Through the looking glass: memory, myth and policing the past

ABSTRACT The legacy of the Northern Ireland conflict continues to weigh heavily on the Police Service of Northern Ireland’s institutional memory and contemporary policing practices. In this paper, we argue that the tension between competing interpretations of organisational memory and the need to ‘police’ the past has contributed to a ‘through the looking glass’ phenomenon as regards the policing past and efforts to ‘deal with’ the legacy of the Northern Ireland conflict. The article opens with an examination of the contested memory of policing in Northern Ireland. We then explore how, on the one hand, from a policing perspective, dealing with the past has been regarded as an opportunity to ‘celebrate’ aspects of the policing past but on the other, has reinforced a desire to ‘censure’ the less palatable aspects of police conduct during the conflict. Relatedly, we then interrogate how this desire to ‘secure the past’ has translated into the argument that existing and prospective mechanisms of truth recovery risk ‘re-writing’ the history of the conflict. Finally, the article considers how efforts at truth recovery have been acted upon and received by contemporary policing bodies. This part of the article lays bare both the practical challenges of ‘doing’ truth recovery within a policing framework and the need to ‘secure’ the memory of the policing past. The conclusion argues for the separation of policing and ‘the past’ and the removal of legacy work from policing bodies whose predecessor force was both a victim and perpetrator of that legacy of political violence.


Introduction
November 2021 marks the 20-year anniversary of the creation of the Police Service of Northern Ireland (PSNI). The transformation of policing, via the recommendations of the Independent Commission on Policing for Northern Ireland (ICP) was a central element of the Belfast Agreement. Against a backdrop of problematic policing practices and a lack of confidence in the force among certain sections of the Northern Ireland (NI) population, the ICP (1999, p. 1) was tasked with providing ' … a new beginning to policing in Northern Ireland with a police service capable of attracting and sustaining support from the community as a whole'. Yet, in 2021, and in the absence of a formal process of 'truth recovery' for dealing with the legacy of the NI conflict, the past remains firmly entangled with present-day policing practices and institutional memory. For example, successive PSNI Chief Constables' have argued that the time, finance and manpower required to service historical enquiries have created an inability to police the present. Conversely, for former members and supporters of the Royal Ulster Constabulary (RUC) such as the Police Federation for Northern Ireland (PFNI) and the Northern Ireland Retired Police Officers Association (NIRPOA), the prospect of fresh historical investigations raises troubling questions regarding reputational damage, their role in the conflict and the possibility of criminal prosecutions. This has translated into criticism of existing methods of truth recoveryincluding public inquiries, inquests and investigations led by the Office of the Police Ombudsman for Northern Ireland (OPONI) and a sustained campaign of opposition to the establishment of a formal truth recovery process.
In this article, we argue that the tensions between competing interpretations of organisational memory and the need to 'police' the past has contributed to a 'through the looking glass' phenomenon concerning the unresolved questions on the actions and inactions of the RUC during the NI conflict. In Lewis Carroll's (1872) famous novel Through the Looking-Glass, and What Alice Found There, the world is reversed. In this fantastical realm, all is not what it seems walking away from something brings you closer towards it, chessmen are alive and nursery rhyme characters are real. We argue that a similar subversion of reality exists regarding the policing past in NI. There, for example, the celebration of the sacrifice of members of the RUC is used to censure allegations of wrongdoing and involvement in partial policing practices; the undeniable evidence of collusion has been met with ever more elaborate techniques of denial and deflection; and a double discourse exists whereby senior members of the PSNI have argued strongly in favour of a formal process of truth recovery but where the organisation has been tasked with servicing historical inquiries, it has repeatedly failed to do so in a timely and transparent manner. Drawing on the collection of over 14 years of newspaper articles, press releases, consultation reports and submissions and policy papers from the relevant policing organisations, 1 this article interrogates this intersection between the 'looking glass' and efforts to deal with the policing past in NI.
Although by no means the first transitional justice (TJ) scholars to engage with policing studies, we use the NI case study to further progress the dialogue between these fields. Existing literature has already joined the dots between TJ and police reform vis-à-vis guarantees of non-repetition (GNR) (McAuliffe 2021, McGonigle-Leyh 2021, yet this has largely viewed the reform process as a GNR in and of itself (Almeida 2019, Mayer-Rieckh 2019. Building on the seminal work of Patricia Lundy (2009aLundy ( , 2009bLundy ( , 2012 from over a decade ago, we advocate a more expansive approach by problematising the role of reformed police bodies in post-conflict truth recovery. In going beyond reform as mere outcome, we, as Diphoor, McGonigle Leyh and Slooter (2021) recently suggest, shift the lens of focus off the police and on to policing.
The article proceeds as follows. By way of background, the paper opens with a brief overview of the NI conflict, the nature and scale of victimisation and efforts to deal with the legacy of the past. The article is then structured by four interlocking themes. 'Memory, Myth and Imagining the Policing Past' explores the presence and impact of two opposing collective memories of policing during the NI conflict. 'Celebrating and Censuring the Policing Past' illustrates how, on the one hand, from a policing perspective, dealing with the past has been regarded as an opportunity to 'celebrate' aspects of the policing past but on the other, has reinforced a desire to 'censure' the less palatable aspects of police conduct during the conflict. Relatedly, the next substantive theme, 'Truth Recovery as "Re-writing the Past"' examines how this desire to 'secure the past' has translated into the argument that existing and prospective mechanisms of truth recovery risk 're-writing' the past. Finally, and in the section entitled 'Truth "through the looking glass"', the article considers how efforts at truth recovery have been acted upon and received by contemporary policing bodies. This part of the article lays bare both the practical challenges of 'doing' truth recovery within a policing framework and the impact of ideological and legal impediments that are traceable to a need to 'secure' the memory of the policing past. The conclusion argues for the separation of policing and 'the past' in those transitional and post-conflict contexts when reformed policing bodies are active participants in a process of dealing with the legacy of political violence that its predecessor force was both a victim and perpetrator of.

Legacy, loss and dealing with the past in NI
The NI conflict began in 1969 and lasted over 30 years. In broad terms, it was fought between Loyalist and Republican paramilitary organisations and British state forces, including the RUC, the British Army and the Ulster Defence Regiment. The conflict resulted in more than 3700 deaths: 1842 civilians, 1114 members of the security forces, 395 members of Republican paramilitary organisations, 168 members of Loyalist paramilitary organisations, and 10 members of the Irish security forces (Sutton 1994). Republican paramilitary organisationsthe Irish Republican Army (IRA), Irish National Liberation Army (INLA) and the Irish People's Liberation Organisation (IPLO)were responsible for approximately 60% of all deaths; loyalist paramilitary organisationsthe Ulster Volunteer Force (UVF), Ulster Defence Association (UDA) and the Loyalist Volunteer Force (LVF)for approximately 30%; and state security forces for the remaining 10%. This final figure of course leaves out the matter of state collusion. As discussed throughout this paper, and central to the contention over dealing with the legacy of the conflict, it is now clear that the security forces and intelligence services had infiltrated both Republican and Loyalist paramilitary organisations and were complicit in multiple murders.
Following the signing of the Belfast Agreement in 1998, NI has been an active site of TJ. However, and in contrast to many other transitional jurisdictions, NI has not had an overarching process of truth recovery. At the time of the signing of the Belfast Agreement including such a recommendation was deemed too politically sensitive. Instead, truth has been pursued in a 'piecemeal' fashion, through, for example, public inquiries; 'right to life' challenges under Article 2 of the European Convention on Human Rights; police-led truth recovery by the OPONI (which investigates historical allegations of police malpractice), and the Historical Enquiries Team (HET), latterly the Legacy Investigation Branch (LIB) based within the PSNI (responsible for a cold-case review of all outstanding conflict-related deaths); and victim-led and civil societysponsored initiatives (Bell 2003). Largely a 'criminal justice' response to a transitional context, this approach to the past has been uneven, incomplete and compromised by the weaknesses of the individual mechanisms (Lawther 2015). In response, victims, victims' groups, non-governmental organisations (NGOs), certain political parties and others have made repeated calls for a full examination of the past.
There have been a number of substantive attempts to establish a formal process of dealing with the past in NI. The latest set of 'legacy proposals' are contained in the Northern Ireland Office (2014). Signed in December 2014 by a majority of NI's main political parties and the British and Irish governments, the SHA makes four specific recommendations on the past. Amounting to the creation of a 'bespoke' legacy process for NI, they are the creation of, one, a 'Historical Investigations Unit'; two, an 'Independent Commission on Information Retrieval'; three, an 'Oral History Archive'; and four an 'Implementation and Reconciliation Group' (Northern Ireland Office 2014). Perhaps most crucially for the subject matter of this paper, the Historical Investigations Unit would be an 'independent body', taking over the past facing work of the PSNI and OPONI, thereby ensuring a separation of policing and the past in legacy investigations. At the time of writing, there has been little progress on the implementation of the SHA. Indeed, the Conservative government's Command Paper (July 2021) has effectively overturned the commitments agreed to the SHA and has proposed a de facto amnesty which would bring an immediate end to criminal investigations and preclude future coronial inquests and civil cases (Northern Ireland Office 2021). This recommendation has been widely critiqued by the security forces, victims, NGO's and politicians across the political spectrum. While it is impossible to predict the future direction of this debate, in what is likely to be a vacuum filled by legal challenges, victims and survivors needs will remain unmet and contests over the role of the police during the NI conflict will continue to dominate the post-conflict landscape in NI.

Memory, myth and imagining the policing past
Defined by Misztal (2003, p. 103) as 'the creative imagining of the past in the service of the present and an imagined future', collective memory is not merely historical knowledge shared by a group. Rather, according to Halbwachs (1992), collective memory is understood to express some eternal or essential truth about the group: it has to resonate with how we understand ourselves, how we see our present circumstances and how we think about our future. By way of background, in this part of the chapter, we discuss competing communal memories of policing in NI. As demonstrated throughout the paper, it is this backdrop of contestation against which competing perspectives on truth recovery have been articulated. At least three main points from the field of memory studies are relevant to this analysis. First, is the circularity between memory and a group or an individual's sense of identity and legitimacy. Novick (1999, p. 7), for example, comments that 'we choose to center certain memories because they seem to express to us what is central to our collective identity. Those memories, once brought to the fore, reinforce that form of identity'. Second, and pointing to the 'living' quality of memory, memory does not have a sense of the passage of timerather, it denies the 'pastness' of objects and insists on their continuing presence (Novick 1999). Third, forming coherent narratives of the past, this 'continuity of conscience' links past, present and future and can be influential in the construction of not just 'imagined pasts', but imagined political communities (Oakeshott 1983, p. 15, Anderson 2006. While those outside the imagined community may be deemed a threat to its values and experiences, bolstering and protecting its kinship may simultaneously involve silencing, denying or forgetting certain aspects of the past (Anderson 2006). This notion of continuity has particular relevance in the context of NI police reform because while police reform may have attracted recruits that would simply not have joined the RUC, the absence of any lustration process also meant that officers crossed over from the RUC to the PSNIparticularly in the higher ranks and intelligence sections mired in historical controversy (Hearty 2017). This crossover presents a quagmire for truth recovery; on the one hand, victims will naturally question the willingness and/or ability of these officers to deliver truth relating to police misconduct during the conflict (Hearty 2017), while on the other hand there is a genuine emotional attachment among these officers to the institutional memory of the RUC that seeks to firewall that force from post-conflict critique (Lawther 2014).
The collective memory of the RUC is shaped by narratives of public acceptability, political neutrality, impartiality and a primary concern with enforcing ordinary criminal law rather than exceptional security measures (Ryder 2000). Corresponding to Loader and Walker's (2001) conceptualisation of 'policing as a public good' dealing with 'unnecessary' and 'unjustified' 'terrorist' activity was hence presented as a 'regrettable necessity' (cited in Mulcahy 2006, p. 151, Mulcahy andEllison 2001). An emphasis on the role of the security forces in securing peace and stability, often by way of death and sacrifice has been further used to support this construction of policing memory. This discourse of sacrifice specifically relates to the 305 police officers murdered during the course of the conflict and those who were injured. It also encompasses the emotional dynamic of loss and trauma experienced by the 'policing family'. The strongest manifestation of this discourse of sacrifice was expressed in the post-ceasefire period when the infallible moral claim was made that officers' sacrifice directly contributed to peace, establishing an ethical dimension to their activities and embedding the notion of the policing 'past perfect' (Booth 2001). It is difficult to overestimate how strongly this notion of sacrifice resonates throughout the 'policing family' still, as shown by recent criticism over a proposal to remove a memorial to dead RUC officers currently housed in Crossmaglen PSNI station that is earmarked for possible closure (BBC 2021).
If favourable policing and Unionist narratives epitomise 'shame avoidance' and a yearning for a return to 'normal policing', then Nationalist and Republican narratives on policing depict the RUC as a sectarian counter-insurgency police force and an extension of the British military apparatus (Hearty 2014). The introduction of the Northern Ireland (Emergency Provisions) Act 1973 which provided the RUC and British Army with widespread powers of arrest and detention and of the policy of Ulsterisation in 1975 which gave the RUC the lead role in a revamped counter-insurgency operation translated into a lived experience of a police force which was overwhelming Protestant in composition and fixated on policing the minority population (Walsh 1983, Hillyard 1988. Indeed, it has been argued that large numbers of people were treated as deviant on the basis of religion, residence and association and that the attitude of the RUC was to treat all members of the Catholic community as potential IRA supporters until they could prove otherwise (Ellison and Smyth 2000).
The excesses of policing experienced by nationalist and republican communities have been particularly influential as regards their collective memory of policing. The creation of interrogation centres at Castlereagh, Belfast and Gough Barracks, Armagh was one arm of the RUC's counter-insurgency efforts. Allegations concerning the mistreatment of detainees by members of RUC Special Branch gathered momentum during 1977 and culminated in the finding of the independent Bennett Report (1979) that 'a degree of ill-treatment was condoned at a very high level'. Outside the police complex, nationalist and republican communities were also on the receiving end of an alleged RUC 'shoot-to-kill' policy, whereby there was 'a police inclination, if not a policy, to shoot suspects dead without warning rather than to arrest them' and the extensive use of plastic baton rounds (Jennings, 1988, Stalker 1988, p. 253, Ní Aoláin 2000. The issue of security force collusion with loyalist paramilitaries is explored below. In sum, these experiences of policing reinforced a 'them' and 'us' attitude towards policing and a belief that the RUC was a partisan police force and heavily biased against the local Catholic community (Hearty 2014).
There are of course 'blind spots' in each perspective. As Assmann (2008, p. 219) argues, 'as easy as it is to remember the guilt of others, it is difficult to remember one's own guilt'. For example, the nationalist narrative on policing overlooks the victimhood of RUC members killed and injured during the conflict, the fact that in Unionist areas there was largely 'normal' policing by consent and that for some RUC officers joining the force was a result of coming from a traditional 'policing family', rather than any nefarious sectarian motive (Brewer andMagee 1991, Ryder 2000). Conversely, the unionist 'usable past' is one in which security force responsibility for the conditions which gave rise to or emerged during the conflict is entirely absent (Booth 2001, Lawther 2013. Perhaps the problem is not a lack of 'truth' about policing in NI but rather that, because truth is inextricably linked to lived experience, there has not beenand cannot be (Hearty 2014)anything approximating a 'shared truth' of policing in NI. These diametrically opposed narratives and collective memories on policing in NI both point to the need for truth about the past and the ideological, political and sociological impediments that are inherent to any such process. The remainder of this article examines these dynamics.

Celebrating and censuring the policing past
The themes of celebration and censure have been heavily ingrained in the construction of selective and one-dimensional collective memories of the NI conflict (Ellison 2000). In this part of the paper, we seek to demonstrate how, from a security force perspective, dealing with the past has been regarded as an opportunity to 'celebrate' aspects of the policing past, but on the other, has reinforced a desire to 'censure' the less palatable aspects of police conduct during the conflict. In part, the use of these dynamics is not unexpected. The essence of TJ is reshaping the collective memory of past atrocities through reinterpreting, not just uncovering past events. The struggle of transitional regimes to change values and normsto delegitimise past abuses, rehabilitate those who were persecuted in the past and change the 'model of the "good citizen"' therefore involves the parallel work of celebration and censure (Dudai 2017).
According to Sumner (1990Sumner ( , 1997, social censure is how dominant groups attempt to maintain hegemonic control through the censure of some behaviours rather than others. Behaviours that are censured become 'negative ideological categories' and function to 'signify, denounce and regulate' those whose actions and practices are considered to run counter to the perceived moral-political orthodoxy of the dominant regime, and where the cumulative effect is to 'mark off the deviant, the pathological, the dangerous and the criminal from the normal and the good' (Sumner 1990, Ellison 2000. Security force responses to allegations and findings of collusion provide a case in point. In NI, the term collusion has been used to describe state-sponsored violence and secret collaboration between state forces and paramilitary groups. The existence of collusion is now a wellestablished fact of the NI conflict (see, for example OPONI 2007). However, given that a central tenant of the RUC's collective memory is its presentation as a 'law-bound, disciplined, visible public service, in order to protect life and property', both allegations of collusion and the prospect of subjecting the forces' activities to scrutiny challenges the 'myth of blamelessness' and undermines the argument that republican culpability is at the root of all that has troubled NI (NIRPOA 2009, p. 19, Lawther 2010, Lawther 2014). Collusion has therefore been framed as a topic of maximum deniability.
A number of elements of this 'discourse of denial' can be identified and are summarised briefly here. First is literal denialthe assertion that something did not happen or is not true (Cohen 2001). Collusion has been literally denied -' … there never was, even at the darkest time of the troubles, any hint of what is often referred to as "institutionalised collusion" and allegations of collusion have been described as a "mythological creature", an "illusion" and merely Republican propaganda' (NIRPOA 2017, p. 6). The second form of denial is interpretive denial, whereby the raw facts of an event are not denied but are given a different meaning from what seems apparent to others (Cohen 2001). The classic form of interpretive denial used by the security forces when confronted with allegations of collusion is isolationthe ascription of wrongdoing to a tiny percentage of 'bad apples' who were immediately dealt with (Lawther 2010). Advantageous comparisons by way of reference to the role of the RUC in upholding the law and defeating 'terrorism' -'under the circumstances, we behaved with great restraint and according to the rule of law. Others would have done much worse' have also been made (News Letter 2016).
The third substantive form of denial is implicatory denial, whereby there is no attempt to deny the facts of an event or their conventional interpretation, but the resulting psychological, political and/or moral implications are denied or minimised (Cohen 2001). Applying this form of denial to allegations and evidence of collusion, supporters of the RUC and Unionist politicians have made attempts at contextualisationthe argument that 'If only you really understood our history, politics, the nature of the conflict, then [the weak version] your judgement would not be so harsh or [the strong version] you would support what we are doing' (Cohen 2001, p. 111). The following statement is indicative of this practice -Context is crucial. The sheer volume of terror attacks, shootings, bombings and civil unrest in the early 1970s in particular, meant that the Police were having to investigate an unprecedented case load whilst at the same time desperately trying to prevent further loss of life. This must be borne in mind when anyone tries to judge the actions of the past by the standards of today. (Ulster Unionist Party 2018a) More recently, deflection, commonly used as a tool to avoid digging too deeply into a subject matter that, consciously or not, is thought to be threatening to one's identity and psyche has been used to shut down uncomfortable conversations on the policing past (Barr and Pease 1990). Strategies of deflection used by the security forces include attempting to diffuse evidence of collusion by responding with a list of wrongs committed by the IRA and using the counter charge of the need to investigate alleged Irish government collusion with the IRA: What about collusion between the Republic of Ireland and republicans: the murder of the RUC officers, Mr Breen and Mr Buchanan; the murder of Lord Justice Gibson; the arms trial in the 1970s; the foundation of the Provisionals; the blind eye turned to on-the-runs, training camps, etc in the Republic's jurisdiction; and the border campaign, particularly in my constituency of West Tyrone and the Castlederg area, and the relative ease with which the IRA could come and go across the entire region? 2 However, as Brogden (1997, cited in Ellison 2000 argues, to prevent censure becoming too lopsided and losing some of its exegetic potential, social censure needs to be balanced by social celebration. According to Brogden, 'The concept of celebration complements the idea of censure by illuminating how, that at the same time that governments and powerful agencies demonise their opponents, they also celebrate their own harmony and normality' (Brogden 1997, cited in Ellison 2000. As detailed above and what we wish to focus on here, is the role of sacrificeoften by way of death and voluntary acceptance of loss and in the name of law and order, the preservation of peace and the maintenance of the Unionin the RUC's self-legitimation discourse (Ellison andSmyth 2000, Mulcahy andEllison 2001). Sacrifice, we would argue, has become a point of celebration and in Brogden's terms, a counterpoint to the techniques of censure discussed above. 3 Members and supporters of the RUC have used a number of mediums to practically and symbolically emphasise the importance of their sacrifice. Their contributions to the debate on dealing with the past are no exception. Eulogising statements on the 'RUC's outstanding gallantry during the onslaught of terrorism' and how, in the words of the Chairman of the RUC George Cross Foundation (RUC GC F) Stephen White, 'As debate about how we, as a society, deal with the past, my focus will be highlighting the RUC's outstanding history of service and its proud legacy' (UUP 2018a, Newtownards Chronicle, 2019) pepper official security force responses to consultations on dealing with the past and associated commentary. Indeed, the RUC GC F's (2018) submission to the NIO consultation on the draft SHA legislation opens with a copy of the RUC memorial poster which details the names of all officers who died during the conflict and sits under the heading of 'Lest we Forget'. In essence, the sacrifice of the security forces has been presented as both a shield of legitimacy and an object of celebration and has thus been used to foreground any discussion on the past.
A number of related objections to truth recovery sit under this theme of protecting the sacrifice of the security forces. One is the argument that while the British government readily made 'concessions' to republicans throughout the peace process and enshrined these in the Belfast Agreement, the sacrifice and service of the security forces has been forgotten and undermined. 4 In part, this critique is reflected in the argument that, from a security force perspective, their victimhood has not been adequately recognised. This is not, however, empirically correct. In addition to the levels of financial compensation noted above, their sacrifice has, for example, been recognised in the Patten Report and institutionalised in the creation of the RUC GC F and RUC memorial garden in East Belfast. Rather, we would argue this argument relates to the ongoing controversy over the legal definition of a victim of the NI conflict. The Victims and Survivors (Northern Ireland) Order 2006 provides the legal definition of a victim of the NI conflict. It is an inclusive definition of victimhood that does not distinguish between how individuals came to be victimised but is premised on the basis of objective needs in the present. In contrast, the dominant framing of victimhood within the unionist and security force community is predicated on a strict demarcation between 'innocent victims' (civilians and members of the security forces) and 'guilty perpetrators' (paramilitary organisations) and on the need for a hierarchy of victimhood (Lawther 2014). The perception that the Victims and Survivors Order provides for 'moral equivalence' between 'those who upheld the law and protected society with those who sought to destroy it' has therefore been a considerable source of controversy (RUC GC F 2018, p. 15-16, see also: NIRPOA 2017, PFNI 2018). When allied to the fact that members of the security forces and paramilitary groups are treated in the same way under the Victims and Survivors Order and would be investigated in the same mechanisms in any formal truth process, the perception of creating 'moral equivalence' has hence become 'an enormous barrier that prevents the police family from engaging with the proposed legacy institutions' (PFNI 2018, p. 10).
Closely related is the critique that the unsolved deaths of members of the security forces have not been prioritised in historical investigations and in particular, redress sought through the criminal courts. The emphasis members of the security forces have placed on retributive justice is directly correlated to a belief in the law and its ability to demarcate 'right' and 'wrong'. There is also the expectation that the state will look after its loyal citizens -'the law is there to be upheld and those who break it should be held to account and, where there is evidence, brought to justice' (PFNI 2018, p. 130). Accordingly, the perceived 'failure' to differentiate between what are viewed as a small number of 'politically motivated' public inquiries compared to the fact that a significant number of cases related to the deaths of members of the security forces remain unsolved is believed to be indicative of the government's lack of regard for their service (Lawther 2014). For some, this has translated into an acute sense that their sacrifice has been subsumed by the demands of political expediency and the creation of unfavourable hierarchies of investigation. From a security force perspective, the British government's failure to prioritise the deaths of members of the security forces has therefore been interpreted as challenging their claim to innocent victimhood, implying culpability and calling into question the meaning of their death. As this discussion demonstrates, the twin poles of 'celebration' and 'censure' have been used by the security forces to ensure that their suffering is celebrated while past transgressions are obscured from view.
Truth recovery as re-writing the past It is against this backdrop that one of the security force's most substantive and longstanding objections to dealing with the past has been structuredthe argument that truth recovery is an exercise to be used by republicans to 're-write' the history of the conflict and place all blame and responsibility for its events with the RUC. Three interlinked critiques have been made. First, that existing mechanisms of, and calls for truth recovery have demonstrated and facilitated republicans' intention of 're-writing' the past 'in a way that would have been understood by George Orwell and praised by Joseph Stalin' (NIRPOA 2017, p. 3). Second, that truth recovery, including public inquiries, inquests and the work of the OPONI and LIB, is imbalanced and state centric 'with a disproportionate focus on the activities of the Armed Forces and the Police' (Democratic Unionist Party 2018, p. 6). This point has frequently been allied to the concern that the inevitable disparity between documents and registration/regimental lists pertaining to the activities of the security forces and paramilitaries would engender a highly one-sided process -This would almost inevitably lead to an imbalance which would serve to help re-write the history of the Troubles and completely detract from the fact that 90 per cent of Troubles related killings were the work of Terrorists and 10 per cent due to the actions of the Police or Army. (UUP 2018b, p. 5) Third, that republicans are neither committed to truth-telling or open participation in a truth process. Indeed, the UUP's (2018a) submission to the NIO consultation on the SHA lists 11 examples of 'dishonesty' on the part of the republican movement. They include Gerry Adams' well repeated claim that he was never of a member of the IRA, Martin McGuinness' use of a republican 'code of honour' at the Saville Inquiry into the events on Bloody Sunday, 30 January 1972 when 13 civilians were killed by members of the British Army, and the involvement of the IRA in the fate of the Disappeared. The submission goes on to argue It is obvious that Sinn Fein are attempting to re-write the history books and give the impression that the IRA did nothing wrong.
Sinn Fein in Northern Ireland is demanding the State reveals all, whilst the IRA invokes a code of omerta that would do credit to the Sicilian Mafia. They will not tell what they know, yet demand others do so. (UUP 2018a, p. 45) These arguments against dealing with the past do not stand up to academic scrutiny. Rather, it is our view that they are part of the desire to 'police the past' and maintain the preferred organisational memory of the RUC. As is made clear in the international literature on TJ and truth recovery, the very point of a truth process is to investigate abuses on all 'sides' of a conflict and contribute to the creation of a more complete historical narrative (Hayner 2011). Indeed, as Hearty (2017) argues, what is often pejoratively dismissed as a 're-writing' of the past may simply be the writing of experiences that have never been acknowledged into 'official' memory. If a truth process is to be a legitimate effort at dealing with the past, by definition, it cannot engage in a partial examination of the past or act as a platform for the re-writing of history. In South Africa for example, the SATRC did not engage in a 'witch hunt' as alleged by some segments of the white Afrikaans speaking population but documented atrocities and attributed blame to all sides in the struggle over apartheid (Gready 2011). Rather, we would argue, to seek truth recovery to validate or justify past actions is to miss the point of dealing with the past in its entirety.
While of course accepting that there will be efforts by republicans to sanitise their own pasts and play up the perpetratorhood of the state and its security forces, the 'rewriting of the past' is perhaps, as Homi Bhabha argued, those from the periphery returning to write their own history (Bhabha 1990, p. 6). This is the view of Sinn Fein's Jim Gibney who has argued that republican experiences of policing were 'an unacknowledged history, ignored in the main and dismissed by those who knew and remained silent' (Irish News 2007). Airing and acknowledging the 'truth' of collusion, torture and other forms of policing malpractice may therefore be thought of as a correction to the historical narrative and a part of the transitional process.
That does not however mean that dealing with the past requires an uncritical acceptance of all counter narratives in the name of unacknowledged history. As Michael Ignatieff (1996, p. 111) has famously argued, the function of a truth process is 'to narrow the range of permissible lies'. Despite police and unionist critique of the OPONI's historical investigations as facilitating a 'rewriting' of the past or a 'witch hunt' of the security forces, the evidence demonstrates that the Office has been engaged in precisely the task of narrowing the range of permissible lies. The following remarks by the former Police Ombudsman Dr. Michael Maguire (2018) illustrate this nuanced and critical engagement with the past: … any position which takes the view collusion was policing policy and as a consequence systematic and endemic is at odds will all the evidence I have seen … . That does not mean there were no investigative failures or that the management of informants was on occasion considerably flawed. But to call this institutional failure 'collusion' in every case is to distort and dilute what actually happened and to introduce a toxicity into the debate which is, as we have seen, profoundly unhelpful.
Given that historical investigations by the OPONI or other past facing inquiries more broadly have not sought to 'celebrate' the role of the RUC and have, in certain cases, shone a light on what they would prefer to 'censure', has however had little impact on members of the security forces attitudes to truth recovery and the claim that truth recovery would become a partial exercise in 'rewriting' the past (NIRPOA 2018, PFNI 2018, RUC GC F 2018. Members of the policing family have variously argued that police officers 'went through and … still are going through a horrific experience at the hands of the Police Ombudsman's Office', while public inquiries have been criticised as a platform from which republicans have been able to propagate a view of state complicity while evading examination of their own pasts (NIAC 2009, ev. 287, PFNI 2018. 5 The Saville Inquiry which exonerated those killed by British Army has for example been criticised as an exercise in 'revisionism' with DUP MP Gregory Campbell going further to argue that 'there have been attempts since the establishment of Saville to rewrite history and punish the soldiers who … were being sent in to respond to the increase in the area of attacks, widespread damage and murders' (BBC 2010a).
A dispassionate examination of the figures on historical investigations and prosecutions legally, factually and intellectually further disproves the theory that a 'witch-hunt' of the security forces is taking place as regards legacy investigations and prosecutions. For example, the now disbanded HET completed 1615 investigations, with 1038 attributed to republicans, 536 to loyalists, 32 to the armed forces and 9 unknown (The Independent 2018). While a small number of convictions did result from the HET investigations, it remains the case that not one single member of the security forces has been convicted to date as a result of a legacy investigation into a conflict-related death (McEvoy 2017). Likewise, since January 2012, 33 cases have been passed to the DPP for decisions on whether or not to prosecute. The DPP has initiated legacy prosecutions in 17 cases. Eight of the cases that went forward to prosecution were against alleged republican paramilitaries, 4 against alleged loyalists and 5 against British Army personnel. Four prosecutions have resulted in convictions -2 republicans and 2 loyalists (McEvoy, Holder, Mallinder, Bryson, Gormally and McKeown 2020). However, the emotional tide of collective memory and a strict adherence to a narrative of blamelessness has mitigated against any other interpretation of historical investigations on the part of the security forces. Their rejection of a truth process therefore effectively equates to a de facto acceptance of a continuing focus on the state in a context in which they claim non-state actors were responsible for 90% of deaths. Continuing this approach to the past arguably means continuing to let non-state actors 'off the hook' (Lawther 2014).
Truth 'through the looking glass' In the absence of a formal process of dealing with the past, the PSNI has been heavily involved in providing a distinctive form of 'police-led' truth recovery in NI. Grounded within the broader TJ themes of 'healing', 'closure' and 'accountability', the creation of the HET in 2005 and historical investigation role given to the OPONI speaks to the need to respond to victims' needs and the capacity of policing organisations to do so. The involvement of the police in recovering truth about the conflict is also reflective of the need to create a police service capable of commanding maximum confidence and 'buy in' from the community. Both the HET and OPONI therefore had the capacity to 'manufacture legitimacy' in policing, assisting in the transition from a problematic past to an organisation that is regarded as accountable, transparent and forward-looking (Wilson 2001, p. 17). However, as we explore below, the competing 'truth claims' of two very different versions of the policing past and the desire to maintain ownership of the past on the part of supporters of the State and RUC has, in certain circumstances, resulted in a 'through the looking glass' obfuscation and manipulation of truth.
The HET was established in 2005 by the former Chief Constable of the PSNI Sir Hugh Orde. Its primary objectives were to assist in bringing a measure of resolution to those families of victims affected by deaths attributed to 'The Troubles' in the years 1968-1998 and to chronologically re-examine all 3268 deaths that had occurred during this period, 2002 of which were never solved. (HET undated) By mid-2013 the HET had reopened 2068 cases which related to the deaths of 2682 people and had completed 1713 cases which related to the deaths of 2209 people (Healing Through Remembering 2013). However, in mid-2012, a number of NI based NGOs who had first-hand experience of family engagement with the HET, raised concerns relating to the HET's independence and effectiveness, particularly in relation to the organisations structural relationship with the PSNI and its staffing policies (Committee on the Administration of Justice 2015). These concerns were reinforced by academic Patricia Lundy following an in-depth period of research with the unit itself. Lundy (2009aLundy ( , 2009bLundy ( , 2012 found that each phase of the HET process included former long-standing RUC officers, including ex-Special Branch officers in investigations; undue influence of the 'RUC corporate memory' and the absence of 'competing memories' and 'alternative voices' in the process of investigation and writing reports; 'cross-contamination' of organisational linkages; inequalities of treatment by the HET in cases where state agencies were involved compared to cases involving nonstate actors; and an inability to identify and trace military personnel allegedly involved in killing civilians and/or eyewitnesses. In light of Lundy's findings, a review of the HET was carried out by Her Majesty's Inspectorate of Constabulary (HMIC). The HMIC report, published in early 2013, was highly critical of the unit (HMIC 2013). It found that the HET had not maintained its Article 2 required independence because it failed to ensure that former RUC members now working for the HET were not involved in 'state involvement' cases; that cases involving state actors appeared to be treated less rigorously; and that 'in cases of state involvement, the HET acts as investigator and prosecutorial decision-taker' (HMIC 2013, p. 87). Following significant funding cuts and a declaration of 'no confidence' in the leadership of the HET by the Northern Ireland Policing Board and widespread criticism of the unit within civil society, the HET ceased investigations in December 2014 (Irish Times 2014). All outstanding HET cases passed to the newly created LIB within the PSNI in late 2014.
The OPONI has faced similar challenges. The OPONI was established in 2000 and is designed to provide an efficient, effective and independent police complaints system. In addition to dealing with day-to-day complaints against the PSNI, the Office is also mandated to investigate historical complaints against the RUC. In 2011 reports into the organisation by local human rights NGO the Committee on the Administration of Justice (CAJ) (CAJ 2011) and the Criminal Justice Inspectorate Northern Ireland (CJI NI) (CJI NI 2011) cast doubt on the capacity of the office to deal with historical cases. Areas of concern included a lack of operational independence between the Office and the PSNI; that reports into historic cases were altered or rewritten to exclude criticism of the RUC; that staff investigating some of the worst atrocities believed the police had acted as 'gatekeepers' to withhold key intelligence from them; and that there was a lack of any clear definition or consistent application of the term 'collusion', with different interpretations of the term being used depending on the circumstances of the case (CAJ 2011, CJI NI 2011. The adoption of a politically fluid and politically expedient definition of collusion has been particularly problematic and is indicative of doing truth recovery through a 'looking glass' that filters interpretations of the past according to the need to preserve a narrative of blamelessness and strict division between innocent and guilty victims. Families affected by historical investigation report issues by the OPONI at the time of the then Ombudsman Al Hutchinson's tenure have, for example, described the organisation as performing 'factual gymnastics to ensure there was no evidence of collusion' (BBC 2011) and of making findings which were patently ridiculous (BBC 2010b). Dr Michael Maguire who succeeded Al Hutchinson as Police Ombudsman in 2012, similarly critiqued 'a cherry-picking approach to reports which is regrettablethe banking of conclusions which are liked and a challenge to the conclusions which are disagreed with' (Maguire 2018). The return to a more robust definition of collusion was one of Maguire's first announcements upon taking Office. 6 At the time of writing, the LIB and OPONI continue to function. Both the PSNI and the OPONI have publicly stated their commitment to the creation of the Historical Investigations Unit as envisaged in the SHA and the separation of historical investigations from contemporary policing practices (The Detail 2015). The financial cost of policing the past, the loss of staff time and the impact of the past on police-community relations have all been influential. However, despite this public commitment to truth recovery, the spectre of the policing past continues to loom large within the PSNI and its commitment to 'doing' truth recovery has not always been followed up through deed. In addition to the preferential treatment afforded to state actors during its investigations as detailed above, the PSNI has obstructed and frustrated truth recovery by delaying the disclosure of necessary information, being overly sensitive with redactions, taking appeals and judicial reviews of certain procedural decisions (CAJ 2015), and it has tried to curtail grassroots truth recovery by seizing materials from investigative journalists making a documentary on collusion (CAJ 2019, p. 4). Pertinent examples include the failure to disclose 'significant, sensitive information' to the Police Ombudsman in respect to his investigation of the UFF gun attack on Sean Graham's bookmakers shop on Belfast's Ormeau Road in February 1992 (PSNI 2019). Five people were killed in the attack amid allegations of collusion with the security forces. Likewise, in respect to the investigation into the Glenanne Gang (a loyalist paramilitary gang which has been linked to over 120 murders and is believed to have included serving members of the police and security services), the LIB for example has been critiqued for pursuing a compartmentalised and decontextualised approach to truth recovery and from 'specifically prohibiting itself from any active investigation of linkages between individual historical crimes and from the active pursuit of new evidential leads/unused opportunities for investigation '. 7 The 'looking glass' effect of the organisation being seen to be committed to 'doing truth recovery', yet not overly committed to the (re)investigation of former state actors; addressing past wrongdoing by the RUC but not damaging the reputation and legacy of a policing institution that it is historically, culturally and socially still connected to; and convincing victims of state violence that it has moved away from past patterns and practices of impunity without contradicting organisational memory tropes of public acceptability and policing neutrality appears to dominate the organisations efforts at truth recovery. The result is neither 'truth' nor 'recovery' for victims and survivors of the conflict.

Conclusion
As illustrated throughout this paper, despite being almost 20 years since the creation of the PSNI, two distinct collective memories of policing during the conflict remain and exert a considerable practical, political, legal and sociological influence on policing in NI. This is particularly acute in respect to efforts to deal with the past and the experience of police-led truth recovery. In critiquing the PSNI's role in 'doing truth recovery' we add to an emerging literature that has framed post-conflict police reform within a TJ framework (McAuliffe 2021, McGonigle Leyh 2021. At the same time, however, we have advanced the ongoing dialogue between TJ and policing literatures by moving the discussion beyond a narrow focus on organisational change (Murphy, McDowell and Braniff 2017), on historical memory (Lawther 2014, Hearty 2017, and on GNR (McGonigle Leyh 2021). We have, instead, merged these foci in critically examining the problematic interplay between policing the past and policing the present (and future) that occurs when reformed policing bodies are active participants in a process of dealing with the legacy of political violence that its predecessor force was both a victim and perpetrator of. This, we have argued, has created a 'through the looking glass' effect in NI whereby truth and efforts at truth recovery have been filtered and distorted to protect organisational memory and facilitate an active policing of the past. The strength and prevalence of the 'looking glass' effect have shown the mismatch between official discourses and policing realities on the ground in NI (Topping 2015, p. 120) and the ongoing challenges that police-led truth recovery poses for 'policebuilding' (McAuliffe 2021) in a society that has seen police reform in the absence of an overarching TJ process. As we have sought to demonstrate, the 'backstage' reality of policing in NI is where, as described by the former Chief Constable Sir George Hamilton, the walls of policing buildings and conversations in staff canteens, the past is dominant (Goffman 1959, Hamilton 2019. It is also, as described by the former Ombudsman Michael Maguire, where some, including current officers, retired officers and police representative bodies 'have difficulty with accountability' (Irish News 2019). Such an approach is incapable (and at times deliberately so) of producing truth for victims and their families, contributing to a full narrative on the role and experiences of the police during the conflict or of promoting cross-community confidence in the delivery of present-day policing. With this in mind, then, it is difficult to disagree with the synopsis of PSNI Chief Constable Simon Byrne that 'we [the PSNI] shouldn't be dealing with it [the past] … other issues … mean we're not best placed to deal with it' (Kearney 2019). Thus, we argue strongly for the separation of policing and 'the past' and the creation of bespoke legacy processes, such as those envisaged in the SHA, to deal with the many unsolved questions of the NI conflict.
Notes 5. This argument implies that the contribution of former police officers to historical investigations has been open and truthful. Evidence presented by the former police ombudsman Nuala O'Loan (2010) and the Committee on the Administration of Justice (2015) demonstrates that this has not always been the case (OPONI 2007). 6. Speech by Michael Maguire, CAJ Annual General Meeting, November 2012, attended by one of the authors. 7. Court finds Chief Constable Breached Article by Failing to Produce HET Report into Activities of Glenanne Gang, Summary of Judgement, 28 July 2017.

Disclosure statement
No potential conflict of interest was reported by the author(s).