such matters tend to be connected to a prison's overall security level, since higher-security establishments generally IJCJ&SD 55 Online version via www.crimejusticejournal.com © 2015 4(1) place more restrictions on prisoner movement and subject them to more frequent searches. However, it is possible to be in 'deep' conditions within a shallow establishment, if segregated within a low security prison or while on remand, and there are subtle but important distinctions between different levels of restriction and supervision (so that, to draw on an example from King and McDermott (1995: 86), prisoners can be highly restricted when moving from one part of the prison to another, but relatively free when on their wings to 'get some distance from staff'). 3 As Downes's comments make clear, depth is also determined by the degree to which the walls of the prison are permeated by external discourses, organisations and personnel. Even the institutions most deeply buried within a prison system may be perforated by visits and phone calls, or normalised by streams of media, technology and information from the outside world. The depth of any one prison is unlikely, therefore, to be uniform. Prisons contain zones and niches (Crewe, Warr, Smith and Bennett 2014;Toch 1992), such as faith centres and classrooms, which feel relatively less 'prison-like', or allow forms of spiritual and mental escape from carceral normality. Such outlets, as well as the presence of members of the opposite-sex (for example, in the form of staff or visitors), animals, forms of nature and aesthetic variety, all contribute to the degree to which prisoners may feel dislocated from the free community. The degree to which the normal rights of citizenship (including voting rights, conjugal or family visits, for example) are extended into the prison environment is also encompassed by 'depth', given that they might also encourage differential feelings of social inclusion.
There are other ways in which a prison can be deep in some respects but not others. As Liebling, Arnold and Straub (2012) found in their study of a high-security establishment in England, prisoners may be deeply suppressed by security and surveillance systems, while at the same time encountering familiar aspects of the outside world through the presence around them of friends and family members from their external lives. Ironically, then, when the carceral net of a society is cast more widely -or, to adapt Wacquant's (2001) terminology, when prison and ghetto begin to 'merge and meld' -for some sectors of the population, imprisonment may, in certain respects, become a less ruptural experience.
A consideration of the depth of imprisonment should mean paying particular attention to the points at which prisoners are submerged into or emerge out of the system. Both symbolically and materially, entry and exit represent stark moments of exclusion and inclusion, when the moral status of the individual is profoundly altered, and the size of the schism between freedom and captivity is perhaps most apparent. Prison systems, and individual establishments, differ in the degree to which they involve forms of ritual denigration, shaming and material dispossession on entry (c.f. Goffman 1961), and converse forms of symbolic reintegration and re-possession on exit (see Maruna 2011). 4 Pratt and Eriksson (2013) briefly contrast reception conditions in Anglophone and Nordic prisons but, since Goffman, prison researchers have paid little attention to the micro-processes of institutional induction: what is done to the incoming prisoner's body, how he or she is addressed, and the expected forms of deference. Even less has been written about rituals of departure: whether prisoners leave the institution with a supportive pat on the back, and even some symbolic certification of penal completion (Maruna 2011), or are seen off by staff with indifference or cynicism ('you'll be back').
Exploring these key chapters of the penal process promises to expose the wider functions of punishment, including the normative elements of state sanctioned opprobrium and stigma. Such considerations are rarely visible in sociological studies of prison life, but are pregnant in assumptions about the less exclusionary nature of Nordic penality (which is said to restore 'full membership' of society to former prisoners [Pratt 2008a: 130]). In highly cohesive societies, feelings of shame and moral self-disgust might be all the greater among those who break legal and social norms, though one might also expect them to differ between those who are within and outside national 'in-groups'. In this respect, then, both the subjective dimensions of the prisoner experience and its differential quality demand greater attention.
Since many prison estates are organised through some consideration of security level, thinking about 'depth' should alert us to the extremities of any penal system. Again, our concerns here should be more than material: not just the limitations on free movement, communication and communion, but also the extent to which practices in these darkest corners of the system are consistent with those in its less extreme spaces. Similar questions should be asked in the shallow end of a prison system: exactly how 'open', how porous, and how congruent with community life are these establishments? As Shammas (2013) describes, open prisons offer transient tastes of freedom, but this remains a long way from freedom itself, and may generate difficulties of its own. Prisoners who spend their days in the community and their nights in prison may experience cognitive confusion over their status, while increased contact with family members may make feelings of powerlessness more rather than less acute. At the same time, the beguiling fruits of liberty -the possibility of drinking alcohol, taking drugs, and developing intimate relationships -become honey-traps of risk. This in itself functions as a distinctive kind of burden -a form of tightness (see below) -passing greater responsibility onto the prisoner to self-regulate his or her desires. Meanwhile, lower staffing and greater freedom of movement combine to create an environment in which prisoners may face higher threats from each other than they do in more controlled environments. Here, we enter the territory of the weight and lightness of imprisonment.

Weight
In clarifying the meaning of 'depth', King and McDermott (1995) noted that many of the aspects of imprisonment that Downes had described using this term were better expressed through the metaphor of 'weight'. This term communicated the almost palpable burden of certain kinds of prison regimes: the sense that in establishments where staff are petty and provocative, where there are endless 'rules-within-rules-within-rules' (Cohen and Taylor 1978), and where standards and conditions are poor, the experience of imprisonment can 'bear down' upon prisoners or feel like a 'weight on their shoulders' (King and McDermott 1995: 89). Such factors had been considered in earlier work, most notably by Street (1965), who differentiated between the levels of degradation that prisoners experienced (that is, the quality of the regime), and the ways that staff exercised their authority. King and McDermott's book, The State of Our Prisons, was an attempt to measure such variations in weight and depth more systematically, employing both official regime data and structured questions to compare five prisons in England and Wales. 5 Of particular interest for current concerns is King and McDermott's (1995) guarded approach to the use of official performance data -including measures such as 'purposeful activity', 'because it includes too much and it regards all activities as equally purposeful' (p. 224) -and their recognition of the need for subjective evaluations of what they called 'reasonably "objective" situations' (p.155), such as cell space, access to legal advice, and (the quality of) work. Liebling's 'moral performance' framework advances this approach considerably (see Liebling assisted by Arnold 2004), drawing on interviews with prisoners about 'what matters most' in prison to develop a survey which measures prisoner quality of life along a range of dimensions without any reliance on official data. Matters such as prisoner safety are evaluated based on the perspectives of prisoners instead of on limited institutional data -such as serious assaults', which may capture only a fraction of violent incidents -in a way that recognises that feelings of safety are about more than just physical security. Prisoners' views are taken to be subjectively important regardless of 'actual' levels of violence.
Also underlying this tool is the claim -substantiated in its subsequent findings -that the primary determinant of the prisoner experience is the direct relationship between prisoners IJCJ&SD 57 Online version via www.crimejusticejournal.com © 2015 4(1) and frontline staff. What 'matters most' to prisoners is how they are treated as humans by other humans, and even 'regime' matters, such as time-out-of-cell and the quality of visits, are shaped and mediated by the mundane ways in which officers and other personnel choose to use their authority. To put this another way, first, the 'moral/relational' elements of imprisonment (respect, humanity, trust, fairness) matter as much as 'facilities' and 'conditions'; second, prison regimes are accomplished by people, so that (to give an example) the access that prisoners have to a prison workshop, the chance that access might be capriciously withdrawn, and the likelihood that work is considered an aspect of personal development is shaped by the wider culture of staff-prisoner relationships. This culture is itself determined by the 'philosophy of self' that staff adopt -how they see prisoners as moral subjects -matters that are not a simple outcome of 'inputs' such as the length of officer training, or even their benign intentions (Crewe, Liebling and Hulley 2011).
The moral performance framework provides a more comprehensive conceptualisation of the experience of imprisonment than anything that has come before, comprising measures of security, safety and staff professionalism -which might best be categorised in terms of 'depth'as well as staff-prisoner relationships and regimes. However, it was never intended as a specific measure of 'weight' so much as an evaluation of the 'interpersonal and material treatment that render a term of imprisonment more or less dehumanising and/or painful' (Liebling assisted by Arnold 2004: 473; emphasis in original). Indeed, little direct reflection has occurred in relation to 'the weight of imprisonment' since King and McDermott's (1995) reappraisal of its meaning. 6 Yet recent research has revisited the concept of weight and questioned assumptions that prisons that are heavy are always 'worse' than those that are light (Crewe, Liebling and Hulley 2014). For while 'lightness' suggests a less oppressive environment, it also hints at some undesirable characteristics. Thus, prisons in which power is under-used, or in which staffing levels are thin, may be experienced by prisoners as insubstantial or deficient. At their 'lightest', such establishments are chaotic and disorganised, and a reluctance or inability among staff to exert authority leads to regimes that are un-boundaried and unsafe. In such circumstances, the problems that develop between prisoners are not -as Downes emphasises in his initial reflections on 'depth' -the outcome of institutional oppression or antagonism. That is, they are not the result of 'pent-up grievances' caused by overcrowding, lack of privacy, or the excessive imposition of staff power (Downes 1992: 18). Rather, they result from deficits in the use of power, which discourage prisoners from exercising self-control and enable exploitation, the acting out of hostilities, and the development of violent informal economies.
Whether a prison that is light is experienced positively or negatively therefore depends on the combination of 'weight' with a related phenomenon: the 'absence' or 'presence' of staff power. 'Presence' has a literal dimension (whether staff are actually there, how they use their power and their ability to 'get things done') and a more communicative component (how they convey confidence, and set boundaries and expectations). Where staff are 'present', prisons that are somewhat 'overweight' may provide forms of safety and support that are lacking from those that are 'underweight': both 'light' and 'absent'. The ideal establishments ('light-present') do so almost imperceptibly, so that power flows unobtrusively, through relationships ('knowing your prisoners') and the legitimate use of authority.

Tightness
In their 1998 article, 'Mind games: Where the action is in prisons', McDermott and King (1995) noted that both staff and prisoners 'seemed to agree that things were not what they used to be' (p. 373). Physical confrontations were scarcer, control and restraint measures were exercised more carefully, and 'bits of paper' had replaced 'physical beatings' as the basis of control. Prisoners serving life sentences were particularly aware of the power inherent in reports written by staff about their behaviour. Such observations anticipated significant shifts in the nature of penal power. In Crewe's (2009)  England, prisoners drew attention to a conventional set of concerns, such as staff treatment and opportunities to engage with the outside world. But many of their experiential metaphors, and their most keenly felt frustrations, related to other issues which could not be captured through notions of depth or weight.
These shifts are connected to the softening of penal power -attempts to make regimes more decent, and staff cultures less authoritarian -alongside a 'compliance project' (Liebling assisted by Arnold 2004), which seeks to 'responsibilise' prisoners, hold them accountable for their own risk-reduction, and make them self-regulate a wide range of activity. In England and Wales, it also reflects changes in sentencing conditions and penal practices (Crewe 2009), including an increase in indeterminate sentencing, which makes forms of psychological assessment and power increasingly central in shaping the prisoner experience. For the system to make decisions about progression and release, prisoners are subjected to distinctive forms of power and knowledge, which assess them according to an actuarial logic, and impose upon them judgments and labels that are highly consequential. As described elsewhere (Crewe 2011a(Crewe , 2011b, these processes are often experienced as highly de-humanising. For most prisoners, their mechanics are bewildering: an opaque form of fortune-telling, with significant consequences in the present. Often they are conducted by off-wing staff whom prisoners barely know, and who they do not feel know them. In reducing subjective complexity to risk-based scores and categories that are relevant to the institutional machinery, they communicate to prisoners a lack of interest in their individuality, and their own understandings of who they are. Cognitive-behavioural and offending behaviour courses, while often welcomed, have similar properties, working not just on prisoners' behaviours but their identities and subjectivities: the essence of who they are. These forms of power are highly intrusive (Fox 1999), demanding the kind of candour that is typically reserved for intimate relationships (Day, Tucker and Howells 2004), but in a context where imbalances in power could hardly be starker.
Here, then, power is psychological and informational. It is grounded in 'paperwork' and computerised files, changing the nature of the prisoner experience. Once documented, comments and labels are highly adhesive -difficult to challenge or erase -and liable to take effect at an unpredictable point in the future. Power hangs suspended. Policies such as differential incentive schemes, mandatory drug testing, and early release schemes have similar effects, not only diffusing power to a wider range of staff (creating new arbiters of treatment and release), but altering its institutional location. Control and surveillance occur from a distance, without the need for direct oversight. They operate through abstract systems of assessment and observation, and prisoners' own recognition of the imperative to monitor a broad range of conduct: their mundane interactions with other prisoners and staff, their 'attitude' and demeanour, and their peer group associations. Responsibility for control and compliance is thus transferred from the institution to the prisoner. Furthermore, in England and Wales, prisoners are expected to 'constructively engage' (Attrill and Liell 2007: 195) in their own penal management. To achieve 'enhanced' status, for example, it is not enough to be docile and obedient: 'The absence of bad behaviour alone will no longer be sufficient to progress through the scheme' (NOMS 2013: 5). Instead, prisoners are required to 'demonstrate a proactive and self motivated level of engagement with the requirements of their sentence plan' (NOMS 2013: 11), and be enthusiastic participants in other aspects of the regime. Being a prisoner becomes considerably more demanding.
The experience of imprisonment is also shaped by the increasing bureaucratisation of the prison and the movement of certain forms of power away from the wings and landings. Prisons in general, and frontline staff especially, have been made more accountable, so that the collective power of uniformed grades has diminished, with decision-making pushed to higher levels in the organisation. The frustrations that result are expressed in prisoners' complaints about the difficulties of navigating the administration: getting clear answers to questions about sentence conditions or the criteria for acceptance onto offending behaviour courses, for IJCJ&SD 59 Online version via www.crimejusticejournal.com © 2015 4(1) example. They are felt as a form of exasperation, where prisoners feel powerless in the face of an exacting and impenetrable bureaucracy which feels unresponsive to their needs and timetables, and indifferent to them as human beings with complex circumstances and individual needs. Much of this frustration reflects the fact that 'the system' appears faceless and disembodied, without a clear 'centre' or a person with whom to argue or negotiate. Thus, prisoners talk about the difficulty of appealing against a system of this kind ('you can never find out who to talk to'), or even finding out where an appeal should be lodged ('you get told "that's a different department"').
Tightness conveys the 'grip' of these aspects of the prison experience. For while new forms of penal power may be 'lighter' and less coercive than other means of control, they are highly demanding. Their focus is broad and constant, and their reach extensive. 'Tightness' communicates the way that power is not simply imposed upon prisoners from above, or by identifiable individuals at specific moments. Instead, its flow is anonymous and perpetual (Foucault 1977), occupying the space all around prisoners, working on multiple aspects of behaviour, cognition and identity. It does not 'weigh down' so much as 'wrap up', entangling prisoners in a web of regulation and self-government. It is more intrusive than 'heavier' forms of penal power, which are psychologically oppressive, but do not seek to refashion the inner self as such. Similarly, the prisoner's autonomy is assaulted in ways that differ from the effects of situational constraints (bars, walls, searches, and so on), or the direct use of staff discipline. Indeed, rather than being constrained and diminished, in the pursuit of obedience, the prisoner's autonomy is, if anything, expanded. Instead of being 'commanded' to behave in particular ways, the aim is to give prisoners space to make decisions, and invite and incentivise them to exercise this autonomy 'appropriately': to develop new forms of subjectivity.
As Barker's (2012) analysis also suggests, these softer forms of penal power are not necessarily experienced as more humane than harder forms. Tightness represents the invasiveness of penal power, the breadth and reach of its demands, and the degree to which it is enveloping, allencompassing, navigable and negotiable. It is difficult to operationalise, or to define in simple terms of 'better' or 'worse', in part because some prisoners will welcome psychological interventions, and because the differences between forms of 'client-centred' and 'systemcentred' rehabilitation are not always clear-cut. Moreover, the stage of the system that is in many ways 'tightest' is often its endpoint, when prisoners are given increasing responsibility but face greater scrutiny, at a time when other aspects of their experience become easier to bear.
Furthermore, prisons can be characterised by 'laxity' or 'looseness' as much as 'tightness'. In general, as they become less rigid, authoritarian and austere in their daily operation, they also become less clear about the terms and boundaries of acceptable conduct. While recognising the benefits of 'softer' regimes, prisoners often complain about the inconsistencies in rule enforcement that result from more discretionary governance. Some establishments are less transparent and predictable than others, or less able to offer clarity about the basis of progression through the system. In other circumstances, prisoners are subjected to the demands of self-governance and psychological assessment, but the institution fails to fulfil its related obligations. Prisoners receive poor quality sentence plans or have to delay parole hearings due to incomplete paperwork. They do not know who their offender supervisor is, or are not directed to the interventions they need in order to progress. It is the bureaucracy, rather than the regime, that is disorganised and understaffed, so that everyday life quality might be high (that is, safe, purposeful, respectful) but prisoners cannot advance towards release.
Here, then, we see the two incongruent life-worlds in which prisoners are psychologically engaged: the present, where 'depth' and 'weight' may be most germane, and the imagined future, where 'tightness' bears more relevance. current frustrations?', for example), since prisoners do not always associate them with the establishment or its staff group. Rather, they relate them to a more nebulous and disembodied 'system'.

Breadth
The final part of this framework is 'breadth', a term associated with earlier concerns about the dispersal of disciplinary mechanisms within non-custodial correctional settings and civil society more broadly. Downes (1988: 187) used the term to refer to the 'penetration of penal control into civil society', and proposed indicators such as rates of breach and recall to custody, the conditions of parole and release, and 'the extent to which penal agents continue to exercise supervision' over ex-prisoners 'above and beyond normal limits' (p. 187). If we think of breadth as the reach and impact of penal sanctions beyond the prison, a second component is the official 'disqualifications and disabilities' that follow a conviction (Garland 2013: 478). Included here are legal restrictions on voting rights, employment opportunities, and eligibility to receive forms of government aid (see LeBel 2012), which 'extend and intensify the sanction in multiple ways' (Garland 2013: 479). To this, we might add unofficial civic disadvantages (the 'negative c.v.') borne by the prisoner post-release, in the form of social prejudice.
These aspects of breadth -the degree to which a society is marked by carceral mechanisms and its social and legal 'attitude' to former offenders -are imposed upon prisoners post-release, into domains that prison researchers typically avoid. Other aspects -the scars and stigmata that exprisoners bear -are inscribed by the prison itself. Some of these forms are corporeal and conspicuous, such as changes in health, appearance and demeanour. Thus, Moran (2012) describes how a poor quality diet combined with rudimentary dental practices leave Russian prisoners marked by missing teeth. As well as obvious symbolic engravings of imprisonment, such as tattoos, prisoners may develop postures and ways of moving through space which betray their carceral origins and may endure for longer beyond the sentence than some cognitive dispositions.
Other forms of 'breadth' are internalised and subjective: first, anxieties about ostracism by family and friends, public stigma, and being branded by the 'invisible stripes' of one's offence (LeBel 2012; Zaitzow 2011); and second, changes in social and psychological functioning, such as levels of intersubjective trust, the capacity for intimacy (Jamieson and Grounds 2005), feelings of self-confidence and existential security. Prisoners become accustomed to particular levels of noise, solitude and company (Moran 2012), and to certain temporal rhythms (for example, the organisation of the day into uniform blocs; the general slowing of time). Once internalised, these habits may persist beyond release, making it difficult for ex-prisoners to settle into family life or new domestic routines.
The permanence and consequence of these disabilities will result from the interaction between the offender's self-perception and the cues and reinforcements that he or she encounters through the eyes and actions of others, during and after the sentence. Documenting the prison's role in branding and diminishing its prisoners, both subjectively and materially, is a critical task for prison research.

Conclusions
Conceptual description helps bring into focus -and find ways of describing -patterns that are otherwise concealed. Among the most important benefits of the framework elaborated here is its potential to provide a vocabulary for the characterisation of penal cultures that moves beyond reductive conceptions of penal 'harshness' or 'mildness'. Such terms may be effective shorthand in broad-brush accounts of penality, but the closer we get to the canvas, the clearer their limitations. Prison systems, to state the obvious, are more restrictive, dislocating, aggravating, intrusive and tortuously bureaucratic in some countries and at some sentence IJCJ&SD 61 Online version via www.crimejusticejournal.com © 2015 4(1) stages than others. But these characteristics do not necessarily correspond in predictable ways. As Barker (2012: 19) notes, the apparently contradictory elements of Nordic penality 'can affect all offenders at different stages of the criminal justice system. … the same offenders could be subject to isolation during pre-trial detention but later sentenced to a low level "open prison"'. King and McDermott (1995: 322) make a similar point, for while they found that 'the determination of a prisoner's security categorization also determine[d] the nature of his experiences in prison', in that most aspects of life were better in lower-security prisons, they also noted higher-security establishments could deliver levels of purposeful activity, safety and regime quality that made them less oppressive than less-secure establishments. 'Depth' and 'weight', in other words, did not always correlate. Indeed, prisoners in England and Wales often comment that, as they progress to lower-category establishments, they lose some freedoms (to cook for themselves, for example) and are more entangled by the kinds of psychological demands that are described above.
Holding these concepts in mind therefore helps us avoid excessive generalisations about the nature and experience of prison systems. They remind us that we should ask, as foundational questions, 'in what way?', 'at what stage?' and also 'for whom?' since the different components of imprisonment are brought to bear upon prisoners in varied ways, To give an example, experiences that are common for female prisoners -losing children, being held in highersecurity conditions and further from home than their male counterparts, and being invasively policed by staff (Carlen 1998;Carlen and Worrall 2004) -point to a distinctive configuration of depth, weight, tightness and breadth.
Documenting these textural qualities requires researchers to clamber inside 'the belly of the beast' (Wacquant 2002), and to remain there for some time. Only through sustained interpersonal engagement can the experiential complexities of living in extreme social and emotional circumstances be learned. Yet, for a range of practical and professional reasons, such studies have become less common (see Simon 2000;Wacquant 2002). Comparative empirical research is difficult. It takes time, of the kind that most academics with teaching and administrative loads (or young families, or other dependents) struggle to find (see Wacquant 2002). It demands commitment and serious labour, whose returns do not always seem proportionate to the approbation given to theoretical work. It is fraught with issues of conceptual and linguistic translation (Nelken 2010). And it is often hampered by the demands of ethics and access committees, not only due to concerns about protecting vulnerable research subjects, but also because open-ended, exploratory research sits uncomfortably with the more quantitative orientation of most gatekeepers and their anxieties about research whose aims are unspecified at the outset.
These anxieties are understandable. As academics, we should neither trivialise nor dismiss the risks and demands that we place on prisons, all the more so when requesting long periods of unstructured presence at times when resources are stretched and institutions are volatile. And yet research of this kind has a considerable pay-off for practitioners, helping them to see beneath the sheen of performance figures (about which they are often sceptical, having seen first-hand how such data can be manipulated) and producing a conceptual vocabulary which enables them to envision their world. In my experience of teaching and presenting to senior prison managers over a number of years, 7 the best-received work has been the conceptualisation of prison staff culture and the use of authority, represented through the intersection of 'heaviness-lightness' and 'absence-presence' that is described above (see Crewe, Liebling and Hulley 2014). Among its key traits, it is aspirational without seeming unrealistically utopian. More importantly, it is empirically derived, based on a study of public and private sector imprisonment that has a significant quantitative component, so that it offers a picture of how prisons operate that is credible and recognisable to those who work in them. Without these quantitative girders, it could not have been constructed, let alone considered sufficiently sturdy IJCJ&SD 62 Online version via www.crimejusticejournal.com © 2015 4(1) by practitioners who -we should not forget -make consequential decisions about policy and practice, and therefore have good reason for wanting to be sure that research 'stands up'.
A good deal of qualitative research (in prisons and beyond) does not stand up very convincingly, just as a great deal of quantitative work is reductive, under-cooked, or risks sanitising issues of profound moral concern. The point here is to advocate methodological pluralism, in which qualitative methods are not simply privileged above quantitative. For all that we need to capture and make visible the complexity of the prison experience and the fullness of its assault on the self, operationalising these complexities allows us to advance our understanding far further. The next step, then, is to translate the framework presented in this article into a set of discrete, measurable concepts, and to continue to refine them in the light of emerging studies of unfamiliar prison regimes, whose terms might require existing definitions to be amended. All of this will not resolve the disjuncture between macro and micro studies of the prison and, to be fully meaningful, any cross-national comparison of penal cultures would need to situate the prisoner experience within a consideration of relative living standards, resourcing, and cultural discourses. But the ideas of depth, weight, tightness and breadth do, at least, offer a set of intermediate concepts that can be connected to external aims, sensibilities and political cultures. Their adoption might therefore ensure that the work of prison ethnographers can be related to broader issues of penality, and that the claims made by penal theorists are held to account by the work of prison ethnographers.