Beyond Counterinsurgency : Peasant Militias and Wartime Social Order in Peru ’ s Civil War

In much of the academic literature on contemporary militias, the focus is typically on their destructive anti-rebel character. By contrast, the perspective of militias as agents of local governance, social reconstruction, and positive transformations is one that to date has been under-researched. Taking a “relational” perspective, this article examines how peasant militias in Ayacucho Department, although initially formed for the purpose of violently opposing Shining Path rebels, became engaged in governing their own “wartime social order” in which they organised, coordinated, regulated, and signified activities and behaviour for the collective good of their local communities. From it we might gain insight into how these peasant militias were able to avoid permanently becoming the predatory sort of militia that much of the academic literature warns about.

Once a neglected topic in academic research on civil wars, scholarly interest in civilian militias, self-defence committees, or civil-defence patrols, as they have variously been labelled, has burgeoned in recent years.A wealth of new research on the subject has enriched our understanding of the forms, functions, contexts, dilemmas, and dynamics of active civilian participation in the violence of civil wars (Jentzsch et al. 2015).From this new body of research we have gained a deeper comprehension of how militias are formed and mobilised (Blocq 2014;Schubiger 2016;Fumerton 2001).We now have clearer insight into the dynamics of their relationships with rebel actors and with the state.We understand better why and how governments use militias for counterinsurgency purposes (Carey et al. 2015).
Militias, it is commonly argued, commit (or are extremely prone to committing) human rights abuses (e.g.America 's Watch 1992;Human Rights Watch 2007a, 2007b;Human Rights Watch 2011;Gourevitch 1988).Some critics suggest that perpetrating extreme violence is part and parcel of the evolutionary process of militias: "whether the militias are established by government or not, governments end up not being in control of the militias because militias develop lives of their own, taking on the predatory ... instincts of the society and government" (Francis 2005b, p. 4; also Aliyev 2016).It is further argued that militias can, in similar fashion to death squads, serve as "subcontractors" in the service of state terror: "This subcontracting allows "legitimate" state authorities (and powerful private actors) to avoid association with atrocities committed on their behalf (Jones 2004, p. 133; also see Warren 2000, p. 235-239).
Furthermore, while the counterinsurgency effectiveness of militias is widely acknowledged, it is often made in conjunction with concerns about what this (frequently abused) "delegation of violence" can mean for the state's monopoly of legitimate violence (Eck 2015;Jones 2004;Mandel 2001;Francis 2005b).Militias are perceived to undermine the state's monopoly of legitimate force, and in so doing (further) weaken state authority (Mandel 2001).Scholars like Francis "argue that the majority of these weak states never had the capacity to exercise domestic sovereignty in the first place " (2005b, p. 20).Hence, resorting to the use of civil militias in war-torn states is like "using fuel to put out a fire" (ibid., p. 19).Indeed, according to Yoroms, the more accurate reading of "the rise of Militia is [that it] is an indication of the failure of the state to address the fundamental issues of human and community securities, which ab initio the state was established to tackle" (2005, p. 31).Finally, the case is frequently made that militia are unreliable and unpredictable, and therefore "constitute a long-term threat to law and order" (United States Army 2006, p. 20, Ch.3;see Mueller 2004, p. 1).
We certainly must not overlook these valid critiques and concerns, many of which were manifested also in the actions of pro-government rural militias during twenty years of civil war in Peru.Nevertheless, it is important to recognise that much of the prevailing sceptical opinion about militias has created an imbalanced understanding of the phenomenon in that it is their "destructive" na-ture that is emphasised, while their socially "constructive," "transformative," and "constitutive" aspects and potentials often go unnoticed, unelaborated, and untapped. 1Comparatively less is known about the significance of militias beyond their relations with states, counterinsurgency tasks, or abusive dealings with the local population.In particular, the social order they bring about, and their aspects as agents of local governance, social reconstruction, and positive transformations within their own communities, have been under-researched (cf.Theidon 2006;Hoffman 2007;Blair & Kalmanovitz 2016).Whereas much of the research has focused on the disorder that militias are often observed to bring, any normalising order that they shape garners much less attention in comparison.
Taking as its point of departure Arjona's observation that "understanding why order emerges in war zones and what form it takes is an important question, as it relates to civilians' experiences of war, armed groups' strategies, and wartime transformation of key aspects of society" (2014, p. 1384), this article examines the case of pro-government rural militias -officially designated as "Comités de Autodefensa" (self-defence committees), and popularly known as "rondas campesinas" (peasant patrols) -in Ayacucho Department, Peru, during the civil war years between 1982 and 2000. 2 The central puzzle it addresses is as follows: Created primarily to perform an anti-rebel security function, how and why did Ayacucho's self-defence committees expand their role to create a wartime social order in which they governed for the collective good of their local communities, particularly during the long transition from civil war to a post-violence society?
By investigating this puzzle we might ultimately gain valuable insight into how and why it is that Ayacucho's peasant militias were able to avoid settling permanently into the role of a predatory and destructive sort of militia that much of the academic literature warns about.

Conceptual framework, method, and data
Militias have been defined in a variety of ways, but mainly in terms of their association with the state and their paramilitary character (e.g.Carey et al. 2013).For Jentzsch et al. militias "…do not require a recognizable link to the state" in recognition that militias can shift their loyalties or follow their own interests that can be contrary to that of the state."The critical issue," they assert, "… is their anti-rebel dimension" (2015, p. 756).The definition of "militia" that best fits my empirical data and analysis is one that I have modified from Seth Jones: "A militia is as an [organised, irregular, armed group drawn from the civilian population] that performs security and governance functions within a state.The primary goal of a militia is population control, especially the establishment of local rule in a given territory" (Jones 2012, p. 4).Unlike most definitions of militia, Jones's formulation mentions a "governance function."This is important, for in the current literature the role of militias in set-ting up local forms of governance is one of the least explored aspects of militias (Jentzsch et al. point 2015, p. 758).
To be clear, I am not suggesting that it is puzzling at all that militias engage in some sort of governance.That armed actors -militias, rebels, militias, rebels, and even criminal groups -also govern, and therefore create some form of order, should not be surprising.Governance is not the exclusive prerogative of the state (e.g.Clements et al. 2007;Boege et al. 2009).The notion of "hybrid political orders," for example, takes a broader understanding of governance provision as involving both state and non-state actors (Boege et al. 2009, p. 17), and has given rise to diverse studies in "criminal governance" (e.g.Arias 2006), "police governance" (Dodsworth 2004), and in "rebel governance" 3 (Arjona 2017;Arjona et al. 2015;Mampilly 2011;Péclard & Mechoulan 2015;Staniland 2012).
Rather, what this puzzle seeks to understand is contingent relational change that shapes institutional behaviour: specifically, how and why it is that shifting actor interactions brought about significant changes in militia behaviour from destructive to more constructive, and in the process shaped a particular social order that was conducive to mending the social fabric of war-torn rural communities in parts of Ayacucho Department.As such, my analysis draws inspiration from Ana Arjona's (2014) theoretical and analytical discussion of the concept of "wartime social order," which she defines as "… the set of rules that structure human interaction in a given community during wartime, allowing for …predictability [i.e.people's expectations about what might happen] to exist (2014, p. 1374).Arjona's research focus on "order" and "institutions" in war zones offers me a stimulating conceptual lens through which to focus on a neglected aspect in militia studies.
The empirical basis of this article is evidence gathered during a total of thirteen months of anthropological fieldwork and archival research in Ayacucho Department in 1997Department in , 2000Department in , and in 2009, and , and from secondary literature such as the Peruvian Truth and Reconciliation Commission's Final Report.A qualitative research strategy was adopted involving the triangulation of data from indepth open-ended interviews, participant observation during militia and community activities, and evidence from a variety of document sources (including a militia commander's personal diary, and regular situational reports written by militia committees for the commanding Peruvian officers of local military outposts) (Fumerton 2002).
During fieldwork, I lived for nearly ten months in the rural district of Tambo, where I interviewed militia officials and rank-and-file members from the District and neighbouring ones, and also participated in important community events (such as civic action programmes, and communal cleaning of irrigation canals).A non-probability sampling method was used to select interviewees according to "particular features or characteristics which will enable detailed exploration and understanding of the central themes and puzzles…."(Ritchie et al. 2003) that I wanted to study in regard to the phenomenon under investiga-tion.These included senior-level self-defence committee members, members of community-level juntas directivas, senior Peruvian military officers and civilian political officials, members of local "mothers' clubs," and knowledgeable individuals from various other institutions represented at the local level, such as hospital personnel and municipal government bureaucrats.
During my PhD research I used the evidence I collected to construct an "event catalogue": "…a set of descriptions of multiple social interactions collected from a delimited set of sources according to relatively uniform procedures" (Tilly 2002, p. 249).In the process of revisiting this body of empirical evidence and supplementing it with data from contemporary and more recent secondary literature, Arjona's theoretical and conceptual ideas regarding "wartime social order" have helped me to identify the themes and questions by which to structure my analysis of the relevant empirical data available to me.The empirical richness of my event catalogue also permitted me to describe in detail the key attributes of the wartime social order that emerged out of the diachronic and dynamic relational interplay between Shining Path rebels, the Peruvian military, and the rural population and their local self-defence institutions.
While Arjona gives me an conceptual lens through which to describe, analyse, and make sense of what the functions, characteristics, structures, and dynamics of militia local governance looked like within this broader wartime social order, it is from Tilly and Tarrow's "mechanisms-process" analysis that I draw to explain how and why social order shaped up that way."Mechanisms" are "events that produce the same immediate effects over a wide range of circumstances" (2007, p. 203).Mechanisms compound into "processes," which are "combinations and sequences of mechanisms that produce a specific outcome" (ibid.).Like Elisabeth Wood (2003), I have attempted to infer my identification of mechanisms from the data.
This article takes a "relational/interactive" approach to the subject, which fits very well the contingent and diachronic nature of the evidence I have assembled (Tilly 2005, p. 19;Kalyvas 2006;Wood 2003).The benefit of combining the conceptual and analytical ideas of Arjona with that of Tilly and Tarrow is that it provides me with the analytical tools (and conceptual vocabulary) to clarify -through descriptive and explanatory accounts -what, how, and why shifting actor interactions have shaped a particular social order.In this way, I hope ultimately to further our understanding of how Ayacucho's peasant militias become agents of constructive social change and reconstitution.

Relational dynamics of violence: Ayacucho, 1980 -2000
The evidence strongly suggests that the emergence and shape of constructive militia behaviour in Ayacucho, and the social order they engendered, were influenced largely by their interactions with state and rebel actors, and by the nature of their embeddedness within the institutional fabric of their own communities.
A detailed chronology of the overall history of civil war in Peru is not required for me to describe the case in support of my argument; general accounts can already be found elsewhere. 4I will instead give a more selective account of the civil war in Ayacucho from a relational perspective, emphasising the dynamic and diachronic shifting of interactions between the main actors.

Rise and deterioration of campesino-Shining Path relations
In the first two years after the Maoist-inspired Shining Path launched a "People's War" on 17 May 1980, the organisation managed to build up a substantial degree of popular sympathy among the rural population (CVR 2003, p.47, Vol. IV;Gorriti 1999;Degregori 1996bDegregori , 1990;;Starn 1995;McClintock 1984).The police's clumsy, heavy-handed attempts to suppress the rebels were largely ineffective.Attacks on police stations escalated between 1980 and 1982, forcing the police to abandon most of their outposts throughout rural Ayacucho.This "generated a power vacuum in the countryside and left in this way a favourable terrain for converting the communities of this zone into [Sendero's] support bases" (CVR 2003, p. 47, Vol. IV).
It was in the second half of 1982 that we begin to see the first ruptures in Shining Path's relations with the campesinos; for the Senderistas came increasingly to resort to violence as they tried coercively to transfigure Andean reality into the communist utopia envisaged by their leadership (Degregori 1987;González 1984).Most provocative of all was their systematic campaign to erase all traces of the state's political authority in the countryside by killing local government officials and community leaders, and then replacing them with their own commissars (Coronel 1996, p. 47;CVR 2003).This roused enormous resentment towards Shining Path, particularly from the victims' kin (Coronel 1996, p. 47;Degregori & López Ricci 1990, p. 332).Whatever degree of widespread sympathy and legitimacy Shining Path had enjoyed until that point was quickly eroded.While violent intimidation worked to keep some peasants at heel, others chose instead to take steps to defend themselves (Degregori 1996a;CVR 2003, p. 288-292, Vol. II).

Privately driven violence within a civil war's master cleavage
In order to appreciate the social context in which the rondas campesinas would later play a crucial role in facilitating the reconstitution of a shattered social fabric (including the reintegration and reconciliation of former Shining Path rebels into local communities), it is important first to understand how the civil war's "master" cleavage created political opportunities for pursuing privately driven violence.
Many of Shining Path's rank-and-file militants and sympathisers came from rural communities and families that had pre-war rivalries and conflicts with other families, and other communities (Taylor 2006, p. 180).Just as Kalyvas (2003) theorises, the civil war's "master" cleavage became intertwined with local conflicts in which Shining Path's militants, their kin, or their friends, were personally involved (see Theidon 2006;Isbell 1978Isbell , 1992)).
During the 1980s in particular, self-defence groups also became involved in this privatisation of political violence as yet another violent actor through which to pursue inter-communal conflicts.With the tacit approval or active participation of government soldiers, some peasant communities used their militias to raid neighbouring communities, at times "with the intention to plunder, rather than to seek members of PCP-SL" (CVR 2003, p. 289, Vol.II; also see Degregori 1987, p. 49;Isbell 1990, p. 11-12;Americas Watch 1992, p. 9).Taking the cue from their military mentors, peasant militias summarily executed those they declared to be Senderistas, or their sympathisers (CVR 2003, p. 289-290, Vol.II).Any reluctance to organise a village militia was promptly interpreted by the security forces as tacit support for Sendero Luminoso (Isbell 1990, p. 11-12;Americas Watch 1992;del Pino 1993, p. 78).Neutrality was not permitted.
The impulse for a militia's formation was often an important factor in its relations with neighbouring communities.In parts of Ayacucho Department where the communities themselves took the initiative to create their own village defence forces -as in the Iquichan uplands of Huanta province, and in the districts of Vinchos, Ocros, and Concepción -formal pacts of cooperation to provide mutual defence were often negotiated between the various communities, thus lessening the risk of violence between the rondas of adjacent neighbours.By contrast, there are some known cases of communities with protracted pre-war rivalries (particularly land-boundary disputes) who, being forced by government soldiers to form a militia, later used their rondas to attack each other (see CVR 2003, p. 115, Vol. VII;DESCO 1989, Vol.1, p. 116-117).
For these reasons, it would be a mistake to view peasants as merely the victims of political violence.Rather, in the course of the war they also tried to take advantage of emerging opportunities to manipulate the political violence in pursuit of private ends, to settle prior protracted vendettas.Denunciation to one or the other armed actor thus became for peasants an opportunistic mechanism by which to settle old scores.Private vendettas and rivalries, protracted land disputes, a squabble over inheritance, or just plain envy, were some of the underlying reasons behind formal and seemingly politically motivated denunciations between neighbours, and even between relatives (Fumerton 2001, p. 475).
By the end of 1982, the social fabric within and between many of the rural communities began to tear apart. 5A cycle of inter-communal and fraternal vengeance gripped many districts of Ayacucho (e.g.Theidon 2000).But worse would come when the Peruvian Armed Forces entered the fray.

Campesinos and the military: shifting relations and uneasy alliance
With the police incapable of suppressing the insurgency and restoring order in Ayacucho, the government declared a state of emergency over the entire department. 6Army and Marine Infantry units were deployed to Ayacucho Department early in 1983 (Tapia 1997).
The military's entry onto the stage brought new dynamics to the political conflict.Distrusting the population, and frustrated by having to fight an elusive and cunning enemy that hid amongst the civilian population, the soldiers unleashed a campaign of indiscriminate and brutal violence on the inhabitants of the department (Tapia 1997).Vast areas of the countryside were depopulated, and thousands of peasants fleeing the violence were forced to resettle into nucleated IDP settlements (CVR 2003, p. 289, Vol. II). 7  Although Sendero's practice of killing ordinary peasants and traditional local leaders was what initially provoked some peasant communities to organise self-defence groups, these mobilisations were largely "isolated actions, local and uncoordinated" events (CVR 2003, p. 289, Vol. II. My translation).The principal catalyst behind the extensive creation of civilian militias in the rural areas of Ayacucho Department between 1983 and 1985 was, in fact, the Armed Forces.
Throughout the rest of the 1980s, the relationship between the peasantry and the Peruvian Army remained a rocky one, for many military commanders and soldiers remained sceptical about the loyalty of the indigenous population.They were opposed to the idea of permitting peasants to form self-defence militias.Even so, by cooperating in joint patrols with soldiers, the self-defence militias helped little by little to improve relations and thereby diminish some degree of the army's "presumption of subversion." The election of a new Peruvian president in 1985 brought a new dynamic to the civil war.In the latter half of his presidency, Alan García (1985García ( -1990) ) and his centre-left APRA 8 party publicly vowed to prosecute any member of the security forces found to have committed human rights abuses (Tapia 1997, p. 40).On occasion, he also expressed what seemed to be sympathy and even admiration for the sacrifices and dedication of Shining Path cadres (see DESCO 1989, p. 173, Vol. I).Naturally, the military came to perceive García as an unsympathetic president, or even as hostile (Tapia 1997, p. 41;Crabtree 1992, p. 112).Consequently, in Ayacucho most Peruvian Army units discontinued their counterinsurgency activities for fear of incurring accusations of human rights violations (del Pino1996, p. 149).In other words, the army units remained in place, but they discontinued most of their active operations and patrols against the rebels.With government forces now content to sit on the sidelines, many rural communities that had mobilised self-defence committees chose to disband them, and accepted or accommodated to the resurgent domination of the insurgency in their midst.
Generally speaking, relations between the peasant militias of Ayacucho and the Peruvian Army would remain strained and inhospitable until the end of the 1980s.On the other hand, the estrangement of the Army from the peasant militias also loosened some of the domination of the former over the latter, creating spaces for action and thereby opportunities for rural communities to assume greater ownership of their rondas campesinas.

Organisational Structure and Membership Participation
To understand the characteristics of the wartime social order that rondas campesinas established in their communities, it is first important to reemphasise two relational processes described in the previous section of this paper.First, Shining Path's strategy of executing or driving away state representatives and traditional leaders left a vacuum of political authority in rural communities.In communities that had organised for self-defence, this left the militia committee as the only remaining institution of authority (see Fumerton 2001, p. 491).The second important process to consider was when the Army withdrew their active support for the peasant militias, and their leading role in the counterinsurgency campaign, in protest of President García's unsympathetic attitude.The result of this change in behaviour was to lessen the degree of direct interference the Army previously exerted in local affairs.
Whereas military officers used to appoint the leaders of the self-defence committees, the withdrawal of the military from large areas of the countryside gave communities the space so they could now elect their own officials in open assembly.When they had the autonomy to organise themselves, self-defence committees drew from prior knowledge and experience and modelled their political organisation after the standard pre-war civilian village governance structure of the junta directiva, which ordinarily consisted of the elected public offices of president, vice-president, secretary of records, economic secretary, and a spokesperson.
It was in the Apurímac River Valley, late in the second half of the 1980s, that civil defence organisations known as the DECAS (Defensa Civil Antisubversiva) grew to become the most organisationally and tactically sophisticated, and the most militarily powerful, of all the militias in the Ayacucho-Huancavelica Emergency Zone (see CVR 2003, p. 290-291;del Pino 1996).Heavily assisted (and at times, closely directed) by a detachment of Marine Infantry that was posted in the Apurímac River Valley, the DECAS were able to organise the inhabitants of the valley into a single interconnected alliance of militias.This consisted of a vast network whereby the militia of each community was interconnected through a hierarchical, segmentary system of command and communication.The apex of this organisational pyramid was the regional committee.Below this were district committees, which in turn were made up of all the militia committees at the local level.The degree to which the system could be mobilised depended upon the gravity of the threat being faced at a particular moment: the graver the threat, the more of the segmentary system that would be mobilised.In this way, the DECAS and the Marines were able to push Shining Path almost entirely out of a sub-region of Ayacucho that previously had been a rebel stronghold for much of the 1980s.
Similar to the other self-defence committees found in Ayacucho's sierra communities, the organisational leadership of the Apurimac Valley DECAS consisted of a committee of elected officials: president, secretary, treasurer, operations commander, and intelligence officer (Fumerton 2002, p. 116).According to DECAS members I interviewed, every physically fit adult, men and women between 15 and 70 years of age, was considered in their peasant community to be a DECAS member, and participation in tasks and activities organised by the committees was regarded as a collective duty to the community.Each member was expected to participate in the tasks of patrolling and community defence, or labouring in public works like cleaning the village water reservoir, irrigation canals, repairing roads, and the like.Men had their specific tasks (mainly patrolling, guard duty, and heavy labour), and women and the elderly had theirs (cooking for the patrollers, mending their clothes, running small errands etc.).When the community was attacked, every able-bodied adult was expected to take part in its defence.
In an environment of social disintegration, it was important both symbolically and in practice to ensure a conscious recognition of the crucial importance of maintaining the solidarity and social order of the community; and instituting active and universal participation for the common good of the community was an effective way to achieve this.When deemed necessary, individuals would be coerced into active participation through fines or forms of corporal punishment.In Tambo District for example, absentees and malingerers were fined, or punished with extra work.In the District's militia archive of official documents, I found a copy of the kind of report that was routinely submitted by the ronda leader to the commanding officer of the local army garrison.It reads: "Verified the following: general registration of the population, armaments, communal faenas, and later proceeded to sanction the slackers who did not participate in the faenas of the programmed tasks, like cleaning their water reservoir and irrigation canal, the construction of their community hall [casa comunal] and latrines" (CCAD-SJB 1995, p. 2. My translation).

Local legitimacy of "rondas campesinas"
The ethos of universal membership, and participation in community tasks organised by its executive committee for the collective good, helped to underpin the militia's legitimacy in the eyes of its villagers, for the militia members were none other than the villagers themselves.Universal membership helped ensure that militia interests largely coincided with community interests.While it is true that some individual militiamen would have attempted to exploit their position within the militia organisation for personal gain, it is not contradictory to say that the system of universal adult membership nevertheless fostered a general sense of obligation to the interests and the very survival of the community in which every rondero was also a member.Most peasants I spoke to during my fieldwork remained sceptical of the extent to which soldiers and policemen, who are not locals, would really put their lives on the line to protect peasants.In this regard, when in a pinch, peasants felt that they at least could trust in their own village militia of which they themselves were the members.They were defending their own family, friends, and community.
Furthermore, as an institution that was modelled after the organisational structure of other local institutions, like the village's pre-war junta directiva, the peasant militias of Ayacucho could easily be absorbed and accepted as a new village institution in the life of the community.Performing such tasks as patrolling or keeping watch at night would not have been entirely new for the peasants, for prior to the war they carried out many of these same activities so as to protect their herds and property against livestock rustlers, robbers, and bandits. 9Many adult peasants would also have been used to having communal faenas -collective work parties -organised by a local authority structure, a responsibility that now was assumed by the self-defence committee.
However, evidence also suggests that in other communities where the local rondas campesinas had been imposed by the military or a neighbouring community upon an unwilling population, the inhabitants could purposefully marginalise their self-defence institution such that it would play a minor and weak role in local governance.

(Re)Constructing community in a time of war
One of the most interesting findings of the Peruvian Truth and Reconciliation Commission was that Shining Path was the principal perpetrator of killings and disappearances, of which peasants and local authorities were its major victims (CVR 2003, p. 56, Vol.I, Ch.1). 10 Peruvian and foreign observers "…had long thought that government forces were the main perpetrators of the violence…."(Landman & Gohdes 2013, p. 84; see Amnesty International 1996).
In a fascinating study of the politics of reconciliation in Ayacucho, Kimberly Theidon vividly describes how Shining Path's ruthlessly violent behaviour affected peasant perceptions."The Senderistas killed people in ways we do not even butcher our animals," was what the peasant villagers she studied often told her (quoted in Theidon 2006, p. 438).The image of the Senderista "enemy" constructed by the peasants therefore stripped them of their human characteristics, portraying them as having "fallen out of humanity," as people "who were only born to kill" (ibid., p. 442).Their demonization and dehumanisation of the Senderistas went even so far as to construct a discourse of radical anatomical "otherness.""We killed them and saw their bodies," one of the villag-ers told Theidon."They had three bellybuttons and their genitals were in another part of their body.I saw them" (ibid., p. 443-444).
It was against this particularly potent backdrop of discursive, moral, and emotional "boundary activation" 11 that the civil war entered a new phase, and with it important dynamic changes in relations between Shining Path rebels and peasant ronderos.As García's presidency and the 1980s drew to a close, the country was deep in economic and political crisis, seemingly imploding from its own flawed policies and leadership mistakes.The victory of Shining Path's revolution now seemed a very real and disturbing possibility.However, the election of a controversial but highly capable president, Alberto Fujimori, in 1990 would turn the tide against Shining Path.The military saw in Fujimori a sympathetic and competent commander-in-chief of the counterinsurgency campaign, which had stagnated perilously under García's leadership.In September 1992 Abimael Guzmán and many of his top lieutenants were captured in Lima, dealing a serious blow to the leadership and morale of his organisation while simultaneously uplifting the spirits of all anti-rebel forces by shattering Sendero Luminoso's image of invincibility.
Most importantly for the peasant militias, Fujimori turned out to be one of their staunchest supporters.Soon after coming into power the Fujimori government extended official recognition to the militias and permitted them the use of specified firearms, thereby bringing clarity and an end to political debates about their legality.This buoyed the will and confidence of more and more peasants to throw their support wholeheartedly behind the rondas campesinas project.
The result of these momentous events was that all across the country, Shining Path's forces (now under the command of their new field commander, "Camarada Feliciano") were steadily being rolled back, beaten on the battlefield, and cornered in their remote strongholds like the Sello de Oro and Viscatán.As fortunes turned against them, more and more Senderistas (often accompanied by their spouse and children) began to surrender themselves to the rondas campesinas.
Yet given the exceedingly virulent enemy image that peasants had constructed of the Senderistas, could Senderistas be transformed into people again?In order to rise above those attitudes and images of animosity, villagers would have to draw from their concepts and practices of communal justice (Theidon 2006, p. 445); and in the process, the rondas campesinas played a crucial role as an institutional vehicle for reintegrating and reconciling former Senderistas into local communities.

Reintegration and reconciliation of ex-Senderistas
"Arrepentidos" was that name that people gave to the Senderista deserters. 12 The first time a small group of rebels surrendered to militias in Tambo District was on a cold and misty Andean morning in December 2001.Comando Zorro, Tambo's militia commander, told me that the deserters surrendered because they could no longer bear the hardships of guerrilla life.All of them were suffering from exhaustion, malnutrition, and various illnesses due to years spent roaming back and forth between the frigid highlands and the humid subtropical ceja de selva.As more and more villages became hostile towards the rebels and mobilised militias, they found logistical support much more difficult to come by.
I was told in my interviews with ronderos and ex-Senderistas that rather than surrender to government soldiers or to the police (who they feared might summarily execute them), the arrepentidos instead gave themselves up to the peasant militias.They sometimes also brought their entire families with them.Theidon reports similar stories told to her during her research among the highlanders of Huanta Province, Ayacucho."They would arrive saying they had been tricked, forced to kill, always walking….'Pardon me,' they would beg the community" (2006, p. 446).
Moved by pity and long-standing traditions of forgiveness and communal justice rooted in Andean Christianity, peasant militias like the DECAS gave repentant rebels the chance to rehabilitate and integrate themselves into the community.Fuist's definition of "ideological performance" as "…how a performer's beliefs, values, and allegiances are displayed for an audience via her behavior, language, movement, use of props, and aesthetics" (2014, p. 430) neatly captures the ritualised way in which the arrepentidos showed and were made to show their contrition, repentance, and new loyalty to the community."[The]… authorities whipped them in public," Theidon was told."They were whipped, warning them what would happen to them if they decided to return [to being Senderistas].Whipping them, they were received here" (2006, p. 447).Participation and working hard in communal faenas was another way for arrepentidos to demonstrate the genuineness of their transformation.One of the most important rituals of repentance, however, was for the ex-Senderistas to prove their sincerity and change of loyalty by joining the rondas campesinas, and participating in their patrols and military operations."Of course we would always make them walk point in all the operations," a famous militia commander told the Lima newspaper Expreso (1989b, p. 4. My translation.).Not only was the point man the most dangerous position to be in an ambush, but by participating in the peasant militia an arrepentido was making a public display of identity and loyalty knowing full well that Shining Path reserved its most ferocious and unforgiving violence for deserters.
Arrepentidos were not always from the village in which they surrendered.Many came from other places, but through the ritualised rehabilitation and (re)integration process were adopted into the community."When they repented," Theidon's informant told her, "they stayed here as though they were from here.Qinan llaqtayarun [becoming fellow villagers], they stayed here and didn't go anywhere else.So they stayed and are here now, without going to the jungle, to Huanta, going nowhere.As if they were from here -they remain" (2006, p. 447).

Eliminating inter-communal violence by constructing new identities
In addition to playing an important institutional role in absorbing repentant exrebels into local communities, the rondas campesinas also helped eventually to eliminate the deadly internecine violence between villages that plagued Ayacucho's rural society in the first half of the civil war.What made this possible was the vast segmentary network structure that I had described earlier in this paper, which the DECAS devised and established firstly throughout the Apurímac River Valley, and then later diffused to other parts of the Ayacuchan sierra.In order to put an end to violent quarrels between communities, the juntas directivas of individual militias made official pacts of peace and collective defence with one another, which was repeated at the district level, and eventually at the highest level between autonomous regional committee headquarters (sede central) of adjacent regions.As part of these pacts, a stronger sede central would often provide training and even some firearms to the militias of a weaker regional committee, as in 1989 when Pichihuillca's DECAS sede central in the Apurímac River Valley signed a pact with the newly formed sede central at Patibamba, in the mountain district of San Miguel (see Fumerton 2002, p. 131-137).
This vast network of interconnected militias was a self-governing system that was conceived, organised, and implemented by the peasants without the direct influence of the state.In other words, it was an indigenous innovation.Administered at various levels by a hierarchy of governing committees, this network helped to (re)build relational bridges between previously belligerent villages.It did so by creating an institutional framework and procedure for peacefully resolving inter-communal disputes.Moreover, this segmented network of militias fostered new, inclusive identities that also helped to heal the fabric of Ayacucho's rural society, hitherto so deeply ruptured by endogenous violence.
Another institutional mechanism that helped to overcome prior experiences of inter-communal violence was the creation within many regional militia networks of a "grupo especial" (special group).A grupo especial is an elite force of militiamen made up of volunteers, one man from every village in the regional militia network (see Fumerton 2002).In other words, every grupo especial was multi-communal in composition.Individual members, known as "rentados" (hired patrollers), were paid a small monthly salary from financial contributions pooled together from all the families in the villages."The rentado is exclusively at the service of self-defence, for the security of each community," Comando Zorro explained to me."He dedicates himself entirely to keeping vigil, to patrolling, and verifying other civil defence committees so as always to maintain the unity of a comité de base [the village militia].There are some communities with few comuneros, so two communities may join together to pay for one rentado."Similar to what del Pino observed among the DECAS' Comandos Especiales in the Apurímac Valley, participation in the Grupo Especial of Tambo District "thus became for the young people a source of work in the midst of economic crisis" (del Pino 1996, p. 154. My translation).
Many members of these elite militia groups had already done obligatory military service in the Peruvian Armed Forces, and so have had military training.These ex-soldiers were called "licenciados," and because of their military knowledge were a match for Senderista fighters when equipped with the very best weapons that the communities could afford to buy: the highly prized Mauser rifles.It was the veteran licenciados of more experienced grupos especiales, or soldiers from the nearest local army base, who gave basic military training to ordinary villagers (ronderos) and to newly formed grupos especiales in other zones, whose members might not have any prior military instruction.
Almost every community within the regional network provided an individual rentado to the grupo especial.This practice of creating a mixed unit proved an excellent way to overcome traditional inter-community squabbles while at the same time creating a new sense of corporate identity, united in the collective purpose of defending themselves against a brutal common enemy.Often, such a group's sense of solidarity and identity became so strong that they even came to wear their own special "uniform," like a tracksuit, or indigenous ponchos all with the same colour.
This step of professionalising community defence can be regarded as yet another way in which militias contributed to the reconstruction of their communities, for through the creation of the grupos especiales the bulk of the adult peasant population were freed up from the daily demands of community defence and able to return to their important agricultural and other livelihood activities, which had all been sorely neglected at the height of the civil war.

"Rondas campesinas" performing citizenship contention
Ever since the early years of the civil war, peasant have actively voiced claims on the Peruvian state; and the historical record shows that rondas campesinas have been the spearhead of a series of "claim making" episodes in the department of Ayacucho.The first of these was the demand that the government recognise the peasantry's right to defend themselves from rebel attacks, particularly in light of the fact that state security forces were proving themselves to be largely incapable of protecting the rural population.The prevailing attitude in government for much of the time was evident in high-ranking authorities like Interior Minister José María de la Jara, who as early as August 1980 had expressed his total opposition to the peasant population's spontaneous formation of rondas campesinas "…because they can be considered 'very dangerous paramilitary organisations'" (quoted in DESCO 1989, p. 367, Vol. I).This became even more politicised as an issue of intense national debate after a small group of Lima journalists, who had travelled to a remote part of Ayacucho, were mistakenly thought by peasant highlanders there to be Senderista rebels, and were killed at the village of Uchuraccay.The ensuing investigation, which essentially paints the Uchuraccaynos as "savage brutes," opened a nation-wide debate on the violence that was rapidly spinning out of control, and created strong opposition to peasants engaging in "vigilante" violence -even in self-defence (see CVR 2003, Vol.V).
However, the evidence also shows that militia leaders were keen to capitalise on any act or utterance from a state official that appeared to express "certification" of their right to self-defence. 13A clear example of this was when General Adrián Huamán Centeno, Political-Military commander of the Emergency Zone for a short period in 1984, openly declared to the press that these rondas campesinas and their communities "…have the right to defend themselves from the assassins [Senderistas] with the arms they have in their hands" (quoted in DESCO 1989, p. 110. My translation).The resonance that these words had on the peasant patrols was immediate, and profound.A few days later, a massive rally of around 20,000 ronderos from the districts of Ocros and Concepción presented the General a signed declaration that they had organised a huge network of mutually supporting militias, which they called a "Frente de Defensa Civil" -a Civil Defence Front (DESCO 1989, p. 110).
The next episode that involved peasant militias making claims on the state revolved around the related issue of the right to bear firearms.An important figure in translating the diffuse claims of peasants into a coherent demand for the right to bear arms was a DECAS militia commander known as "Comandante Huayhuaco."However, it was the Expreso, a Lima newspaper, which made this possible.For a time, the newspaper turned him into a national celebrity through which to tell the story of the DECAS militias and their heroic fight against Sendero Luminoso in the Apurímac River Valley -and most importantly, to advocate for arming them (e.g.Expreso 1989a).One article in particular that the Expreso published with, as its sensational title, Comandante Huayhuaco's claim of "Denme 500 fusiles y libero Ayacucho en un año" ("Give me 500 rifles and I will liberate Ayacucho within a year") caused a heated debate at the national level. 14But it also brought much attention to the activities and circumstances of the peasant militias, which hitherto were not widely known in the national capital.In December 1989, President García, made a surprise visit to the Apurímac River Valley to distribute 200 shotguns personally to the DECAS militiamen.In the presence of the national press, García shook Comandante Huayhaco's hand, gifted him a revolver, and proclaimed him as his very own "personal representative" in the counterinsurgency struggle being fought in the valley (IDL 1989, p. 11).It was largely a symbolic gesture; a random act done as his presidency was coming to a close.Nevertheless, DECAS militiamen and other ronderos everywhere in the Emergency Zone saw his action as yet again "certification" of their claim, this time from the highest political level (see Fumerton 2002, p. 135-141).
Coming at the very end of his presidency, García's act of distributing arms to rural militias was never transformed into an official government policy to arm the rondas campesinas.It would be during the regime of his presidential successor, Alberto Fujimori, that the peasant militia system would finally be given formal government recognition.Decreto Legislativo No. 741 and No. 740 were promulgated by the Fujimori government in November 1991, thereby legalising the existence of peasant militias and their right to bear firearms.The militias were also baptised with a new name: Comités de Autodefensa y Desarrollo (Committees of Self-Defence and Development -CADD).
With legal recognition came stricter government monitoring and control of the self-defence committees.From the mid-1990s onwards, the military -by now having recognised and embraced a population-centric "hearts and minds" counterinsurgency approach, and operating under the motto "Más Desarrollo, Menos Represión" (More Development, Less Repression) -was made responsible for administering certain bureaucratic formalities and procedures in the operational practices of the CADDs. 15Now that the militias possessed firearms, such controls and protocols were designed to make civil defence patrols more accountable to the military for their actions, thereby preventing the rise of local militia warlords, and curbing their potential for committing further human rights abuses (Fumerton & Remijnse 2004;cf. Blair & Kalmanovitz 2016).
In my interviews with CADD commanders and patrollers, most were adamant in the view that D.L. 741 was a declaration of the state's obligations and responsibilities towards the peasant militias and their communities; or in other words, a recognition that peasant lives mattered."Aren't we also Peruvians?" a campesino friend once rhetorically asked me."Don't we have as much rights to protect ourselves with guns as the rich men of Lima?"For the campesinos I interviewed, this piece of legislation gave them a stronger sense of having a real stake in contributing to the future wellbeing of la Patria, for which every rondero death was now wrapped in a new discourse of "sacrifice for the fatherland." The third important episode of claim making emerged from a profound change in the peasantry's self-perception; namely, a rise in their self-esteem during the 1990s, as compared to the start of the civil war when the word "campesino" was often taken to imply subordination and discrimination.Peasants in Ayacucho came increasingly to believe that the CADDs, and not the state's security forces, were the ones truly responsible for defeating Sendero.As such, they began to see themselves no longer as simple victims, but also as the true "citizen" heroes who were giving their lives to defend Peru.
The promulgation in 1998 of Decreto Supremo No. 068-DE/SG, which stipulated that CADD members who had suffered temporary or permanent physical disability, along with the widows of deceased militiamen, were eligible to receive monetary indemnity was welcomed by peasants as recognition of their sacrifices in the war against Shining Path.Similar legislation to indemnify members of the armed forces, police, and civil servants had existed for much longer (Garcia-Godos 2008, p. 71), and so D.S. 068 reinforced the belief that with its promulgation peasants could now expect in this regard to be treated as equal citizens, more or less.In the course of the rest of the decade, and in the consciousness of ordinary peasants, the self-defence experience helped to forge a closer association between themselves as "citizens with rights" and the state.And it is with this stronger and more confident consciousness of themselves as "citizens with rights" that a delegation of militia commanders travelled to Lima in 1999 to press further claims on the Fujimori government in regard to the implementation of D.S. 068.

Closing reflections
Security provision does not work in isolation from other aspects of society.It is always linked to wider social fields (e.g.Hoffman 2007).How this occurs, and what these linkages and these wider social fields might constitute, is what this paper has attempted to illuminate.In this article I have attempted to shift attention to a more unfamiliar and under-researched area by exploring the socially constructive and re-constitutive aspects of militias.Specifically, I have explored how Ayacucho's peasant militias, initially formed for the purpose of violently opposing Shining Path rebels, came to play a key part in governing and rebuilding their communities.In the process they became as much political actors (making citizen claims on the state) as counterinsurgent ones.As a way of conclusion I will now pull the various narrative and analytical threads together to give an answer to this puzzle, and reflect upon the broader implications of this perspective.
This paper has taken a relational perspective that highlights the changing configurations of interactions -cooperative or conflictive -between actors.These shifting relational configurations shaped the strategic choices and trajectories of the rondas campesinas as an institution that came to shape a new wartime social order for their communities.Significant in this interaction, as confirmed by the Peruvian CVR, is the fact that rebel victimisation of peasants remained high throughout the entire civil war whereas, by contrast, the military changed its behaviour from a kinetic, indiscriminate, and highly repressive counterinsurgency approach, to a population-centric "hearts and minds" strategy.The unpredictability of momentous relational changes during the civil war underscores the contingent nature of these interactive configurations: the choice made by some peasant communities to take up arms against Shining Path was not an inevitable one, even under the pressure of coercion. 16As such, the existence of Ayacucho's militias cannot be fully understood unless placed within the specific context of what Theidon calls "the thick of regional histories" (2006, p. 442).
The larger context of civil war presented opportunities and obstacles that shaped the emergent action of these militias.The evidence heavily points to Shining Path as being the actor primarily responsible for creating the oppor-tunity for peasant self-defence committees to assume governance functions within their villages.Shining Path's dogmatic adherence to its Maoist-inspired strategy compelled it to destroy both governmental and traditional authority structures, and to execute local "enemies of the people."The strategy achieved its intended result, and left a governance vacuum.Yet in as much as local leadership was still needed to organise and to coordinate the collective activities and affairs of community life, it fell upon the village militia (which became the only functioning local institution remaining) to assume the general responsibilities of village governance.
The social significance and meaning of Ayacucho's rondas campesinas went so much deeper than their anti-rebel character, or their engagement with the Peruvian state.Within the shifting dynamics of civil war violence in Ayacucho, I have described how rondas campesinas invented and administered their own wartime social order in regard to local practices of communal justice, and in so doing began to heal the torn fabric of rural society by rehabilitating, reintegrating, and reconciling former Shining Path guerrillas.These interactions with former enemies were deeply performative, symbolic, and ritualistic in nature.In addition, militia leaders devised innovative ways to inhibit intercommunal violence: namely, by amalgamating all the villages within a region -some had been erstwhile enemies -into a single network of interdependent communities under the authority of a sede central (central headquarters).The result of these organisational innovations was the eventual de-activation of violently antagonistic boundaries between rival villages, while conversely fostering greater social solidarity through the provision of mutual security and the creation of an inclusive regional militia identity.Within these relations we see at work the mechanism that Tilly and Tarrow call "attribution of similarity," which they define as the "identification of another political actor as falling within the same category as your own" (2007, p. 215).Does Ayacucho's Comités de Autodefensa y Desarrollo present an anomalous case in the study of militias?I do not think so.The governance function they assumed, and as I have described here, is certainly not unique when we take a longer historical perspective.Prak, for instance, examines what he explicitly calls "civic militias" in late medieval and early modern Europe, and demonstrates that (within a setting similar to rural Ayacucho) where the state lacked the monopoly of violence, "…both the idea and practice of the citizensoldier remained a vital ingredient in the socio-political structures of society" (2015, p. 96).Prak's observation resonates in my case study.Many of Ayacucho's peasants regarded militia service (which they see as a form of military service that demonstrates loyalty to the state) as the basis for claims to citizenship status.Scholars have long recognised that military service has been an integral component of citizenship (see Burk 1995).This is why CADD members claim that their armed defence of the state during the civil war entitles them to full citizenship rights, like the right for compensation for their losses.
Why the CADDs might appear to be an exceptional case in the militia literature seems to me to do more with scholarly perspective than with an empirical reality.Recent research, however, has begun to counterbalance the pernicious perspective towards non-state armed actors in general, and militias in particular.For instance, in their study of the conditions for warlord rights and legitimate authority, Blair and Kalmanovitz recognise that many accounts "…neglect the possibility that armed nonstate actors may legitimately perform state-like functions, preserving social order and protecting civilians from harm" (2016, p. 429).Hoffman's ethnographic study challenges the assumption that war brings a breakdown of sociality.He therefore invites a rethinking of the meaning of Civil Defence Forces (CDF) in West Africa, concluding that " the patronage networks which dominate everyday existence have not been replaced in wartime, they have simply become militarized" (2007, p. 660).
Finally, what we have seen in this article is that the right kind of oversight from the government greatly contributes to militias becoming agents of positive social reconstruction and transformation.In this case, the willingness of the Peruvian government and its military to change how they practiced counterinsurgency had a significant impact on the dynamics of violence, and on their relations with the peasantry.As Stanton explains, "if governments choose not to target civilians, they limit violence among both their regular forces and militia forces.The fear that militias will act outside of the control of the government, targeting civilians even when the government does not, is not borne out in the evidence" (2015, p. 901).Throughout the 1990s the Peruvian military proved successful in supervising the CADDs of Ayacucho, and in so doing making them more accountable for any recourse to violence.This kind of tight governmental oversight helped to ensure that Ayacucho's militias did notindeed, could not -degenerate into death squads and criminal groups, as has happened in other settings (e.g.Fumerton and Remijnse 2004;Spencer 2001) Acknowledgements: I am grateful to Martijn Kitzen, Corinna Jentzsch, Chris van der Borgh, Nora Stel, and Beatrice de Graaf for their invaluable comments on previous drafts, and to Claire van Nieuwburg for her assistance in assembling the empirical data.I also would like to thank the Department of History and Art History, and the Centre for Conflict Studies, at Utrecht University, for providing me with financial assistance to participate in a "militia workshop" held at University of Notre Dame on 20 October 2016, and at the conference "Paramilitaries, Militias, and Civil Defense Forces in Civil Wars" at Yale University, 19-20 October 2012.

Notes
1. Blair & Kalmanovitz (2016) make similar observations with regard to the literature on warlords.2. Although as Blair & Kalmanovitz (2016) explain, warlords can also organise and recruit militia, I would argue that the Peruvian Comités de Autodefensa (CADs) differ typologically from such warlord militias in a number of significant ways.Whereas warlords are "subnational strongmen who seek autonomy within geographically demarcated territories" (2016, p. 428), I show that certain state-controlled mechanisms prevented the rise of CAD warlords.Furthermore, unlike warlords, CAD commanders did not explicitly "resist the imposition of state rule over the territories they control" (ibid.)but rather often sought alliances with units of the Peruvian armed forces.This is not to deny, however, that there are also characteristic similarities between the two.3.At first blush Arjona's (2017, p. 8-9) concept of "rebelocracy" seems to typologise Ayacucho's CAD militias quite adequately -and she would probably classify militia social order as such.However, the crucial details that would appear to disqualify the application of "rebelocracy" to their governance/community practices and ensuing wartime social order are (1) their pro-state/anti-rebel dimension, (2) the absence of authoritative dominance by a rebel organisation through which to extract "massive obedience" (Arjona 2017, p. 13) from CAD communities, (3) in reality no clear categorical separation between militia members (as combatants) and able-bodied adult community members (as civilians), for in principle they were one and the same; and (4) the pro-state orientation of the CADs, particularly in regard to their political "claim making" behaviour.I suspect that "rebelocracy" is perhaps too narrow and too loaded a term for the analytical scope of variation that Arjona subsumes under the concept.In any case, this issue warrants more careful conceptual contemplation, which I am unable to pursue any further in this article for reasons of space limitations.4. See the CVR 2003 (volumes I, IV, and V in particular); Degregori 1996aDegregori , 1990;;Gorriti 1999;Stern 1998;Starn 1995. 5. See CVR 2003, Vol.VIII, Chapters 1 and 2. For an excellent analysis and discussion of how war put into motion a process of internal cultural and moral destructiveness within peasant society in Ayacucho, see Theidon (2000).6.In the Emergency Zone, the constitutional rights and liberties of its civilian inhabitants were suspended, and political control of the department passed from the hands of a civilian official to a senior military officer (DESCO 1989, p. 352;Amnesty International 1991, p. 16-17).7. "IDP" refers to "Internally Displaced Persons."The aim of this tactic was to deprive roving guerrilla bands of any logistical support from rural communities.Additionally, concentrating the rural population into settlements adjacent to an army garrison made it easier for soldiers to keep the inhabitants under surveillance.8. APRA is the acronym for the Alianza Popular Revolucionaria Americana (American Popular Revolutionary Alliance).9. Community policing and organised anti-crime patrols were activities that had existed in other parts of Peru for decades."Peasant patrols" were already an observed phenomenon in northern departments of Peru since the late 1970s, thus predating the CADS that were organised in Emergency Zone as part of the counterinsurgency effort (see Gitlitz & Rojas 1983;Starn 1999).10.According to the Peruvian Truth and Reconciliation Commission (CVR), 53.68 per cent of deaths and disappearances reported to the CVR can be attributed to Shining Path (CVR 2003, p. 56, Vol. I, Ch. 1).11. "Boundary formation" refers to the "creation of an us-them distinction between two political actors" (Tilly & Tarrow 2007, p. 215).12. "Arrepentidos" can be translated as "repentant ones."13.Tilly & Tarrow define "certification" as "an external authority's signal of its readiness to recognize and support the existence and claims of a political actor" (2007, p. 215). 14. De Althaus & Salgado 1989. 15.These included such mechanisms as obliging CAD leaders to write up a weekly report of their activities, and requiring CAD patrol commanders to obtained a signed declaration from the leaders of every community they visit stating that while in that village the patrollers did not commit any abuses or crimes (Fumerton 2002, p. 199-200).16.Indeed, many rural communities in Ayacucho responded to Shining Path in a variety of ways other than forming defence groups (Isbell 1988(Isbell , 1992;;Coronel 1995;Coronel & Loayza 1992).
. Instead, Ayacucho's DECAS and CADDs became instrumental in enabling the return of peace and stability to their respective regions.Mario AntonioFumerton <m.a.fumerton@uu.nl> is Assistant Professor at the Centre for Conflict Studies, Utrecht University, where he teaches various graduate and undergraduate courses.His research interests include social movements, contentious politics, insurgency and counterinsurgency, militias, performance and performativity.His publications include Kurdistan's Political Armies: the Challenge of Unifying the Kurdish Peshmerga Forces.(Wladimir van Wilgenburg co-author; Carnegie Endowment for International Peace.16 December 2015), and "Insurgency and Terrorism: Is there a Difference?" (Isabelle Duyvesteyn co-author; in C. Holmqvist-Jonsäter and C. Coker, eds.The Character of War in the 21st Century, Routledge, 2010).He has conducted fieldwork research and given consultancy in Peru, Philippines, Sri Lanka, Afghanistan, and Iraqi-Kurdistan.