Trans-Scalar Ethnographic Peace Research: Understanding the Invisible Drivers of Complex Conflict and Complex Peace

ABSTRACT Dynamics of peace and conflict emerge in the complex non-linear circulation of influence across scales, from the local to the global and back. Recently proposed methodologies for examining these dynamics, whether quantitative (primarily focusing on the macro), or qualitative (primarily focusing on the micro), are distinctly limited. Both struggle to trace influences across the full range of scales, and neither incorporates mechanisms to examine the consistency or divergence between implicit concepts which underpin expectations and experiences of conflict, peace, and peacebuilding. Trans-Scalar Ethnographic Peace Research, however, focusing specifically on these implicit concepts, facilitates examination of complex conflict dynamics across scales.


Introduction
At its heart, this article is about methodology, but not in the sense of describing practical steps for carrying out research. Instead, the article examines two closely interrelated challenges within the field of Peace and Conflict Studies (PCS), and why their interaction demands a particular approach to data collection and analysis. These two challenges are complexity and radical alterity. While complexity has been quite frequently deployed in recent PCS literature, and sensitivity to local culture has been consistently highlighted as key to the success of peace interventions, this article argues that the field as a whole has not fully wrestled with the deeper implications of either of these challenges (see Chandler 2018;Brigg 2018). More importantly, the field has failed, thus far, to recognize the multifaceted challenge posed by the interaction of these two; that posed by radical alterity within complexity. This article seeks, therefore, to provide an initial examination of the problematic implications of this challenge for research methods deployed in the field, and to explain why and how a trans-scalar Ethnographic Peace Research (EPR) approach would respond to this challenge.
While nobody would argue that international peace interventions always have negative impacts in local settings, it is certainly recognized today that they often do have unpredictable and problematic results. Indeed, this article will argue that such and across multiple scales. The article then proceeds to describe the key strengths of EPR for the study of such dynamics, both at specific scales and then for the circulation of influence across scales, and reflects on some of the challenges to trans-scalar EPR at each scale and across scales.
The final section summarizes the argument and concludes the article by highlighting the two main strengths of a trans-scalar EPR. These are, first, that it is an inherently inductive and post-positivist approach for examining the underlying conceptual foundations of peace interventions that remain invisible to many other methods but nonetheless generate unpredictable outcomes. And, second, that it is applicable at each specific scalelocal, national, international and globaland can allow scholars to track how these concepts interact across scales by examining both their subtle composition in the minds and worldviews of the actors involved, and how they are manifested in actions, processes, and institutions in the world. In this way, as I will show, trans-scalar EPR can allow us to better examine and understand the invisible drivers of complex peace, conflict, and peacebuilding dynamics.
Considering causation: Peacebuilding amidst violence or violence amidst peacebuilding?
The central theme of this Special Issue evidences the growing recognition within PCS, and particularly among scholars interested in post-conflict peacebuilding, that building peace in post-conflict contexts is more complex than was once thought (Chandler 2013;De Coning 2016). Just as the cognate area of Transitional Justice recently recognized the importance of paying attention not only to justice mechanisms administered in the post-conflict phase but also in the 'during-conflict' phase (Loyle and Binningsbø 2018), so peacebuilding scholars have started to realize that peacebuilding often occurs amidst violence and not only after violence ends. But this paper will push this argument further in highlighting also how peacebuilding efforts can sometimes inspire violence, and may in fact, even require violence of different kinds.
In many cases of liberal peace interventionism such dynamics are clear. The liberal peace, as composed of the three pillars of democratization, the rule of law, and marketization (Paris 2004;Doyle 2005), for example, can very clearly contribute as much to conflict and violence as it does to justice and peace. The democratization pillar, for example, as implemented in consociational arrangements in Northern Ireland, Bosnia-Herzegovina, Burundi, or Lebanon, has evidenced a tendency to institutionalize and further reify ethnic divisions. Such power sharing arrangements were seen as central to peace processes in the immediate post-Cold War period, and are still in vogue among many liberal peace proponents. However, the conflict resolving potential of such arrangements has been in doubt for some time (Selway and Templeman 2012), and as such this embodiment of the democratization pillar of the liberal peace is recognized as a potential contributor to long-term inter-group conflict dynamics (Aitken 2007).
The same is true for interventions intended to promote the rule of law and human rights. Many interventions related to the rule of law, such as Security Sector Reform (SSR), are perceived as unfairly benefitting particular groups. In most such cases individuals belonging to one group are considered more amenable than others to service in the new security structures and choices regarding inclusion and exclusion can become reasons for continued intergroup grievance (Perito 2008, 6). Furthermore, establishing institutions of law inherently requires the transmission of more litigious traditions that benefit certain groups over others. The establishment of the legal right to property, for example, does not benefit everyone equally, but privileges those with the resources to influence legal decisions, whether through legal (via the purchase of legal expertise) or extra-legal means (via intimidation or corruption) and so can foster marginalization and disempowerment which may inspire further conflict (see, for example , Sieder 2003;Millar 2017).
Similarly, marketization often seems to give rise to conflict, particularly over the inherently limited resources and opportunities made available when markets are opened to foreign investment, competition, and expropriation. Millar's recent study of an industrial agriculture project in post-conflict Sierra Leone illustrates exactly this, in that the resources invested inadvertently promoted conflicts between groups within the local context; between elders who accepted the project and youth who felt their land had been sold out from under them, between urban and educated youth employed by the project and those rural youth excluded from those benefits, and between the rural communities who felt that the political and traditional elites had made promises regarding development and profit that they had no intention of keeping (Millar 2016). Such conflict promoting dynamics, driven by the marketization of post-conflict economies, are echoed in many cases around the world and illustrate the manner in which processes intended to promote peace can function instead to inspire, or even require, the violence of exclusion and expropriation (see Richani 2005;Miklian 2017).
Instead of peacebuilding amidst violence, these examples might be instances of violence amidst peacebuilding, but could more accurately be described as peace interventions fostering, and perhaps even requiring, the violence of marginalization, exclusion, and expropriation. My argument here is that these unexpected sub-optimal outcomes emerge because these interventions are based on concepts (of authority, legitimacy, empowerment, justice, development, dignity, security, etc.) which are assumed to be universal by those on the global and international scales who design, fund and plan such interventions but which, in reality, vary radically from place to place and so foster unexpected and potentially negative dynamics at the local scale. Of course, this argument assumes a concept of violence itself as not only direct, but also structural and even cultural (Galtung 1969(Galtung , 1990); a definition not always accepted by scholars concerned with the rule of law, security, governance, or conflict management. But if we are examining the relationship between peacebuilding and violence, and considering how and why violence circulates across scales, then we must be sensitive to how the many forms of violence (direct, structural and cultural) interact and are implicated across scales; how policies of expropriation at macro scales can beget direct violence at micro-scales, or how disempowerment or exclusion at micro-scales can beget violent political conflict at macro scales.
The examples above are representative of common interventions applied in dozens of post-conflict cases over the past 25 years. That such interventions can foster new conflict dynamics is well established in the literature (as noted above), and these 'downstream' effects are prime fodder for critics of the liberal peace and authors within the 'local turn' tradition. But I note these examples specifically to highlight what is usually left unexplored in that literature; the radical alterity underpinning the often negative effects emerging from complexity across scales. At this point it is perhaps necessary to clarify specifically what I mean by complexity, by scales, and by radical alterity.
As will be further noted below, complex systems are those characterized not simply by complication (a system requiring expert knowledge to understand and predict), but the emergent self-organization of novel phenomena that cannot be predicted (Cilliers 1998, ix). Such systems give rise to unpredicted and unpredictable outcomes that need not reflect the intent of any particular actor or institution (Miller and Page 2011, 4), but instead reflect the operation of negative and positive feedback mechanisms giving rise to non-linear effects. As such, complexity theory has become particularly helpful for those struggling with the unpredicted outcomes of peace interventions by global and international actors into local spaces. It is for this reason that this paper uses the language of 'scales' as opposed to discussing these issues as occurring on separate 'levels'. As articulated by Hameiri and Jones, borrowing the idea of scales from political geography avoids falsely reifying levels of analysis such as the international and the local, and allows a dynamic interaction of influence across scales (2017). It is this dynamic interaction that particularly led me to link the scalar analysis with radical alterity, as it is the manner by which concepts central to peace and conflict differ across scales which leads to unpredictable or emergent experiences.
To articulate this as clearly as possible, actor's internalized concepts (norms and ideas) about what is acceptable, appropriate, or expected 'underlie and are expressed in the ways humans live' (Keesings 1981, 68-69). These internalized expectations form the basis for individualand eventually also institutionaldecisions about how to respond to dynamics of violent conflict, including decisions made to leverage the various kinds of capital available to actors within their environment, at a specific scale or across scales, for the purposes of waging conflict or building peace. At the global scale, the pillars of the liberal peace can be seen to embody such implicit norms, which then become materialized in specific interventions, such as those described above in the spheres of politics, law, and economics. But, and this is the key, they are implemented in contexts underlain with radically different conceptual systems.
The idea of 'radical alterity' is much developed and debated within Anthropology, and is primarily deployed as a means to reacquire the idea of substantive otherness for a discipline whose central concept (culture) has been, some argue, essentialized and stripped of substance (Povinelli 2001;Vigh and Sausdal 2014;Graeber 2015). A 'strong' form of the construct usually carries implications of radically different worlds (Hage 2012) or even natures (Viveiros de Castro 2015); i.e. it implies not only that different cultures hold to different views of the world, but that they actually experience different worlds. But here I use the construct in a 'weaker' form more akin to that of Fowles (Alberti et al. 2011, 906), which instead recognises that debates about the nature of the 'real' world (or, potentially, worlds) are best left to physicists and philosophers, but which retains the notion of incommensurability central to the basic idea of radical alterity. I therefore use the term 'radical alterity' to focus on the incommensurability of concepts of the world, and not of the world itself. The broader purpose here is to make clear how the sub-optimal outcomes of peace interventions result from the divergence and difference between such concepts across the global, international, national, and local levels; i.e. from radical conceptual alterity.
This should make clear how decisions by powerful actors in global institutions such as the World Bank or United Nations (macro scale) engage with and exhibit non-linear influence on experiences within local settings (micro-scale). But it also indicates how the most local of processes (even down to the psychological) engage with and can exhibit non-linear influence on higher scale processes. But it is key also to realize that what we usually see as 'global' norms, constructs or concepts, are themselves the outgrowth of just another local that has made a claim to universality (Björkdahl et al. 2016, 4); a micro-norm now expanded or mutated into a macro-norm. This Eurocentric order has achieved hegemonic status over the last 150 years, and in this way a particular micro has become what we think of today as the macro, and the discrepancy between that 'local turned global' and the settings into which it intervenes is the central driving force of violence amidst peacebuilding today.
Unfortunately, most PCS scholarship pays little, if any, attention to internalized concepts, tends to un-root the global from the European local, and so focuses on the influence of a dissociated global on national or local dynamics within conflict affected states. But this is more a product of disciplinary training than it is of the phenomena under study. Most such scholars have themselves internalized a conception of causal influence as topdown and wielded primarily by actors at the higher scales, of institutions as the dominant causal mechanism in conflict and peace (see Millar 2014a), and of institutions of European cultural origin specifically as bearing a historically evidenced truth and legitimacy. Such a perspective seems wedded to the positivist enlightenment tradition, which sees the world as observable, measurable, knowable, and, eventually, controllable via planned intervention. There are alternative perspectives within PCS, of course, including those which see interpersonal interaction, the development of trust, and the realignment of conceptions of the 'Other' at the micro-scale as key to changes in inter-group relations, to reformation of policy, and to national reconciliation at more macro scales (Allport 1954;Fisher 1990;Lederach 1997). But it must be noted that this inversion of the causal logic is just as limited in a complex environment characterized by the non-linear entanglement of the micro and macro as opposed to linear vertical relationships.

Peace, conflict, and complexity
This more complex perspective requires PCS to develop more subtle theories regarding and more appropriate methods for studying relationships involved in peace and conflict dynamics, and the field has been moving in this direction for some time. It has long been recognized that in a globalized world conflicts are not restricted to bounded spaces. On the contrary, conflict dynamics within states are incentivized and driven by larger systems (economic, political and social) in which conflicts are nested (Millar 2020a), as has long been recognized in the literature regarding 'internationalized' or 'regional' conflicts (see Patrick 2011), 'new wars' (Kaldor 2012), and 'conflict systems' (Keen 2006). In such an environment, it is wholly inappropriate to identify either conflict or complexity at one scale. It is more accurate to recognize the system as spanning across scales and, thus, to analyse and understand conflict via methods that can trace influence across scales. But while key PCS thinkers have skirted around the issues of complexity for quite some time (see, for example, Wright 1951, 209), until very recently scholars have not engaged robustly with complexity theory. As a result, earlier references to the concept failed to really inject the necessary substance and nuance into mainstream PCS theory. Examples of this can be found, for example, in cases where 'complexity' was deployed largely to explain the inability to fully understand what was occurring (Azar, Jureidini, and McLaurin 1978), or where it was used as a measure of the number of actors or issues involved in a particular conflict (Galtung 1996, 90-92).
An interesting exception is Boulding's early description of 'the ecological complexity of the total system, which means that pressure applied at one point will produce often completely unexpected and counterintuitive effects elsewhere ' (1978, 352). This contribution exemplifies a far more-subtle conception more closely in line with what complexity theorists such as Robert Jervis would later describe (1998), and which presage the development of new research methods in PCS today. Those who have recently turned to complexity theory as a tool to understand peace and conflict dynamics, for example, deploy a distinct repertoire of concepts such as non-linearity, interdependence, coevolution, decentralized control, self-organization, non-equilibrium order, and adaptation; all developed outside the field to understand the dynamics observed within complex adaptive systems more generally (see Moffat, Bathe, and Frewer 2011, 60). This turn to complexity theory has been facilitated by the empirical realization that contemporary conflict is not a function of binary pairs of violent actors, and the methodological approaches recently proposed to address the challenge of complexity have largely been developed to understand relationships between more than two variables and across scales.
Central to these efforts has been a strong critique of the hegemonic methods in the fieldvarious forms of linear regressionand an admission that this approach is not appropriate for the study of complex systems. The various problems of linear regression are most starkly outlined by Schrodt, who summarizes his critique by stating that 'for many problems commonly encountered in political analysis, linear models aren't just bad, they are really, really bad. Arguably,' he continues, 'it would be hard to design a worse set of potential side effects ' (2014, 289). The various problems that he identifies, he argues, are largely the result of assuming the applicability of a statistical method designed appropriately for use with data collected via random samples or true experiments, to 'large swaths of political science where they do not apply, notably pretty much the whole of IR ' (2014, 293). These noted limitations of statistical regression have inspired others to develop alternative approaches for PCS which can better model complexity, such as the recent special issue edited by Dorussen, Gartzke and Westerwinter, who argue that within PCS specifically 'dyadic data routinely violate the independence of observations assumption' within linear regression (2016,(283)(284).
However, while some PCS scholars have taken up the challenge of responding to Schrodt's critique, the various attempts to develop quantitative alternatives to linear regression which can better analyse the complex interactions of actors and institutions involved in peace and conflict dynamics across scales do not include any consideration of underlying concepts and exhibit no ability to consider the challenge of radical alterity within complexity. The attempt in such work is clearly and appropriately to develop means to assess and understand non-binary and often non-linear relationships across scales, but unfortunately these efforts rarely consider any concepts, actors, processes, expectations or experiences below the national scale (the sub-national or local) or above the international scale (the regional or the global). Articles by Minhas, Hoff, and Ward (2016) and Warren (2016), for example, develop dynamic models that provide novel quantitative methods which can respond to complexity theory, while both Warren (2016) and Gartner and Regan (1996) creatively address such dynamics across multiple scales. However, none of these efforts place focus on divergent conceptions across scales. In short, the underlying concepts that motivate expectations, actions, and experiences are left completely outside the scope of analysis.
It is in such studies, well-meaning and innovative as they are, that we see how many quantitative methods create blind-spots in the analyses of peace and conflict dynamics across scales, by allowing underpinning concepts to become subsumed by quantitative measures at higher scales. In such efforts, concepts that differ across scales and contexts, and which are central to peace and conflict (concepts, for example, of authority, legitimacy, representation, etc.) are subsumed within measures of 'democratization' at the national scale. We witness the same dynamic in studies that measure the health of economies via the Gross Domestic Product (GDP), which, while they may be nuanced today via inclusion also of the Gini coefficient or the UN Human Development Indicators, still provide no means to provide more fine grained assessment of the radically variable beliefs or concepts which underpin and legitimate institutions of law, taxation or markets at higher scales (concepts, for example, of empowerment, progress, dignity, development, etc.). In this way, the odd combination within PCS of a blindness to the underpinning conceptual motivators of intervenors' decision, combined with an almost exclusive focus on the power of institutional solutions at a macro-scale, serves to exclude any analysis of how such underpinning motivators foster problematic emergent micro-scale dynamics (see Millar 2014b). This is not to argue that scholars working within these traditions would deny in principle that sub-national or 'locally rooted' concepts and norms influence the national or global scales from below. It is simply to note that such influence tends to be recognized in narrative, made secondary in theory, and become almost completely marginalized when subsumed methodologically into quantitative measures at national or international scales. So, for example, when such studies 'measure' religious faith with dummy variables (1 if Catholic, 0 if not), or political culture with the polity score, they are accepting in theory the importance of faith and ideology, but then refusing the potential for radical alterity across settings and scales in their need to deploy a pre-existing operationalization of those concepts. Such measures ignore most of the important nuance and diversity at the micro-scale that make such concepts so important in peace and conflict dynamics. Furthermore, and perhaps even more problematically, the hegemonic concepts underpinning the beliefs and expectations of actors and institutions at the higher scalesi.e. of those who fund and plan peace interventionsremain unquestioned by such scholarship; retaining the power of assumed universality in their exclusion from the process of analysis. The efforts described above, therefore, while laudable for what they do achieve, are nonetheless limited.
Unfortunately, similar critique can be levelled against efforts made recently by qualitative scholars in response to the same recognition regarding the multiplicity of actors and agents and the complex and often unpredictable results of interaction across scales. Bara, for example, and similarly citing the limitations of binary logistical regression, introduced a novel Qualitative Comparative Analysis approach to take account of 'conjunctural causation' and 'equafinality' in the emergence of ethnic conflict (Bara 2014, 699), while Visoka justifies his 'Peace Figuration' as an approach for arriving at 'reality congruent' assessments of what he describes as inherently complex processes characterized by unintended, uncontrollable, and unpredictable outcomes (2016). Indeed, Brusset, de Coning and Hughes (2016) have recently co-edited a full volume of contributions seeking to provide tools for collecting and assessing data for the evaluation of peacebuilding interventions within complex contexts and with complex impacts. But these efforts too, make limited effort to explore and understand the conceptual foundations of beliefs, expectations or experiences at either micro or macro scales. Other efforts, such as Millar's ethnographic work (2014a) and Firchow and Mac Ginty's more quantitatively aligned 'Everyday Peace Indicators' Project (2017) specifically examine concepts of peace, justice, development, and security at micro-scales, but they do not put equal emphasis or substantial focus on the underpinning concepts motivating peace interventions at higher scales. This is a substantial problem for the field because neither of these traditions (quantitative or qualitative) is yet providing any means by which to collect and interpret data which will allow us to understand underlying concepts at different scales and how the divergence across scales is contributing to unpredictable or sub-optimal results; focusing instead only on how actions and processes interact across the scales (as is the focus of process tracing, for example). While others may argue that the quantitative approach has some hope of operationalizing and collecting reliable and generalizable data regarding concepts at each scale, and so try to collate and analyse data from and then across various scales, there are endless challenges to (a) operationalizing the underpinning concepts pertinent to peace and conflict dynamics, (b) collecting valid quantitative data on many of the specific scales, and (c) analysing that data to reliably trace influence circulating between scales. Similarly, many qualitative methods based on documentary analysis, elite interviews, process tracing, and comparative case analysis, have some potential to track the interaction of concepts across scales (and this is true also of multi-level mixed-methods approaches), but even these do not demand an examination and understanding of the divergent conceptions, experiences, or expectations of actors at each scale (see, for example, Hirblinger and Landau 2018). As such, many qualitative studies too fail to assess the underpinning concepts which are foundational to the dynamics of conflict, peace, and peacebuilding. However, as the next section will show, if used appropriately, trans-scalar EPR could respond to both of these challenges.

Trans-scalar ethnographic peace research
EPR should not be considered a new form of research. The approach follows in the footsteps of scholarship within Anthropology which has long examined many different dynamics pertinent to peace and conflict. This includes many descriptions of culturally specific processes of conflict resolution and mediation (Dillon 1976;Hamer 1980;Podelefsky 1990), research into the local dynamics and politics of transitional justice and peacebuilding (Avruch 1998;Wilson 2001;Honwana 2006;Vigh 2006;Bräuchler 2015), and studies of the local experiences of violence and recovery (Nordstrom 2004(Nordstrom , 2007Das 2007;Theidon 2013). The recent effort to promote ethnography as a tool for PCS scholarship is not, therefore, an attempt to develop a new method, but to highlight the value of this existing tradition for the study of the conceptual foundations of our beliefs, expectations and experiences of conflict, violence, peace and peacebuilding. Indeed, EPR should not be seen as a specific methodology at all, but as an approach incorporating various methodologies, each of which supports this focus. As is the case with other 'approaches' thereforequalitative, quantitative, realist, constructivist, or interpretivist -EPR allows for the use of many different methodologies for sampling, collecting, and analysing data (Millar 2018a, 257); something Hennings recently described as a 'method repertoire' (2018,632).
But even given this methodological diversity, there are key characteristics which set EPR apart from the quantitative and qualitative approaches described above and which allow it to focus on the underpinning concepts that motivate action. Specifically, there are two characteristics which should be required of any research that claims to be EPR; (a) thick description and (b) the ability to explain 'how' and 'why'. The first, thick description, is associated with traditional Anthropological ethnography (Geertz 1973) and requires that EPR must seek 'to provide a nuanced and deeply textured presentation of the local setting and lifeworld' in which the dynamics of violence, conflict, or peace occur (Millar 2018a, 259). Even most qualitative PCS research does not attempt to provide this. The second leads from the first, in that this thick description seeks to provide a 'contextualized understanding of 'how' and 'why' events unfold as they do and are experienced as they are' (Millar 2018a, 259). Other methodological elementsincluding both methodological diversity as noted above, and another lynchpin of Anthropological ethnography known as reflexivity (Davies 1970; Anderson and Stewart 2016)are not required, but facilitate EPR. They do this by enhancing the ability to interpret the implicit concepts that are foundational to expectations and experiences of conflict, violence, peace, and peacebuilding at and across scales.
It is only by developing a deeply textured understanding of context and culture that we can come to interpret the concepts that underpin and motivate expectations and experiences. Such understanding can only ever emerge from close observation and engagement in the field, which allows the researcher to then form theories about the concepts underpinning beliefs and ideas regarding conflict, peace, and peacebuilding. As such, EPR is an inherently inductive and post-positivist research endeavour, which certainly sets it apart from deductive quantitative approaches, but also from many of the qualitative studies carried out in PCS today. The EPR approach, therefore, is key to examining and interpreting the concepts which fundamentally underpin beliefs about, expectations for, and eventual experiences of peace interventions.
But it should also be apparent that EPR is centrally important if we want to understand the complex interaction of actors, institutions, and processes across scales. To study how influence circulates, in essence, it is not sufficient to examine and interpret the concepts, expectations, and experiences only of the recipients of peace intervention, which has been the primary argument of the EPR literature to date (see Millar 2014a, 2018a), but we must also examine those at all of the higher scales involved in the planning, funding, design and implementation of such interventions as it is often the inconsistencies and invisible incoherence across scales that generates unexpected effects. But can EPR be used at various scales? And, perhaps more importantly, can it be used to trace the entangled dynamics related to peace and conflict across scales?

Local scale EPR
The first scale at which ethnography has evidenced its value for analysis of complex dynamics is at the micro-scale. As this is largely the mainstay of traditional Anthropological research there is really no limit to the number of examples that could be provided. But even restricting ourselves only to recent publications pertinent to peacebuilding we can identify exemplary studies such as Shaw's study of memory and trauma in Sierra Leone (2002), Das's research regarding local recovery from violence in India (2007), or Theidon's analysis of the local processes of reconciliation in rural Peru (2013). All such studies follow in the best of Anthropological tradition in their detailed focus on the most local of sociocultural, political and economic interactions, relationships, traditions and norms to provide 'thick description'. They are, first and foremost, exemplars of the level of fine-grained analysis and analytical reflexivity necessary to interpret action within a complex sociocultural environment. As such, the value of EPR for the examination of conflict and peace dynamics at the local and sub-national scalesincluding of course the underlying and invisible concepts pertinent to such dynamicsshould be the least controversial claim made in this paper.
However, there are also substantive challenges to EPR at this scale and, indeed, it is quite instructive to note that these 'exemplary' studies are by scholars who would be considered Anthropologists and not PCS scholars. Indeed, while their work is referenced frequently by PCS scholars, particularly those working in the countries and regions of their studies, those cited above do not publish their work primarily in PCS journals. They did not go initially to their field sites to study processes related to conflict or peace, such as mediation, conflict resolution, reconciliation, or peacebuilding, but were each studying other issues within the local setting which eventually gave rise to studies of topics which were of interest to PCS scholars. As Anthropologists first, they spent time learning the language pertinent in the setting, studied other elements of local culture, and spent years learning to interpret the interactions and communications they witnessed around them while in the field. To put it quite bluntly, as Anthropologists they were able to commit a level of time and attention to their fieldwork rarely available to PCS scholars, and certainly not to those coming from most of the disciplines contributing to the field today, such as Political Science, International Relations, Law or Economics.
However, EPR does not require mimicry of the traditional Anthropological form of fieldwork and the same length of time that has traditionally been required. Instead, it must be based on a 'recognition of the importance of culture in shaping how individuals see and experience their world and a willingness to try to understand alternative experiences' (Millar 2014a, 6). In this sense fluency in the local language and multiple years in the setting observing tangentially related social engagements may not be necessary for later interpreting local concepts, expectations and experiences pertinent to peace and conflict. However, it should still be clear that more time, more language skills, and more experience with a diversity of cultural practices at the local scale will always be better, as such knowledge and experience is exactly what provides the contextual information necessary to interpret local concepts foundational to experiences of conflict, violence, peace, and peacebuilding. This means, in short, that while years of fieldwork may not be required, substantial time on the ground in the setting is still necessary, and extended fieldwork is central to EPR efforts (Millar 2018b). The challenge for EPR at a local scale then is twofold: the first relates to the willingness of non-Anthropologists to engage with and accept a post-positivist perspective which centralizes the unobservable or implicit concepts usually marginalized (a challenge which is true for all scales); and the second is the more practical question of the willingness of the dominant disciplines in the field to provide their students with the skills and time necessary to conduct such research.

Intermediate scale EPR
At the intermediate scales, as noted above, the first challenge still holds in that scholars must still accept a post-positivist perspective. However, to some extent the second challenge may not be so daunting at this scale, which should make EPR at the intermediate scales a very attractive prospect for many PCS scholars. By 'intermediate' I mean the scales between the local and the global (centring on the national but including also the sub-national and the international). At these intermediate scales EPR can be extremely valuable for the same reasons as at the local. It is imperative that research is able to provide the necessary contextual detail to interpret the underpinning concepts which govern peacebuilding policies and practices at these scales. Specifically, institutional ethnography may be the ideal tool for examining and understanding such dynamics in the form of the norms, discourses, operations and power dynamics within institutions at these scales. Institutional ethnography has been developed more within Sociology than Anthropology and has to a great extent been a tool for critical analysis of the internal culture of institutions. Work by scholars such as Smith (2005) and Devault (2006) evidences clearly the potential for this form of research to uncover the often unseen or opaque dynamics that govern collective behaviour within institutions.
There is far less institutional ethnography generally than traditional Anthropological ethnography, and even less conducted specifically to feed into the literatures on conflict, violence, peace, and peacebuilding. But a handful of studies evidence the value of institutional ethnography for peace research, including work by Richmond, Björkdahl, and Kappler (2011) and Klein (2018). In addition to such studies of individual institutions at the intermediate scales, it is also important to note that this scale is particularly interesting because actors and institutions at these scales have a great diversity of capacities and purposes, depending on the context. At these scales you find both actors with substantial power and influence, such as national politicians or business leaders, as well as actors with very little power, such as some civil society institutions. EPR, therefore, becomes as much about examining the interaction between such actors and the variability of and inconsistencies between their concepts, expectations and experiences, as it is about these interactions within individual institutions.
The challenges posed to EPR at this scale, as noted above, are potentially less extreme than at the local scale. While the need to accept a post-positivist perspective remains (and this is challenge enough for many) in most post-conflict contexts there are at least fewer linguistic and social barriers to engaging with actors at the intermediate scales. In most such contexts regional and national elites will speak a European language (often the language of the prior colonial power) and they are often much more amenable to discussing issues pertinent to peace intervention processes than are many actors at the local scale. Indeed, it is at these levels that you find many people who believe themselves imminently qualified to contribute to PCS studies and talk about their ideas and actions. That said, actors at these scales may also be far more capable of misdirection and deception, more tied into networks of power and privilege, and more willing and able to manipulate the researcher for their own ends. Thus, EPR at this scale may pose different methodological challenges.

Global scale EPR
Finally, the highest scale of all, above even the international, is the global. It is at this scale, as already discussed, that the Western centric order has achieved hegemonic status over the course of the last 150 years and has reigned as the unspoken (although often resisted) 'universal' since the end of the Cold War. This, as noted above, is a European 'local turned global', or as Tsing described it, 'an engaged local' … 'charged and changed by [its] travels', which, even as it is changed, is engaged in lands far past its home by actors and institutions unaware of its boundedness; who are 'notoriously bad at seeing the limits and exclusions of their knowledge ' (2005, 8). But it is for exactly this reason that EPR at the global scale is so necessary. For it is the many interrelated concepts and ideas of Europe and its colonial inheritors in the 'Global North' today which govern how powerful institutions think about, plan for, and administer peace interventions. It is exactly these concepts, taken as given by the powerful, which then interact with and interpose themselves on the many scales downstream. It is exactly the unobserved and unthought nature of such 'universals' that gives them their hegemonic power (see Bourdieu 1977) and which, at the same time, requires EPR at a global scale to examine, interpret and make evident the actually constructed nature of these apparent realities which so forcefully influence all of the scales below.
While there are even fewer ethnographic studies which can serve as exemplars for a global scale EPR, there are a handful that stand out as potential examples. Tsing's already cited research into how concepts, practices and institutions travel between and are engaged within different contexts (2005) could be joined, for example, by the earlier work of Arjun Appadurai, whose now classic Modernity at Large (1996) examined the work of imagination and creativity in shaping global culture, and James Ferguson and Akhil Gupta's ethnographic study of the operations of neoliberal governmentality (2002). While few in number, therefore, there are examples to be followed (see also examples in Burawoy and Kerdery 2001). But while these examples provide inspiration for an as yet unrealized global scale EPR, the challenges to such studies are quite substantial. While on first impression such research may be about analysing global patterns or aggregate data, in reality each of the examples above is rooted just as much in the fine-grained analysis of particular localities and specific practices as they are in broader global patterns. As such, these studies require many of the same skills as ethnography at the local or intermediate scales, but also the ability to identify, analyse, and explain how those local particularities are related to and interact with broader national, international and global patterns. While the studies noted as potential exemplars hold out the best hope for a trans-scalar EPR in that they evidence the huge potential for ethnographies to span these scales, they also illustrate the grand challenge of such work.

Trans-scalar EPR
And this, in many ways, epitomizes both the promise of and the most daunting challenges to EPR for the study of complex interactions between the radically varying concepts influencing expectations and experiences of conflict, peace and peacebuilding across scales. While EPR has been developed thus far as an extension of the 'local turn' in PCS, and has, as such, been focused on local concepts of conflict, violence, peace, and peacebuilding, in reality this is only one small sliver of the deeper interactions that need to be examined and understood. For EPR to reach its potential and contribute to the development of a trans-scalar peace (Millar 2020b), it must also reach beyond the local to examine the unobservable concepts which serve to underpin and legitimate decisions, actions, and policies at higher scalesthe national, international, regional and globaland then examine how these concepts interact across scales when embodied in the decisions, actions, policies and institutions they motivate.
EPR, in this light, might be seen as a deeper form of the kind of trans-scalar case studies recently discussed by Hirblinger and Landau (2018), in that it is interested not only in the decisions, actions, policies or institutions themselves, but in the concepts that underpin and motivate them. Ideally, therefore, EPR would be deployed at each scale and then identify the interaction of concepts through their manifestations as actions, institutions, or processes and read these interactions for the interplay of narratives and power across scales. We see hints of the potential for such studies in the work of Keen (2006), Duffield (2007), or Kaldor (2012), which actively telescope their analysis back and forth between the local, national and global scales to construct narratives about contemporary conflict. But what they gain in breadth they lose in the depth, thoroughness, thick description and contextualized interpretation of the motivating concepts which are evident in studies such as Tsing's 'Friction' (2005), and are more emblematic of the breadth and depth of analysis that is necessary for a truly trans-scalar EPR.
For all its potential, however, this can only be accomplished by preparing PCS scholars for the often daunting, sometimes dangerous, and always time-consuming work of conducting ethnographic studies on the ground, at the various intermediate scales, and even in the global arena. As it stands today few PCS schools, programmes or departments prepare students to conduct extended periods of ethnography or actively encourage them to consider this as an option for their PhD research (not to mention the various other disciplines noted above which train scholars who eventually go on to contribute to the discipline). Furthermore, funders consider large-N quantitative studies to be the more rigorous, informative, and generalizable form of research, thus funnelling resources away from EPR efforts, and many of the most influential journals in the field rarely publish ethnographic work, thus limiting the influence of ethnographic outputs. Without substantive reform, PCS as a field will be forever unable to extend EPR out from the local to identify and trace the complex patterns of interaction and influence motivating negative experiences of peacebuilding interventions or even generating further conflict and violence across the local, national, and global scales. The challenges are, therefore, substantial. But in examining Tsing's work, for example, even pessimists may be tempted by just what could be learned if such a process were applied to the study of contemporary peace and conflict.

Conclusion
As has been described, in peacebuilding today ethnography is, for good reason, primarily associated with examination of 'the local' or 'everyday' impacts of peacebuilding interventions (Millar 2014a(Millar , 2018a. This makes sense given the substantial theoretical contributions of the 'local turn' and the more recent 'ethnographic turn' (Mac Ginty and Richmond 2016, 223). However, as this article has demonstrated, the nature of conflict and peacebuilding today requires that EPR be extended out from the local in order to better examine the complex interaction of concepts pertinent to peace and conflict dynamics across the local, national, international, and global scales. This is particularly so given the recent recognition in the field that has inspired this Special Issue (that peacebuilding regularly occurs amidst violence), and the three challenges facing the field, of complexity, radical alterity, and, most importantly, radical alterity within complexity.
As I noted above, the recognition that peacebuilding occurs amidst violence is actually not clear enough about the relationship between the two phenomena. I suggest instead that we admit to ourselves that in many cases violence occurs because of peacebuilding, or in fact, that peacebuilding in its dominant form seems often to require violence of different kinds. This is true in no small part because violencecultural, structural, or directis fostered by the frictions generated through the interaction of the divergent concepts underlying peacebuilding institutions, interventions and processes at different scales; generated as those frictions are by radical alterity. Although it is not often recognized, radical alterity is itself a major contributor to the complexity of contemporary conflict and peace dynamics, in which a variety of actors and institutions interact across scales and with non-linear effects.
As noted above, the recognition of complexity in peace and conflict dynamics has led to a number of new methodologies being proposed, and these have been both quantitative and qualitative. But both have generally been quite limited in their ability to examine and understand the pertinent complex interactions. This is partially because these approaches have largely restricted their interest to specific scales, with quantitative approaches applied largely to interactions at macro-scales (national and above) and the qualitative approaches mainly applied at the micro-scales (sub-national). But it is also because almost all of these studies limit their purview to the observable, measurable, knowable, and, eventually, controllable aspects of peace intervention. They are, as such, largely blind to the fundamental role played by the concepts (and by extension the expectations or experiences) which underpin our beliefs about, support for, or resistance to actions, institutions and processes pertinent to peace and conflict dynamics. While some, such as Millar (2014a) and Firchow and Ginty (2017) have started to focus on the underlying concepts of conflict, peace, security, etc. in local settings, they have so far not extended this study to the higher scales involved in peace interventions and have not yet developed mechanisms for studying the interaction of concepts across scales.
EPR can overcome both of these limitations. First, it is founded on an inherently inductive and post-positivist approach based not on a priori theory development and testing but on the inductive development of insight and interpretation based on thick description within the scope of study. It is centrally concerned with examining and interpreting the underlying and often invisible concepts which motivate or underpin beliefs, actions and processes. As such, it stands quite firmly at odds with the positivist methodologies, whether quantitative or qualitative, recently proposed for the study of complex conflict dynamics across scales which have no means by which to explore these hidden aspects. Second, and again unlike the other recent proposals, EPR is applicable at all of the pertinent scales (local, national, international and global). It opens up, therefore, the possibility for PCS to examine the underlying concepts and motivations of actors and institutions at each scale. But, most importantly, it allows scholars to also track and trace the manner in which these concepts interact across scales by examining both their subtle composition in the minds and worldviews of the actors involved, and how they are manifested in actions, processes, and institutions in the world. Trans-scalar EPR would seek, in short, to read these interactions for the interplay of narratives and power across scales. In this way, trans-scalar EPR can allow us to better examine and understand the invisible drivers of complex conflict, violence, peace, and peacebuilding dynamics.

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

Funding
This work was supported by Leverhulme Trust: [Grant Number Leverhulme Research Fellowship (2018-2019].