Edinburgh Research Explorer Changing the subject

This article reports on part of a project that introduced philosophy programmes to a number of Scottish prisons. It centres on the deployment within these prisons of McCall’s (1991) Community of Philosophical Inquiry (CoPI). It provides a rationale for, and analyses the participation structure, of CoPI, setting out how its communicative constraints and demands provided prisoners with novel means of reasoning and engaging in dialogue with others and with oneself. In interviews conducted with a sample of participants, they described how the critical listening to, and reasoning with, each other in CoPI tutorials had allowed them to develop greater self-awareness and a more reflexive understanding of their own thinking and actions. Findings are framed within sociocultural theorising on literacies, learning and identity. Drawing on Holland, Lachicotte, Skinner & Cain (1998) account of identity and agency, we show how CoPI afforded participants a new positionality and discursive practices.


Background
The wider project There have been moves in prison education within the last two decades to broaden out learning from a narrowly instrumental focus on basic literacy and numeracy skills to a more expansive understanding of literacy education and to subjects that may be more intrinsically engaging. This move to provide a wider curriculum has included a range of arts projects (Tett, Anderson, McNeill, Overy & Sparks, 2012) and initiatives to provide philosophy discussion groups in prisons (Szifris, 2014(Szifris, , 2016.
The current article reports on one such initiative. The findings presented here originated from a project, Philosophy in Prisons: Critical Thinking and Community of Philosophical Inquiry, that involved partnership between the School of Philosophy, Psychology and Language Sciences and the School of Education in the University of Edinburgh to create and deliver programmes of philosophy in four Scottish prisons as an alternative form of educational experience for those prisoners who chose to attend education classes. Access was established through discussions, between the University of Edinburgh and the Education department of the Scottish prison service. Links were then made between researchers and prison teachers in four Scottish prisons; and students were invited to attend the six-weeks programme of one-hour tutorials. In Scotland, students attend education classes by choice, so the students involved in the project were self-selecting. The classrooms in which the tutorials took place were inhabited by the university tutor, the prison teachers and the students; prison officers remained outside of the teaching rooms and there was no video surveillance. A total of 50 participants across four Scottish prisons were involved in this project.
The current article sets out to: • situate the project within the wider context of education in prisons; • delineate the interactional constraints, norms and affordances that characterised the very specific literacy practices of the Community of Philosophical Inquiry tutorials; • present and interpret the gains that a sample of participants reported through taking part in these tutorials, principally in terms of developing: more reflective dialogue with others and self, greater self-awareness and a more reflexive understanding of their own thinking and actions. Nolda (2014) has observed that adult education research has been 'more interested in the political and discursive aspects of power, and less in the interactional ones ' (p. 98). Guided by such a general concern and a belief that it is particularly pertinent to pay attention to interactional aspects of power within settings such as prisons, we felt the need to provide a fine-grained analysis of the interactional order and communicative practices that are distinctive to the CoPI approach. Accordingly, a section of the article draws on relevant insights from pragmatics and sociolinguistics to achieve this objective. The intention is to bring into sharp focus the particular character of the CoPI tutorials with their specific interactional moves and forms of argument. It will be argued that what occurs here is the provision of 'tools', of ways of structuring and controlling one's interactions with others and self, that may, to a degree, become internalised and encourage a more measured, reflective response to situations. In summary, the aim is to delineate the practices of thinking, relating and being within the programme; and then to present findings concerning how participants represented their deployment of these practices. First though it is necessary to situate this project within relevant literature.

Prisons and education
The UK prison context In July 2019, the population of UK prisons was approximately 82,676; 8205 of whom were in Scottish prisons. Since 1990, the Scottish prison population has risen by 60%. (House of Commons Library, Briefing Paper Number CBP-04334). The rise in numbers of the prison population has brought with it a complex set of problems relating to overcrowding, self-harm, prisoner assaults, violence and mental health (ibid.). In Scotland, the vision of 'Unlocking Potential -Transforming Lives' agenda has informed a new Learning & Skills (LS) approach which has opened up more opportunities for university education programmes to be delivered in a growing number of prisons in Scotland. (Scottish Prison Service, 2014) The Scottish Prison Service publishes its commitment to 'Improving the Delivery of Purposeful Activity' (SPS Annual Report & Accounts 2017-18), key indicator of which is to 'care for prisoners with humanity' (ibid, appendix 8b) in order to reduce reconviction and improve educational attainment. Key to the goals of the Learning Centres in Scottish prisons is increasing literacy and improving students' skills profiles. Currently, literacies learning is built on a social practices model, which acknowledges that literacy is a contested term that changes over time and can refer to multiple social communications and conventions across various contexts. This goes beyond the traditional 'functional literacy' approach to include the development of skills to express ideas and opinions, make decisions, solve problems and acquire and employ a range of communication skills that contribute to citizenship (Literacy Principles and Practices Paper, Scottish Government).

Why philosophy in prisons?
The opening paragraph of this article has pointed up the moves that have been made in prison education to extend out learning from a narrow focus on basic skills to a wider conception of literacies and the inclusion of subjects that may prove more personally engaging and rewarding. Here, Tett et al. (2016, p. 172) have flagged up the potential rewards to be gained from arts and humanities programmes that operate 'from an individual's 'strengths', rather than their 'deficits ' (p. 173). This was an important consideration in launching a philosophy programme within prisons. Support for taking ahead work on philosophy in prisons can also be found in the substantial research and development work that has been undertaken by Szifris (2017) in a Scottish (2014) and in English prisons (2016. The philosophy programmes studied by Szifris took a somewhat different form from the one reported in this article (see, Szifris, 2014, p. 8;Szifris, 2016, p. 36) but also made extensive use of discussion. In her 2016 article, Szifris observes that 'by starting from the point of a person in society, as opposed to an offender with deficits to be addressed, philosophical dialogue … allows the individual to see themselves and their place in the world, from a different perspective (p. 33)'. This potential for philosophy tutorials to position participants primarily as inquirers, rather than as offenders, strikes us as being a key matter.
Interviews with individuals involved in these studies revealed their enjoyment of the philosophy sessions (Szifris, 2014, p. 25) and the opportunities they afforded for 'positive personal interaction with peers' (Szifris, 2017, p. 419). There were self-reported gains in listening, communication and reasoning skills (Szifris, 2014, p. 3), including the 'skills of rational debate' (Szifris, 2014, p. 17). Gains were also reported in self-reflection, selfquestioning (Szifris, 2014, pp. 28-29) and in understanding other people's perspectives (Szifris, 2014, p. 3;2016, p. 35). Summarising her findings, Szifris (2017, p. 421) noted that they 'demonstrated a clear relevance of philosophy to the self-understanding of participants with prisoners highlighting the role of the dialogue as well as the subject matter in encouraging self-reflection, providing structure to their opinions, and providing language for alternative self-definition.' Szifris has framed her findings in relation to theories of desistance from crime. As following sections reveal, our own concern is not with such matters and we draw on sociocultural accounts of literacy, learning and identity as opposed to the literature of criminology.

Community of philosophical inquiry
Philosophical Inquiry: critique and response Subsequent sections give an analytical account of the CoPI approach around which this project centred. It has been employed in a range of adult education settings, but had its origins in work conducted at school level. The general movement of philosophy for, and with, children has recently been the object of a trenchant critique by Biesta (2017). Accordingly, it is appropriate to engage with this critique before turning to CoPI itself. Biesta views philosophical work in schools as having focused on the analytical-logical tradition of Western thought, giving insufficient attention to more phenomenological/existential traditions. In his own words, 'it feeds [children and young people's] thinking but does not really reach their heart or touch their soul' (p. 18).
A key point of Biesta's critique is the charge that philosophical work in schools is driven by an instrumentalist purpose, imbuing skills in critical thinking required for navigating the 'uncertain world … of global capitalism' (p. 418). While such an instrumentalist orientation might fit well with the neoliberal agenda of 'lifelong learning', it is clearly antithetical to the humanistic and emancipatory traditions of adult education.
This critique has, however, met with a spirited response. Cassidy (2017), among others, makes the counter-claims that philosophy with children 'demands an encounter with others' (p. 488) and that dialogue 'is central to this way of living, of being in the world, and, arguably, to being'(ibid.) She states that: 'The ego, in PwC, is not at the centre; the community and the dialogue take precedence' (p. 489). As the following pages will establish, the CoPI approach adopted in this project, does not have a narrow instrumental focus but very much centres on promoting communally embedded, open, dialogic enquiry.

CoPI: 'philosophising'
In McCall's Community of Philosophical Inquiry the key objective is to engage participants actively in philosophical reasoning, in 'philosophising' (McCall, 2009, p. 2), rather than on conveying substantive content concerning philosophy. To engage in the activity of philosophising is quite distinct from the practices of studying traditional philosophical texts and philosophers. Within a community of inquiry, participants engage in thinking aloud together, driven by a framework for dialogue that encourages them to think together in a connected fashion and that demands rigorous attention to critical reasoning (Kennedy, 2004;Simenc, 2008;Daniel & Auriac, 2009). A key reason for employing McCall's framework lay in the potential that it appeared to offer participants to 'change their mind about the ideas they hold, maintaining that the philosophical assumptions that underlie our thinking and judgements are fundamental … to who we are' (McCall, 2009, p. 86).
Her approach works to unearth possible error and explore contradictions in thinking. Thus, one of the important goals of CoPI is to strip away what cannot be the case rather than to seek out the 'truth'. Some of the key features of the practice challenge participants: • to illuminate the concepts, ideas and theories that emerge from the dialogue • to collaboratively deconstruct and/or reconstruct ideas that form from initially puzzling questions to reveal possible error in thinking • to recreate 'self' through the community of inquiry and the relationship formed with others in the group • to develop engaged and active 'citizens' through rigorous, logical reasoning • to recognize and evaluate analogies • to recognize fallacies • to remain conscious of emerging judgments • to remain alert to alternatives (adapted from McCall, 2009) In a CoPI, the chair leads and manages the dialogue, structuring the process by: imposing time frames; deciding on questions; indicating who can speak; asking for responses. The chair models such skills as: summarizing; clarifying; seeking examples and definitions; pointing out contradictions; connecting different ideas. Ideas are conveyed through everyday, common language without specialist terms and with no appeal to 'authority'. This is seen to be a leveler for the community of participants and helps to create a thinking forum of 'equal' participants, (a way of being that is quite distinct from the hierarchical relationships customarily found in prisons, on the 'block'). Participants learn how to summarize a previous point, or how to seek clarity. As the chair presses participants to unpack ambiguity or challenges assumptions, so eventually they learn to do this for themselves and each other. The chair retains a unique perspective outside of the dialogue and is able to attend to the dynamics and mood of the group.

CoPI: participation structure
CoPI has a number of rules of engagement that act to create a very particular structure of participation. Before the start of each six-week programme, the participants were asked to choose another name for themselves and wear it for each other to see. This established a key feature of the tutorials -individuals were invited to distance themselves from their 'criminal' identities. One of the 'rules of play' was a constraint on what the participants could contribute: i.e. they were not allowed to talk about themselves or introduce anecdotes about the events that had led them to be incarcerated. This critical distancing was intended to help participants free themselves from their assumptions and give full attention to the arguments at play within the inquiry. Taking on a constructed identity aimed to provide the detachment required for the participants to wonder aloud together, involving the 'imagination as well as critical analysis' (McCall, 2009, p. 81).
Turning to another rule of engagement, participants in a CoPI tutorial are told that they must follow the format of: This key constraint ensures that the participants make clear whose contribution in the group their argument has developed from and what they believe they have actually heard, before going on to articulate their own thinking and reasoning. It can be seen to be central to the 'participation framework' (Goffman, 1981) that characterises McCall's CoPI, and it is therefore worth unpacking its functions and effects. It ensures that the new contribution made by a speaker does relate to the preceding turn and develops the consideration of a particular topic. Here, following Rawls (1989), Malone (1997, p. 50) notes that 'the more indexical the talk is -that is the more it is tied to the situation -the more compelled a listener is to pay close attention.' The preceding speaker is directly included in the new utterance, which can be seen to be a means of building a community of inquiry. Nofsinger (1991) has observed that in formulating (p. 121), i.e. giving: the gist of what some other participant(s) said … the person who formulates what somebody else said is displaying a certain understanding of that earlier talk. This is an important source of alignment in everyday conversation and other types of interactive talk. … The assumption is that perhaps the best way to align one's understanding with what someone else intended to convey is to formulate that other participant's talk and monitor his or her response. (p. 121-122) In addition to acting to align participants' contributions and thereby build a degree of intersubjective understanding, this CoPI rule of engagement allows disagreement to be cast in an 'impersonal' form, tying disagreement to a reason or position that the speaker needs to construct. Agreement also needs to be justified rather than being a simple act of affiliation with the preceding speaker.
Returning to the role of the chair in shaping a particular structure of participation, it is crucial to the development of the dialogue, channelling it within a tight, logical structure. The chair's role is played out by identifying the philosophical assumptions within the contributions of the participants and finding questions that 'bring this underlying philosophy to the forefront' (McCall, 2009, p. 10). The chair's role is to ensure that the participants engage in genuine dialogue that is 'philosophical in content; that there are two or more philosophical ideas at work; that contrasting philosophical ideas are juxtaposed and set against each other; that arguments are given to support ideas and that there is a movement or development of the ideas and arguments' (McCall, 2009, p. 91).
By pursuing these purposes the chair ensures that in the often combative atmosphere of prisons participants are required to make an argument, 'use reasons, evidence, claims and the like to "make a case" ' (Nosfinger, 1991, p. 146) rather than have an argument. The chair acts to create an interactional order which fosters the pursuit of contrasting positions, of reasoned disagreement, while attenuating the possibility of acrimonious personal confrontation.

CoPI: texts and questions
In each inquiry, participants sat in a circle and were given a text, (print and/or image), which was read out to ensure that no one would be excluded from the processes of thinking. The texts served as conceptual springboards, not as objects for analysis in themselves. They included extracts from philosophical works, e.g. Heidegger, What is Called Thinking? (1968); from literary texts, e.g. Robert Frost's The Road Not Taken (1920) and academic text extracts, e.g. on epistemology and the problem of doubt. Other texts were visual (see the war image on p. 12). For each tutorial, the text was selected by the chair for its potential, as a key element of CoPI, to elicit contrasting ideas on a specific philosophical topic. Topics covered included: The Nature of Knowing; Mind, Brains and Computers; Morality; Free Will and Determinism; Time Travel. The chair collated the questions that arose from participants' sense of curiosity or puzzlement as they first engaged with a text, writing them up on a whiteboard with the names of the questioners alongside. For example, the questions asked by the prisoners in response to the Frost poem, The Road Not Taken were: What is the path?
Who decides what the path is?
Are we pre-destined to repeat the mistakes of the past?
Where does the path lead to? Do we really have any choice in this life?
Why do we think we are free?
By choosing one path, are we fated?
The discussion then began with the question selected by the chair which was deemed to have the potential for lively philosophical exploration. Rooted in the key question selected by the chair, discussion continued for approximately forty-five minutes. Participants engaged in exploring and evaluating the implications of each other's arguments in order to contribute to the development of the dialogue and to the construction of a more complex configuration of a particular philosophical problem.
As a further illustration of a text and its accompanying questions, one tutorial explored questions concerning morality: is it objective, subjective, or relative? The group were given the following image as an initial stimulus to thinking. This text clearly had an effect on the participants, some of whom were imprisoned for murder and/or violent offences, who used it as a means to open up and share views about morality that were very close to the bone. The text elicited the following questions from one of the groups: Is killing in war ever justifiable?
Can taking someone's life ever be defended?
Are some people born murderers?
Does society create murderers?
In war, do some people have to die for most people to live in peace?
Are soldiers murderers?
Is the law always right?
Should people sacrifice their lives for something they believe in?
[190] Bovill & Anderson The Discourse of CoPI It will be clear from the preceding account that CoPI has a much wider scope than a narrow instrumental concern with the development of communication skills and critical thinking. Its concerns are well encapsulated within Gee's conceptualisation of capital "D' Discourses (Gee, 2008). Gee distinguishes between primary discourses individuals acquire in their immediate family and community of origin, and secondary discourses that they then go on to acquire in different contexts. He defines a 'Discourse with a capital 'D' '(p. 155) as: composed of distinctive ways of speaking/listening coupled with distinctive ways of acting, interacting, valuing, feeling, dressing, thinking, believing, with other people and with various objects, tools, and technologies, so as to enact specific socially recognizable identities engaged in specific socially recognizable activities.
Here then the connection between engaging in a set of literacy practices, such as CoPI, and ways of valuing, being and enacting an identity are brought out very clearly. We will move shortly to describe how the participants in this project perceived their encounter with this particular secondary Discourse. First though it is necessary to outline the approach and methods of interviewing and analysis.

Methodology and methods
The preceding section has set out the potential gains in reasoning and in relating to others and to oneself that may flow from engaging in CoPI tutorials. The question remains, however, of whether the participants themselves perceived any benefits from their participation. How did they react to the programme? Did they view it as having any value for their lives? How would they themselves characterise its effects? To investigate these questions, interviews were held with the prisoners. This interview study was grounded in an interpretative phenomenological approach (Finlay, 2014) 'which construes people and the world as inter-related and engaged in a dialogic relationship that constructs [multiple versions] of reality' (Shaw, 2010, p. 234).
In the context of a prison, close attention clearly needs to be paid to ensuring that consent to participation in research is indeed freely given. The self-selected participants in the tutorial groups, (who gained no extrinsic benefits for their participation), were all given a clear overview of what the course and the research project entailed. They completed ethical consent forms that established their agreement to be part of the course and separately to take part in a group interview. Teachers in the prisons assisted the participants in this process. It was also made very clear to the prisoners that they could opt out of the interview part of the project at any time, but no participant chose to do so. No prison officers were present during the CoPI sessions, or in the interviews; nor were there any surveillance cameras.
After each six-week block of tutorials in each prison, participants were interviewed in groups of four; the time constraints on being able to have contact with the prisoners on an individual level and the number of prisoners involved influenced the setup of the interviews. It was also seen as of distinct value to have the prisoners engage in smallgroup interviews to allow for a conversational style of discussion, which could be viewed as less intimidating than individual interviews and allow for both individual and collective perspectives to emerge. Consonant with an interpretative phenomenological approach, the interviews were treated as a 'social encounter in which knowledge is actively formed and produced impl[ying] that the interview is not so much a neutral conduit or source of distortion, but rather a site of, an occasion for, interpretive practice' (Holstein & Gubrium, 2011, p. 151). The interviews were semi-structured and allowed for the emergence of key themes in each round of interviews. Interviews lasted for approximately 40 minutes.
Participants were asked a range of questions which encouraged them to reflect on their views and beliefs about their experiences of 'doing' philosophy in this way. These included the following: What did you think of the course over the last six weeks? What did you think about adopting a different name for yourself and leaving your 'identity' at the door of the classroom? You were asked to contribute to the discussions using a specific format (I agree/disagree with X when they said Y because…). What did you think about this? Do you think that participating in the CoPI method influenced the relationships between the group members, or not? One of the key features of the tutorials involved you listening critically to each other in order to help you make connections between each other's thinking. How did you feel about this? Do you think you developed your reasoning skills through the six-week programme, or not? In doing this course, did you learn anything about your self?
The findings reported here derive from the analysis of interviews with men in one site where, in line with the focus of this article, the course was firmly centred on CoPI. The ages of this group of 12 participants ranged from around 18 to 65. They had all been convicted of serious offences and were in the main serving long sentences.
The semi-structured interviews were transcribed and analysed through the lens of an interpretative phenomenological approach which requires an orientation to the data that is 'profoundly dialectical [where] researchers need to straddle subjectivity and objectivity, intimacy and distance, being inside and outside, being part of and apart from, bracketing the self and being self-aware' (Finlay, 2014 p.124). Scrutiny of the assumptions, process and outcomes of the analysis was guided by Smith, Flowers and Larkin's (2009, pp. 180-183) articulation for interpretative phenomenological analysis of Yardley's (2000) principles for assessing the quality of qualitative research: 'sensitivity to context; commitment and rigour; transparency and coherence'; and 'impact and importance'.
Key themes that emerged from the analysis of the interviews were: opening up the mind/gains in reflexivity; developing through the CoPI constraints; reasoning, selfquestioning and identity; enacting a different culture of thinking, relating and being. The following section provides a succinct, illustrative account of these interconnected themes.

Findings 'Opening up your mind'
It has been noted that a central reason for adopting CoPI tutorials was the possibility that they would constructively challenge participants to reflect on their own world-views and encourage them to see themselves from a range of different perspectives. The interviews provided evidence that these goals were being achieved. Almost all participants agreed that CoPI tutorials 'disturbed' them from their own insular views and brought them to develop a more reflexive standpoint: 'I really enjoyed it. It made you look at different people's perspectives… Sometimes you modified your own views or strengthened your own opinion. It gave you confidence in your argument through thinking that way, kind of stepping back and being able to reason about how you came to your decision'.
One participant described the experience of reflexivity as an 'opening up of your mind in an isolated environment, being in prison'; and this motif of 'opening up' their own minds to their own and each other's perspectives recurred throughout the interviews.

Developing through the CoPI constraints
The participants identified several elements of the CoPI 'rules' as key to the dynamics within the group and to change: the use of a new name, (leaving their previous name and criminal 'identity' outside the classroom); and adopting the interaction format of 'I agree / disagree with … X … when they said … Y because … Z'. These constraints emerged as central to growth into a more expansive sense of self. The participants found that adopting a new name and engaging in a form of critical objectivity had the effect of 'leaving your ego aside … and it allows you to also not necessarily think about something from your own perspective, but from the point of view of a question'. One individual encapsulated the groups' perceptions of the value of using different names for the tutorials in the following terms: I agree that using alternative names was a good idea because if you were maybe having an argument -no arguing in a fighting kind of way -but having a discussion argument with somebody, it wasn't a personal argument that you would have with the individual, the argument was with the idea, as opposed to a person against a person … It led you to be more understanding and more open.
This nudge towards critical distancing and the development of the ability to remain objective throughout challenged their perceptions of themselves and contrasted with their experiences of disagreement in the 'halls' where 'if you disagree with somebody it is an attack on them and they react in violence…In the halls, [it is like] I disagree with you because I am the Alpha male'. Taking part in the philosophy classes was viewed as inhabiting a secure space where the participants learned how to disagree with each other. This was seen to be a 'good exercise in how to be a social animal'; and is in line with Szifris' finding that philosophy in prisons challenged the 'machismo' image of a group of alpha males vying for dominance (Szifris, 2018). Indeed, the men involved in the CoPI tutorials described how the name they adopted and the framework that governed contributions to the discussion were 'like having a different identity. In that one hour session, you're no yourself … It's like you're new, a new individual and having that philosophical debate … you're opening up your ears and listening to other people's opinions'.
This participant also identified the demand in CoPI for critical listening as key to fostering critical reasoning and a more open orientation to others. He described how the participation framework in CoPI acted to break down narrow, self-centred, selective listening: because you were having to say ' I agree with Mr A when he said …', because you were having to say exactly what you agreed with rather than just nit-picking: 'I agree for such and such a reason.' When you were having to say, 'I agree with X for reason Y', you actually had to listen to them, because a lot of the time, we're guilty of only listening to what we want to hear. This observation echoed another individual's perceptions of the power latent in CoPI to bring about a more open orientation to others. His more open orientation can be seen to involve an expansion of the contours of the self: I think a lot of the trouble I've been involved in in the past, especially violence, has been through misunderstanding or being misunderstood has led to a lot of violence, including what I am in for now … In CoPI it's good to hear people having different views without actually feeling as if, well, [they are] actually different from me just because they have a different view on something … A lot of stuff and violence that I've been involved in in the past has been through misunderstanding, you know.
Thinking before talking, thinking before acting The participation framework and reasoning demands at the heart of CoPI influenced participants' perceptions, and ways, of communicating with each other: 'Through this course and in this way of thinking, you actually think more about what you're saying before you're making statements, and you'll be more understanding if somebody else has maybe come to a conclusion with their reasoning and thinking behind it.' The idea of thinking-before-acting was articulated clearly by one participant who stated why developing such dispositions becomes important for all incarcerated members of society: Because that's the way you've got to [think about your own thinking] … because you're no right all the time and you have to think about what .. the way you conduct yourself and you have to think about all kinds of things. You have to think about how you are going to get through your sentence and what you're going to do when you get out. If you don't think about that, you're coming back in.

Rethinking the self
The ways of thinking and of relating together during these philosophy tutorials allowed participants to experience a different way of being, both individually and as part of a group. One of the rules of play in a CoPI tutorial is denying participants the opportunity to talk about themselves and requiring them instead to attend to the development of the conceptual discussion. One prisoner described this experience as 'a good thing because it took like individually ourselves, what had happened in our life, taking that away from us and made it more of a group thing … instead of being like individuals we had all like false names which made it a wee bit more about the topic we were talking about and less about ourselves. I enjoyed that.' CoPI can be seen to provide a space that to at least a limited degree detaches individuals from their past selves. One participant saw this as an important element of CoPI: 'I can see that being anchored in the past can be damaging for some people, being anchored in the person that you think you are.' Through engaging in the 'constructed world' (Holland et al., 1998, p. 7) of CoPI participants were led to question their longheld assumptions and consider other possibilities. They found that CoPI 'gives you selfawareness because it opens up topics that you've maybe looked at through one perspective, and no really seen that person's point, or a person's view.' On this theme one individual observed that: I think it does improve your self-awareness, it improved mine anyway, where it makes you think. It just gets you thinking more deeply about the meaning of life and why we are here. I think that is the interesting part in philosophy because it makes you curious as well … I think more curious about learning.
The degree of freedom that the CoPI tutorials opened up for the participants to inhabit a different way of being was captured by several of them in the image of casting off 'a mask'. They discussed the way in which engaging together in a community of philosophical inquiry allowed them to cast off the masks they felt they had to wear in other parts of the prison to 'suppress' their feelings: I would say ... in the prison environment you're quite often wearing a mask sort of thing, you're no letting people into your thoughts, your feelings … Whereas within this group [CoPI] you kind of let that mask go … You didn't feel as though it had to be hidden or explained …. Because you were explaining as best you could on a topic and leaving it to be open to question … Whereas a lot of the time, other times, you will not even bring opinions up because of that environment [the halls]: 'he's going to think this'; or 'he is going to think that, he'll say this'; or 'he'll say that'. … But in the whole philosophical reasoning, you learn more to put that to the side and it was the topics … If you have a strong point or something, you state your point because it wasn't a personal attack or vendetta or something you were trying to preach to someone else … it was what you believe or thought, your view, which was open to others to agree or disagree with.
A few observations can be made concerning this statement. Drawing on Gee (1990), Janks (2010 has claimed that 'The difference between discourses is productive. As individual human subjects enter into new discourses, they acquire alternative and additional ways of being in the world that is, new social identities.' It can be argued that such a process can be inferred from the above quotation. Gee himself (2008) states that the use of a particular Discourse can be 'liberating' (p. 177) 'if it can be used as … a "meta-Discourse" … for the critique of other literacies and the way they constitute us as persons and situate us in society' (ibid.). It can be argued that CoPI can be seen to be serving this purpose in this instance. At the same time, it is essential not to overestimate its liberating effects. This individual and his fellow-participants still needed to survive in the very different Discourse of 'the halls'.

Questioning oneself
Some participants made statements that revealed how engaging in critical dialogue with others could carry over into a challenging, self-questioning dialogue with oneself. For example, one participant described how CoPI allowed for 'a personal attack on yourself' as part of the process of inquiry. The way in which debate in the group could spark a challenging questioning of oneself is vividly conveyed in the following quotation: Because there are a few things in there and I would think to myself after somebody has said something, I'm like: 'I wonder why I've thought that way for so many years about some certain thing?' I've been thinking like that for years and then somebody says something and I'm like, 'Fuck! I was wrong!' You know? Such a process of critical self-examination opened up the potential for change. In this vein, one individual talked of how: I would say that I mean, like everybody else, I think I'm right all the time -even when a judge tells me that I've been wrong and I've got the jail for it. I still think I'm right … but it has been said that by us in here, saying different points of view, like 'what have I done?'; and I do think extremism is a bad trait and it does lead to jail if you do take things to extreme and it's [CoPI] taught me not to be as extreme as I have been.

Enacting a different culture of thinking, relating and being
It was evident from the interviews that a culture of thinking, relating and being emerged throughout the course which participants viewed as unique. They saw the actual enactment of a different way of thinking and acting as distinguishing CoPI from other educational interventions.
I mean within this environment, in a prison, it's a useful strategy teaching habits to cultivate because it's basically what the [other educational] programmes are trying to get us to do, to change our ways of thinking to consider more … I would suggest that in that kind of environment [a CoPI] it's actually … because you're practising it as you do it, and you're adopting the behaviour you're not just listening to somebody -'Well, you need to do this.' -you're actually practising the adoption of a new behaviour. … This view was re-expressed by another participant who 'found myself, my behaviour changed in the way in which I converse with somebody by the actual doing of that particular system compared to how I would act on the outside or on the wings'.
Another participant agreed that the experience of CoPI resonated with him and his sense of self in ways that he had not experienced in other education programmes designed to 'teach you to think in a certain way.' He pointed out that some courses are pursued by some students to 'jump through hoops to get to opens and things like that' whereas 'coming to philosophy' was like 'getting taught how to use a gym and choosing to use the gym.'

Access
The findings of this study point up a number of issues concerning education in prisons and how best to assist individuals to gain the forms of thinking and relating that may allow them to construct a more positive future. The opening section of the article set out the critique that has been made in recent decades of a narrowly instrumental approach of providing prisoners with basic literacy skills. By contrast, the CoPI approach to philosophising introduces a set of literacy practices of argument and analysis embedded in a supportive set of social relations. Consonant with at least certain of the tenets of critical literacy, it encourages participants to interrogate multiple perspectives (Vasquez, Tate & Harste, 2013, pp. 11-12) and to disrupt the commonplace (pp. 8-10): 'things that seem normal need to be rethought'(p. 8). Janks (2010Janks ( , 2014 has argued that 'critical literacy work has to pay attention to questions of 'power, diversity, [and] access' (Janks, 2014, p. 5). On 'access', Janks poses the questions: ' Who gets access to the language of power and its prestige variety? Who gets access to high-status knowledge?' These questions concerning access would seem to be particularly pertinent to education in prisons. The philosophical discussion approach would appear from this study, and from Szifris's work (2017) to be one means by which individuals can gain access to a greater fund of cultural capital. A wider challenge for education in prisons also comes into view here: what knowledge, skills, and literacy practices 'should' those who are imprisoned have access to if they are to become critically literate citizens?

Participation structure
A section of this article has delineated central features of the participation structure that is to be found in CoPI tutorials. To our knowledge, preceding literature on CoPI has not analysed the pragmatics at its heart in the same level of detail. This analytical exercise was intended to make a wider point. In prisons, where relations of domination and submission can have a detrimental effect on more discursive approaches to education, it would appear to be worthwhile, as in this article, to give close attention to how the dynamics of a group can best be structured to create more equal and productive interaction.

Enabling constraints
The Findings section has set out the substantial gains that this group of participants reported in terms of: 'opening up your mind'; thinking before talking, thinking before acting; rethinking the self; questioning oneself; and enacting a different culture of thinking, relating and being.' The findings revealed that CoPI appeared to have acted to create, what can be termed as, 'ontological disturbance'. The participants felt that taking part in a CoPI challenged them to question long-held assumptions and worked at stripping away the illusions and contradictions in their own thinking. In this way, their own constructs of identity were challenged. These findings are in line with McCall's claim that when 'assumptions are altered through the practice of CoPI, the participants change or recreate themselves -maybe in a very small way; but however small, it is fundamental. And to be involved in this fundamental change of oneself 'with' others is to be in a unique relationship with these others.' (McCall, 2009, p. 86).
At first sight there would seem to be a paradox here in that an educational intervention which placed considerable constraints on interaction had led to an expansion in possibilities for thinking, relating and being. What was at stake here can be interpreted with the aid of Holland et al.'s (1998) 'processual understanding of identity and agency' (p. vii). Building on the sociocultural theories of Vygotsky (1978) and Bakhtin (1981) they look at how specific cultural discourses and practices instantiated within particular contexts provide the 'tools' for building a self: 'processes or traditions of apprehension which gather us up and give us form as our lives intersect them' (p. 41). They also recognise the shaping power of the subject positions that the participants in any social encounter occupy (p. 41). They note that: 'people develop different relational identities in different figured worlds because they are afforded different positions in those worlds. ' (p. 136) There can be seen to have been a marked shift in the positionality of the participants during the CoPI tutorials. They were forced to bracket their past history, including their offending, and to take on the role of inquirers; and, as they observed themselves, the participation structure of CoPI brought about more equal relationships in contrast to those of domination and submission on the 'halls'. At the same time, their engagement in CoPI drew them into new practices of argument and interaction: 'a set of constraints' also provided 'a set of possibilities for utterance' (Holland et al., 1998, p. 171).
Using CoPI in prisons, the contribution to knowledge is clear: the interviews revealed that at least to a certain degree the discursive practices of CoPI were acting as 'tools' for internal dialogue and enhanced self-control. As Holland et al. observe: 'We achieve self-control, albeit of a very limited sort, by the mediation of our thoughts and feelings through artifacts. We learn how to control ourselves from the outside, so to speak (Vygotsky 1978); we learn how to position ourselves for ourselves.' (Holland et al., 1998, pp. 63-64) These tools for reasoned debate with others and with oneself can be viewed as valuable resources for 'alternative self-definition'. (Szifris, 2017, p. 421) At the same time, as we have observed earlier, it is important not to exaggerate the potentially transformative effects of the Discourse of CoPI when students in a prison need to survive within the very different Discourses of the halls. Future research in adult education programmes of this kind will, however, contribute to our developing understanding of the dynamics involved in the teaching and learning of excluded members of society.