Guest post by Dr Ben Crewe, Deputy Director of the Prisons Research Centre and Reader in Penology, Institute of Criminology, University of Cambridge. Ben tweets @crewebencrewe. This is the final post of Border Criminologies' themed series 'Penal Policymaking and the Prisoner Experience', organised by Ben.  

Over the last year, I have been conducting a piece of research in Close Supervision Centres – the units within high security prisons that hold prisoners too dangerous or difficult to manage to be held elsewhere in the prison system. In the course of this project, I have spent a lot of time during long lunch hours sitting in staff rooms, writing up notes and eating sandwiches while officers take their break. I have watched quite a lot of episodes of Storage Hunters and Bargain Hunt, and taken part in some long discussions about holidays, Brexit, Donald Trump and the best kind of domestic vacuum cleaner. Officers have sometimes asked me about my life, but rarely about my research. Sometimes they have asked me very little, preferring to sleep, zone out in front of the television or discuss the internal politics of the prison. This is their back-stage, and I feel quite conflicted about whether it is a legitimate source of data.

Photo: Ben Crewe
In these ambiguous research zones, officers have also talked about prisoners, sometimes sympathising with their situations and showing insight into their lives, but more often sounding off, using language that is sometimes very distasteful. Prisoners have been routinely described as ‘dickheads’, ‘shitbags’ and ‘fruitloops’. In these situations, officers have been at pains to emphasise to me that this kind of venting is not to be taken at face value, and is the same kind of ‘gallows humour’ that is found among almost all frontline occupational groups. This is something that I can understand and appreciate, particularly in this very extreme and intense part of the prison system, while at the same time it strikes me that the things that we say but don’t quite mean still mean something. The hard part is knowing exactly what that something is, and whether it needs the kind of direct confirmation that, as researchers, we often seek out.

I have eaten my lunch alongside staff over a number of research projects over the years, and I think I have learned as much about staff cultures this way as I have through formal interviews. When we talk about prison staff culture, we are often referring to how officers use their authority, their attitudes to rehabilitation, and so on, and there is an established literature in this area (see Crawley 2004; Liebling 2011; Liebling, Price and Shefer 2011; Tait 2011; Crewe, Liebling and Hulley 2014). But culture is broader and deeper than this. It is the lifeworld in which behaviour is embedded, a lifeworld – in this case – of DIY, petrol prices, and holidays in the sun, reflecting a certain kind of occupational position, marked increasingly by a sense of economic insecurity.

Culture is also the set of schemas that shape people’s practices. Asking very directly about these submerged norms and tacit understandings is ineffective in part because culture of this kind – ‘the way we do things around here’, our pre-conscious dispositions – is not easily accessible to the people who are embedded within it as an object of description. Bourdieu argues that taste fits like a glove, so that we experience our preferences for certain forms of art or music or clothing as almost entirely natural and sub-reflexive even though they are socially structured and somewhat predictable. Asking people to articulate what they take-for-granted is expecting quite an advanced form of self-analysis. Very often, we do better to learn from what people say when they are not being asked, and from the times when incidents and accidents put their practical schemas at strain, and therefore bring into relief what, for the most part, is hidden and implicit. To put this another way, ‘Do you trust prisoners?’, asked in a formal interview, is a less effective route to understanding trust than ‘why did you let this prisoner out, rather than that one?’, based on the observation of a particular interaction.

The most illuminating lunchtime discussions have been about matters such as care, trust and authenticity. One day, sitting on institutional chairs on the wing, I started a conversation with one officer, Graham, about whether he cared about the prisoners on the unit. He was slightly ‘old-school’, someone for whom the language of care did not come very naturally, but who prisoners regarded as ‘a do-er’, someone who – to draw on Sarah Tait’s (2011) work in this area – delivered ‘care on the side’ through his reliability and professional commitment. What he cared about, he explained, was his colleagues and the running of the unit. He cared – he said – in the sense that he gave prisoners what they were entitled to, and tried to reduce their complaints. Described in this way, care was somewhat limited and instrumental.

Graham’s comments seemed inconsistent with his behaviour, which, to me, looked - at the very least – like a form of practical care, and often more than this: a recognition of prisoners’ humanity through a commitment to meeting their needs. But Graham did not provide me with the kind of direct expression of this position that a literalist might have wanted.

Over lunch, in the company of several of his colleagues, he took the discussion further. ‘There must be things you want to ask us’, he said. ‘What about the stuff you asked me about earlier?’ And so began a kind of informal focus group about care and compassion. Staff said little initially that I had not heard many times before in discussions of this kind: ‘you develop relationships, but you have to detach a bit’; ‘you can’t let compassion affect your work’, and similar statements. And then one officer revealed that: ‘you do sometimes want to hug prisoners’; that the reason you don’t is that you’ll be the subject of security information reports, or because of what the prisoner will then think. But once she had hugged a prisoner who was dying: ‘I thought “sod the rules”’, she said; ‘it was human to human’. This was a unit in which officers often challenged each other, and reflected on their and others’ decisions. On this occasion, no-one questioned her sentiments, but nor did they support her. The silence felt like reluctant recognition that care was there, somewhere, and it was okay.

In the fieldnotes that I completed later that day, I wrote the following:

‘The fact that it was initiated by Graham was not coincidental, I think. I wonder if it had played on his mind after I talked to him about it on the wing. The conversation was held without bluster … I think the staff were genuinely thinking through their own behaviour. It is taboo, to some degree, to say you care for prisoners, but they were describing care and I’ve seen them display it in their behaviour too’.

My encounters with ‘trust’ were similar. Uniformed staff reinforced to me that while CSC prisoners could – to quote – be ‘superficially charming …. they’d ‘slash your throat in an instant if they needed to’ … ‘that’s why you can never trust them’. But they clearly trusted prisoners enough to play pool with them, to sit with them drinking tea, to be alone with them in an art workshop, to let me wander about around them; and in all kinds of minor ways, they trusted them sufficiently to do what they said they were going to do. When one prisoner’s laundry needed to be moved to another room, another prisoner – the most litigious and challenging man on the unit – assured the officer that he would inform his friend what had happened. ‘I’ll make sure he’s not suspicious’, he said, and the officer thanked him. This was a rather limited form of trust, but it was trust nonetheless.

At lunchtime, I asked officers if they trusted this particular prisoner more because he had a legalistic mindset, and was therefore a stickler for people doing what they said they would do. Did they see him as a ‘man of his word?’ ‘No’, they said. ‘the golden rule is that you never trust them’. But, in certain ways, it was clear that they did.

Photo: Ben Crewe
I’m still not sure that I have worked out exactly what I have learned from these conversations. But the inconsistencies between attitudes and behaviour, and the evident discomfort that officers felt when asked about these matters, has sensitised me to their importance, and to certain aspects of prison officer culture: the cultural norms against talking expressly about care, and the difference between the ways that staff define trust and the ways that they unknowingly demonstrate it. Whenever we find discrepancies between what people say and what they do, there is an opportunity to explore culture, the prevailing norms that mean that we cannot admit to what we feel or act in the ways we would prefer to, or that steer our behaviours in ways we are not fully aware of. ‘Asking’ might tell us about attitudes, but not about behaviours; observing might tell us about norms and practices, but not necessarily about underlying beliefs. And if we ask officers about care or trust without observing how they interact with prisoners, we are likely to under-estimate their presence in practice; conversely, if we observe officers displaying care and trust, without talking to them about what these terms mean, we are blind to the subjective meaning of these acts.

None of what I am saying is radical, particularly to anyone trained in a more orthodox ethnographic tradition, which emphasises the importance of absorbing culture rather than seeking it out through direct interrogation. To an anthropologist, or a human geographer, an afternoon spent hanging out, talking about sport or shoes or supermarkets, is productive, even when it does not directly address the research questions at hand. One of the many benefits for me of working in a multi-disciplinary team over the last two years is therefore being reminded that my preference for mainly interview-based data collection is only one mode of knowledge acquisition, and not always the mode that is most illuminating.

Any comments about this post? Get in touch with us! Send us an email, or post a comment here or on Facebook. You can also tweet us.


How to cite this blog post (Harvard style) 

Crewe, B. (2018) Asking and Observing in Prison Research. Available at: (Accessed [date]).