‘It’s about keeping children safe, not spying’: A governmentality approach to Prevent in primary education

From its inception in 2015, the Prevent Duty has required educators, and other members of the social sector, to exercise ‘due regard’ in preventing pupils from being drawn into terrorism, irrespective of the age of the child. This article explores how primary educators have understood and implemented this preventative security policy in their schools. Analysis is based on a survey of 345 primary school educators and 37 semi-structured interviews with primary school educators and Prevent Education Officers from the West Midlands. Through a lens of governmentality, we shed light on how this mandate has been broadly interpreted and exercised by educators within and outside the school gates. In so doing, we contribute to debates on the puzzling acceptance of Prevent in education, on the process whereby educators identify threats, and on the securitisation of educational spaces in a risk society.


Introduction
In January 2021, the Prevent Duty (PD) made headlines after a 4-year-old was referred to the scheme after talking about a videogame at his after-school club in the West Midlands. He was the latest of hundreds of primary-aged children referred to Channel: in 2016-2019 alone, 624 children aged under 6 years and 1405 children aged 6-9 years were reported to schemes for the prevention of violent extremism (Stein and Townsend, 2021). Our study examines an original survey of 345 primary school educators, 32 semi-structured interviews with primary school educators, and 5 interviews with Prevent between July 2018 and July 2019, including an online survey of 345 primary school educators, 32 semi-structured interviews with primary school educators, and 5 interviews with PEOs. Through an understanding of governmentality (Foucault, 1991), we discuss the ways in which primary school educators understand and operationalise PD. Governmentality captures much of what is at stake with PD, including the importance of knowledge and surveillance as part of how that knowledge is produced. Analytically, the concept of governmentality allows us to understand PD as a technique of power, as one means by which populations are understood and managed, and to examine how it shapes the conduct of populations (see Rose, 1999: 20). We consider this production of the selfregulating subject as accounting for the acceptance of PD as part of educators' safeguarding roles, moving away from previous research into the subjectivities produced by PD (Bryan, 2017), and focusing instead on the ways in which PD has been successfully embedded in education. This process can form a nebulous landscape, as techniques of power are often disguised as a kind of pastoral protection and support (Gutkowski, 2011: 350). However, we hold that governmentality allows us to understand how PD operates as a series of techniques and prescriptions that shape subjects' behaviours within a securitised world. Reading PD through an analytical framework of governmentality allows us to explore its use as a soft preventing/countering violent extremism (P/CVE) measure and its normalisation. Governmentality allows us to account for a growing sense of comfort with PD by positioning it as a technology of power that normalises the deputisation of educators as security agents through their established safeguarding, educational, and community roles.
The remainder of the article proceeds as follows: first, we review the literature on PD in education and track the conversation surrounding the dynamic nature and implementation of PD (Thomas, 2017: 309). We subsequently propose a theoretical framework grounded in governmentality to explain the operationalisation of PD as a soft P/CVE measure and its normalisation among educators. Then, we outline this study's methodology. The following sections discuss our empirical evidence on educators' understanding of the Duty and how they enact it. The conclusion maps our contribution to existing research on PD reiterating that governmentality accounts for the ways in which educators understand PD, appease their worries and perplexities, and enact PD in primary education.

The PD in education
The introduction of PD in education has generated a vast academic literature critiquing its rationale (Jerome et al., 2019;Moffat and Gerard, 2020;O'Donnell, 2016;Thomas, 2017), implementation, and wider implications. The majority of studies focus on secondary and post-secondary education, and only two existing studies reflect on the ways in which PD is understood and enacted in primary education (Bryan, 2017;da Silva et al., 2020). In this study, we contribute to the wider literature on PD in education by presenting the first large-scale empirical study on how primary school educators understand and implement PD in their daily practices and analysing this innovative data through the lens of governmentality.
Our study addresses significant gaps in the literature about PD in three further respects. First, key contributions have explored the 'banalization' of PD by practitioners in the education and healthcare sectors (e.g. Heath-Kelly and Strausz, 2019: 91), despite criticisms from some sectors of society (mainly academics and civil society organisations). A consensus is emerging that this normalisation is due to the UK Government's framing of PD as part of safeguarding: It sits alongside long-established duties on professionals to safeguard vulnerable people from exploitation from a range of other harms such as drugs, gangs and physical and sexual exploitation. The duty is designed to help ensure that vulnerable individuals who are at risk of radicalisation are supported as they would be under other safeguarding processes. (HMG, 2018a: 35) Children safeguarding encompasses protection from maltreatment, prevention of impairments to their health or development, ensuring the provision of safe and effective care, and enabling them to achieve 'the best outcomes' (HMG, 2018b: 5). The framing of PD as safeguarding in the education sector has been investigated by Busher et al. (2019) and is also central to the debates on PD in education explored in the collection edited by Jerome and Busher (2020). Busher et al. (2019) and Jerome and Busher (2020) argue that previous experience of safeguarding allowed educators to synthesise the additional PD statutory guidelines into their previous knowledge -it became, to many, just another form of safeguarding (Busher et al., 2019). We propose that approaching PD through governmentality can help explain educators' acceptance of PD by identifying PD as a technology of state power (Foucault, 1991), which has successfully embedded into primary education in the West Midlands and made security and risk governance possible through the responsibilisation of educators.
Second, a fundamental critique of PD in education has been the extent to which it singles out British Muslims as the source of risk (Kundnani, 2009;Martin, 2018). In particular, critics of the counter-radicalisation mentoring programme Channel suggest that it embeds discrimination through its overwhelming focus on individuals identified as being at risk of radicalisation into Al Qaeda-or ISIS-inspired violent extremism (Heath-Kelly, 2013;Kundnani, 2014;Martin, 2018). Since 2011, the number of PD referrals of young people to Channel grew by over 75%, with those from the educational sector more than doubling and disproportionately involving young Muslims (Open Society Foundation Justice Initiative, 2016). Sian (2017: 7) traced an Orientalising logic through PD, arguing that 'what we see in Prevent is the persistence of positivist ideas used to aimlessly classify and categorise particular individuals and groups in pursuit of countering extremism'. In her view, the toolkit Learning Together to be Safe, 'reinforces a particular construct of a Muslim "threat" and embeds a nebulous danger that Islamic extremists lurk both inside and outside the UK' (Sian, 2015: 186). However, the UK government has argued that PD and Channel address all types of radicalisation, pointing to a sharp increase in the proportion of far-right-related referrals and interventions in recent years.
Third, extensive scholarship has expressed concerns about the spread of security measures through education (Davies, 2016;Durodie, 2016;Taylor and Soni, 2017) and the creep of security into a broader social framework in the name of the extraordinary and existential threat of international terrorism (cf. Heath-Kelly, 2018 on health). PD and its applications in education appear as part of an expanding education-security nexus (Davies, 2016: 6;Durodie, 2016: 22). Furthermore, in exploring the experiences of lecturers in higher education institutions, Awan et al. (2019: 138) identified this securitising spread alongside the neoliberal trend of the responsibilisation of the individual, particularly nonstate actors in security regimes, and argued that the 'deputisation' of higher education staff hindered their ability to function as voices criticising and resisting the state.
Our study addresses these three gaps through an innovative, theoretically grounded approach to PD in education. Specifically, we propose that Foucault's (1991) concept of governmentality provides a critical lens through which to examine the ways in which primary educators internalise and perform PD in their daily work. The following section lays out our theoretical approach.

Understanding Prevent as a technology of power
In this study, we consider PD through the lens of governmentality, as this concept describes a productive project whereby populations are subtly managed and regulated. This means two important things: that the exercise of power is productive both of subjects (Kwon, 2013: 10) and of knowledge (Awan et al., 2019: 72), and that power is exercised not through the direct control or manipulation of individuals, but rather through the ensemble of institutions, states, laws, and norms that more subtly regulate and manage populations. This helps us to situate PD within a governmentality framework, as PD is one means by which this management and regulation is performed. Governmentality also highlights the ways in which PD, and policies like it, emphasises the pastoral nature of this kind of governance (Gutkowski, 2011). The concept of governmentality sheds lights on the ways in which human action is influenced, coerced, or prescribed towards a desired social norm (Inda, 2005) that works to manage risk within a population, in this case counter-radicalisation efforts.
Governmentality, sometimes referred to as the 'conduct of conduct' (Inda, 2005: 1) or 'art of government' (Awan et al., 2019: 64), has been applied to studies of education policy both in the context of the United Kingdom (Awan et al., 2019;Bryan, 2017;Farrell and Lander, 2019;Sian, 2015) and beyond (Flores, 2013;Mitchell, 2006;Suspitsyna, 2010: 469). Farrell and Lander (2019: 469) understand governmentality to imply a combination of 'government' and 'mentality', using it to articulate a 'mentality of rule'. Flores (2013: 265) argues that governmentality is 'the process by which subjects and populations are made to fulfil their role in the governing structure through particular relationships of power and knowledge'. In the case of PD and its application in primary education, governmentality helps to explain how certain forms of knowledge surrounding risk (what it is, who it impacts, and how it can be ameliorated) and normative behaviours of Britishness are produced and legitimated. This implicates the addition of Fundamental British Values education to the curriculum as well as the juxtaposition of P/CVE to safeguarding concerns. Awan et al. (2019: 65) argue that to understand the operation of PD, we must examine the body of 'expert knowledge' that has emerged around P/CVE, a regime which requires knowledge to manage the population, both in terms of cataloguing potential risk and in producing its own knowledge of correct social norms. Knowledge serves multiple functions in governmentality: knowledge of the population itself and knowledge that legitimises certain actions. Flores (2013: 265) emphasises that such knowledge can be produced in institutions outside the government; for instance, knowledge of risk can be produced in schools, which highlights the subsumption of civil society within P/CVE (Awan et al., 2019: 72). Taking surveillance as a main feature of PD, Sian (2017: 7) presents PD as a technology of power 'whereby civil society and public sector bodies, have been drawn in to observe, spy upon, and control the Muslim body' -thus producing knowledge of one group within the population. This argument is in keeping with understandings of PD as policing an insider/outsider boundary, through what Miah (2017b: 132) refers to as the 'racial governmentality' of PD.
The production of the teacher as a governmental subject is important for understanding how PD has successfully embedded in education. Bryan explored the experiences of three school leaders in the implementation of PD, arguing that subjects act within 'the conditions of possibility for individual agency' (Bryan, 2017: 218-219) produced by policies such as PD, wherein the actions of this 'governmentable subject' serve these regulatory regimes. The governmentable subject, then, is one that consents to and participates in these regulatory regimes, not passively but as 'an active . . . and responsible subject' (Sokhi-Bulley, 2013: 234). This coercion towards self-regulation (or 'self-government') forms part of what Kwon (2013: 9) refers to as 'affirmative governmentality'. Important for this discussion is the light that governmentality sheds on regulation and self-regulation, normalisation, the production of deviance through the prescriptions of 'normal' (or, fundamental) behaviours, and the role of surveillance and knowledge production. Desired social norms are performed through everyday practices, forming 'a process of normalisation' (Ettlinger, 2011: 538), and these desired outcomes are not neutral, but rather appear natural or common-sensical. We suggest that the concept of governmentality allows us to connect 'abstract societal processes with everyday material practices' (Ettlinger, 2011: 538), showing how the reproduction of norms by 'citizen-subjects' (Rose qtd. in Ettlinger, 2011: 538) elides the political project behind P/CVE regimes.
Through governmentality, we can view PD as a technology of power that shapes educators (and indeed, students) into agents of security that are aware of and broadly accepting of their position. By technology of power, we refer to PD as a specific tool through which individuals gather, categorise, and produce the knowledge necessary for P/CVE, and through which they are encouraged to self-regulate. As a technology of power, PD produces a field of what is knowable -it categorises individuals according to parameters of risk and keeps them under surveillance to control (though not necessarily to halt) that risk. It creates the conditions of possibility for, in this case, the actions of educators and their relationships to the state security apparatus (see Bryan, 2017). PD requires educators to gather and relay knowledge of threat and risk, and to disseminate knowledge of proper conduct. To understand PD through governmentality, then, is to examine the ways in which PD produces understandings of threat and risk and normalises certain understandings of what constitutes proper conduct.

Methods
Our findings result from a research combining a large-scale survey with semi-structured interviews carried out between July 2018 and July 2019. The survey questions were adapted from Heath-Kelly and Strausz's (2019) survey on PD in the National Health Service. Following the format of Heath-Kelly and Strausz's existing survey allowed for the collection of comparable data across different social sectors, creating the opportunity for a future structured comparison of our respective results. Beyond closed questions, we included free-text boxes to give participants the opportunity to reflect, comment, and elaborate on their responses. 1 The survey was sent to all primary schools in the West Midlands via the official school emails listed online, and via the PEOs' networks. To encourage participation, we offered a £10 Amazon voucher to participants who completed the whole survey. Between February and July 2019, 345 primary school educators answered the survey. Table I in the Appendix provides descriptive information on survey participants.
The semi-structured interviews included 32 primary school educators and PEOs 2 in the West Midlands. The questions encouraged the narration of participants' experiences and perspectives on PD, enabled probing for more information and clarification of answers, and allowed participants to elaborate on areas of particular interest. The interviews took place in a private room in the participants' school or via telephone, when a meeting in person was not possible. Participants were able to talk for as long as they wished, with the interviews lasting between 20 and 60 minutes. All interviews were recorded, fully transcribed, and anonymised with participant consent.
This article draws on descriptive statistics of the survey data (carried out through the R package) and on a qualitative analysis of the interview transcripts and free-text comments based on thematic analysis (Braun and Clarke, 2006). The first author started by thoroughly analysing each interview, annotating, and coding each participant's transcript fully before starting the next one. Thus the coding process was both inductive and driven by the data, allowing an analysis grounded in the 'lived' experience of participants in the case, leading to a unique understanding or potential theory of the case (Simons, 2009). According to established procedures (Yin, 1989), the second and third authors conducted a mini audit of the transcripts and summary documents and agreed with the coding and themes identified. This exercise enhanced the coherence of the analysis, enabling the emergence of new themes and refining of the existing ones with the aim of determining that the final themes were representative of the data. The same approach was also applied to the survey's free-text comments.

Primary educators' understanding of PD
The majority of primary school educators in our study accepted that PD is compatible with the education sector. As Figure 1 in the Appendix shows, 65.2% of survey respondents agreed with the statement 'the Prevent Duty belongs within education'. A key way in which PD has been successfully embedded in education is through its alignment with existing safeguarding requirements (HMG, 2018b). Much has been written about the framing of PD as safeguarding in secondary and further education (Busher et al., 2019;Jerome and Busher, 2020). Corroborating the existing literature, understanding PD as safeguarding was key to how the majority of educators in our study deemed PD as compatible with the education sector and rationalised its position in schools. Our wider data underscores two processes which allow primary school educators to navigate the complexities and contradictions related with the implementation of PD in primary schools: the embedding of PD into existing professional habits through its conflation with safeguarding, and educators' acceptance of the requirements placed upon them by a risk society, which will be discussed below. This evidence suggests that while most primary school educators in the West Midlands accept PD, this is by no means acritical. Many of them felt that the Duty should spread beyond education, as safeguarding could be a community initiative.

The embedding of PD as safeguarding and surveillance
Questioned about the application of PD in primary schools and early years settings, a Home Office spokesperson explained that 'prevent is first and foremost about safeguarding, and through this referral, the child will be able to receive the vital support they need' (Stein and Townsend, 2021). In keeping with this message, most of our interviewees (n = 27) situated PD firmly under safeguarding when asked for their initial thoughts ('What do you think of the Prevent Duty?'). Similarly, 67.2% of survey respondents agreed with the statement 'The Prevent Duty is just safeguarding. It is the same as safeguarding pupils from domestic abuse and sexual abuse' (Figure 1 in the Appendix).
Situating PD under safeguarding turns radicalisation from a national security issue into a form of abuse. This view of radicalisation as an issue of personal security was supported by several free-text comments: I think that the Prevent Duty is safeguarding. As professionals we are looking to support vulnerable young people and ensure they do not become exploited in anyway [sic]. Radicalisation is just the same as any other form of harm where a person is exploited. It was further illustrated by an interviewed headteacher: I have just done child protection training with a group of staff here this morning and obviously we talked about Prevent. I think it is just another part of safeguarding, in the same way that if you saw a child with a bruise, or if a child disclosed something to you. If you are concerned that a child might be being radicalised or exposed to extremism and says something or does something -it might be graffiti on their book or it might be an actual disclosure, might be something they wrote in their English book -if you are concerned you pass on your concern like you would with any other child protection concern. (E18, headteacher, Birmingham) In the context of primary education, PD-as-safeguarding softens PD, depoliticises, and desecuritises it, as 'radicalisation to extremism' is juxtaposed with 'other' forms of harm.
Framing PD as safeguarding is one way in which the state can introduce national security concerns into the wider social fabric in depoliticised ways, as safeguarding is already a key part of educators' work. Governmentality allows us to understand this as an important shift, as PD is both presented as a common-sense addition to ongoing safety interventions by educators and as a policy that individualises responsibility in a way that softens controversy. The alignment of safeguarding with PD also reverses the traditional understanding of safeguarding as primarily about individuals, and focuses on safeguarding the wider population from certain individuals (Heath-Kelly and Strausz, 2019: 93). A lens of governmentality helps make sense of this variation, showing that safeguarding occupies a particular space related to the desire of the state to manage risk in the population. It aligns with the regulatory aims of the state to 'make live and let die' (Foucault, 2003), or to manage and control risk but not necessarily to eliminate it entirely.
On the one hand, the majority of respondents endorsed the assumption that PD is underpinned by a duty of care: 'It's about keeping children safe, not spying' (survey respondent). On the other hand, a minority of responses indicated discomfort with the introduction of PD in primary schools. More than one in (22.6%) five survey participants expressed disagreement with the proposition that PD should be encompassed by wider safeguarding. Survey comments indicated that some educators situated PD and its safeguarding role within the broader security matrix -'It is Safeguarding but it is also part of the wider national security picture-prevent is just a small part of this' -with one respondent emphatically but simply responding 'It is not JUST safeguarding!' Thus, the voice of primary school educators who took part in our study suggested that identification with safeguarding only partly desecuritises PD.
Indeed, as Figure 1 in the Appendix shows, 36.2% of survey respondents agreed with the statement 'The Prevent Duty is a form of surveillance'. Yet, respondents who equated PD with surveillance often filled out the free-text box to voice that 'Surveillance is keeping an eye on people to make sure they are safe' and 'Surveillance is good to ensure all of our safety'. These responses indicated the degree to which PD and the wider P/CVE discourses have been broadly accepted and normalised by the participants. They illustrated clearly primary school educators' acceptance of counterterrorism measures, specifically the need to screen for threat, and the extent to which PD has been fully internalised as part of their professional and social duties. A lens of governmentality allows us to shed light on the context in which P/CVE has become a necessary and legitimate intervention to the point of overruling individual misgivings. One survey respondent reflected that surveillance is 'necessary in today's society to ensure people are kept safe'. This comment exemplifies how far the logic of preventive security and the requirements placed upon educators by a risk society have been internalised by primary school educators.

The internalisation of educators' responsibilities for preventing violent extremism
Our survey respondents displayed a continuing concern with violent extremism: 88.4% of survey respondents disagreed with the statement 'The defeat of ISIS in Iraq and Syria will end the threat of terrorism' and 70.4% agreed with the statement 'I worry that right-wing terrorist attacks will become more frequent' (Figure 2 in the Appendix). Most of the primary school educators who took part in our research were confident that PD is an effective instrument to tackle the diffused and ubiquitous risks of violent extremism. This confirms that risk and its management has formed a critical part of the P/CVE landscape.
Thus, 81.7% of survey respondents agreed or fully agreed with the statement 'I am confident that the continued development of PD will reduce the threat of terrorism' (Figure 2 in the Appendix). Their free-text comments illustrated the extent to which educators endorse the logic of preventive security at the heart of a risk society: I am confident that the Prevent Duty has prevented attacks that the general public are unaware of. (Survey respondent) It will never be eradicated, but I feel that at many levels acts of terrorism have been stopped and credit should be given to the vigilance of schools, the police and all anti-terrorist work. (Survey respondent) Despite being alert to the possibility of irreparable damage intrinsic in a risk society, the perceived direction of risk in primary education was atypical of its wider understanding in P/CVE. Our research participants did not generally consider children to themselves be a risk, but rather at risk. According to educators in this study, understanding PD as part of safeguarding combined with the young age of primary school children led to the perception of pupils as potential victims of radicalisation (as opposed to perpetrators). Survey respondents repeatedly described their pupils as 'vulnerable', 'influenced', and 'at risk': If children are in danger from radicalisation this is the same process as grooming or any kind of controlling or coercive behaviour. It all affects the a [sic] child's ability to thrive and to reach their full potential. (Survey respondent) A deputy headteacher considered that '70% to 80% of our children are fairly innocent' (E15, deputy headteacher, Birmingham) and a headteacher agreed that 'it's really hard for children to be nefarious and secretive' (E23, headteacher, Birmingham), to which an assistant head added: Generally, at the age of our children they probably are less susceptible to Prevent issues, so we're not . . . if we get a trigger about a worry there is usually something else behind it, it's unlikely. (E2,Assistant Head,Walsall) In this vein, and echoing Gutkowski's (2011: 350) observation that techniques of power function as means of pastoral care, PD was described as crucial to help children: it's educating them about tolerance and diversity and about their self-esteem and self-worth, so that they don't become vulnerable groups that are targeted for recruitment. (E27, Teacher, Sandwell) The awareness created by PD was perceived as crucial to equip children with tools to resist 'an extremist voice trying to drag them in a certain way' (E13, Headteacher, Birmingham), but also to support their families and wider community: It is there to prevent extremism, extremist behaviour, er . . . looking at children that might pick up on other members of the family, er . . . and say things about, perhaps, visits or behaviours within families, er . . . and it's to safeguard children, and families as well, from harm through that. (E19, Headteacher, Walsall) As this quote illustrates, our research participants reported that they feel responsible for P/CVE initiatives. In explaining this feeling of responsibility, they referred to their expertise and their professional habits, pointing out that educators are able to notice and to deal with 'suspect' behavioural changes, as stated by different survey respondents: 'changes in behaviour can be spotted in school and teachers know how to talk to children' and 'teachers often get to know children very well and are able to see the changes in them to prevent radicalisation before it becomes too late'.
Existing research is still contending with the puzzling acceptance of PD in education and social care. Our empirical evidence confirms that primary school educators in the West Midlands are remarkably accepting of PD. We contend that a lens of governmentality helps make sense of this evidence, revealing PD as a technology of power that responsibilises educators for P/CVE through their established safeguarding and educational roles. The internalisation of PD has been facilitated first by embedding it into existing professional habits through its conflation with safeguarding. Second, educators have accepted the logic of preventive security and the requirements placed upon them by a risk society. In this context, one survey respondent spelled out this shift in professional identity and practices: 'I consider my role to be that of a protector of children as well as an educator'.

Primary educators' operationalisation of PD
The previous section has presented our respondents' understandings of PD in education and unveiled the Duty as a technology of power perceived by the majority of respondents as both beneficial and urgent in the context of a risk society. Our data show that, despite the discomfort of a minority with PD, most primary school educators in our survey saw PD as 'the responsibility of everyone who deals with children' (survey respondent). To shed light on the materiality and actioning of their responsibilities both within schools and -most interestingly -outside schools, this section will explore the operationalisation of PD by primary school educators in their quotidian. Bryan (2017: 218) emphasises a lack of research into 'how schools are generating their own knowledge and discourses in relation to radicalisation, and how they are designing their own practices'. We address this gap and show how our participants actioned their responsibilities for precautionary risk governance through the curriculum, their raising of PD-related safeguarding concerns (and referrals to Channel) and their role in the wider community.

Actioning PD in the curriculum: The normalisation of fundamental British values
In some primary schools, PD has found a comfortable position within the existing curriculum on Fundamental British Values (FBV). Much like PD itself, the promotion of FBVs is expected to build 'resilience to radicalisation' by empowering children to resist and challenge extremist views (Department for Education, 2015). Research has identified the problematic nature of teaching Fundamental British Values (FBV) in a plural society (Farrell, 2016;Farrell and Lander, 2019;Taylor and Soni, 2017). This was connected to PD as a form of 'soft' surveillance that attempted to create the 'right' kind of citizens and subjects (Farrell and Lander, 2019: 469). Others warn that the FBV framework could be abused by the far-right, as 'framing them specifically as British values' could implicitly exclude other groups (Busher et al., 2019: 447).
Through governmentality, we can understand that, much like PD, FBVs are part of the normalisation that allows for security governance from afar by codifying and normalising what is determined to be appropriate understandings of national identity and what is excluded from it. Indeed, Bryan (2017) highlights the importance of shoring up a sense of 'Britishness' through the emphasis on 'British values' as a 'counter-narrative' to extremist views. FBVs would accordingly ensure a shared sense of British identity and belonging to a coherent 'British' society, particularly in environments where a lack of social cohesion is viewed as generating conflict (Keddie, 2014: 540). They specifically and explicitly instruct students (and, by extension, their teachers and families) in what are deemed to be the prevailing norms of British society.
The teaching of FBVs was related to embedding PD in the curriculum by 10 interviewees directly, with some more oblique references including the reflection: 'we teach children about what's right and what's not right in conventional society' (E23, headteacher). The delivery of FBVs was seen as 'respecting one another, and different cultures, and how to live together' (E9, Deputy Headteacher, Walsall), which, in turn, was seen as preventing radicalisation towards violent extremism. One headteacher from Birmingham actually considered that PD strengthened the delivery of FBVs: You know, we knew that British values have to be taught, we were already doing some of it, but, I mean, we just needed to make it much clearer. I think there is a lot more of a sharper focus on what's being done now because of the Prevent Duty. I think people feel a responsibility. (E13,Headteacher,Birmingham) Despite not being asked directly to reflect on either FBVs or PD's place in the curriculum, some free-text comments in the survey highlighted a relationship between PD and FBV. One respondent reflected that through PD 'teachers are being trained to look out for action and thought that is opposite to British Values'; and another reported that PD 'belongs with society culture and British identity'.
Data from the survey and the interviews underscored that some educators see PD as contributing to shaping British citizens. A handful of interviewed educators mentioned their responsibility in leading children towards becoming 'responsible citizens of the UK' (E21, inclusion manager, Coventry). Similarly, according to one survey respondent: [Prevent] is about educating children of their rights and to be safe from harm. There are children who are not taught to be accepting of others and because of that are narrow minded. Education is key for children to safeguard themselves and be global citizens.
Only a handful of educators flagged up the tensions in this approach. For example, a headteacher warned that: Even calling it British values annoys me a little because surely some values are values to everybody and you can almost alienate people by saying they're British. (E13,Headteacher,Birmingham) In a different approach, one survey respondent noted PD's prescriptive approach to values and its normalisation in the curriculum and broader society and expressed concerns that: we are know [sic] creating a culture where we are intolerant of anyone who has intolerance. Freedom of thought is being attacked. British tolerance is intolerant of anyone who is intolerant to it. (Survey respondent) Besides these two voices, our data suggested that the notion of PD as strengthening beneficial FBVs has been internalised and normalised by primary educators in the West Midlands.

Actioning prevent: The prevention of catastrophic risk
Beyond enacting PD through the manifest curriculum, educators discharged what they deem as their responsibilities for preventing violent extremism by raising internal queries flagging up suspicious individuals to the school leadership. These queries can be escalated into Channel referrals when they indicate a serious safeguarding concern. However, thousands of primary-aged children were referred to Channel between 2016 and 2019 (Stein and Townsend, 2021). Indeed, the vast majority of survey respondents were confident or very confident in referring a query about radicalisation (90.9%). Many referred to the quality of their PD-specific training to explain their confidence in understanding and identifying the signs of radicalisation. In fact, a lack of specialist training or expertise in radicalisation beyond the Home Office Workshop to Raise Awareness about PD (WRAP) did not concern educators. In line with other research (Bryan, 2017;Heath-Kelly and Strausz, 2019), 96% of survey respondents and a majority of interviewees (n = 20) indicated that they were satisfied with the existing training opportunities available to them.
In the survey, we questioned what kinds of scenarios would be considered traumatic, threatening, or worrying enough to warrant a PD-related safeguarding query. Figure 3 in the Appendix summarises primary educators' responses to seven scenarios, and underscores the internalisation of the logic of preventive security as responding to an all-pervasive and potentially catastrophic risk of terrorism. The overwhelming majority of respondents reported that they would raise a safeguarding query in four scenarios: when a student or staff member made violent threats against the British Armed Forces (Q4.6; 84.9%); when a student or staff member watched video clips of beheadings (Q4.4; 87.5%); when a student drew or wrote about weapons (Q4.9; 79.1%); and when a student or staff member expressed anger about immigration and non-British cultures (Q4.8; 67.8%). When faced with a scenario where a student or staff member made hateful statements against an ethnicity, sexuality, or another minority group, more than half of the participants chose not to answer the question. The majority of the 120 participants who responded (69%) agreed that this occurrence should be flagged up.
The survey data suggested that primary school educators were aware of the ambiguities and contradictions related with implementing PD in primary education. However, the free-text comments highlighted that concerns, doubts, and ambiguities were most often pushed aside in favour of raising a query: It depends on the child, but I would be concerned if they were obsessing over something like that [weapons]. I would raise it, even if it turned out to be nothing. (Survey respondent) This was clearest in the free-text comments regarding the two most ambiguous scenarios: when a student or staff member expressed anger about British and American wars overseas (54.9%); and confidence in telling the difference between someone experiencing radicalisation and someone interested in the politics and wars of the Middle East (56.8%). Some survey respondents offered reflections along the lines of 'the person may be angry about violence in general and may be a pacifist, which is certainly [sic] not a bad thing. But this would depend on the context'. While others worried, 'I do not know enough about middle Eastern politics to know what was radical and what wasn't'. This insecurity informed the securitisation of their responses to political discussion and opinion: this would be something to monitor-it could just be a person's opinion however it could also lead to more. (Survey respondent) I'm not sure I would recognise the difference [between an interest in politics and radicalisation] so I would make a referral either way just in case. (Survey respondent) This applied to a variety of scenarios, where survey respondents reiterated the need to 'be vigilant: suspect the worse'. When analysed through the lens of governmentality, these comments demonstrate the pervasiveness and normalisation of the logic of preventive security in primary schools. As a survey participant summarised it in the free-text comments, 'better to be safe than sorry'.
The free-text comments and evidence from the interviews also revealed the importance of educational institutions (particularly of their established hierarchies) in reassuring educators that they are discharging their responsibilities under PD. Time and again, survey respondents reported that they raise PD-related safeguarding queries 'just be on the safe side. Allow the DSL [Designated Safeguarding Lead, sic.] to make the decision'. This approach was corroborated by the experience of the participating DSLs. On the one hand, a deputy DSL from Walsall shared that the guiding principle in relation to safeguarding concerns is to 'raise it anyhow' (E3, Deputy DSL, Walsall). Another DSL from Birmingham said that staff members approach him because they know he is 'very fluent in the Prevent Duty' and he knows what the best next steps are, erm, to be able to make sure that any concerns are acted in a timely fashion and effective way as well. (E25,Teacher,Birmingham) Compliance with these principles was underpinned by an overwhelming trust in the school system, deemed by 90.1% of survey respondents 'to make sensible and appropriate decisions about referring pupils and staff to the Local Authority through the Prevent Duty'. In this vein, 95.9% of survey respondents agreed with the statement 'My school should refer a person to the Local Authority (Channel Panel), if they are sufficiently concerned about radicalisation'. Similarly to PD itself, Channel referrals (the step following an internal referral under the PD) were both presented as safeguarding and as directly preventing the ubiquitous risk of violent extremism. The majority of survey respondents saw Channel referrals as directly preventing terrorism (60%): That is why channel panel exists, to help prevent terrorism, if they weren't aware nothing can be done to stop events from occurring. (Survey respondent) This was not without unease in all cases, as one survey respondent added: Schools are for educating, not to police however if a parent was showing signs of crime and serious plots to kill then you would have to. This is a slippery slope into thought police and thought crime. (Survey respondent) Another respondent made a similar reference in reflecting on PD as surveillance: inevitably we will lead to the things George Orwell predicted in 1984. But since we have created the culture that has caused the terrorist to rise up (Gulf wars that were un-needed and have created ISIS etc) we need to do somewthing [sic] to monitor and stop evil people. (Survey respondent) These responses illustrate particular anxieties around PD -the Orwellian 'thoughtcrime' -suggesting that, despite successful internalisation of the Duty as safeguarding, tensions persist around the deputisation of educators as security agents. As the survey respondent above suggested, these tensions are largely resolved by firmly locating PD (and the resulting Channel referrals) as part of safeguarding vulnerable children. As a survey respondent reported, there may be a 'need to investigate adults around them who are exposing children to terrorism'. Indeed, 10 of the interviewees articulated that safeguarding children meant building a 'bigger picture' of what is happening at home, as younger children's behaviours 'stem from copying family members or saying things that they've heard family members say' (E19, headteacher, Walsall). Reportedly, this bigger picture informed previous PD referrals: We've only had one situation where there was a concern about a family because a child said something that the brother had done, and we had concerns in that way, so we referred it to Prevent. (E17,Deputy Headteacher,Birmingham) Actioning PD in the wider community: Preventive security and surveillance As discussed, the primary educators who took part in our research reported feeling responsible for the successful implementation of PD in education. The logic of a risk society, where PD is deployed to mitigate the risk of irreparable damage from violent extremism, is normalised to the point that several educators in this study extended their duty of care to whole communities. This included an acceptance of the extension of the surveillance implications of PD into the wider community, through a broad understanding of safeguarding responsibilities (Heath-Kelly and Strausz, 2019).
Thus, survey respondents reflected that their role as educators entails 'positions of responsibility in the COMMUNITY (sic) not just school' and that 'as members of the community should all be vigilant to potential risk and that should not be limited to the hours of work'. Thus, the overwhelming majority of survey respondents agreed that 'it is totally appropriate for education professionals to apply their Prevent training outside work hours' (n = 82.9%). Educators in this study affirmed going beyond what they perceived to be expected of them: over half of respondents believed that they are not expected to apply their PD Duty training outside work hours -in the local community, but, many would willingly enforce PD in their local community. This is exemplified by survey respondents' comments: Any professional should be mindful of the safety of any child they meet. Just because you clocked out on a Friday does not mean you ignore an abused child you see on Saturday. Children's Advice & Support Service would still be available.
In this context, survey respondents described actioning PD outside the school gates as 'a social duty', as acting as 'a good citizen', and as being 'accountable citizens of the UK'. These comments validate our description of PD as a technology of power that normalises the deputisation of educators (and the wider citizenry) as security agents through their established roles both within schools and -most interestingly -outside the school gates. Aradau and van Munster (2007: 108) predicted that the War on Terror would lead to the spreading of technologies of preventive security in wider society, through a new prudentialism that responsibilises individual citizens to remain vigilant to threat. In this article, through extensive survey and interview data, we show that the embedding of PD in primary schools is one example of this spread. Our findings corroborate existing literature in showing that the governance-by-risk of PD has found a comfortable position in primary education. The first section, focusing on understanding PD, explored the extent to which educators have internalised and normalised PD as part of their professional responsibilities, and the mechanisms whereby this was made possible. The second part, focusing on doing PD, considered the operationalisation of PD by educators both within the school and in the wider community.

Conclusion
Viewing PD through the lens of governmentality also allows us to contribute to previous research into PD in education in three respects. First, we account for the acceptance of PD illustrated by Busher et al. (2019). Governmentality helps explain the framing of PD-as-safeguarding, which normalised it and rendered it less problematic. Our focus on primary education also casts an interesting light on this safeguarding emphasis: children are presented as being at risk of potential radicalisation due to family or external influences. Educators in our study therefore presented PD as protecting children from externalities and moulding them into good, resilient British citizens. This echoes Gutkowski's (2011: 350) reflection, as discussed above, that technologies of power are often 'disguised as being in the best interests of the individual, the citizenry, or both'. It also shows that the discourse of threat and irreparable damage surrounding P/CVE is understood to necessitate the spread of preventive security measures into education. In this context, some educators afforded PD an extraordinary position because of its gravity and its place within broader national security frameworks, enhancing the overall sense of compatibility between PD and primary education.
Second, considering PD as a technology of new prudentialism sheds invaluable light on the process whereby educators identify threats (cf. Martin, 2018). PD has powerful implications for the norms and conceptions of 'good citizen' conveyed to children through primary schools' curriculum. On the one hand, this evidence corroborates studies denouncing the victimisation of British Muslims as a result of PD. On the other hand, our evidence shows that primary school educators are keenly aware of the threat of right-wing terrorism. This divergence might be due to the specific context of the West Midlands, where the threat of right-wing extremism is particularly acute. It may also suggest thatas time has allowed PD to become increasingly embedded in education -educators have endorsed the logic that 'the potential terrorist could be any of us' (Aradau and van Munster, 2007: 103-104).
Third, our study contributes to debates about the securitisation of educational spaces by analysing PD as a technology of power that produces educators as governmentable subjects (Bryan, 2017). While a significant minority of our research participants reflected on the contested nature of PD and expressed discomfort with PD as a form of surveillance rather than safeguarding, most participants seemed accepting and willing to implement it because of its understood necessity and legitimacy in the context of risk. The latter reported feeling responsible for preventing violent extremism, both inside and outside school, suggesting that they have all been, to varying degrees, responsibilised within the preventive security regime. This captures the dominance of particular forms of knowledge in preventive security, where the need for PD appears so pervasive and common sense that any form of criticism is treated suspiciously. Through their adaptation to and of PD, educators make preventive security possible in a context where, as one survey respondent stated, equating the implementation of PD with surveillance 'is a typical argument rolled out by those who have something to hide'.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was supported by the British Academy Postdoctoral Fellowship (pf170092) and by the Leverhulme Early Career Fellowship (ECF-2015-026).

Supplemental material
Additional supplementary information may be found with the online version of this article.

Content
Appendix: Demographic and survey data. Table I: Survey respondents. Figure 2: Educators' understanding of the Prevent Duty in primary education: Survey data by number of respondents. Figure 2: A risk society: Survey data by number of respondents. Figure 3: Educators' confidence in raising a referral and seven scenarios: Survey data by number of respondents.

Notes
1. The survey is available upon request. 2. Prevent Education Officers (PEOs) work with schools to increase compliance with the Prevent Duty, help shape local strategy, and deliver interventions to strengthen resilience to all forms of extremism.