The Power of Narratives in Conflict and Peace: The Case of Contemporary Iraq

ABSTRACT Presenting original data from interviews conducted in Iraq between 2013 and 2019, this paper charts the evolution of conflict related narratives of the ‘other’ among members of the Sunni confession immediately prior to, during, and after the rise and fall of the Islamic State (IS). It charts the evolution of these narratives through three escalatory phases (victimhood narratives, divisive narratives, and violent narratives) and three de-escalatory phases (nuanced narratives, reconciliatory narratives, and unifying narratives). It concludes with reflections on the lessons this case can provide for identifying the best moments for Conflict Resolution actors to intervene in such violent conflicts.


Introduction
Conflict resolution literature has consistently recognised the importance of narratives in the construction of identity, the development of in-group/outgroup distinctions, and, as a result, the promotion of conflict dynamics (Funk and Said 2004: 2). A number of approaches within the field have focused on deconstructing conflict promoting narratives and fostering more inclusive narratives that might serve to foster peace and reconciliation between groups (Kelman 1990, Burton 1987, Saunders 1999, Fisher 2001. However, while we certainly recognise the importance of such approaches, this paper inverts the perspective to some extent, and focuses instead on the idea of shifting narratives or the emergence of less antagonistic narratives as indicators of when intervention might be more possible or fruitful. Using novel data collected by an intervention practitioner, the paper exhibits the value of listening to local actors to understand the best windows for peace intervention. This data was collected as part of preparatory work and actual peace engagement primarily with Sunni Arabs in Iraq and covers the divisive years between 2013 and 2019. While there is a longer history of ethnic and religious division in Iraq, in the years following Saddam Hussein's ouster in 2014 these divisions were further aggravated. Iraq represents a complicated religious and ethnic mosaic similar to Lebanon or Syria. Religiously, the Shia majority forms about 65 per cent of the population while the Sunni minority forms about 32 per cent, in addition to Christian, Yezidi, or Shabak minorities. Ethnically, the Kurds, who are themselves mostly Sunni, form about 20 per cent of the population, while 80 per cent of the population are Arabs, with a smaller minority of Turkmen. In the years discussed in this paper the violent rise of and resistance to the socalled Islamic State further complicated this picture, as it resulted in the effective partition of the country into three separate and mutually hostile entities; areas controlled by the Shia dominated Iraqi government, by Sunni dominated Islamic State (IS), and by the government of the Kurdistan Autonomous Region. As we will describe, over this relatively short period of time historical events gave rise to three different kinds of narratives which defined the relationships between these groups; victimhood narratives, divisive narratives, and violent narratives. These narratives legitimated and promoted the violent conflict that would ensue.
However, the paper will then also explore how narratives among this group evolved, from inflammatory expressions and justifications of violence (and even terrorism), towards increasing self-reflection and expressions of reconciliation. Over time, and in response to the violence which occurred, different forms of narrative emerged; first more nuanced narratives, then the first hints of reconciliatory narratives, and finally, unifying narratives. As will be discussed, these alternative and pro-peace forms of narrative opened the door for attempts at conflict transformation from within and supported efforts by international actors to engage with and facilitate such attempts to deconstruct the in-group/out-group dichotomy. The paper, therefore, explores the complex interweaving of historical events and vocalised narratives in the construction of conflict promoting in-group/out-group dynamics. The empirical focus is on how those historical events gave rise to specific narratives in the case of Iraq, while the theoretical contribution focuses on the role that narratives may also play as indicators of when intervenors can have the most positive impact on conflict transformation.
The paper is divided into four sections. The first section introduces and discusses the key scholarly work related to in-group/out-group dynamics, the role of narrative in conflict, and the conflict resolution mechanisms we believe can benefit most from paying close attention to emerging narratives in the field. The second section provides a brief description of the methods used to collect the data presented in the paper. The third and longest section then presents the empirical data from the case, describing the 6 different forms of narratives that were observed, primarily but not exclusively, among Iraq's Sunni confession. The fourth section then concludes the paper by relating the findings back to the literature discussed and illustrating the most pressing implications.

Literature Review: 'Othering' and the Power of Collective Narratives
As a field, Conflict Resolution (CR) has benefited enormously from the ability to incorporate and learn from a variety of disciplines such as political science, economics, law, social-psychology, and sociology. Such an eclectic mix, however, has given rise to many sub-fields. This includes scholars and practitioners who work in arbitration, mediation, negotiation, track-II diplomacy, intergroup dialogue, peacekeeping, peacebuilding, conflict transformation and reconciliation. But while the field is obviously quite diverse, there are a handful of central ideas which contribute to and provide foundation for most of these sub-fields and their related practices. One of those central ideas is that of the in-group/out-group dichotomy and the related notions of the 'other' and the process of 'othering'.
In The Functions of Social Conflict (1956), Lewis A. Coser argues that conflict performs key functions for social groups. First among these central functions is the reification of the boundaries between groups, which provides further strength and stability to the group. In reifying the boundaries between groups, for example, conflict serves to strengthen group control mechanisms, solidify group identity, and provide the 'in-group' ('we' or 'us') with an 'other', or an 'out-group' ('them'). This 'other', therefore, is the group against which the in-group compares itself, and it is this 'other' which serves as the target for aggression in conflict, whether violent or otherwise. While there are additional concepts which have become central to the field -Galtung's idea of the 'Positive Peace', for example (1969), or Lederach's more recent 'elicitive' approach (1997) -few ideas have affected more areas of the field of conflict resolution than this idea of the 'in-group' and the 'out-group' (Millar 2012: 717). But explaining how conflict contributes to this formation and reification relies on further theory. This brings us to the role of 'narrative' and how identity is formed through the presentation of events. Vamik Volkan argued, for example, that collective or shared traumas are transmitted, even across generations, through simple processes of storytelling and myth-making, and that the memories of such traumas form the foundations by which identity groups then define themselves in opposition to the 'other' (2001). Rice and Benson describe, for example, how historical traumas on the Island of Ireland (the Irish Famine and the Great Hunger) structure the narrative dynamics among Catholics in Northern Ireland and fuel conflict dynamics (2005), while Mahood and Rane describe how similar narratives of trauma are used in extremist recruitment propaganda (2017). Such collective narratives of trauma (of past invasions, massacres, defeats), as well as shared heroic myths (of military victories, visionary leaders, etc.) can then be appropriated by political elites to galvanise the in-group and foment violent action against the 'other' (Mertus 1999). As noted by Ron and Maoz in their study focusing on ingroup narratives in the context of Israel-Palestine, 'contesting narratives play a crucial role in ethnopolitical conflicts, with each side adopting a narrative that justifies its own claims, demands, and position while delegitimizing those of the other side' (2013: 281).
Many classic cases of 'intractable' conflict evidence such dynamics, indicating further how conflict occurs and serves to further reify pertinent ingroup/out-group divisions via emerging narratives of trauma and heroism (see Bar-Tal 2007: 1435. To return to the case of Northern Ireland, for example, one of the pivotal cases of the past half century, the Catholic and Protestant groups are not only defined by alternative historical narratives, but the division is itself reified by the process of conflict; the manner in which the groups live in segregated spaces divided by the prominent use of 'symbols of local community identity' (Sluka 1996: 385), such as 'flags, anthems, murals, badges, bunting and graffiti' (Sluka 1996: 381). Such symbols act as 'public manifestations of group identity' (Brown and Macginty 2003: 84). and, when combined with the actual physical separation, they add to the othering effect of the conflict (Leonard 2010: 333). Similar phenomena are evident in cases around the world, from the influential cases of Israel/Palestine, Cyprus, South Africa, Rwanda, or the former Yugoslavia (Saunders 2003, Halperin 2008, right through to cases that few in CR commonly engage with, such as within Indian cities (Mehta and Chatterji 2001).
This perspective, of course, is quite pessimistic. But the theoretical contribution of this literature can also be hopeful. It suggests that narrative can play a central role both in creating and then in recreating identity groups. If in-and out-groups can be constructed, then they can be deconstructed, if they can be reified, then they can be tempered. Indeed, the goal of many CR practices has largely been to break down or deconstruct the in-group/outgroup dichotomies which have been reified in conflict between groups. Allport's early development of 'intergroup contact theory ' (1954), for example, formed the foundation of many practices which would later become central in the field and which work to break down inter-group animosities (Schofield 1979, Pettigrew 1998, Saunders 1999, Miller 2002, Dovidio et al. 2003). Osgood's description of a process for 'Graduated Reciprocation in Tension Reduction' (GRIT) was similarly developed to lessen the tensions between two opposing groups in conflict (1962; see also Cook et al. 2005: 122).
This issue is of more interest to certain sub-fields of conflict resolution than it is to others. Theorists or practitioners of inter-group dialogue (Burton 1987, Saunders 1999, conflict transformation (Lederach 1997), or reconciliation (Fisher 2001, Nadler andShnabel 2008) are clearly interested in deconstructing such dichotomies. Whether reconciliation, for example, is considered an individual psychological process which helps to realign individual cognitive and emotional understandings of the relationship between groups ( Bar-Tal and Benink 2004: 34, Kelman 2004, Maoz 2000, Fisher 2001, Saunders 1999 or as a collective psychological process 'removing conflict related emotional barriers that block the way to ending intergroup conflict' (Nadler and Shnabel 2008: 39; see also Bar-Tal and Benink 2004: 34), both respond to Coser's conception of social conflict as driven by the in-group/out-group dichotomy and of reconciliation as about the deconstruction of that dichotomy.
Further, the active engagement and reconstruction of narratives is key to such CR approaches. Again, we see this clearly in reconciliation practices, which, at least in their largest scale in the form of Truth and Reconciliation Commissions (TRCs), focus on the public-performance of truth through presentations which are thought to reframe the existing narrative. To those scholars who see conflict as always occurring between an in-group and an out-group, this process of 'truth-telling' is 'a collective storytelling therapy' (Millar 2015: 245) which creates a new nationally shared narrative above the level of either party to the conflict. What might be called a meta-narrative or a 'collective memory' (Chapman andBall 2001: 15, Sooka 2006: 319) which, because it is shared by the whole population (by those in both groups) provides a new way to minimise 'the number of lies that can be circulated unchallenged in public discourse' (Ignatieff 1996: 113). This new narrative, in turn, provides a foundation for more peaceful coexistence between the groups. The foundation of the new shared nation, in this sense, is necessarily dependent on a new shared narrative.
But such processes have, in recent years, come in for extensive critique. TRCs have been attacked for their politicisation (Wilson 2001), for sometimes re-traumatising victims of past violence (Millar 2015), for their focus on the national as opposed to the local histories and dynamics of conflict (Robins 2011), and for their foundations on largely Western or European traditions of psychological healing (Pupavac 2004). Similarly, other practices implemented to deconstruct group narratives, such as intergroup dialogue processes, have been critiqued for inappropriately psychologising and individualising conflict dynamics (Erasmus 2010) and for their inability to overcome the broader structural conditions of conflict, or what Pettigrew described as the 'generalization of effects problem ' (1998: 70). Such critiques, therefore, call for a very sensitive and responsive approach to interventions seeking to influence narratives. They show how important it is to trace the emergence and spread of narratives in order to understand how and why new narratives emerge. However, as this paper argues, tracing the emergence of new narratives among key groups within the conflict context can also help indicate precisely when interventions to shift such narrative might be more successful.

Methodology
The data presented in this paper is quite novel, in that it is not the result of a predesigned theoretically informed research project, but was collected instead by the first author in their roles first as a humanitarian worker and later as a mediation professional. These roles provided the interviewees with the freedom to express themselves freely and with the clear intent for their statements to be transported to a wider international audience. There are two specific time periods and general geographic areas from which this data was collected. First, interviews were conducted between October 2013 and December 2014 in Baghdad and bordering areas of Al-Anbar province and among members of the Sunni confession who supported the Sunni-led uprisings in 2013-2014, which led later on to the establishment of the socalled ´Islamic State´ during the same period. These interviews provide insight into the initial conflict promoting narratives circulating in the Sunni population at that time. Analysis of these more wide-ranging conversations led to the selection of the various quotes presented below for the insight they provide regarding conceptions and framings of the 'other' among this community during this tempestuous time.
The second set of data, describing the later wave of more peace promoting narratives, were collected between April 2018 and May 2019 from Sunni representatives from Ramadi and Mosul -two sites at the centre of Sunni thought. These data were collected not during one on one interviews, but during dialogue initiatives conducted by the first author. As such, these data are more akin to that which might be collected during focus group interviews and, again, the analysis of the wider data collected through this process led to the selection of quotes specifically for the insight they provide regarding how this group were describing and thinking about the 'other' at this specific time. In addition, because this dialogue process included individuals from both sides of the conflict, this data also includes reflections from representatives from the holy Shia city of Najaf. These data, as will be illustrated below, show the similarity of the unifying narratives emerging at this specific time. Some of these statements were collected during conversations, others during filmed interviews, and others during addresses to the audience of the dialogue initiatives.
All interviews were conducted by the first author in Arabic and translated for the purpose of this article. Mostly, the interviewees were male Arabs, Sunni or Shia, depending on the location, and an between 35 and 65 years of age. Within that category, about 30-40 persons were interviewed in the period 2013-2014, and most of them belonged to Sunni tribes. In the period 2018-2019 about 20-30 persons in the age group 35-60 were interviewed, about two thirds of whom were Shia and one third Sunni. Interviewee's backgrounds were rather diverse, ranging from tribal and religious representatives to more progressive civil society personalities such as academics, journalists, NGO leaders, or other opinion makers. In this period interviewees also included about 10-15 males and females aged 18-30, one of the females in Mosul was Christian, another two males were Yezidi.
As it was collected initially not for research purposes but during intervention practice, there are clearly limitations to this data. The form of data collected across these two waves is quite distinct, with individual interviews in the first wave and a more focus group like process in the second. As such the distinct differences between the forms of data that each of these methods produceleads to a problem of comparison between the two time periods. On top of this, while comparison between Sunni and Shia is possible in the latter phase, it is not possible in the former, and so verification of the narrative dichotomy in the earlier period is not possible. Further, in neither case were participants randomly sampled or even selected via a purposeful or convenience sampling technique and, indeed, there was no formal 'instrument' used to collect the data and, as such, questions posed to participants were not uniform.
As a result of all of these challenges the data collected from participants and presented below is clearly not generalisable and must be taken instead as indicative of the narratives that were present in the wider society in these various sites and in these two recent historical periods. While further more structured and rigorous research would therefore be necessary to confirm our findings, we nonetheless feel that the data collected, analysed and presented is sufficient to provide some indication of how these narratives were emerging and circulating in Iraq during these time periods, while also responding to recent calls within the field for more communication between practice and theory and more engagement across the scholar-practitioner divide (Paris 2011, Nathan et al. 2018. As such, while we recognise the limitations of these data, we hope that the indicative findings and development of theory they allow exhibits the strengths of a collaborative process that bridges the scholar-practitioner divide.

The Context of Iraq
Iraq has a long history of conflict and intergroup antagonism that has been well traversed by scholars (Allawi 2007, Abdel-Razek andPuttick 2016), and it is not the aim of this article to add new knowledge to this historical discussion. However, for the reader to be able to put the development of narratives around the dramatic events of 2014 into context a short overview of the factors contributing to it will be provided. Critical to this context is to understand that former Iraqi President Saddam Hussein was himself a Sunni Arab and throughout his regime he had focused his power-base upon relevant tribes and cities in Sunni-majority areas around the capital city of Baghdad, located in the centre of Iraq and the city of Mosul in particular, which contributed as many as 300,000 to the Iraqi security forces (Hamilton 2008). At the same time, Hussein largely sidelined Shia tribes and religious centres in the Southern provinces; a dynamic which was only heightened by the Iran-Iraq war between 1980-1988 in which parts of the Iraqi Shia minority were seen to have supported the enemy.
The US-invasion and overthrow of Saddam in 2003 reversed this dynamic. Most importantly, all members of Saddam´s ruling Baath party were expelled from the system, including not only a majority of bureaucrats but also the security forces (Isakhan 2011). A significant number of Sunni actors were thus legally excluded from any positions of authority, while another significant number chose to boycott the newly developing system of government. This new approach, as expressed in the 2005 Constitution, developed a system of quotas referred to as muhasasa in Arabic (Al-Amin 2016), basically conferring power to the Shia majority and marginalising the Sunni minority who had previously wielded power. This majoritarian governance led to a struggle for power and resources which turned into violence and resentment amongst the two main confessions; Sunni and Shia. Adherence to one of these sects was used either to claim power and benefits if you were Shia or to call for resistance if you were Sunni. In this context the historic imbalances between the two sects were readily exploited by politicians to strengthen their power bases through polarisation supported by partisan media. The sectarian violence which occurred between the Shia and Sunni populations around 2006-2007 increased this divide further in the years prior to 2014.

The Power of Narratives
As if nothing had been learned from previous brutal periods, Iraq's leaders continued their polarising policies until the next disaster emerged in 2014, when the Sunni minority rose up violently against the Shia dominated government. As is true in most conflicts, the experience of violence and the toxic political discourse both fed from and then further contributed to popular narratives that emerged within the country and served to frame the pertinent in-and out-groups. As will be illustrated below, these narratives leant initial support to conflict escalation in the form of victimhood narratives, divisive narratives and violent narratives, before then reversing this pattern in the emergence of nuanced narratives, reconciliatory narratives, and, eventually, unifying narratives.

Victimhood Narratives
When analysing the narratives among Sunni citizens in Baghdad province and the border areas of Anbar province, which, as mentioned above, are regions previously favoured by the Saddam regime, it was clear that many felt like they were no longer part of the same country or society as they remembered. They presented narratives that framed the two confessional groups as antagonistic and the Sunni as being victimised by the new situation. Parts of the Sunni population, for example, had started to equate the Shia population as a whole with the US-supported Shia-dominated central government. 'The invaders promised democracy and freedom before destroying our country, only to give all power to the damned confessional dictatorship', 1 one said. Such extremely simplistic narratives, whether rooted in malice, ignorance, or some semblance of reality, completely ignored the fact that large parts of the Shia population in the South of the country were also at odds with the central government and fought against US-troops and government security forces.
Others, perhaps in an effort to attribute the responsibility for this internal societal division to foreign powers rather than their neighbours and fellow citizens, preferred to place most, or even all, of their grievances on neighbouring Iran. In this spirit some theorised, for example, that 'The US only attacked Iraq in order to pass the power over to Iran'. 2 The people giving voice to such theories likely had no evidence to support them. But a narrative need not be evidenced and there was some tenuous truth behind such ideas. Iran was certainly trying to influence Iraqi affairs for the benefit of its own foreign and security policies and not for the benefit of the Sunni population which had been the main support for Saddam Hussein, Iran´s arch-enemy.
In addition, the Sunni population started recalling the seemingly better times during the past regime under Hussein, comparing it to the situation under Al Maliki and framing the Sunni as the victims of this transition of power. As one interviewee stated, 'the Shia now have all the power. Under Saddam there weren't such differences between Sunni and Shia. Al Maliki is worse than Saddam'. 3 Or another, who argued: By God, Saddam was a tyrant. But at least he repressed everyone the same. One knew what to expect. At least there was some sort of equal system. Saddam once came here. In two cars, no security convoy, no army, nothing. Al Maliki wouldn't even dare to come here with half of his army. Because this confessional government of today doesn't do anything else than oppress the Sunnis. 4 The concentration of power among the Shia also translated to a domination of the security forces, who often behaved like foreign occupants, harassing local Sunni populations and heightening the sense of victimhood. A vicious cycle developed, whereby increasing attacks against security forces in Sunni majority areas were answered by mass arrests of young Sunni men. Meanwhile Sunni extremist forces like the Islamic State in Iraq and the Sham (greater Syria), or ISIS, grew ever more powerful. Its attacks were increasingly directed against the civilian population, with car bombs in popular places in Baghdad becoming frequent, and the victims of this violence were mostly Shia.

Divisive Narratives
Following this trend, the security forces also became the subject of increasingly divisive narratives among the Sunni population, which went so far as to justify retributive attacks on the security forces. Indeed, even acts of terrorism were downplayed in such narratives. All that seemed to matter to many was that someone was defending them, or at least taking revenge on their perceived oppressors. That this 'revenge' was often carried out by a brutal terrorist organisation indiscriminately targeting Shia men, women and children seemed secondary, even though the collateral damage done by ISIS´ attacks on the Sunni citizens was also significant.
The following, for example, is from a community leader in the provincial Abu Ghraib district in Baghdad province. It was collected in December 2013, just a few days before a bigger clash was to happen: 'This army is not Iraqi, it´s a confessional army. The army of Al Maliki, not the army of the Iraqis. And everyone always talks about terrorism! Which terrorism?! There is no terrorism, that´s all just politics!' 5 It is arguable that geographic distance contributed to this statement, as the indiscriminate ISIS attacks took place almost exclusively in the more urban areas of Baghdad. Such statements, however, portray a division in Iraqi society that was widening as the enemy became the entirety of the 'other'; in this case, the Shia. As is common in such conflicts, de-humanising the 'other' served to justify attacks on innocent people (Vaes et al. 2012), as was occurring more frequently.
We can also see this divisiveness in verbal attacks against the way of life of the 'other'. The traditions of the Shia population, for example, started to be viewed with suspicion and were even ridiculed by members of the Sunni population. Traditions which had been essential elements of Iraqi society and its cultural heritage became something alien, illogical and unintelligible in the 'otherizing' narratives of the Sunni minority. Previously innocuous distinctions between Sunni and Shia traditions became flashpoints for contention. One example regards the Arbaeeniya, the mourning period for the death of the Shia-Imam Hussein. As one Sunni interviewee stated: A few days ago my cousin and his friend went to Baladiyat (a Shia-majority neighbourhood in Baghdad). They started joking and laughing about something and a man with a black beard came, shouting whether they were laughing because Imam Hussein had died. 6 By the second half of 2013 protest camps had been erected in Sunni strongholds West of Baghdad. Only a spark was needed to light the fire. At the same time the narratives voiced by Sunni interviewees indicated that some kind of violent change was about to occur. It is not hard to understand that many welcomed this thought. As one interviewee stated: While oil is bringing money without end to the government, there are no services for the people here. Every year there are inundations but not enough drinking water. No functioning security apparatus and yet they lock up our sons for years without reason. The State says it´s fighting terrorists, but lets the militias do whatever they want. By God, this will not always stay like this. Something will have to change or else this country will break apart and we will fall back to the days of darkness. 7 The last sentence of this statement clearly expresses the idea that even change with an uncertain outcome was viewed as a better option than the status quo. Unfortunately, the change that was soon to come was even worse than the status quo of late 2013.

Violent Narratives
There were, of course, others amongst the Sunni population who were worried about the growing movement, concerned about the cult of violence. Like a Sahwa militia leader in one of the hot-spots of Al Qaeda activity in the countryside near Baghdad, who expressed his dread at what was going to come: These God-less people of Al Qaeda are destroying our societies. They come with customs not related to ours, not even to Islam. They only know death, they only want to fight and kill. More and more of our sons join them, thanks to the damned government. My nephew is with them. Once we managed to persuade them to a meeting. Me and the other Elders. We tried to explain them that a destroyed checkpoint will only be rebuilt and that three new soldiers will follow one dead one. Death will only be followed by death. But they didn't care. They said their day would be coming soon. 8 Then, in the last days of 2013 security forces attacked the home of an influential Sunni parliamentarian in Ramadi, arresting him and killing his brother. Soon after, they moved to dismantle a protest camp in nearby Fallujah, about 30 kilometres from Baghdad. It is possible that the Al Maliki government was seeking to take advantage of the holiday season in the Western world to crack down on its opponents. However, it was quite obvious that others had been waiting for this, the last provocation. Obviously well organised violent attacks were staged in Falluja, Ramadi and other smaller cities just across the Euphrates river from Baghdad, where tribes united in the Tribal Revolutionary Council rose openly against the government. The army replied in force and soon frontlines were established just a few kilometres west of Baghdad.
When interacting with the Sunni population in areas close to this frontline a sense of relief was recognisable. Finally, the longed-for change had come. Their 'brothers' had stood up against the oppressors. A sense of pride was rising amongst the Sunni population, and this pride seemed to blind people to the fact that the terrorist fighters of Al Qaeda were amongst the rebel forces, if not, in fact, in the lead from the beginning. In their narratives, however, interviewees emphasised the fact that the uprising was led by the tribal population of Anbar. The formulation 'sons of the tribes' was commonly used in this context and became a forceful part of the narrative of this first chapter of the 2014 uprising. Interviewees even ridiculed the idea of the uprising being connected to terrorism by equating Al Qaeda/ISIS/Daesh with the Shia militias. As one interviewee stated, "Daesh! Which Daesh? All of them sons of the tribes. Sunni tribes. They are only in their right! Which Daesh -you mean Maesh!" 9 (Maesh, as an abbreviation of 'militias of Iran in Iraq and the Sham').
It is not completely clear when and how Al Qaeda/ISIS gained control of what started largely as a popular Sunni uprising in reply to the actions of a repressive Shia-dominated government. But by June 2014, when a few hundred ISIS fighters drove tens of thousands of Iraqi security forces out of Mosul and swept soon after through all of northern Iraq right up to the northern borders of Baghdad, it was obvious that the Iraqi Sunni population was no longer in control. At this point all restraint was lost. The sense of the 'other' and of its dehumanisation, in effect, reached its culmination. Hundreds of captured Shia soldiers were massacred in cold blood and in triumph.

Nuanced Narratives
Not all Sunni were able to understand these developments. During a conversation in Abu Ghraib district, for example, many participants expressed confusion in response to the fall of Mosul in June 2014. Confusion existed, for example, about the identity of the fighters perpetrating these acts of violence: 'Where are tribes? Has anyone said anything about tribes? Those are all foreigners with Daesh' said one participant. 10 "Nonsense, without the tribes Daesh can´t do anything. Everyone knows that! The tribes control everything. Daesh is only on TV", said another. 11 'In reality this is all orchestrated by Al Maliki himself. So that he can incite the Shia against us. From the beginning this was planned by him!' said a third. 'Better Daesh and the tribes than the Maliki-army!' 12 In their immediate reaction to the violence, the population in the northern Sunni provinces welcomed the invaders, taking them at first for tribal fighters liberating them from Shia occupation. By the time they realised who had actually come, it was too late. Triumph changed to disillusionment and the narratives, in turn, shifted from triumphant to critical, or even regretful. In July 2014 (a month after the ISIS invasion), a local journalist and Internally Displaced Person (IDP) from Tikrit exhibited this quick reversal of narratives: Yes, most people welcomed the fighters. They thought Daesh were tribal fighters coming to save us from the Shia repressors. They were actually doing just that, only that they were in fact Daesh. I could see that soon. Many of them I knew. Some since I was young. With some of them you already knew back then that they were going to become Al Qaeda. And after they became Al Qaeda they of course became Daesh. Soon we realised they were all just criminals. 13 It seems to have been a slow realisation that the invaders were not led by tribal heroes fighting for a just cause, but by brutal terrorists fighting for power. On 29 July 2014, ISIS´s leader, Abu Bakr al Baghdadi, declared a worldwide caliphate and renamed the movement Islamic State (IS), thus leaving no doubt who was in charge, where this 'uprising' had led, and what it was now about. This realisation greatly impacted the popular narratives as it made the situation clear to the urban population. As one interviewee stated: There [in bigger cities like Mosul and Tikrit] people are not used to this style of life that they are forcing upon us. A friend of mine had a DJ-shop in Tikrit. He had to run away like me when Daesh came. 14 On the other hand, the more conservative population in rural areas had an easier time accommodating themselves with the new rulers and their narratives seemed unaffected in the first period of the invasion. As one interviewee said: What problem should I have with them. All they want from me is that I live like a good Muslim. I have been doing that all my life, so they are definitely better than these oppressors from Baghdad! 15 Others, who were not under IS rule but exposed to the actions of Shia militias, wished for IS, even despite the obvious brutality displayed by them. For example, a tribal leader in the Abu Ghraib district bordering IS-territory in August 2014, argued that: Most people would prefer to have Daesh, like in Mosul. Now with these [Shia] militias it´s even worse here. They are like the devil. Completely unpredictable, doing whatever they want and no one can tell them anything, not even the army. 16 It seems that brutality was not enough to deter a desire for their presence until it was witnessed directly.

Reconciliatory Narratives
But narratives broadly changed when the brutality became too hard to conceal and the scope of it was no longer justifiable for ordinary citizens. A few months after IS invaded, during the last months of 2014, many Sunni became disillusioned by the violence and narratives started to grow increasingly self-reflective. One interviewee stated, for example, that 'Many say that others are responsible for the mess. Some say Iran, some Saudi Arabia, some Qatar. But I believe in fact us Iraqis are responsible'. 17 Indeed, already at that time narratives started to focus on the more fundamental problems Iraq faced, and the antagonistic sense of the 'other' seemed to have moved to the background. One interviewee lamented how: The country is full of oil. But the people are not benefitting from that. That´s the real problem of the country. If all had a small part of that and a few greedy [individuals] wouldn´t eat up everything then no one would care about who is Sunni or Shia. Then there wouldn't be any Daesh and Abu Bakr Al Baghdadi would be sitting somewhere alone in a cave like Usama bin Laden back then. 18 Other interviewees followed suite. One said that 'I believe most people are good, but the issue is like with a glass of water. One drop of poison and everything is poisoned'. 19 This new reconciliatory tone emerged organically, as if the excessive violence had finally woken people up. Coincidentally the same local tribal leader who had earlier lamented Shia rule and Al Maliki's power (see footnote 3) stated almost one year later that 'in our culture we say when the Sunni is in peace, the Shia is also in peace'. 20 As the brutal reality of IS rule became ever more obvious, so narratives continued to evolve in this more reconciliatory direction. It had become clear that violence and antagonism had led nowhere good. The realisation that their uprising had been hijacked for a completely different purpose finally tipped the balance for Sunni citizens.
These reconciliatory narratives were further strengthened by the call to arms by the Grand Ayatollah of the Shia Marjaiya following the fall of Mosul in June 2014 and the subsequent formation of Popular Mobilisation Units (PMUs). Even though the call was mostly answered by Shia citizens of Iraq´s southern provinces, it was directed at all Iraqi citizens, and it proved an important step in encouraging reconciliatory narratives. In this call, IS was portrayed as an external enemy that targeted both Shia and Sunni populations and the Marjaiya called on all Iraqis to fight against it. In its wake many of the cities occupied by IS fell swiftly, until finally the ultimate battle for Mosul, Iraq ´s second biggest city, was decided towards the end of 2017. There was some fear at this point as the Shia militias that had played a brutal role during the sectarian violence of 2006-2007 were instrumental to these victories. As one high-ranking Iraqi army commander noted in November 2014: In the long run there is a problem. Some do come to join our [Iraqi army] structures but no one here knows what to do with so many unexperienced recruits. Therefore, many of them go directly to the [Shia] militias, and what they are capable of doing everyone knows . . . 21 However, as another high-ranking Iraqi army commander stated 'One can´t hold the whole Sunni population of the country to account', 22 and the Iraqi army countered fears of retribution against the Sunni population by deploying the militias and PMUs as far away as possible from the Sunni population. Overall, the experience of IS's repressive rule, and its reframing as an external power antagonistic to Iraqis, both Sunni and Shia, helped to shift narratives away from antagonism and towards reconciliation.

Unifying Narratives
Since the fatal year of 2014, Sunni and Shia have replaced antagonistic narratives with narratives of national unity. In their approach they seem selfreflective and understanding, as expressed by one influential representative of the Shia Marjaiya in April 2018: If my brother and father had been killed by our security forces, I myself would definitely also have joined Daesh. 23 An example of this newfound unity was the exchange of visits by delegations of important religious and civil society representatives from Ramadi, Najaf and Mosul between January and May 2019. Ramadi, as the capital of Anbar Province, was one of the first places where the uprising in 2014 took hold, Mosul was the informal capital and largest city of the IS-Caliphate. Najaf on the other hand is the holy city of the Shia confession, the 'capital of the Shia', as one representative from Mosul phrased it. All Najafis asked said that they had not been to Ramadi or Mosul at least since 2013. But the fact that they might be targeted by extremist cells in these cities did not deter them from their mission.
One memorable interview at this time was with a Shia religious representative who stated that he had studied in Mosul in the seventies. He went on to say that years later, in the nineties, he was arrested and tortured, along with his brother and father. His brother and father ended up being executed in front of his eyes. When asked how he was able to return to Mosul and talk about reconciliation in a city that had staunchly supported the Saddam-dictatorship, he simply said: 'Those back then were Saddam´s men -and these people today have nothing to do with that. I know how to establish that difference '. 24 The Sunni representatives from Ramadi and Mosul were similarly apprehensive before visiting Najaf. Likewise, most had never been to Najaf before. One participant from Mosul, referring to the fact that her last name was the same as one of the first Sunni caliphs fighting the first Shia Imams, shared the following story: When the guests from Najaf came to Mosul, I was scared of mentioning my name to them. And this is my city! However, when I arrived here in Najaf, I was greeted so warmly that now I want to tell my name to everyone. To show that we are one country, no matter our names, our confession, religion or ethnicity. 25 Whether by coincidence or by design, the messages passed in all three meetings (in Mosul, Ramadi and Najaf) were concrete, problem-oriented and seemed to echo the change towards unifying narratives, which focused on economic issues and the government's efforts to remedy these problems.
Addressing an audience of university students and civil society representatives at Ramadi University during the event, an influential civil society representative from Ramadi argued that 'There is no real problem between Iraqis, our problems are mainly economic . . . Where are the efforts of the government to strengthen reconciliation in our society?' 26 One student and youth leader from Mosul university explained: You know, what happened was a trauma for everyone in Mosul. However, there was a negative and a positive side. On the one hand many were killed, many became widows or orphans. On the positive side, this catastrophe united us. Daesh targeted all of us, not just one sect or one religion. So, we all had to be together against them. 27 It seemed in 2019, five years after the events of 2014, that there was a willingness within Iraqi society, amongst both Sunni and Shia, to reject sectarian conflict and present unified narratives by focusing on issues of concern for all Iraqi citizens. There seemed an awareness that societal unity is an urgent matter. As one civil society representative in Ramadi stated; 'We need compromise and we can´t only stay in our place. Otherwise sooner or later conflict between us will return'. 28

Discussion and Conclusion
In the above analysis of the narratives heard mainly among members of the Sunni confession in Iraq, we clearly witness an evolving interpretation of the relationship between the Sunni and Shia communities. As described, prior to 2013 the Sunni community had already started to define itself as victimised by the newly dominant Shia majority. They came to identify the Shia as empowered and emboldened by the US intervention, and as using their newfound power specifically to undermine the Sunni community. These perspectives formed part of a victimhood narrative. Following this came quotes which not only described those of the Shia faith as the tormentors of the victimised Sunni community, but started to identify Shia as inherently dangerous, as violent, and even as terrorists. This form of divisive narrative served to de-humanise the enemy and set the stage for violence to erupt if and when sparked by a catalyst event.
This event came on the last days of 2013, when, as described above, the Al Maliki government set to work dismantling Sunni protest camps, which led to the rise of a violent revolution in Iraq's western province of Anbar. This revolution was initially led by the Tribal Revolutionary Council and was seen by many Sunni as a rightful response to the violence and indignity forced on the Sunni community by the Shia majority government. As such, in the articulation of violent narratives, the Sunnis interviewed defended this revolution as carried out by the 'sons of the tribes' and defied claims that this was an external intervention led by Al Qaeda. The evolution of these narratives, from one focusing on the victimisation of Sunnis, to then dissociating the Sunni from a de-humanised Shia, and eventually to one which openly supported a violent revolution against the Shia government, evidences how such narratives relate to each other and evolve over time, and how they shift in response to historical political and economic events. But such dynamics are evidenced also in the evolution towards more conciliatory narratives.
As described, once the role of external actors in the form of Daesh fighters in the 2014 violence became apparent, a new set of more nuanced narratives started to emerge. These narratives started to deconstruct the clear 'us' versus 'them' binary evident in the victimhood, divisive and violent narratives, and instead started recognising that there was some diversity within the Sunni perspective or position regarding the revolution. The extreme brutality of IS throughout 2014 led to a call to arms issued by the Grand Ayatollah of the Shia Marjaiya in June 2014, which was heard by both Shia and Sunni as a call to defend Iraq against this external 'other'. The declaration of the IS caliphate by Abu Bakr al Baghdadi from the city of Mosul on 29 July 2014, was a turning point; a moment when it became clear exactly who would be governing this new political establishment and how. Many Sunni started to question the movement at this point. While some felt that rule by IS was still better than by the Shia, it was at this point that more reconciliatory narratives started to emerge, with some now turning to recognise Daesh as the 'other' and both Sunni and Shia as Iraqi victims of this externally driven movement. This, in turn, evolved into the more unifying narratives described in the final section.
The evolution of these narratives follows the classic model of the conflict cycle (see Figure 1), with the first three forms of narrative (victimhood, divisive and violent) on the left or conflict escalation side of the curve, and the second three forms of narrative (nuanced, reconciliatory and unifying) on the right or conflict de-escalation side. The top of the classic model is represented by the '(Hurting) Stalemate'. As usually articulated, the 'mutually hurting stalemate' is the point at which all parties to the conflict realise that they are locked in a conflict from which they cannot benefit more than they might from peace (Zartman 2001). This is the moment when both parties are more likely to be open to being drawn into the process of engagement, brought to the negotiating table, and to make concessions which might bring an end to violence. While the mutually hurting stalemate is too simple a frame by which to analyse the de-escalation witnessed in the Iraqi narratives, it does nonetheless highlight the importance of a point of inflection; a critical moment of change that can inform potential intervenors of the ideal time to act.
In the Iraqi case this critical moment appears to have been two-fold; first, the call to arms issued by the Grand Ayatollah of the Shia Marjaiya, and second, the declaration of the IS Caliphate under Abu Bakr al Baghdadi. These two moments served to drive home the realisation among many Sunnis that the violence was not in their favour, was not defending the 'us' against the 'other', and was, in fact, undermining a more general Iraqi security, identity and sovereignty. They were the first cracks in the foundation of the older conflict promoting narratives, and the first steps towards the more conciliatory narratives that would replace them and initiate deescalation of the conflict (Kriesberg and Millar 2009: 24). The case illustrates the importance of these pivotal moments as it evidences how such turning points emerge within conflict dynamics and highlights how critical it is that intervenors be ready to support the transition to more pro-peace narratives as they emerge. Key questions remain, however, which include: 1) How can external actors assess if and when such critical moments will occur? 2) If a critical moment appears to be occurring, can external actors encourage parties to take advantage of them? 3) If such a moment is not occurring, is it ethical for external actors to promote them? Answering each of these questions requires more data from more cases. But this case seems to hint at some answers. First, this case indicates how difficult it is for international actors to assess conflict narratives in real time. Scholars are very good at assessing historical narratives post hoc, but current or contemporary analysis is more challenging. You cannot know what actions will give rise to a critical moment until the moment has passed, either having resulted in substantive change or not. It is also very dangerous, therefore, for external actors to try to force parties to respond in a particular way to an event which the external actor believes should be a pivotal turning point but may not be. For the same reason, it seems unethical for external actors to try to engineer such moments; particularly if such engineering carries the risk of aggravating the conflict itself and of reifying further antagonistic narratives. Instead, this seems another case where external actors are best placed to serve as facilitators and assistants as opposed to drivers of peace.
When shifts in the nature of narratives emerge, international actors can certainly provide the necessary skills, capacities, and resources to help local actors take advantage and further develop more conciliatory narratives. But, importantly, in order to be available, to be trusted, and to be there when they are needed, conflict resolution, conflict transformation, peacebuilding, or reconciliation organisations must already be engaged, they must have built rapport and trust with local communities, and they must understand the local dynamics of conflict enough to play a supportive role in taking advantage of such moments. This requires, therefore, a long-term commitment to support the parties to the conflict, and particularly actors within the setting who are interested in and capable of supporting peace and of identifying the right moments to act. In short, identifying and acting on these pivotal moments must be locally driven and externally supported.
In conclusion, the data described above both provides new and interesting insights into the subtle evolution of conflict and peace related narratives within a particular context, while also reaffirming some of the central theoretical insights in the field regarding othering, the importance of narrative, and the more recent focus on bottom-up or locally driven CR processes (Lederach 1997, Mac Ginty andRichmond 2013). It also serves to highlight a more subtle or complicated issue for international CR actors, and this is the balance that must be struck between hands-on and hands-off engagement or between being proactive and being reserved. It also indicates worryingly bureaucratic issues regarding funding and impact assessment. How exactly can external actors acquire the funds and support to stay engaged and prepared on the ground while ostensibly waiting for ripeness to occur? Will funders require activity that is more directly hands-on or proactive, that can be assessed as impactful, or are they prepared to support CR institutions simply to be ready to facilitate activity from the bottom-up when it becomes possible? These are further questions for future research and analysis.

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